source
stringlengths 620
29.3k
| target
stringlengths 12
1.24k
|
|---|---|
How to work with supercapacitors? I'm thinking of using a supercapacitor to power a device with low power consumption. I have one 1 F capacitor rated at 5.5 V and I plan to charge it using input voltage of 5 V. What precautions should I take when charging it? I'm afraid that it will overload my power supply if I just connect it directly. Any other tips would be appreciated too. <Q> I have had several students who have used super cap chargers . <S> They provide a lot of nice features like variable current limit on the charging, the particular one that I linked to will also make it so you could use 2 caps with lower voltage on each one and it will make sure you don't over voltage one. <S> The only down side is the chips for these are VERY small and can be difficult to solder if you aren't very experienced. <A> You are not only going to fry power supply, but also damage your supercap <S> - it's internal connections cannot withstand high peak currents. <S> Charge it with some 0.1A current using current-limiting resistor in the simplest case. <S> Same applies for discharging - never short leads. <A> Buy a CC/CV power supply, I picked up a 3A one for £40. <S> I've used them before and they won't damage themselves, even with a continuous short. <S> You can set the CV control to about 5 volts and the CC control to the maximum charge current - 1 or 2 amps. <S> Initially, current will stay high, but as the voltage rises the supply will drop the current automatically. <S> Plus, the CV control ensures the cap won't ever be damaged by a high supply voltage. <A> Step 1: Specify the constraints on the charging behavior that you want, based on the characteristics of the capacitor and/or your power supply. <S> (e.g. current limit of not more than 0.2A or 0.5A or 0.1A or whatever) <S> Step 2: <S> Build a circuit that will meet those characteristics. <S> If the current is low, you can just do this with a resistor. <S> It turns out if you do the math that charging a capacitor through a resistor from 0V to some fixed voltage, always wastes exactly 50% of the energy involved, so it's not hugely efficient, but it's not horribly inefficient either. <S> If you need higher efficiency, you need a switching power supply of some kind. <S> The ones appropriate for this application would produce nearly constant current into the capacitor with a voltage limit. <A> If it were my power supply, I'd have one that can handle dead shorts with some sort current limiting, and not worry about it. <S> This is inefficient, as power will be dissipated within the power supply. <S> Inductors resist changes in current, but eventually settle into acting like straight wire. <S> You'll want something like shown here <S> http://www.hammondmfg.com/153.htm with the laminated core and metal mounting bracket. <S> What size? <S> To get a rough idea... <S> The fundamental equation for inductors is dI/dt = <S> V/L. rearrange this for dI = <S> (V/L)dt and imagine capital Greek deltas instead of <S> the infinitesimal lower case 'd's. <S> The change in current is from zero to the initial inrush current. <S> Let's put dt at 1 sec, since, although we're not trying to build a resonant circuit, the characteristic frequency for a circuit with 1F and 1H is 1 radians/sec. <S> Suppose you get a nice hefty 1H coil. <S> The initial voltage drop across the inductor is 5V. <S> So we get 5 amps. <S> In real life, the ESR of the capacitor and the internal resistance of the power supply, as well as its ability to deliver only so much current, means you probably wont' get that. <S> To limit the current to say 1A means a 5H coil. <S> The least efficient way to limit current is with a resistor, or a current regulator such as an LM317. <S> But this is likely much cheaper than a 5H coil, unless you already have a coil like that in a junk parts box (these were common in old 1960s/1970s TVs) <S> Supercapacitor charging circuits can get pretty involved for applications that harvest small amounts of energy, where not one microwatt should be wasted, where >90% efficiency in charging is vital. <S> In some applications, the charging current might be only a small factor greater than leakage current. <S> A good read with technical details: <S> http://www.energyharvestingjournal.com/articles/using-a-supercapacitor-to-manage-your-power-00001921.asp
|
If you're worried about inrush current harming the supply, put an inductor in series with the capacitor.
|
Advice making simple solar shed vent from salvaged panel I have a 10'x10' shed that gets hot as hades inside, and want some ventilation involving pushing air through a hole I would punch in a wall near the roof peak and screen over so my friends the wasps don't move in. With a gas-powered mower and weed whacker, gas cans, and spare propane tanks for the bbq and camping lantern, it gets thick with fumes, and I'm uncomfortable with how hot it gets in there. I have myself a salvaged solar panel in a weather-proof frame, which in its former life charged a battery that then ran a string of little LEDs at night. It being March, I don't know what to expect from summer sun in terms of juice production, but I did test with a single 60w incandescent bulb in a reading lamp pointed directly down at the panel. The multimeter registers 6.5v/1.3a at a distance of about 6", and with the bulb just about directly over the panel, just barely 9v/2a. This is direct from the panel; the battery and charging bits have been removed. They had stopped charging and/or holding a charge, which is how I inherited this. I have a fan in this vein that came from an old external hard drive enclosure which puts out enough air, I think, to do the job: en.wikipedia.org/wiki/Centrifugal_fan It's marked 12v/0.38a. It would be inside the shed blowing through the opening I make. Solar shed vents sold for this purpose are $40+ and reviews say they all just plain suck and are worthless. What I'd like advice on is how to take this panel's output, which I can't imagine ever getting above the 9v/2a number, and probably hoping for the 6.5v/1.3a is optimistic, stepped up to 12v at sufficient amps to turn this fan. I found the items below which to my novice eye seem like they might be steps in the right direction: http://tinyurl.com/457sr8t http://tinyurl.com/4vkqapo Am I out of my mind? Could a device like one of the above take the panel's output and put out enough to get me 12v/0.38a if I get enough direct light? Is there some other, better way to go about this based on the panel I have? <Q> That was a long question :-)You are not out of your mind and think in the right direction. <S> In fact you can get twice much amps you need for your fan. <S> So yes, get one of these modules, and add some extra capacitors on the output (some 10'000 uF) to help it deal with FAN's pulsed current consumption & startup current. <S> You can even get way bigger fan, cap will help to start it and it will work even if you have just 30-40% of the power it needs. <S> You just need separate switch for DCDC board and FAN - so that you turn DCDC first, and after 1 second - FAN (so that cap is charged). <S> And finally, direct sun gives much much more juice than 60 Watt lamp. <S> So you really need more powerful fan & more powerful DCDC. <A> Thank you for taking the time to reply. <S> That's a big help. <S> Two questions: <S> Could you please clarify what you mean by " You just need separate switch for DCDC board and FAN - so that you turn DCDC first, and after 1 second - FAN (so that cap is charged). "? <S> I understand the general idea of charging the capacitor to give the fan the boost it needs to start, and I think you are saying panel -> <S> dc/dc converter - <S> > <S> capacitor -> <S> fan is the wiring sequence, but I don't understand the "separate switch" part and the "after 1 second" part. <S> I was able to start and run a smaller 12v/0.17a fan directly via the panel's 9v/2a output (see below). <S> Do you think is there a way with caps that I could start and run the 12v/0.38a fan the same way if the caps gave the fan the initial push? <S> This might mean not having to buy anything (I have a variety of caps on hand). <S> I did try to wire it the panel and the bigger fan together, but it wouldn't start, and even with a push (I spun the fan, both directions) I would not run. <S> I'm confused about how 9v/2a could start and run a smaller 12v fan, but could not run the bigger 12v fan even if I started it manually. <S> Thanks! <S> More to add just to contribute the to world of Google search results for someone else at my level that this might help: BarsMonster is absolutely correct about the sun versus my lamp... <S> I took the panel outside in moderate sun in March in North America and it generated the same 9v/2a I'd seen with the 60w lamp right on top of the panel. <S> I take this to be the max the panel can produce. <S> This is not what I expected as an electronics novice. <S> I thought sunlight would be faint compared to the 60w bulb practically sitting on the panel. <S> Go sun! <S> I was able to connect the panel directly to a much smaller fan (cpu fan from the old days) that wants 12v/0.17a, and 9v/2a would start and run that fan w/o a converter or capacitors etc. <S> So I guess the boost from the capacitor BarsMonster described could get the larger 12v/0.38a fan going, then what the panel puts out could keep it moving, albeit more slowly than full speed. <A> (This was a month ago - did you get it working?) <S> They look like they're rated for sufficient power as well. <S> The only concern might be what happens when the output drops below that which can run the fan but not yet zero; if the fan is stalled, does it dissipate current as heat in the fan or DC/DC converter? <S> Depending on the converter, intelligent ones will go into "brownout" mode and switch safely off in this situation.
|
My reckoning is that the capacitor won't help, but the DC/DC converters are exactly the sort of thing you need.
|
What are these plug-like cable endings called? I've been looking for things like the plugs displayed on these orange wires in this photo: What are they called? I imagine that this is a part that I could buy somewhere. <Q> All they are are stiff peices of wire soldered to stranded Wire with a molded rubber case. <S> you can buy them from adafruit , also called bread-boarding cables, jumper cables, (premium) wire jumpers. <A> I'll bet that they can easily be located in different sizes, etc. <S> elsewhere. <S> http://www.sparkfun.com/products/9194 <S> This page refers to them as 'leads' and gives some DIY instruction on making them stackable: http://www.instructables.com/id/Stackable-Jump-leads-for-BreadBoards/ <S> And again, for the DIY type person, on page 6 of this catalog is a crimp-style terminal. <S> http://www.molex.com/catalog/web_catalog/pdfs/C.pdf <A> Patch cords? <S> Unless I'm missing something, functionally, they're the same as wires. <S> I wouldn't bother with them and would just use solid-core 22ga wire for solderless breadboard. <A> Those look like single row breakaway headers, like these: http://rndwarehouse.com/breakaway-headers-1x40-straight-singlerow-extended-height-p-213.html?zenid=83og3qmqjb3g10obpv92sbavo7 . <S> I've done this in the past as it's the cheapest way to get non-solid core jumper wires for breadboarding. <S> In order to make them, you get a 40 pin header like the one I linked to, and then carefully break them apart, either with clippers or two pairs of pliers. <S> You want to solder your stranded wire to the short end and then use shrink tube over the connection. <S> The long end plugs into your breadboard. <S> EDIT -- looking at the picture again <S> , they aren't breakaway headers, but I would still recommend going that route if you're okay with a little soldering.
|
OK, they are known as 'Jumper Wires' Here is a link to some that SparkFun sells.
|
Do RGB LEDs actually behave like three separate LEDs? I am trying to operate a lot of RGB LEDs for a project. I came up with a crazy scheme which only works if the RGBs work exactly as three separate LEDs. What I want to do is connect the R, G and B lines of all the LEDs with the same source of PWM and connect their grounds to separate pins(which have tri state logic). So when I want to turn on a pin I will give the PWM for the color on the PWM line and drive its ground pin low and ground pins of other RGBs to high impedance (theoretically separating them from the circuit). Will such a circuit work? <Q> Yes, but they are not full separated LED's, as @Joby Taffey stated they share a common anode or common cathode. <S> In order to use the circuit you said, you'll probably need to get one common anode one and connect the CA pin with the PWM output. <S> The other 3 pins (cathode for R, G and B) will be switched with an transistor for example. <S> You could use an common cathode, but in this case, with an PWM duty cycle of 100% (turned one) the corresponding led would be turned off. <S> One problem I'd see is the refresh rate. <S> You can't turn on the 3 leds at once because you want to control each color individually. <S> The solution is multiplexing. <S> You should use an enough high PWM frequency and enough low multiplexing frequency, so the multiplexing will not interfere with the color intensity while trying to avoid flicker with multiplexing. <S> And keep in mind that the output luminance will be only one third of the tri-led maximum. <A> Sometimes they're common anode, sometimes common cathode. <A> RGB leds are common anode of common cathode as Joby has stated. <S> your way will work but will take up 4 pins (R, G, B, PWM) depending on if your pins can sink/source the current. <S> I would use the below example depending on your specific RGB led, and drive each led line with PWM. <S> Common Cathode example <S> As a side note, you could hook up the example below and drive Vcc with PWM and drive each pin high or low to turn on and off colors
|
Yes, RGB leds are usually 3 coloured LEDs in a single package with 4 wires.
|
Simple way to wire a stoplight I have a retired stoplight - with real lightbulbs. I am looking for a simple way to wire it to have it automatically move RED-GREEN-YELLOW cycle just as the ones on the street. I have manually switched it but was looking for something more "automatic" that I can put together as a learning project with my 15 year old. I saw this answer: Most Simple Stoplight circuit But I am not sure about the power requirements for the circuts - the bulbs are actual bulbs as used and are 120V and I think 100 watt. Ideas or suggestions? <Q> Funnily enough, I was just making a stoplight myself. <S> A triac is a bit like a diode switch that is able of switching mains voltage, and when coupled with an optical isolator, can be driven (safely) by low-level logic voltage (from a microcontroller, ttl, or anything else you like). <S> They cost about 99 cents each (plus $2.15 for a typical optoisolator) so its a very cheap solution. <S> Here is a prototypical TRIAC control circuit, you could require additional capacitors/resistors depending on load type and the characteristics of the particular triac, so see their datasheets. <S> The input resistor is set at the right value for the diode inside the optoisolator to receive the correct current from your logic source <S> The more expensive solution is to use a solid-state relay, which will accomplish approximately the same thing for a bit more. <A> In order to drive that load, you would need relays or power transistors. <S> You cannot drive relay coils directly from uC, you would need transistors to get more current (uC can supply not more than 20mA). <S> Another option is to use high-voltage BJT/FET transistors, but you also cannot drive them directly from uC. <S> So all you need is 3 power transistors or relays, and 3 low-power transistors. <S> This way you can switch nearly unlimited load. <A> Well, since your main criteria is simplicity, I'd say use the classic method. <S> A 1 RPM motor, a plastic shaft and rotor, some un-etched circuit board, and two wipers--one onto the center of the rotor and one off the rotor's edge. <S> Commutate the mains onto the board with the rotor, leads come off the edge of the board to the lights. <S> Divide the board into sections: Half of the board goes to the red light, 1/8th to the yellow, and the rest to green. <S> Cut the copper away from any mounting points and cover/box for safety.
|
I found a microcontroller (see: arduino for a cheap, premade dev-platform) with a triac for each light to be the cheapest, simplest way to achieve what you want.
|
Code based tools for drawing schematics Is there any code based tools for drawing schematics. I only need to draw some simple diagram, including some 16pin Microcontroller, wire, power supply... I'd like to use code to specify the pin and components, and let the software do the layout for me. Kinda like define a graph with graphviz dot format. Some thing I have in mind is: define a Microcontroller pin1.label="Vcc" pin1.goto=other components How can I design a circuit using a text based format? <Q> <A> You could use TeX http://www.ctan.org/tex-archive/graphics/circuit_macros/ <A> Try Fritzing . <S> It's an all in one simplified schematic and breadboard designer with a straightforward XML format. <S> http://fritzing.org/support-us/developer/fritzing-sketch-file-format/ <A> SchemDraw and lcapy are two Python schematic drawing libraries. <A> gschem has a simple text-based file format. <S> But if you want to automatically place the wires making connections on the schematic, then you are asking for autorouting, similar to a pcb layout package. <S> I don't know any software package that implements this. <S> A schematic is a visual way to represent a netlist. <S> It is for humans. <S> The computer (and the electrons) just care about the netlist. <S> So you could design the pcb from the netlist and skip the schematic step altogether. <S> Schematics with big FPGAs or processors end up as netlist anyway: each pin is shown connected only to a named net. <S> If that is what you are planning, there's no need to draw a schematic. <A> I think that you're going to struggle to find or make such a package. <S> To represent connections, there are plenty of netlist formats; several EDA packages use text-based formats for internal representation, and almost all can output a text-formatted netlist. <S> This is probably the easiest third of what you want. <S> I'm not aware of any that have a syntax that's easy to write and keep track of. <S> The labeling and definition functionality is typically contained in a part library. <S> Again, there are text-based formats for this. <S> Some manufacturers publish text-based pinouts of all their components to be used in generating a library. <S> This functionality should probably be separated from the netlists, i.e. you'll do #import and then instantiate one. <S> Laying these components out in a sensible way is the last third of the problem, but it would be 99.99% of the effort. <S> As Mark said, schematics are for humans. <S> There are some basic rules, like voltages are higher on top and signals/data flows left to right <S> that help make schematics more readable, but there's a lot of information contained in the arrangement of symbols that would be difficult, if not impossible, to represent in a netlist or other code format. <S> Autorouters are a very high-dollar code project, and autoplace would be, if they could get it to work well. <S> Those tools are used on PCBs, which don't generally contain semantic information in their placement. <S> If you can generate a schematic from code that's almost as readable as a human-designed schematic, then you'll be a millionaire in no time.
|
You should take a look at skidl which uses python to design and test a circuit and it will create the netlist https://github.com/xesscorp/skidl
|
Recomendation for a digital inverter made of discrete components I have a digital circuit in which I need only one logic inverter, and both PCB size and power consumption are serious constrains. So I thought about using a pair of complementary SMD MOSFETs (the classic coupled gate configuration) instead of an IC with many ready-made inverters. Which MOSFET would be more recommended in this case (both for small size and low gate-drain current)? <Q> You can now get individual SMD inverters like this one , which is just one of the six gates found in a classic 74AHCT04 style package. <A> You'll never beat an IC for size with discrete components; tcrosley's link only draws a few µA and the footprint is a SOT23-5, which is only 2.1x2.4mm. <S> However, if simplicity or education is your goal, you can do fairly well with a simple FET inverter. <S> With a logic-level PNP and NPN, you can make a push-pull totem-pole circuit that will have similar power requirements to an inverter IC. <S> You can also replace the top transistor with a resistor: <S> This creates a simple inverter in a small space. <S> When the input voltage is above $V_{TO}$, the output will be connected to ground through the transistor. <S> When the input voltage is small, the transistor will be off, and the output will be pulled up to $V_{CC}$ through the resistor. <S> There will be power dissipation in the resistor when 'driving high' (really a passive operation for this circuit), and current will flow through the resistor when you're driving low, so make the resistor as large as your load capacitance, switching speed, and input impedance allow. <S> This circuit can sink lots of current, but sourcing current requires a smaller resistor. <S> If you need to source current, use a PNP and flip the design upside down. <S> For best performance, use two transistors. <A> When I have a digital circuit and I discover I need "just one more inverter", I usually put in an entire chip -- perhaps a 74HC132 -- because usually a few minutes after I get that inverter wired up, I discover that I need yet another "just one more inverter". <A> There are literally hundreds of different kinds of individual transistors and transistor arrays currently on the shelf of my favorite suppliers that would work.(By "work", I mean "has a threshold Vgs is small enough that you can drive it with digital logic, rather than something that needs 6 V or more to turn on"). <S> Randomly-picked examples: BSS84 p-channel FET; and 2N7002 n-channel FET, each in a discrete SOT-23 package <S> Diodes Inc. DMC2004DWK-7 dual complementary MOSFET pair in a single SOT-363 package
|
A pair of complementary MOSFETs in the standard static CMOS inverter arrangement should work fine.
|
Usage of the #pragma pack(1) compiler directive on embedded applications I have recently come across this #pragma pack(1) preprocessor directive and was wondering why it is being used? I Googled the usage, and found it has other options such as push,pop etc. Has anyone used this on their embedded application? I would like to know some examples of how/why you have used this directive and on what kind of processor? What are the pro's/con's of using this directive? Thanks. <Q> It may save RAM - which is often precious in a microcontroller. <S> Packed structs also allow for casting directly over memory buffers for data interchange: <S> void f(void *buf){ <S> struct <S> ip_header *hdr = <S> (struct ip_header *) <S> buf; hdr->dst = 0x8000001 <S> ;} Be careful where you use #pragma pack . <S> It's globally scoped (as it's in the preprocessor), so forgetting to turn it off will affect any #include files. <S> If you only mean to pack certain structs, consider using GCC's __attribute__ ((packed)) . <S> However, due to alignment issues, packed structs can impact performance. <S> For example: When packed into bytes: struct{ uint8_t a; uint32_t b; uint8_t c; uint8_t d:}; Will be stored as (not accounting for endianness): a, b0, b1, b2, b3, c, <S> d <S> However, many architectures require 32bit accesses to be aligned to 32bit addresses. <S> With the packed struct, the machine will have to make several byte accesses then stich them together again. <S> Faced with the above struct without packing enabled, the compiler could reorganise it as: b0, b1, b2, b3, a, c, d <A> Another reason to pack structures on byte boundaries is to ensure the alignment of the members when transferring data between different processors. <S> I often need to transfer data structures between an MCU and a host PC application. <S> The PC will pack structures on 32bit boundaries unless instructed to pack them on 1 byte boundaries. <S> A PIC24F MCU will pack structures on 16bit boundaries unless instructed to pack them on 1 byte boundaries. <S> By instructing them both to pack their structures on 1 byte boundaries, it ensures the data is in the same place when accessing it on either end. <S> Without it, you would need to pad the structures with reserve bytes so the data members would align properly. <A> The traditional use where I've seen this being used is for reading information from files. <S> For example, you can define a struct whose members match those of a BMP file header, and then read all of the header in one swift read operation. <S> OK, so BMP might not be the best example (its header does not have alignment issues on 32-bit systems), but you get the idea. <S> I suppose this is just as useful in the embedded world.
|
#pragma pack(1) ensures that C struct items are packed in order and on byte boundaries.
|
Wifi enabling my Arduino I'm having a lot of trouble finding wifi enabled arduino shields. All the places I've seen them are out of stock for some reason. Would this thing work? How would I hook it up? http://www.rovingnetworks.com/wifly-gsx.php Basically I want to build a network enabled (wireless) robot with the arduino. So you can see it's clearly better if I can get a wifi shield. Edit: To follow up on this post, I did end up getting the Wifly shield from Sparkfun. I was unable to get it working with the Arduino (Duemilanove) I had, and I ordered an Uno on a hunch, and it works on that! Connects to APs and can host an ad-hoc network. Unfortunately no security options exist for ad-hoc. But I'm okay with that. <Q> wit a custom firmware installed. <S> a google search for "Arduino Fon" will give you some examples. <S> Jim <A> Here is the later version of the WiFly Shield - US $70. <A> These modules are the basis of most WiFi shields for Arduino. <S> If you are reasonably comfortable with your soldering ability, you could dead bug a module like this without too much difficulty. <S> Just connect all the powers and grounds properly, and bring TX/RX to your Arduino. <S> That said, if you dig a little deeper, you'll find the shields in stock at vendors like SparkFun, Futurlec, Adafruit, etc. <S> If one vendor is out of stock, others will have it. <S> If you're just looking at using WiFi for remote control, another option might be ZigBee. <S> WiFi might be a little overkill for simple remote control, unless you really want the novelty of controlling the thing through the web. <A> http://www.watterott.com/de/Arduino-WiFly-Shield <S> i looked for it on the sparkfun page, but it says deprecated. <S> Watterott ships internationally so you could order there, eventhough its here in Germany. <S> I hope i could help.
|
Depending on how comfortable you feel with Linux, a really cheep option can be to use a cheep wi-fi router (such as a La Fonera) Watterott sells a nice Wifi shield called "WiFly Shield" its about 80€, get it here:
|
How to choose between little endian or big endian for the processor settings? We have just selected a new processor for our application, and one of the project options available to me is to select the Endian mode: Little endian or Big endian (under Big endian two more choices, BE32 and BE8) Can you please tell me what factors I should consider to make this choice? I was under the impression that a processor is either Little endian based or Big Endian. Does this mean that based on the option selected, the compiler would fix the endianess to suit the processor, if required? Thanks a lot. <Q> Incidentally, I know of an argument that historically favored little-endian format: in a number of CPUs, most notably the 6502, code could run faster using little-endian format than would have been possible with big-endian. <S> Consider, for example, an instruction: $1234: LDA ($8A),Y ; <S> Load byte from Y register plus the value stored in $008B-$008AAssume $008B-$008A hold $5432, and Y holds $10 <S> The execution sequence is: $1234 read $B1 - Fetch opcode$1235 read <S> $8A - Fetch operand$008A read $32 - Fetch LSB of target while adding $01 to $8A$008B read $54 - Fetch MSB of target while adding $32 to $10$5442 read $XX - Fetch byte from target <S> The MSB and LSB of any address to be read or written must be either based on a computation done in the last cycle with values already in the CPU, or else must be the last fetched value verbatum. <S> The third cycle can read the first byte of the target address (i.e. $8A) without delay; if the processor wanted to read the second byte first, it would have to waste a cycle computing the address. <S> The fourth cycle can compute the LSB of the target address while fetching the MSB; if there's no carry out of the LSB, the LSB (computed) and MSB (fetched from memory) will be ready simultaneously. <S> If the 6502 were big-endian, it would be necessary to waste a cycle on every indexed access (instead of only losing a cycle when indexing crosses a 256-byte boundary). <S> I'm unaware of any processing advantage to big-endian format beyond the fact that it was arbitrarily chosen as the "standard" network byte order. <A> It's kind of an overdone argument these days - either works well, if you're writing code that actually cares <S> and you want it to be portable <S> you're going to code any dependencies out anyway <S> so why might you choose? <S> Big Endian: almost all network protocols are BE - this is probably the main reason to choose these days Little Endian: there are a few data tricks (like passing an int and treating it as a char) that are available - but none of them are portable so best avoided <A> The time that engineers spend thinking about endianness is not free in a commercial project. <S> On an ARM processor, little endian is the path of least resistance. <A> What processor is it? <S> It is generally the case that there is a preferred mode that everyone uses. <S> For instance, the Microchip C32 compiler for the MIPS-based PIC32 uses little-endian format. <A> It only matters if you're interfacing with the outside world i.e. if data has to be passed out from the processor or read into the processor. <S> If you are not careful, you end up with flipped-order bytes. <S> As others have mentioned, networks use big-endian format. <S> So, that is something to consider. <S> Another possible consideration is when you have a multi-processor setup and they need to share data. <S> Also, some processor architectures can switch endianness in software.
|
Unless you have an overriding justification, definitely choose little endian, as this is ARM's traditional endianness.
|
Grounding a soldering iron At school our soldering irons don't have grounded tips. I need one as i'm soldering MOSFETs in to my PCB. So what solutions could I use to ground my soldering iron tip? <Q> it somewhere to your iron, and make sure you 0Ohm resistance between other end of the wire and the tip. <S> Then connect other side to your grounding - if you don't have one, you should call for electrician to make one - there is no universal way to do it yourself. <S> But personally, I solder FETs without any grounding, and noone ever failed. <S> You should only be really worrying about grounding when you work with fragile very high-frequency ones - they indeed are too easy to kill, unlike the majority of FET's and IC's. <A> I often like to solder live on a circuit with a grounded low voltage psu, but if my iron is grounded touch the + rail and a short occurs, so I use a 1 Meg resistor in my iron earth leg to prevent this. <S> It will still discharge stray static voltages. <S> It also saves time in keeping to disconnect the psu when soldering on test parts. <A> Actually, the MOSFET won't care whether your iron is grounded, it will only care if the iron has the same potential as the circuit you're working with (and your body, if you happen to hold the the circuit you're soldering with your hands). <S> Many soldeing irons provide a plug (banana or similar) which you can connect to the ground of your circuit to eliminate ESD. <S> This only makes sense to do if you already wear a grounding strip , since your body has much more capacitance which can accumulate electrostatic charges, and most parts are killed by touching them with your hands, not your iron. <S> If you need to solder a particularly sensitive component with a cheap non-grounded iron, consider disconnecting the iron from mains power before doing so. <S> Such irons tend to have poor insulation from mains, constantly leaking a small current to whatever circuit you touch with it. <S> You may be totally unaware of it (you don't feel a few uA), but it may be enough to damage some sensitive low-voltage parts.
|
No rocket science there - in the simplest case you just take wire, and mechanically connect(no soldering obviously :-) )
|
Supply power to mini-ITX from car battery Two related questions have been combined here as the answers and concepts overlap: (Q1) We are building a robot that will be controlled by mini-ITX PC. The power source for robot will be detached car battery. Question is how to power mini-ITX mother board (and HDD) using this battery? What devices are needed? Edit: I think I need DC-DC converter that converts from battery voltage and converts it to 12V and stabilizes it. Right? (Q2) Originally from here I have a linux box powered by a "deep cycle" battery, and I"m wondering, would it be more efficient / possible for me to convert (or replace) the AC power supply in the unit with a DC one so that I don't lose energy in the conversion from AC power to DC, optimizing the use of power from my battery. DIY link: http://www.instructables.com/id/Convert-an-ATX-Power-Supply-Into-a-Regular-DC-Powe/ <Q> Mini-ITX does not standardize a power supply. <S> If your board takes 12V, then yes, you need a DC-DC converter that converts the battery voltage to 12V. However, according to Wikipedia , a 20 or 24-pin standard ATX power connector is conventionally used. <S> This connector has a number of independent supplies for 3.3V, 5V, and 12V. <S> Here's the PDF standard <S> if you want to develop your own - It's not trivial! <S> You'll need to know the power requirements of your mini-ITX board to determine how many amps each supply needs to provide. <S> However, there are a number of ready-made solutions. <S> If you have an inverter available, a standard ATX power supply will work fine. <S> mini-box.com has been debugged (good or bad, your choice) on this site before; their m-series power adapters are specifically designed for in-car use. <A> Yes, and you can buy special power supplies for the job. <S> They are sold as "auto" or "car" power supplies, and usually include circuitry to send an ACPI shutdown signal when the ignition is turned off. <S> that are not much bigger than the ATX plug that goes into the motherboard - ideal for Mini-ATX and ITX systems that don't need much power. <A> For a DC/DC converter that does exactly what you want <S> see: http://dren.dk/carpower.html <A> Minibox offers a number of supplies for motherboards, take a look at this and others at the site. http://www.mini-box.com/M1-ATX-90w-Intelligent-Automotive-DC-DC-Power-Supply <A> Yes. <S> You would expect to get efficiency gains. <S> Efficiency-design capabilities of designers of new and old equipment and Cost / efficiency manufacturing tradeoffs which have been made will have significant impact on what you can achieve. <S> (Efficiency of "what the market will accept" and "best we can do" implementations varies substantially.. <S> Whether it's worthwhile depends on cost versus other options. <S> To go 12VDC - 110 <S> VAC <S> - PSU/+12/+5/... Requires energy losses in the upconversion to AC - probably in the 10% to 20% losses range depending on how hard you try. <S> Probably not less than 5% <S> no matter how hard you try. <S> Making 5/12/... from 110 VAC or 12 VDC will incur losses however done. <S> 110 VAC supply should do this with 5% to 20% losses <S> BUT that's an estimate <S> - I've not measured a typical PC supply to see. <S> Easily enough done. <S> Making 5/12/... also requires a switching power supply. <S> 12 -> 12 is lossless IF battery version of 12V is acceptable. <S> As battery can droop to 10 - 11 V it MAY need boosting and as it cab go over 12 V <S> you may need buck boost. <S> Say 5 - 15% losses range. <S> 5% only with great care. <S> 12 <S> -> 5VDC requires a standard buck converter. <S> Say 10% losses. <S> Other voltages say 3.3V and CPU voltages of 1.XV. <S> Say 10 -15 %. <S> Could be 5% with great care. <S> So overall estimates: <S> 12VDC - 110VAC - 12VDC etc at (80%-90%) <S> x 90% = <S> ~ 70% - 80% overall. <S> 12 <S> VDC -> psu at say <S> 80% - 90% overall <S> Difference is perhaps break even (80/80) to as good as 90/70 <S> =~ 30% improvement. <S> Liable to be somewhere in between - say a 15% to 25% gain. <S> This translates directly either into increased run time or reduced PV panel capacity if solar powered. <S> Or greater safety factor. <S> If you do this in a solar powered system then also using an MPPT PV panel controller would make great sense.
|
You can also get some very nice little tiny power supplies
|
Regulator cooling in a confined space I've designed a simple PWM RGB LED slow fader to be used as a garden lighting effect. My circuit works great, but I under-estimated the amount of heat that is generated by the 7805 linear regulator. It's mains powered, with a 6Vac transformer and a 5V 1A linear regulator. If all the LED's are at full brightness then it draws around 600mA. I've mounted my circuit board in a plastic enclosure with a transparent lid, and it's rated at IP67 (and I want to keep it that way!). I've put quite a small heatsink on the regulator, and it takes around 3 or 4 hours of continuous use to get to a temperature that is just about too hot to touch, I'm guessing around 70-80°C. My plan is to give this to my Dad for him to use in his garden, but obviously I don't want it to melt or catch fire. My questions are: Is this an acceptable temperature for it to operate at? Is it likely to get any hotter if left on for longer? I didn't want to test this as I didn't want to damage it, but the datasheet says the operating temp is max 125°C so I guess it would be ok. What can I do to make it run cooler, given that I don't want to drill vents into the enclosure and ruin it's IP67 rating? If it does happily operate at a high temperature, do I need to be worried about heat conduction through the PCB tracks into other components that may be damaged? Will it melt the solder? <Q> The regulator should have thermal limiting (check the data sheet for details), and will shut down if the temperature gets too high. <S> It won't damage any other parts, and definitely won't melt any solder! <S> Using a larger heatsink will reduce the temperature, of course. <A> You probably have too high of an unregulated voltage coming out of your transformer. <S> Ideally, you want to have the VDrop across the linear regulator be pretty much <S> just the regulators drop-out at full load. <S> What is the voltage on the input to the linear reg? <S> Also, as Leon Heller was mentioning above, there are some pretty easy-to-implement switching voltage regulators, particularly Nation Semiconductor's Simple Switcher Regulators. <S> You can not really breadboard them successfully . <S> OTOH, if you're willing to spend a little bit more, you can get drop-in switching replacements for something like the 7805. <S> A number of manufacturers sell <S> little <S> 7 <S> 805-sized <S> switching regulator modules , which even have the same footprint. <S> However, they're generally about $10-$15 dollars. <S> Edit - Does anyone know why my dollar sign symbols are vanishing? <A> I fixed my overheating linear regulator just by adding a big chunky heatsink. <S> I lifted the regulator out of the PCB, attached hook up wire to the leads, and mounted it on a big chunk of metal I cut from an old power supply. <S> I've run it for about 4 hours <S> and it's only slightly warm... <S> problem solved!
|
A switching regulator will be a lot more efficient and will just get slightly warm. You can also use a low-drop-out regulator to reduce the dissipated power (the 7805 is not LDO). However, for any switching regulator like these, to make them perform properly, you really need either some really tight hand wiring, or a custom PCB.
|
Arm development board with lots of PWMs I am looking for a Arm development board for a new project I am working on (64 RGB LED POV Globe). POV globe speed questions I have lots of experience with the Arduino boards but I would like to take a step up to a more powerful board for this project. Also I have been wanting to play with ARM for a while and this seems like a good project to start on. Requirements Not insanely expensive, <= $250 preferably Lots of PWM outputs at lest 10 idealy 24 Program able by the USB port. 32k of chip memory Nice to have Ethernet, Not that I would need it for this project but most of my other projects require it. SD card or other persistent storage, Logging, loading of settings files, ect... The first board I looked in to was the beagleboard as it has a huge community and lots of resources but it does not appear to have any PWM outputs? Next I looked in to mbed-NXP-LPC1768 from sparkfun But from as far as I can tell from the spec it only has 6 PWM pins, and I was looking for more of a complete board with power supply, ect.. My question: Suggestions on a Arm development board with lots of PWMs? Am i missing something with the PWMs on the beagleboard? It seems strange that the board would not have any I/O <Q> TLC5497 has 24 channels of 12bit PWM. <S> Just shift the data in serially and latch it. <S> http://focus.ti.com/lit/ds/symlink/tlc5947.pdf <S> You can control one of them with three Arduino pins (or possibly less, using Roman Black's Shift1 protocol - <S> http://www.romanblack.com/shift1.htm ) <S> They can also be strung out in series to control as many as you like with the same three Arduino pins (check the example on page 1 of the datasheet). <S> The only drawback is that as you add more chips in series of the same serial stream it'll take proportionally longer to shift all the bits in for each refresh. <S> If you don't like this performance cut, you can use another pin as another serial stream and use the same CLK and XLAT pin, and cut your refresh time in half again (by running half the chips off of stream A and the other half on stream B). <A> The OMAP chip used on the Beagleboard does actually have PWM capability, but you will have problems implementing it as there don't seem to be any applications using it. <S> Here is one of their boards being used to control an LED cube. <S> Ethernet support is available on one of their boards. <A> A LPC2917/ <S> It's grouped in 4 segments though, so you'll only be able to choose 4 different frequencies of PWM, but you can control the duty cycle of each individual pin. <S> For the purpose you're trying to achieve, that's all you need. <S> Atmel has plenty of PWM too on the AT32UC3L032 chips and are not as expensive as the NXP chip if you would decide to make your own hardware. <S> It also comes in a more friendly package, which is nice. <S> I haven't seen a real development board looking quickly, so USB programming is not possible with Atmel chips without using one of their programmers. <S> The XMEGA64 and 128 has the same capabilities, but are 16-bits proccesors. <A> Some of the Luminary Micro ARM Cortex-M3 devices have up to 8 PWM outputs. <S> http://www.luminarymicro.com/products/motion_control_pwm.html <A> Atmel sam3x8e (used on the Arduino Due) has 12 PWM outputs, loads of memory, USB, the works. <S> It is a bit expensive, though. <S> You'll have to pay at least 10 USD. <S> The AT32UC3L032 may be a better option if you want cheap.
|
LPC2129 has up to 24PWM outputs. The easiest way to implement a large number of PWMs is to forget the ARM and use an XMOS chip - they are often used to control large numbers of LEDs. Low-cost development boards are available that may be controlled via a USB port.
|
types of resistors and inductance So while trying to interface a 5v SPI to a 3.3v SPI device, using the inline resistor method outlined below, http://www.sparkfun.com/tutorials/65 I discovered that at high clock speeds (above 200khz), my clock signal was being corrupted by the inductance in the resistor (with an oscilloscope, the clock looks signal looks triangular shaped) I am using carbon film (through hole) resistors for prototyping and I heard that these can have high parasitic inductance and not suitable for high frequency use. Is this information correct? What types of resistor instead should I be getting (digikey link would be appreciated)? Right now, I'm just working on a protoboard but eventually I would like to manufacture this prototype on a PCB using surface mount components. It seems like from what I read about surface mount resistors, the inductance is not typically a problem for applications under 20Mhz. <Q> In order to observe a clean square wave at 200kHz, the rise-time needs to be around 250ns (about 10% of the pulse width). <S> The sparkfun tutorial shows the clock coupled through a 10k resistor. <S> Now, to get 250ns risetime, you need an RC time constant of 114ns giving a capacitance of about 11pF. <S> So the sum total of the circuit input capacitance, the stray capacitance and critically, your 'scope probe capacitance must be less than 11pF. <S> I reckon this is your problem. <S> If the waveform looks OK at the driven end of the resistor and the input capacitance of the driven device is low, you might be OK but you won't be able to prove it without a very low capacitance 'scope probe. <A> I'm taking you at your word that inductance is the problem, though I find it dubious at a mere 200 kHz. <S> Some axial thru-hole resistors are often created and trimmed by spirally cutting the element over a form which can give it some inductance. <S> This is usually seen in metal film resistors, though possibly in some other film-type resistors as well. <S> Perhaps more fool-proof, some parts suppliers (Digi-Key does at least) have filters <S> so you could just select "Non-inductive" and be done with it. <S> You are quite correct that SMT resistors have very low inductances. <S> I'm not sure of the trimming patterns, but whatever it is, it's done in a plane, so at most you'll have a turn with an extremely small area. <A> When looking at fast signals it is usually beneficial to use the 10x setting on the probe, for the reasons discussed in the other posts.
|
Carbon composition resistors, however, should possess very low inductances as it's more of a block of material.
|
Making Etchable PDFs in linux Hey I am trying to make a etchable pdf for a project of mine. It is supposed to be a 2x9inch board. But I cannot find a good software to make this. Eagle's free version will not allow me to make this size and I cannot understand gEDA. Is there other simpler option? <Q> Eagle's professional edition layout tool can do 64 x 64 inches. <S> Since it seems that you know Eagle, that's the simplest option. <S> It is not the cheapest option, but only because it's not free (as in beer). <S> Like most powerful tools, gEDA has a significant learning curve. <S> It's not impossible to understand; it's just not as intuitive as Eagle. <S> Keep trying. <S> As with many open-source projects, features come before documentation, so many of the tutorials are out of date. <S> I suggest you follow the geda-user mailing list to get the most current information; it's quite active and the devs want to help! <A> <A> Not an ideal solution, but DesignSpark PCB is reported to run fine under Wine. <S> From their site: What are the limitations on designs? <S> There are no intentional limitationson designs. <S> Unlimited schematic sheetsper project, up to 1m squared of boardsize and no limits on layers allow youto get your creativity flowing withoutrestraints.
|
KiCAD is the best you're likely to get: http://www.lis.inpg.fr/realise_au_lis/kicad/ On Linux, gEDA, KiCAD and Eagle are the best games in town.
|
Is it important that vias are plated on a PCB? As in the title. I've noticed several PCBs of mine have vias which are not plated - they do not have a characteristic "gold" shine to them. I suspect this is because I was pushing the limits of drill sizes - my fab specifies 12 mil and I used 15 mil (20 mil including plating.) I've checked a few with a multimeter and they do have continuity, which is good, but I've only checked a few, so have no idea if there might be a few dodgy connections because of this. Is it something to be concerned about? Board passes DRC as specified by fab. I've seen quite a few high density boards like motherboards and graphics cards without plated vias. Here's an older board of mine which shows a similar problem to my newer boards. Note that some vias look gold, while others don't look plated at all. <Q> I Think you are using the term plating incorrectly. <S> Plating will decrease <S> the diameter of the hole, not increase it. <S> The larger dimension (the pad surrounding the via hole) is called the Annular Ring . <S> All the fabs I have worked with generally want a 0.005" per-side annular ring, or the via diameter + 0.010". <S> You definitely have some really hairy registration issues on the board you posted pics of. <S> It may work, but you're really pushing it. <S> Generally, you never want the via hole to break out through the edge of the annular ring, which is happening a few times in the picture you posted. <S> Anyways, Registration refers to the accuracy between a fab-house's etching and drilling process. <S> Basically, if a fab house etches a circle in the coper of a board, and then drills a hole in the middle of the circle, how close is the drilled hole to the center of the copper circle? <S> Remember, drilling a board and etching <S> it are separate processes, and involve the board being unmounted and remounted in different equipment. <S> Generally, you can get as good registration as you are willing to pay for, and it looks like your boards are from a pretty budget board-house. <S> You need to allow enough annular ring that you never wind up with the via hole too close to the edge of the pad. <S> This is generally specified by the fab house (they should have a minimum annular ring spec on their board requirements). <S> However, it is important to remember that they may run your boards anyways , even if it does not meet their minimum required specs. <S> The board house may run the boards anyways, and just refuse to fix any issues if they don't work out. <S> This is particularly common in China, where the general philosophy seems to be "Let the Buyer Beware". <S> Anyways, I think the reason you are finding your vias a bit odd looking is that you have tented your vias , which is the practice of covering the annular ring and the hole for the via with soldermask. <S> With 0.015" vias, you will occasionally get a contiguous layer of soldermask over the hole, and they will look different. <A> Get your boards tested. <S> It costs a bit more, but it's worth it. <S> You won't be able to see the plating inside the vias unless you use a microscope. <S> Hi-tech boards like motherboards and graphics cards always use PTH. <A> If you're talking about the annular ring outside of the via <S> hole-- Check your gerbers. <S> You might have specified solder mask right up to the hole. <S> If a few of them test good, they're probably all good, especially at 15 mils (vs. the 12 mil lower limit of the fab). <S> Of course, if this is a really critical board, you should have paid for 100% E-testing ;-)
|
You won't be able to see the plating inside the hole without a microscope or some other really good magnification.
|
Safe Powering Methods - Working with Children I'm carrying out a research project this year at uni, which will include creating small electronic sensory devices for children (under 7) with autism. Obviously, safety has to be taken very seriously. Only relatively small amounts of current will be used - probably powering simple sensors on an Arduino board - but seeing as my knowledge of all things electronic is pretty limited, I thought I'd try and get some suggestions first. From the research I've done, I found PoE at littlebird: http://littlebirdelectronics.com/collections/freetronics/products/4-channel-power-over-ethernet-midspan-injector - which looks like a pretty good solution, seeing as I want the ability to control the device remotely, over a local network. A benefit also being that I can keep the number of wires poking out of the device to a minimum and also not have a direct lead going to a wall socket. So far, I'm really only at the stage of planning the 'core' of these devices, which would be an Arduino board with limited sensors attached, possibly using the internal parts from a Wii remote. Because I don't yet have confirmed participants, I don't know exactly what sensory needs I'll be working with (each autistic child will have specific sensory issues) and therefore exactly how much power I will need for the devices. I will aim to keep it at a minimum though, simply for safety reasons. <Q> A PoE switch is normally used to power IP phones over an Ethernet cable without having to have an additional power adaptor. <S> It sounds like you're doing a something similar with your devices. <S> The midspan injector will certainly be a cheap way to add power to an unpowered Ethernet port, with the added advantage of being able to control the supply voltage. <S> You can get PoE switches with everything built in, so I also recommend looking at these as it might be cleaner and easier to use one, but it might cost a bit more. <S> You could also make your own version of the midspan injector for even less. <S> Here is a table of the pinout for RJ45 ethernet with and without power: <S> Make sure that when you split out the supply from the ethernet on the Ardunio end you have something to reduce the voltage to 9-12v. <S> If you're using an injector then you can control this and supply a lower voltage, but if you use a standard PoE switch you need to be aware of this. <S> What type of sensory devices are you going to be connecting to your PoE ports? <A> You probably have a human research review board to pass you decision thru (at least in the us ). <S> You still need to isolate from any other electrical source, probably including your network. <S> Check with the review criteria. <A> The Power over Ethernet (PoE) is likely most suitable if the devices are network-aware, otherwise one Ethernet cable is pretty much equivalent to a (low current) power cord or a USB cable. <S> For any devices that to be "worn" or "attached" by test subjects, I would strongly consider an un-tethered design, powered using rechargeable batteries. <S> For small signal (i.e. sensors, data logging, no motors, very simple LED lighting) <S> this option should be easily economic and practical for up to 24 hour periods or longer. <S> NiMH would be my first choice due to low cost and wide availability, with Lithium as a second. <S> Just ensure you use an appropriate charger for the battery type, and things should just work. <S> Note that most rechargeable batteries such as 'AA' size cells, only provide 1.2 rather than 1.5 volts <S> , so 4 of them is not sufficient for 5V needed to stably power an Arduino, while non-rechargeable (i.e. disposable) cells such as alkaline or zinc-carbon would. <S> With electricity you need to be aware of both voltage and current. <S> The Arduino itself takes an input range of 7-12 DC, up to 500mA <S> (I don't have a reference on average / typically current, but would guess around 100-125mA or less), from an external DC power source. <S> The USB port can also be used to draw 5 volts, up to 500mA from a powered USB hub or powered USB port. <S> Using a low-cost low current (e.g. 250-500 mA output ) AC-DC power adapter ( wall wart ) would be a default method if there is non-trivial power consumption, or needs to continuously operate for a long period of time. <S> A modern switch mode power supply (SMPS) based unit can be had for a modest cost, and is light weight, being able to dispense with the need for a large power transformer encased in it. <S> Combined with the power limiting capabilities already included in the Aruino (UNO) from the resettable fuse <S> (PTC, I believe) used for USB power source, and/or the linear voltage regulator used to regulate power from a AC-DC power adapter which also includes over-voltage, current limiting, and short circuit protection unless you need addition power requirements (e.g. motors, high power LEDs) you can use the protection built into the Arduino as sufficient for electrical shock/burn protection.
|
You'll need some RJ45 sockets and a power supply, and just simply connect the data pins straight through, and the supply to the output sockets to the power pins. Batteries are often the simplest safe power supply. One thing you need to be careful of is that 'standard' PoE is 48v, which will drop to as much as 32v depending on the length of your ethernet cable.
|
Why does radar processing always require FIR filters? Digital signal processing in radar applications is usually done using finite impulse response filters. Why is that the case? Wouldn't the use of an infinite impulse response filtering be much faster and feasible since we are talking about processing in the gigahertz range? <Q> Radar often uses phase as part of the signal. <S> After down converting you are going to have a signal that is normally "relatively" low frequency(kHz compared to GHz). <S> When designing a Digital filter you can get very good brick wall filters . <S> The major difference is that FIR filters can be symmetric and have linear phase delay. <S> Linear phase delay means a flat group delay which correlates to the phases being preserved. <S> When phase is a major part of your signal(in CWFM radar it makes up the position part of the signal) <S> any phase delay can cause a incorrect reading. <A> AFAIK they are not really doing the processing in the gigahertz range (how many chips do you know that can perform several multiplications and additions for each sample at multi gigahertz sample rate). <S> The only way it can work is by up- and downconverting (using mixers) to baseband and operating on that. <S> Returning to your question: IIR filters are fast but their performance is as good (or rather as bad) as their analog counterparts. <S> FIRs on the other hand can have any phase and frequency response if you are willing to pay for lots of taps. <S> OTOH long FIRs can be quite efficiently computed using FFT overlap-add methods. <S> In general check The Scientist and Engineer's Guide to Digital Signal Processing book. <S> It is a great book about DSP with lots of practice and not too much theory. <A> Both answers are very good, but I hope my 2cents and a different perspective will add some value. <S> I am in the middle of having to do essentially the same thing as a radar system. <S> Baseband versus RF <S> First off, Radar can be viewed very similar to wireless communication systems in that you have a baseband signal that you modulate, transmit, and then demodulate. <S> Often phase shift keying is used in comm systems. <S> In PSK, you are trying to detect a phase shift, usually between -pi and pi and based off of that phase shift you would determine the bit that was being sent. <S> In radars, the delay/reflection time creates what can be viewed as a phase shift. <S> The only problem is that the delay in a radar is more then likely going to be greater then the 2*pi span that you can detect normally. <S> In order to get a greater range you have to turn off your signal and wait long enough for you to receive the reflected signal. <S> In Radars, you could use an envelope detector and/or an RF mixer. <S> An RF mixer is the same as a multiplier. <S> DSP <S> The key at this point is to remove all noise and determine the time delay between when you transmitted and where your current signal is at. <S> Just like the comm systems, this delay can be viewed as a phase shift. <S> Because of this you want your filter to have a linear phase response otherwise you wouldn't be able to tell (at least not easily) <S> what delay your signal actually had. <S> Here is an example of the frequency response of an FIR filter <S> I just made: <S> As I am sure you can see, the phase graph is a straight line, so it is linear. <S> Now here is an example of an IIR with similar passband and stop band specifications: <S> The phase graph isn't so linear this time. <S> Now it is possible to reverse this phase effect in your delay time calculations, but it usually isn't worth the effort, instead people just use an FIR filter.
|
FIRs can also be quite easily tuned on-the-go to get adaptive filters .
|
What software is used to design and simulate Mac computers, iPod and iPhone circuits? Apple computers are great. I love my MacBook, however, I haven't been able to find any decent circuit design and simulation software that runs on a Mac. This brings me to a question: If Mac OS X has no professional electronics design software, then Mac electronics are designed on PC's? This would be interesting news for "MacFanaticsPCHaters" and for all of us trying to develop hardware on a Mac. <Q> "What circuit design and simulation software runs on a Mac?" <S> Some of those tools, such as McCAD, were originally designed to run on a Macintosh. <S> After Mac OS switched to a Unix base (starting with Mac OS X), quite a few tools originally designed to run on Unix were ported to the Mac. <S> "What software is used to design and simulate mac computer, ipod and iphone circuits? <S> " <S> I do not know , but this reminds me of a funny story. <S> One story ( perhaps apocryphal ) as repeated by Gordon Bell in 1997: "Seymour Cray ... <S> When asked what kind of CAD tools he used for the CRAY1 <S> he said that he liked #3 pencils with quadrille pads. <S> He recommended using the back sides of the pages so that the lines were not so dominant. <S> When he was told that Apple had just bought a Cray to help design the next Mac, Seymour commented that he had just bought a Mac to design the next Cray." <S> According to the Cray Supercomputer FAQ , Apple's Cray did a lot of work in the design of the metal molds used to produce the outer case of Macintosh computers -- the Cray simulated the 3D flow of plastic and rendered it as Quicktime movies. <S> Although I hear that wasn't the original reason Apple bought the Cray. <A> They (Apple) most likely use PCs and Altium , OrCAD , or something of that type. <S> Given the tight integration with the enclosures in many Apple products I would think that Altium is the best fit. <S> Most professionals aren't into the entire 'fan boy' attitude. <S> You use whatever is the best tool for the job. <S> Apple doesn't even go after the workstation market from what I've seen outside the graphics/video/content creation fields. <S> In an example of what I mean, for a very long time www.micosoft.com was hosted on Linux servers. <S> It wasn't moved to Windows servers until they developed a web server OS that didn't suck (it took them awhile). <S> For hobby use <S> Eagle and DipTrace should be sufficient and run on Macs. <S> You can always just boot up a Windows VM and use whatever you want. <S> Also, although Altium doesn't work all that well, it uses Direct3D for rendering and VM performance <S> isn't all that great yet. <S> I am not sure if OrCAD is using any 3D rendering nowadays. <A> That is an interesting question but I imagine that Mac hardware developers would rather work on some Sun workstations running some crappy commercial Unix than a PC. :P <S> Chip design is still mostly done on Unix AFAIK. <S> For the hobbyist: <S> You can you Eagle for PCB layout and schematics. <S> There is (or rather was, but can be still downloaded) <S> MiSugar for simulation. <S> I have also used ngspice from command line and MacSpice for plotting but they do not offer any schematic capture. <A> Many years ago, my Apple Hardware Design friends (including chip designers) used to design stuff using Unix workstations. <S> Now they use Macs (which are Unix workstations, come to think of it), and Xgrid to distribute the insanely large simulation jobs. <S> I run <S> LTSpice on my Mac <S> using WINE . <S> I also run MacSpice , EagleCAD and KiCad . <S> Last time I checked the gEDA tools, they didn't immediately compile and work. <S> I'm about to try out iCircuit , which looks more basic and costs $10, but has a nice UI.
|
The list of software tools for electronics design includes a bunch of tools that run on the Macintosh. There is a free Java simulator on which iCircuit is based, but it's not a Mac UI.
|
Is it possible to fire a spark plug from batteries I wanted to create a remote fired trigger for a spud gun so I can trigger it remotely. Ideally I'd like to use a spark plug to fire because it should work well for gas mixtures but the trigger is pretty small/simple. Basically it's a a little remote box with a push button trigger that runs off of batteries. It has a light, arm switch, and push button that triggers a relay but I currently don't have anything connected on the relay switch end. Would it be possible to charge up some caps and get enough juice to bridge a spark plug with a small battery source (4 AAA or AA batteries) or would I need to step it up to something bigger (lipo or lead acid). Note: I have used caps for filtering before but never for charging/discharging. Update: Just wanted to clarify. I know that a gas grill push-button igniter is a viable option but I'm looking for something more reliable. Ie, an igniter that I'm 100% positive it'll fire a good spark when I push the button. Check out "Tater gun Fires a Hunert yards" to see what I mean by unreliable. <Q> The answer for automotive spark plugs is no with a normal battery. <S> Only thing with that though is that your battery (9v) won't last long, as its "Juice" will be all gone in a few shots. <S> Hope this made sense :) <S> just tell me if not.... <S> You could make a spark though with your normal everyday disposable camera.... <S> Something like this: http://www.angelfire.com/80s/sixmhz/camera.html <S> P.S: If you used a larger battery (more volts) then you would have more juice :) <A> As mentioned previously, automobiles use ignition coils, a transformer (inductors) in an autotransformer configuration like <S> so: Image from " <S> How to use an Ignition Coil [in potato cannons]" Another possibility, though there are some problems with it, would be to just open up a circuit with some large inductance in it, which will cause high voltages (and sparks) across high resistances (e.g. a spark gap). <S> To do this, you'd need some switch that had a voltage rating much greater than your spark gap, which may be difficult to find. <S> From what I've heard though, pneumatic air cannons are much more controllable, and even better, they can be triggered using sprinkler valves which work at about 24 V (I use (3) 9 V batteries) <A> Here's what I did, I put a spark plug in the end of the spud gun, disassembled a $10 ebay stun gun. <S> soldered wires to the "legs" used the wire <S> leads to the spark plug (one to casing ground metal and one to the end of the plug) push button and viola!! <S> a few second spark!! <S> most of these stun guns are rechargeable too so no carrying batterys!! <S> One note of advice though prolonged use of the wire will cause burn spots and eventually lead to a short so use wire rated for spark plug use not some cheap speaker wire type stuff. <A> We have a gas oven. <S> That uses a piezoelectric crystal and a 1.5V battery to light a spark. <S> It's not a spark plug per se, but perhaps you could use a similar setup. <A> You can get electronics gas grill lighters. <S> Look for replacement types. <A> Maybe you could use onse of these: http://www.ebay.com/itm/Boost-High-voltage-Generator-DC-3-6V-to-40kV-Booster-Ignition-Coil-Power-Module-/231527277338?hash=item35e819e31a:g:cdcAAOSwqu9VIfeq Edit: " <S> high voltage generator power module 40 kv" is a good search term (if the link wouldn't work anymore) <S> 40kV should generate a spark on the spark plug.
|
But with a 9v battery by stepping up the current using a ignition coil (which ramps it up thousands of volts) then you will be able to use a spark plug.... Capacitors themselves can't step up voltages beyond doubling or inverting, to do anything else you need inductors.
|
I took apart an LCD module. How do I fit it back together again? To repair something, I had to disassemble it. Unfortunately, it had an LCD, and that used an elastomeric connector to connect to the LCD. Now I have an LCD, a PCB, and a spare elastomeric connector. I can occasionally get the LCD to work if I put them together, but it will not stay there, and the connection is very dodgy. Is there an easy fix? <Q> As Toby Jaffey said in his post: <S> You need to ensure the strips are clean and dust free, line them up perfectly then maintain pressure. <S> Be careful not to stretch or bend the strips while installing them as the contacts are often very small pitch. <S> This is probably the way to go. <S> I would comment that in the picture it does not look like the contacts are very small pitch. <S> I disagree with the other posters who have suggested using graphite to enhance the connection. <S> I would not put graphite or any other conductive powder/substance on the pwb or connector. <S> In order for electrical contact to be made, pressure must be continuously applied to the LCD/elasto/PCB "sandwich". <S> This is normally accomplished by the case, which is screwed together, pressing in on the structure from both sides. <S> The following procedure should work if you can pull it off. <S> You might need help from a friend or some kind of jig... <S> Sandwich <S> the elastomeric connector between the LCD connector and the board pads. <S> You will need to keep enough pressure on the sandwich to keep the connector from wiggling around. <S> Screw the case back together while making sure the LCD/elasto/PCB sandwich doesn't come apart. <S> This is assuming that you still have the case, and you will be reassembling it... <S> Test it <S> Note: the LCD connector is not visually distinctive. <S> Referring to the photo you have attached, I would guess that it is the translucent region at the bottom of the LCD module. <S> You can just make out the "ghosted" strips which are possibly electrically conductive. <S> Or maybe the clear strips are. <S> You will figure it out. <S> You probably could use an ohmmeter to test continuity if you are very careful and gentle. <S> Anyway, the black squares on the connector edges are the conductive parts of the elasto connector, and should be lined up with the PCB pads and the conductive LCD "pads". <S> I once took apart an electronic guitar tuner which was having display problems, and to my great surprise encountered one of those elastomeric connectors. <S> I had never seen one before. <S> It totally freaked me out when it fell off <S> and I could not see any adhesive or other fastening agent to reattach it with... but using the above procedure <S> I put it back together <S> and it's working, once again... <A> No, there is no easy fix. <S> I worked on a product with an LCD and two 100+ pin zebra strips. <S> It looked something like this: <S> Image Link Expired <S> In our design, there was a metal can with tabs poking through holes in the PCB. <S> These tabs were bent to maintain pressure on the zebra strips. <S> This picture shows the idea: <S> You need to ensure the strips are clean and dust free, line them up perfectly then maintain pressure. <S> Be careful not to stretch or bend the strips while installing them as the contacts are often very small pitch. <A> Once upon a time I took something apart one step too far. <S> I had situation like that LCD. <S> To put it back together, I used a very soft graphite pencil - 5B - to coat the contacts on both sides - LCD and PCB. <S> It actually helped, though the end result was still dodgy, it was quite a bit less dodgy. <S> This kind of pencil can be found at any artists' supply shop. <A> Same problems with LCD display on Beckman 3020 <S> DMM's I used to repair 100's of the units this way, <S> Strip down the LCD Display, Clean up the actual LCD with a cotton bud soaked in Isopropanol, leave to dry, then get the elastomeric strip drop this in Isopropanol, leave this for not a long time <S> say MAX five minutes, remove this carefully with miniature tweezers, place on a dry tissue for a time of one hour, place this back into the display, note clean up the Print on the module with cotton bud soaked also in Isopropanol leave to dry and replace all parts back, this will cure most problems, the Isopropanol actually expands most and leaving this to dry will shrink the unit back to nearly normal size, not to leave this strip in for too long as it might go out of shape, throughout the Beckman 3020 DMM they modified the Display unit to cure a known issue. <S> Philip 39 years in Electronics Testing <A> Make sure both the PCB and the strips are clean of any debris - so that there are no gap irregularities. <S> Also, you need to let the trapped micro air gap get squeezed out -- if you're lucky, the elastomeric strip will "seat" with time. <A> I do a lot of these zebra strip reconnections and it is always problematic. <S> The strips often leave silicone rubber debris on the transparent glass electrodes and on the PCB contacts. <S> This must be removed with solvent and perhaps an eraser. <S> Chlorinated solvent is good and an old typewriter eraser is good. <S> Careful with the transparent contact and if the PCB ones are gold plated. <S> Big problem is the zebra's PCB side takes an imprint of the bumps and valleys of the copper contacts on the G10 surface. <S> These ridges and valleys won't go away unless they are sanded off with sandpaper. <A> Isopropanol didn't work for me to clean the rubber remains of the zebra stripes on the LCD glass.. <S> Instead I used Nitro thiner on a cotton bud <S> and it worked like a charm!!!
|
Third problem is that the elastomeric strip takes a compression set with time so when it goes back together it has less clamping force to overcome dust and irregularities. Also, you need to make sure you have the right amount of clamping pressure. As for graphite -- graphite lubricating powder is available at home depot -- it's a dirty dusty mess, but takes the squeak out of door hinges -- and perhaps might work for this application, as suggested by others.
|
What device can measure the amount of water though a line? My dream device is a way to quickly make tea. I can use most of the parts from a Mr. Coffee tea maker but the last part is a way to start the flow of water, measure say, 1/2 a gallon and then stop the flow. It can't be hard . . . ice makers do it all the time. How does an ice maker do it? <Q> I'd use a load cell to measure the weight of the container, and turn off the water flow when the contents reach half a gallon. <A> Hobbyist supplier futurlec.com has in interesting-looking range of flow sensors, also. <A> You might consider an ultrasonic flow meter . <S> But, I'd either use a calibrated timer or prefill a chamber of known volume on demand (using a sensor to show when it's full), then empty the chamber into the cup. <A> Put the vessel that is receiving the hot water on a sprung base, and attach a microswitch to the base that is pressed down as the vessel fills with water that cuts the power to the pump (or heater if your using steam pressure to transfer the water) when the base is pushed down far enough. <S> You'll have to fiddle around with the sprung base and switch to get it to trip at the moment that it has been filled to the desired level. <A> Most appliances use timers or level switches as mentioned before. <S> To answer your specific question, you can use one of these . <S> I haven't tried one yet, but the price is so low (flow sensors tend to be expensive!!) <S> I can't resist. <A> Some water inlet valves include the pressure regulator, so you can buy one part and get both the regulator to give you a consistent flow, and a solenoid to turn the flow on and off. <S> You could try one similar to this . <A> Proteus Industries makes electronic flow meters. <S> They came to mind because I once helped swap one out in a high powered laser exciter; one was used to detect if the flow of water cooling the system was inadequate, and would shut down power.
|
Cheap appliances typically use a water regulator and a timer. Some coffee machines detect water level using a magnetic float inside the glass/plastic tank and a hall-effect sensor outside it.
|
How to do circuit analysis using Matlab? I often hear of people using Matlab for circuit analysis, but I never actually figured out how it is done. I assume that there is something more to it than just setting up equations by hand and solving them in Matlab. I'm looking for a good starting point. <Q> I use MATLAB quite a bit for circuit analysis. <S> Sometimes I prefer it to spice, other times I prefer spice, depends on my mood and requirements. <S> These are the following steps: 1: take the Laplace transform of the circuit 2: obtain the transfer function 3: <S> plot/analyse using MATLAB functions. <S> bode, impulse, freqresp and so on. <S> The trickiest part I find is to take the Laplace transform and derive your transfer function equation. <S> There are many examples and text books on taking a Laplace on the Internet. <S> Briefly the aim here is to get the equation in the form of $$H(s) = <S> \dfrac{as^2 <S> + bs + c } { <S> ds^2 <S> + es + <S> f} <S> $$ where \$a\$ to \$c\$ is the numerator and <S> \$d\$ <S> to \$f\$ the denominator in the example presented below. <S> To do this convert all you passive elements into complex impedances. <S> Thats is <S> C = <S> 1 <S> /sC <S> R <S> = R L = <S> sL <S> Next derive an equation for your circuit in the form of Vout/Vin. <S> For a simple low pass filter in the form of: Vin -------R-------------- Vout | C |------------------------------ <S> this would yield: \$ \dfrac{V_{out}}{V_{in}} = <S> \dfrac{sC}{R <S> + sC}\$ <S> Write the above equation in the form of num and den for MATLAB: <S> num = <S> [C 0];den = <S> [C R]; Then follow on using any matlab function you like to analyse the transfer function (bode), pole zero diagram and so on. <S> Below is an example of filter I was recently playing with and trying to tune the values: R1 = <S> 20e3;C1 = <S> 235e-9;R2 = 2e3;C2 = <S> 22e-9;num <S> = <S> [2*R2*C1 0];den = <S> [C1*R1*C2*R2*2 (2*C1*R1 + C2*2*R2) 2];g = tf(num,den);P = bodeoptions; % Set phase visiblity to off and frequency units to Hz in optionsP.FreqUnits = ' <S> Hz'; % Create plot with the options specified by Pbode(g,P);%[num,den] = <S> eqtflength(num,den); % Make lengths equal%[z,p,k] = <S> tf2zp(num,den) <S> % Obtain zero-pole-gain form <A> I some time use scipy <S> (a numerical toolset for python) to do circuit analysis. <S> And yes, that typically involves solving the circuit equations by hand first. <S> This is mostly helpful when doing tolerance analysis and sensitivity analysis on the circuit. <S> There is a book on the subject "Tolerance Analysis of Electronic Circuits Using MATLAB" that provides some examples of how to carry out the typical analysis on some common circuits. <S> It's not really a replacement for something like SPICE, but is useful when trying to design for good production yield over all component tolerances, or to account for component drift over time and temperature. <A> It includes RLC components, switches, electrical machines, etc. <S> You can create your own component and modify any parameters of the library components. <S> As you can combine your circuits with any Simulink blocs, any Simulink solver or any Matlab function, this tool is very powerful. <S> There is no need to solve the circuit equations first because you work in the Simulink environment. <S> It is originally oriented for power systems <S> but I think you can use it for any electronics circuit. <A> for simple RLC circuit with any topology(series and parallel) we can use " <S> rlcdemo". <S> It's good gui for analysis filters(LPF-HPF-BPF-BSF) <S> rlcdemo Analyzing the Response of an RLC Circuit <S> This demo shows how to use the Control System Toolbox(TM) functions to analyze the time and frequency responses of common RLC circuits as a function of their physical parameters. <A> You can use a program created in Matlab called SCAM (symbolic circuit analysis in Matlab), and is here: https://www.swarthmore.edu/NatSci/echeeve1/Ref/mna/MNA6.html <A> Besides SCAM in Matlab, there is also a slick online symbolic circuit analysis tool at CircuitNAV , which uses netlist files (from LTspice, Micro-Cap, TINA-TI, PSpice, etc) as the input and generates the algebraic solution for each circuit parameter. <S> CircuitNAV also provides a demo and a tutorial .
|
You can use the Matlab Simulink Simpowersystem toolbox to make circuit analysis.
|
Why is it so problematic to have close to zero standby power consumption? Each electronic device consumes electric power when it is "idle" unless it has a mechanical switch. I can understand that for example a TV with a remote control needs to "be ready" to receive a command from the remote control. But even a cell phone charger consumes power when it is connected to the outlet and not connected to the phone. For example, Nokia claims that one of its new chargers consume less that 30 milliwatts when not connected to the phone and they say it is very cool. I don't understand - the charger is a very simple device, what does it do with those 30 milliwatts? Why can't this standby consumption me made lower when we already have microprocessors with gazillions of transistors fitting onto a plate size of fingernail? What's the fundamental problem here? <Q> The mobile phone charger is a power conversion circuit which changes your power line voltage (110 or 220V) into something that is useful for your mobile phone (probably 5V). <S> To do this it needs to have some electronic circuity inside which has to be powered and it has to function even if there is not phone around <S> so it can detect one when you connect it. <S> The charger could be merely a mechanical device like the power socket itself but it would then require all the charging circuity to be inside your phone. <S> Unfortunately it is quite big and relatively heavy <S> so it would be inconvenient to carry it around all the time. <S> Regarding the actual 30mW figure: if instead of mW you consider the currents involved you arrive at around 300μA (30mW at 100V). <S> This also means a resistance of \$330\,\mathrm{k\Omega}\$. <S> It is quite difficult to work using resistances higher than and currents lower than this while still having to sense the moment when somebody plugs the actual load. <S> OTOH 30mW is really, really small. <S> The vampire current draw problems are not as important as many believe. <S> If you want a good review of many aspects of this then I suggest reading "Sustainable Energy – without the hot air" , especially the chapter on this topic <A> It is very hard to make a PSU that can efficiently provide a couple of mW for standby as well as several Watts for actual use, so it's not too bad that Nokia managed to get standby consumption down to 30 mW for a charger. <S> The only way to be more efficient would be to have a separate PSU just to handle the standby consumption of the main PSU, but that could double the cost of a small charger, so it's unlikely to ever be done. <A> Another point not yet mentioned is that energy-conversion devices (be they electronic, mechanical, chemical, or whatever) lose energy through a number of mechanisms. <S> Some mechanisms waste energy proportional to the amount of energy being converted, while others waste energy largely independent of the energy being converted. <S> A device which could convert 0-100W of power with 0.1W of waste would appear to be 99.9% efficient when used to convert 100 watts, but less than 1% efficient when being used to convert 1mW. <S> In reality, most devices lose energy through a combination of mechanisms, some of which are proportional to the amount of energy converted, but there are design tradeoffs. <S> For example, suppose the above device is used for a minute a day, and one could change the design to reduce the "constant" energy loss to 0.05W in exchange for accepting a 50% loss in conversion efficiency. <S> Saving 0.05W continuously would make up for the loss of 50W during the minute of use, but dissipating 50 watts for a minute in a small device would cause it to get very hot, which might cause problems in and of itself. <A> There are several problems. <S> But the most obvious one is that every consumer product has some kind of standby mode. <S> Don't forget, when your PC is off, it will easily draw about 100mA from the +5V. <S> A ATX PSU has a special +5V Standby line, which can deliver up to 2A according to the specification. <S> This is all just circuitry to monitor whether the PC needs to be turned on, wake up LAN, etc. <S> If so, it will probably active a 'bigger' supply to power the whole thing up. <S> Also, a switching power supply gets a peak effiency closer to it's maximum rating than to it's minimum rating. <S> A controller needs current to operate as well. <S> It needs to have a oscillator (generate a reference signal to PWM from), feedback, etc. <S> Low duty cycles aren't helping neither, because little energy gets powered. <S> 30mW is not much. <S> If you assume they would use a perfect AC to DC transformation, you will still use only 2.5mA at 12V. <A> Power Integrations have recently introduced the LinkZero-LP - a range of zero consumption integrated circuits specifically to eliminate AC-DC consumption when the phone is disconnected from the charger - whether the charger is connected to 115Vac or all the way up to 265Vac. <S> http://www.powerint.com/en/products/linkzero-family/linkzero-lp
|
For a charger I could imagine that most of the power is wasted to some monitoring circuit to see if a phone connects.
|
Best way to produce 50 VDC, 20 mA from 5 V Ok, so I need to generate 50 V from 5 V, with a load of 0 to 20 mA, for a mass-produced device. There are simple boost converter circuits like this LT3467 example: but it seems these are limited by the maximum voltage on SW (in this case 42 V). Are there any that can go directly to 50 V? If not, there seem to be several possibilities: external FET with higher voltage rating charge pump after the boost converter autotransformer instead of inductor 1: Is adding an external FET sufficient to get up this high? It seems like it, though I don't know how much current this can supply: ( from here ) 2: Can also use a charge pump / voltage doubler after the boost converter, without an external FET, something like this (with the FET built into the boost converter IC): ( from here ) Then the SW voltage can be half (or less) of the output voltage. One advantage of this is that it produces intermediate voltages, which might be useful. In this case, should the feedback be before or after the voltage doubler? 3: Other examples show an autotransformer replacing the inductor, and no external FET: ( from here ) ( more here ) So, uh, what's the simplest/cheapest/best way to do this? What are the advantages and disadvantages of each solution? Which external components cost the most? Which solution can supply the highest current? Which has the best output regulation? <Q> Try the Nat Semi Simple Switchers. <S> Here are some examples <S> I found with their interactive selection tool. <S> I used 5V to 6V input, 50V output and a current of 50 mA. <S> You should be able to design a suitable circuit on-line that meets your cost and performance requirements. <A> This will give you the most current, good efficiency, and the least issues. <S> Other solutions, like a typical boost topology in your first two diagrams can work, but you have to watch out for any duty-cycle limitations on the converter chip. <S> Switched capacitor converters can work for currents of 20 mA, but are noisy and don't regulatewell with a 10:1 increase in voltage. <S> I've seen these used for generating +48v in commercial products, but I wouldn't recommend them. <A> Obtaining or wiring your own flyback inductors is more difficult than just buying a medium-voltage MOSFET. <S> Just be sure that you use a chip with adequate output since an open drain output of a typical intergrated-switch DC/DC converter will not work with an external MOSFET. <S> I am not sure how you want to use the 10V switched capacitor circuit. <S> All these are non-isolated circuits so you cannot just add the 10V to the 40V gnd-to-rail.
|
The best solution, but not the easiest, is to use a flyback converter topology. From the ideas you provided I think the one with the external switch is the easiest.
|
8/16 bit microprocessor for hobby project I'm evaluating different processors for a small hoby project. My first choice was the Z80 but it seem to be hard to come by nowdays. Whats left for us that wants an 8/16 bit CPU clocked on 10-20 Mhz with an external memory bus ? <Q> Maybe you should let go of the idea that for small hobby projects 8-bitters are always the best solution. <S> Nowadays 32-bitters, esp. <S> Besides, in your comment you mention 512 KB internal memory, and you won't find that easily on pure 8-bitters; they tend to have a 16-bit address bus which means 64 KB maximum. <S> If you think an ARM may suit you, you may have a look at mbed : <S> "This mbed Microcontroller is based on the NXP LPC1768 with an ARM Cortex-M3 Core running at 96MHz, 512KB FLASH , 64KB RAM and lots of interfaces including Ethernet, USB Device and Host, CAN, SPI, I2C and other I/O. <S> " <A> Interfacing with an external memory module is best done with a hardware peripheral known as an External Memory Interface (tricky, huh?). <S> An External Bus Interface can also be adapted for the purpose. <S> These peripherals are abbreviated EMI/EBI on manufacturer and distributor parametric search engines. <S> Using Digikey's parametric search, for example, I note 209 different parts have an EMI or EBI (with a few other distinctions, like onboard Flash, solderable packages, and less-than-reel quantities) . <S> The following is a selection of the more well-known processors from that set: AVR Xmega <S> Arm Cortex-M0 Arm Cortex-M3 Arm 7 <S> Arm 9 <S> Coldfire <S> STMicroelectronics <S> ST9 ZNEO (Zilog 16-bit successor to Z80) 8051 <A> There are some PICs with external memory interfaces that can address up to 2 Mbytes (code or data): http://ww1.microchip.com/downloads/en/AppNotes/00869b.pdf <A> It can address 24 bits of memory and comes with plenty of peripherals. <S> It's compatible with Z80. <S> Here are the microprocessors (but the microcontrollers have memory busses too and are actually cheaper): <S> http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=2556260&k=ez80 <S> They also have development boards. <A> ARM is the way to go, lots of various sizes of ram vs flash. <S> Every combination of size speed, power, etc. <S> Mainstream tools, etc. <S> For an external memory interface Cirrus has ARM products. <S> When you say 512KB I assume that is just for data <S> and you need flash and/or ram for the program as well. <S> Depending on the instruction set the amount of program space can/will vary dramatically. <S> The next question is how fast do you need to get at this memory, you could probably go with some spi or i2c and use pretty much any micro if it doesnt have to be lightnight fast. <S> Granted if you were looking for a handful of megahertz 8/16 processor then for the same price you can get into a 50-80mhz arm and probably get similar memory speeds to the 8/16 micro with a serial memory. <S> For the price of an arduino you can get a GameboyAdvance <S> , 17mhz ARM7 256K plus 32M of onboard memory, then for a little over half the price of a second arduino 512KB of ram, plus another 16KB plus 32MB of flash. <S> You get a display and serial port as a freebie... <A> For that kind of application it boils down to two main product lines: PIC (Microchip) and AVR (Atmel). <S> Both have many models from the very basic (the ATTiny13, for instance, is 10 MHz and costs about $1) to the quite sophisticated (barely overlapping the entry level ARMs). <S> All of them support SPI/I2C via software or hardware, and those with enough interface pins can be connected to external memory. <S> For those who don't, there's also the option of using a SPI memory. <S> Also there are many options of tools for development & debugging, many of them free or low-cost.
|
ARM, are everywhere and they have become serious competitors for the often older 8/16-bit parts, both price-wise and toolchain-wise. ez80 series is available both as microprocessor and microcontroller.
|
Is it normal for LM317 to give a higher voltage without a load? I just built a small 5.1V regulator with LM317, using 220 and 680 ohm resistors. However, I measure 8.6V at the output without a load connected. Is this normal, i.e. under a load (about 120 ohm) it would drop to 5.1V? The input is a 9V/300mA unregulated wall wart, which reads about 13V without a load. <Q> You need a minimum load of 10mA for it to be stable. <S> A tip is to add a LED and run it on 5mA. <S> You'll know whether the supply is on, and it will add additional load to make it stable. <S> Together with a LED <S> you get more than the required 10mA. <A> For example 240 ohms and 75 ohms gives a 5.25V output and draws 16mA. Tweak the resistors to get it perfect. <A> The regulator works mostly like a resistor which is automatically adjusting itself so if there is <S> too little current the voltage drop on the regulator will be too small to regulate the output. <S> It is possible to make regulators that work without load <S> but they tend to suffer from high quiescent current.
|
One way to ensure a minimum load is to make sure the feedback network draws the minimum load. Note that the adjustment resistors of 220+680 ohm also give some load, of about 5.6mA.
|
how might i make a mouse who's L and R buttons flip depending on what side of the keyboard it's on? I was thinking trying to make a little prototype of a wireless mouse that was left-handed or right-handed depending on which side of the keyboard it's on. How might I modify a mouse so that the signals from the button clicks cross if they are on oppisite sides of the keyboard? I'm thinking some kind of small component that knows if it's with a keyboards-width-distance of something attached to one side of the keyboard could toggle whatever mechanism does the reversing. Is this maybe doable? <Q> a manual switch would be the easiest, but you could use IR LEDs on each side of the key board and 2 IR receivers one on each side of the mouse, then it just detects which side the IR led is lit up on and changes the buttons to the correct configuration. <A> I'm guessing that if the mouse is on the other side of the keyboard it's being held by someone who's left handed? <S> You may be able to get away with having capacitive sensors in key areas - one set where right-handed people might put skin (fingertips, palms, etc) and another where left-handed people would. <S> If there's not a lot of overlap then you might be able to judge whether a person is left or right handed based on where their hand touches the mouse. <S> Man that's probably patentable if it works. <A> You could do a hall effect sensor in the mouse with a magnet embedded in the mousepad or desk. <S> Depending on which side of the mousepad is face-up, the hall effect sensor will return a positive or negative value reflecting the north or south side of the magnetic field. <S> You could then switch the left and right mouse buttons based on the values coming from the hall effect sensor. <A> I'd suggest a simpler solution: check if the movement sensor of the mouse can detect when the mouse is lifted up (Don't know it <S> that's doable, but I believe it is), if not, add a little sensor to the bottom of the mouse, <S> that checks if the mouse sits on ground. <S> Now, when the mouse is lifted up by, say 15cm, you put the mouse in a 'unset' state, the first button that is pressed, after the mouse lands on a surface, defines button orientation. <S> Example: Mouse sits on right of keyboard: <S> Lift mouse, and put it back at left side of keyboard. <S> Press <S> right mouse button: <S> Mouse is now a left-hand mouse (right button = Button 1). <S> If the laser - that most mice nowadays have - or the LED led is capable of detecting the distance from the surface (or movement away from those), you could implement the solution completely in software on the driver side. <S> This might even work with standard mice. <S> To minimize accidental button switching and allow "swiping" the mouse several times over the same surface, you's simply activate the 'unset' state after the mouse has been in the air at an altitude of ~15cm and for more than 2sec. <A> This answer assumes that the important thing is not whether the mouse is on the left or right of the keyboard, but whether the left or right hand is on the mouse. <S> If you look at the mouse in your hand, you will see that there is a hole between the thumb and index finger, where there is no hand contact with the mouse. <S> So put a touch contact in that place on both sides of the mouse. <S> When the left hand is on the mouse, the palm will rest on the right touch contact, and vice-versa for the right hand. <S> Reading of touch contacts is a commonly integrated function into modern microcontrollers. <S> TI app note . <S> So a small micro can read the two touch sensors and decide which hand is on the mouse, and swap the buttons with a 74hc4053 or similar. <S> Of course, a commercial design would integrate all functions in the mouse's microcontroller. <A> You could put an IR LED mounted on the mouse powered by it's internal battery. <S> You could then put a IR receiver mounted on one side of the keyboard facing to the side. <S> Then build a simple IR recieve circuit that detects the presence of the mouse one side, or the non-presence of the mouse meaning it must be on the other side. <S> The IR reciever on the keyboard doesn't have to have any special keyboard interface, as it's the mouse <S> that's wireless and not the keyboard <S> then just have your circuit wired to the serial port on your computer, then perform the button switching in software depending on the signal received on the serial port.
|
If the mouse case is thin enough, you can put the touch contact as a piece of conductive foil inside the case.
|
Power for POV display I want to build a small POV display , but how is the power (and if possible a communication line) transmitted to the spinning part? I've looked at some POV project, but they usually don't explain that part much... <Q> Here's some instructions on building a rather ingenious slip ring using ball bearings. <A> You need to use some form of slip ring and carbon brushes. <S> The slip rings are mounted on the spinning shaft which is a ring of brass that makes contact with sprung brushes that supply the power. <S> They are similar to a commutator on a DC motor, except they are a continuous ring instead of segmented. <S> You'll need two slip rings for power and GND unless it's possible to use the shaft itself as GND. <S> In that case, you can get away with just one. <S> But this would mean you'd need to mount all of your control circuitry on the spinning assembly, otherwise you'd also have to have extra rings to supply individual LED circuits. <S> A neat and cheap alternative to using carbon brushes is to use tool clips that are mounted on a piece of insulating nylon that surround the rings. <A> If you are building a small display, you can also do it with a round circuit board with circular traces that rotates coaxially with your rotating part with two fixed "brushes" (small brass springs) <S> that each drag on one of the circular traces. <S> Like a flattened out version of how a DC motor commutator works. <A> The slip ring is the most obvious solution, but it's prone to wear and noise. <S> An alternative may be a videohead from an old VCR. <S> They contain coils that act as a transformer to transfer the video signal from the magnetic heads in the rotating part of the head to the fixed part. <S> Very quiet and low wear. <A> Please check out the list of POV displays .Many <S> POV displays are "open-hardware" and so should have extensive documentation on exactly how their power and communications work -- if one doesn't have enough detail, try another one. <S> Nearly all POV displays that I've seen working use slip rings for power.(I've seen one working commercial POV display that uses an alternative to slip rings for power, and I plan to build a spinning POV display without slip rings that transfers power in a completely different way, but it's currently highly experimental).Power on the stationary side <S> goes through the slip ring to a large capacitor (to ride out power glitches from bounces and non-conductive dirt in the slip ring) at typically 8 to 12 V, which supplies power to a voltage regulator that powers the microcontroller and the chips that apply power to the LEDs. <S> Many POV displays have no communication between the stationary part and the spinning part -- any user-interface buttons are on the spinning part, and you have to stop the motion, push the buttons (with a difficult-to-read linear LED array for feedback), then re-start the motion. <S> Some POV displays have a "communications" slip-ring or two for serial communication between the spinning part and the outside world. <S> Alas, the slip ring adds glitches that are difficult to ignore. <S> In principle, one could send data from a stationary coil to a spinning Hall effect sensor near the rotational axis, but I haven't seen that actually working. <S> Some of them use infrared communication between the stationary part and the spinning part, much like IrDA or TV remote controls. <A> I'm surprised to be the first one to mention this, but: Build a generator! <S> Attach a magnet to the fan casing, and length of copper cable wound in a manner so that the field lines from the the magnet pass through those windings perpendicularly as they pass the magnet, rectify the resulting voltage, and store the energy in a large capacitor for use with a voltage regulator. <A> Besides having a lot of "chatter" and there being limited power handling capabilities in the contact another alternative is to use a rotating transformer. <S> The primary and secondary can be used on the stationary vs the rotating platform and transmit energy across an air gap. <S> One key aspect is to ensure that there is not a lot of "run out" in air-gap as this presents itself as a variable reluctance. <S> Here is one instance taken from Wikipedia: and another taken from here <S> These can even be used to transmit signals ... <S> How I've made them in the past is to use pot cores Which are like inside out toroids. <S> Here is a picture from dexter magnetics: <S> You use two bobbins to wind your primary/secondary on. <S> cast them with epoxy and then mount the two open faces close together.
|
Many POV displays have a Hall effect sensor on the rotor that passes a fixed magnet, so the spinning microcontroller can compensate for the actual rotational speed.
|
AVR and external crystals I have an atmega16 which I need to clock to at least 16MHZ. How do I know what fuses to set? The fuse calculator gives you a choice of low, medium and high frequency. What does that mean? Is there a limit to how fast the chip can go despite you put a crystal in? <Q> The low, medium and high frequencies are explained in the datasheet. <S> For ATmega16 this is in Table 4 on page 26: <S> In general you should read the information about fuse settings in the datasheet since there is a lot of information there that just did not fit into short descriptions shown to you by the fuse calculator (they are more of a quick remainder then a reference). <A> I'd call 16 MHz high frequency, in terms of setting the AVR fuses. <S> Details are in the data sheet, in the System Clock and Clock Options section. <S> Note the startup time options, you shouldn't use 6 CK for a 16 MHz clock. <S> Your last question is meaningless. <S> The processor speed is determined by the oscillator frequency, up to the maximum frequency specified (16 MHz). <A> I've had good luck with this online AVR Fuse Calculator in the past. <S> Might be useful for ya. <A> Also If you don't have a good reason not to I usually just pick the longest startup time (SUT fuses) option to give everything time to settle at startup. <S> This would be the <S> should be the 16K CK + 64ms option for the ATMega16.
|
The fuse setting descriptions come from Atmel and they sometimes contain the MHz ratings but this depends on the processor model. Yes, 16MHz is definitely a "High Frequency" setting.
|
How to build a constant magnetic field source for a "railgun"? I'm trying to build a device similar to railgun. It will have two rails and a conductive movable bar in-between. At one side, I'll provide voltage source and a resistor. If I put it in a strong enough magnetic field, it will move. The idea is that the device will work as demonstration of forces behind an electric motor, so I don't require high speed for the bar. Here's a sketch: So, I've got voltage source E, resistor R, bar whose length is A and magnetic field B. If my calculation are correct, force when bar isn't moving should be $F_m=IAB$. I don't require much force, so I guess that 5 mN would be more than enough. The power supply could be a problem. I should be able to get up to 1.3 A without much problems, but anything higher will be problematic, since I can't get a good match between resistor power dissipation and limitations of my power supply. Since I can't get a perfect constant field, I'm going to have to compromise. I expect the field to drop linearly with A, but on the other hand, force will also get stronger linearly with A, so the differences should cancel themselves out, if I manage to do everything right. Now about the field source itself. Most obvious sources are natural magnets or electric magnets. I can get natural magnets of various sizes but I don't have their magnetic field strengths. I found several types of magnets in local electronics stores. Their compositions are listed as AlNiCo500, NdFeB and SmCo5, so which one would be a good choice? Another option would be to make an electrical magnet, but I'm unfamiliar with them, so I don't know where to start. I'm also open to other means of generating a magnetic field, if I missed some. <Q> There is a configuration of two coils, known as a Helmholtz coil, that is supposed to produce a nearly uniform field. <S> The idea is to have two coils that are separated by a distance about the same as the diameter of the individual coils. <S> Going with this, the coil diameter would obviously have to be about the same as the length of your rails, so would be rather large relative to the rest of the apparatus. <S> But you might not need an exactly uniform field anyway. <S> If you had a field that was non-uniform, but symmetric across a plane that divides the mid-line of your rails, the forces on the movable bar should be in-line. <S> It might be possible to produce a field like that by using two rectangular coils, same width as the rail separation, with one half above the plane of the rails and the other below. <A> Perhaps some neodymium permanent magnets (NIB magnets) will be adequate. <S> Bill Beaty has posted a list of places that sell neodymium magnets . <A> If you series connect the current in the conductor and the current in the electromagnet the force will be proportional to the square of the current while the magnetic field (no saturation) and velocity (low back emf) remain in the linear regions. <S> The force will accelerate the bar out of your 'gun' at a rate that will be rather high once the friction is overcome (unless you use some form of damping) and the experiment will be very brief. <S> I think you are actually looking to make a DC linear motor and <S> not a coil gun (magnetic attraction) or a rail gun (self induced repulsive magnetic field). <S> Friction and the yaw errors (such as suffered by overhead gantry cranes) on your conductor may make for a rather unreliable mechanical arrangement. <S> The force required to overcome static friction between electrical contacts (that will try and weld together with a reasonable current) can be significant. <S> Have a look at the following pictures that include rotary ideas of Homopolar motors like the Barlow Wheel to show the same concept. <S> Most of the successful implementations used mercury as the contact and this allowed reliable starting with very low static friction and a reliable electrical contact (especially if the electrode had been wetted by the mercury). <S> Just to add for those new to such things, mercury compounds are often rather toxic to life forms in a number of ways, elemental mercury while safe'ish combines readily with many other elements and compounds to form toxic compounds and most are very hard to eliminate from the environment or clean up in the classroom, there are a couple of other metal <S> alloys <S> that remain liquid at room temperature that might be suitable for these experiments <S> but I have not seen such described to date.
|
The choice of magnetic field is up to you, the field does not have to be uniform in any real way unless you are trying to do calculations to determine constants, in this case look at a Kelvin balance .
|
High Resolution ADC for Noisy Sensors in Variable Conditions Intro In response to this question about adaptive amplifiers , It was recommended that in order to deal with variable conditions, it may be more economical to simply use an ADC with higher resolution so that I don't need to worry about amplification and I can do scaling in software. Overview I'm trying to design a data acquisition circuit for body mounted textile-based stretch sensors. The textile varies resistance as it's stretched (about 1 order of magnitude, 10k\$\Omega\$-100k\$\Omega\$ with 30% stretch). The exact ranges will change depending on how the textile is cut, whether it's soaked with sweat, the temperature, how old the material is, how it's mounted, etc. The entire thing needs to be as small as possible because it's mounted on the hand, so minimizing the number of components is a big plus. Moreover, I'd like the circuit to be reusable for other applications that may have worse performance. For instance, if I use a cheaper version of the textile, my resistance range may be as bad as 100\$\Omega\$ to 300\$\Omega\$. Signal Path [textile] -> [Wheatstone bridge] -> [lowpass] -> [instrumentation amp] -> [ADC] -> [AVR] Requirements So, I'm looking for an ADC that will meet my requirements. The ADC should be: 16bits+ As easy to use as possible: much better if there is interface code already written for AVR/Arduino... ...yet at the same time as comprehensive as possible: I've seen some ADC's with lowpass filters and PGA's built in – all the better as long as it doesn't make configuration a pain 8+ channels, or if it's easy enough to implement, 2x 4+ channels. EDIT: If I'm using a Wheatstone bridge, perhaps I want 8 differential input channels (so 16 channels)... I don't think operation voltage matters... (best if not above 5V) Surface mount Doesn't need to be cheap (it's a one-off) SPI vs. I2C doesn't matter I think... 100+ Hz Research So far through Googling, I've found the following chips: Linear devices offer various 16-24bit delta sigma ADCs, some of which I've seen recommended: http://parametric.linear.com/html/no_latency_delta_sigma_adcs?p=5312974 Microchip has a range of options, some of which I've seen recommended: http://www.microchip.com/ParamChartSearch/chart.aspx?branchID=11022&mid=10&lang=en&pageId=79 Analog devices have a number of comprehensive data acquisition chips with amplifiers and filters (no need for external signal processing stuff): http://www.analog.com/en/analog-to-digital-converters/ad-converters/ad7783/products/product.html http://www.analog.com/en/analog-to-digital-converters/ad-converters/ad7715/products/product.html http://www.analog.com/en/analog-to-digital-converters/ad-converters/ad7709/products/product.html I haven't looked at the TI chips yet... and the following tutorials: http://arduino.cc/blog/2010/11/29/tired-of-a-10-bit-res-hook-up-a-better-analog-to-digital-converter/ (LTC2400) http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1275676171 (TI ADS8341) http://forums.adafruit.com/viewtopic.php?f=31&t=12269 (MCP3424) http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1248751435 (LTC2410) Voltage Reference? Finally, some people have recommended a precision voltage reference, such as the Analog Devices REF19x series. Do you think this is necessary? Resolution is definitely important for me. Conclusion Let me know if you have any recommendations! I'm also not sure exactly what I'm looking for, so tips on how to decide are also appreciated. <Q> ADS1256 from TI has eight single-ended 24bit channels with high-impedance input buffer and PGA. <S> OpenEXG project has PIC code to interface it (they use two channel version ADS1255, but it should be the same). <S> If you want differential inputs, then there is ADS1298 , with 8 channels, PGAs and A/Ds, internal reference, plus ECG/EEG circuitry which you can ignore. <S> I am not sure you can find any example code for this one, though. <S> If you are looking for resolution, then precise, low noise reference is a must. <A> A maybe unconventional idea, I am curious what you guys think about it: One order of magnitude seems a large enough change to measure it directly in a voltage divider circuit. <S> You could then use a smaller ADC and vary the current through the sensor. <S> A filtered PWM voltage source + a voltage follower (may be one NPN transistor if you are thigh on space) may drastically improve your dynamic range. <S> You could use one or two of these and switch the voltage when measuring different sensors. <A> You might also follow this variable gain stage with a charge integration circuit so that you can gain fine tune signal sensitivity by adjusting the "exposure" period. <A> If you have enough compute power for the sample rate you need, consider digital filtering. <S> A Savitzky-Golay filter, f/ex. <S> You can change algorithms easier than you can change parts; <S> By pushing some of the filtering onto the software, you can probably use a lower spec part than if the part itself had to be more noise tolerant or do all the filtering; <S> You'll learn a bunch more about your inputs and and what you need from them and can make a better informed parts choice, if you do, in fact, need a higher spec part. <S> Software and skills are readily transferred to your other applications! <A> Why not turn it up to 11, and just use the TI ADS1262 . <S> It's a 32-bit ADC, with 11 inputs and a PGA! <S> With 32-bits, you can pretty much sample anything. <S> And it's not even that expensive. <S> What's more, if you're only making one of these, just get a free sample . <S> Another option is to use a PSoC. <S> These are microcontrollers containing re-configurable analog and digital blocks, which you can use to make up all kinds of functions. <S> You can choose one with a 16-bit ADC, a PGA, a DAC and a digital filter, to make your own auto-ranging, auto-trimming, over-sampling, digital filtering, ADC! <S> Programming these things it a doddle, as you simply draw out the schematic you want, choosing pre-defined functions from a list. <S> Then write some C code, and you're away.
|
If your main worry is to have a wide dynamic range for any given "sensor", you might consider using DAC's (or even just MPU-pin controlled voltage sources) to adjust the amplifier offset/gain to alter the system performance for different materials.
|
Cheap method to drill neat holes in plastic casing? A good deal of projects I see have their inputs and outputs neatly aligned in to plastic enclosures. I wish to do the same, however, I do not own a drill or anything similar. What is a cheap but effective way to drill neat holes in my cases? <Q> A hand-held tapered reamer ( http://www.google.com/images?q=tapered+reamer ) is a good tool to have for this. <S> A "step drill" bit is even better. <S> For starting holes, a small hand drill (non electric, http://www.google.com/images?q=egg+beater+drill ) gives much better control than an expensive electric drill anyway; try a flea market or junk store. <S> A small hex shank drill bit can also be used by hand mounted in a bit-driver. <S> For rectangular holes, a hand nibbler ( http://en.wikipedia.org/wiki/Nibbler ) is great, starting from a hole made by your reamer or step drill. <A> You need a drill, if only to make a pilot hole. <S> I use a hand-held tapered reamer to enlarge the hole to the desired size. <S> It leaves a very neat edge, which is difficult to achieve with a large drill bit. <S> The technique works with aluminium as well as plastic. <A> An old, cheap soldering iron can be used to melt the plastic. <S> Always start to melt from the internal side, so that the displaced plastic stays there. <S> After it cools you can easily pull this excess plastic out with a plier (the melted plastic is weaker and gets loosely attached to the rest). <S> The problem with this, and other cheap solutions, is that you must have a very steady hand. <S> I'd prefer to use a Dremel rotary tool with work station. <S> It's not that expensive these days (about 60 + 40 dollars) and will guarantee much better results; not to mention that such tool has MANY more uses. <A> On some softer cases, I've used scissors to make holes. <S> The basic idea is to take one blade and carefully keep pushing against the plastic until it penetrates. <S> After that, start rotating until you get desired hole size. <A> I purchased a toner refill kit once, it came with a cheap soldering iron and a copper fitting cap attached instead of the normal tip, to make the hole in the toner cartridge, to add the toner. <S> (It would require drilling a small hole through the center end of the cap, and finding a screw that fits in place of the soldering tip.) <S> I have used it many times to make holes in plastic. <S> Some plastics make fumes when melted, use in a well ventilated area. <A> The simple answer is to buy a really cheap drill on eBay. <S> You could make your own hand drill by purchasing drill bits and a tap handle to turn them. <S> It would work, but seeing that there are electric drills on eBay <S> Buy It <S> Now for less than $10, it's really not worth it. <S> The local used goods place (e.g., Goodwill in the US) often has used power tools even cheaper than this. <A> Get the right size drill bit(or one size smaller), and a wire nut. <S> Screw the wire nut to the end of said drill bit. <S> for more holding power, use epoxy JB weld. <S> Then you have a hand(finger)held bit that will last as long as you don't lose it. <S> Or use pliers to twist it and break loose the epoxy. <S> In which case it would no longer be a finger bit, but a pliers bit.
|
You can also melt a starting hole for your reamer hole with a hot nail or a worn out soldering iron tip in a pinch.
|
Best way to drive a 20 kg object on a turntable? I need to rotate an object about 20 kg's in weight, 360 degrees. I'm thinking of using a stepper motor with some kind of rubber wheel pushing along the rim of a lazy-susan. Geared or belt solutions are probably better, but I don't reckon I need the accuracy. Mostly I am unsure how to spec which stepper motor I need? This is the given spec for a stepper motor I can get locally: HIGH TORQUE HYBRID STEPPING MOTOR, WITH DUAL / DOUBLE ENDED SHAFT12VDC, 0.31A, 38E5/PHASE, 1.6KG-CM TH, 6-WIRE, 120G-CM TD Any comments welcome. <Q> If you are not trying to be accurate (as you allude to) but just continuously rotate something you could probably just use a normal DC motor with rubber drive wheel (have you considered using the belt/motor from an old vinyl record player turntable?). <S> If you need slow rotation you may need a gearbox (model shop, old toy, meccanno, lego?), or need to PWM (Pulse Width Modulation) drive the motor (remote control car motor controller?). <S> Another idea is an old variable speed electric drill/powertool, you could rig the trigger to a specific speed fairly easily (cable tie?). <S> This will all depend on how quick-and-dirty vs robust/reliable/long-term you are aiming for and whether you need variable speed and/or direction control <S> (things you don't specify). <S> Stepper will give you the most control, but is probably the most complicated. <S> Match the complexity of your solution to the complexity of your problem. :) <A> Provided that you will have good bearings (and very low friction), even the lowest powerful steppers would work, even ones salvaged from DVD-drives. <S> So you would need to focus on mechanical part & bearings, not motor. <A> A disco ball rotating gearbox motor has a scary amount of torque but turns rather slowly. <S> These motors are similar to those found in microwave oven turntables but with even slower output. <S> The starting direction may be random on cheaper models. <S> I would introduce a compliant coupling between the motor and turntable if direct drive or use a belt drive to allow for the start-up inertia. <S> If you do not need to be exact with speed or position then a stepper is overkill. <S> A simple cam and limit switch to stop it after it has rotated the full 360 degrees and a push button to bypass the cam to start a new turn.
|
Using a timing belt on the outer perimeter of the turntable to engage with a stepper (or gear motor) is a robust way of delivering torque that could jump a few teeth when starting without much damage.
|
How does this rectifier work? In a car battery charger, I found a strange rectifier. Can someone explain to me how does it work? Here I have a transformer which is unmarked. By measuring resistance between its output terminals, I determined that the plus cable is connected to the center of the transformer. The two outer connections both have a diode connected as shown on the image. To me it looks similar to this: rectifier, but the diodes are turned backwards. I checked their orientation several times with two different multimeters and I'm sure that I've drawn them correctly. <Q> Maybe these illustrations will help: <S> Let's assume that the start and end of the primary and secondary windings are such that the 'starts' are at the top of the picture and the 'finishes' <S> are at the bottom. <S> When primary current flows from the top of the picture to the bottom, the top of the primary winding is at higher voltage than the bottom. <S> This will induce a voltage in the secondary winding with the highest potential at the top of the winding and the lowest at at the bottom (and a potential somewhere in between, at the center-tap). <S> You can tell quickly that D1 will be reverse-biased, because the voltage at the top is higher than at the center-tap. <S> Current will flow however it can, which will be out of the center-tap, through D2 and into the bottom of the winding. <S> When primary current flows from the bottom of the picture to the top, the reverse condition holds true in the secondary: D2 will be reverse-biased, and current will flow from the center-tap through the load and through D1 back to the winding. <A> By center tapping you have just picked a reference that has two different sides that at any one point will oscillate in reference to it. <S> When the top side goes negative in reference to the center tap it will conduct from the load. <S> The same applies for the bottom side. <S> The number of turns on each side will determine what voltage peak you can receive. <S> The center tap just makes it like you have 2 transformers each with have of the loops of the larger transformer <S> it could be by ignoring the center tap. <S> With a capacitive load and regulator this is easily turned into a DC voltage. <S> Why does diode direction matter? <S> It is just going to change which of the pins is the negative reference. <S> The graph you loaded in shows the rectified hills which are all positive, because of the diode direction they are all negative. <S> They need a negative reference for something in your car. <S> I hope this helps. <S> Let me know if there is something in this answer I can expand upon to help. <A> It's the same, although I've never seen a full-wave rectifier this way round before. <S> Electriclly it makes no difference which way it's done, but it seems more 'natural' to connect the centre-tap of the transformer to the circuit's reference rail - which is usually the negative. <A> The diodes conduct on alternate half cycles (the two ends of the secondary winding give a phase reversal), so that the circuit behaves like a full-wave rectifier with four diodes. <S> It's a popular circuit, because it's cheaper to manufacture than one using a single winding secondary with a four diode bridge rectifier. <A> From a practical point of view this half bridge arrangement is exellent if you do not want the trouble to isolate the nut of a power diode. <S> This speeds up assembly and cools the diode more effectively because of the metal-to-metal thermal coupling with nothing in between. <S> Downside is that a ground referenced positive supply needs reverse direction diodes. <S> A negative ground referenced supply can however use standard direction diodes where the diode nut is the cathode. <A> One thing I haven't seen in the other answers: <S> This is a centre-tapped rectifier. <S> Its biggest positive is that you only need 2 rectifiers to do this. <S> Because of this, these were used when solid state rectifiers were still very pricey or you had to use valve rectifiers. <S> Its not used often today because the cost of creating two secondary windings in a transformer is much higher than adding 2 more rectifiers to create a full-wave rectified signal from a simple transformer.
|
It uses a centre-tapped transformer to create a full-wave rectifier.
|
Any problems with "floating" an oscilloscope? I have a Chinese scope with a two-pin power cord and a plastic body, and measure megaohms between its ground terminals and the actual earth, so I believe it has an internal isolation transformer (I suppose I could open it to confirm...) I've used it to measure small floating voltages before without problems (connecting the ground to +5V relative to earth, and the probe to a current sensing resistor also around +5 V), but now I want to measure power supply voltages where the ground of my scope will be connected to a rectified wave at maybe 200 VAC relative to Earth. Any problems? Does an isolation transformer make everything good? I know that what you're supposed to do is use two separate probes and use the Math function to subtract them, but that doesn't work in practice because the common-mode voltages are so much higher than the differential voltage. ... Oh. Besides the, um, safety reasons. The whole scope will then be at 200 VAC relative to Earth ground, so if I touch one of the scope's BNC connectors and a real ground at the same time I'm in trouble. I guess that's a big problem, but is it the only problem? Floating the device under test with an isolation transformer would not have this problem? But that transformer would have to be a lot bigger to handle the high power. Does it have other problems due to parasitic capacitance, etc.? Is there any trick to using the differential math function method by reducing the common-mode voltage with a capacitor or large resistor, but not connecting it directly enough to produce a safety hazard? <Q> When we needed to do something like that where I worked many years ago, when testing power controllers that had a direct mains connection (no transformer) under load with a scope I specified a high-power isolation transformer in an "earth free" test area. <S> The setup cost a lot of money, but was necessary for safety reasons, and worked very well. <S> People sometimes remove the scope ground connection in those circumstances, but it isn't a good idea. <S> In your case, I'd use an isolation transformer and an earth-free area. <A> Floating the device under test (DUT) is the safe bet, you've mentioned the reasons. <S> Your scope appears to be a class II piece of equipment (no protective earth = <S> PE connection). <S> Usually, stuff like that is tested with a high voltage (some 1...2 kV) between mains and secondary, so this is what the scope's mains transformer should be able to take. <S> (Just for clarity, in case it's not obvious: In your case, as seen from the sope's power supply, secondary means everything around the scope inputs and interfaces.) <S> The high voltage is just used for production testing and is by no means meant to be a working voltage that you should continuously apply. <S> However, to be sure, there should be something in the manual about this issue. <S> AFAIK, it is mandatory that scopes be grounded even if powered with a power supply that doesn't have a PE connection, or there must be a specification. <S> (The Tek THS7020 battery scope is an example, but this one has isolated inputs that can even be floated with respect to each other). <S> Of all non-battery-powered bench type scopes with ordinary (non-insulated) inputs, this is actually the first one I've seen sold as a class II piece of equipment, i.e. as one that doesn't have a PE connector. <A> What you describe is sometimes done but you should be extra careful and always verify the polarity of things (a screwdriver-like device with a small incandescent light inside — how do you call it?). <S> Also never move (or even touch) the scope probe while the circuit is "live". <S> Always disconnect the power to the DUT if you want to change your connections.
|
If the scope has 2 wire supply then it is not earthed and has to have an isolated supply inside.
|
Very high current very low voltage application I'm not an electrical engineer. I have an unusual need and would like suggestions as to how to proceed. I have an application that needs very high currents at very low voltages. I'm trying to heat a thin strip of aluminum to temperatures near 100C quickly and then control the temperature. My calculations suggest about 150 to 200 amps at less than 0.1 volt. About the only idea I have is using a toroidal transformer with maybe 1000 turns and thick copper bar passing through the center of the toroid connected to the aluminum strip. I can use either AC or DC current and if AC, I'm not necessarily limited to 60hz. Any and all good suggestions are welcome. <Q> Power supplies from arc welding devices are quite capable of what you describe. <S> Just be aware that you need really really thick wires for 100A. <S> The problem will be worse with high frequency AC because of the skin effect. <S> As far as the regulation is concerned you should be able to use a simple PID regulator with a thyristor output for the soldering gun. <S> The arc welding supplies may be more difficult depending on their internal construction. <A> How thick is that strip? <S> Aluminium is very reactive and oxidizes instantly when exposed to air, leaving an insulating coating on the aluminium. <S> The resistance of this coating may be much higher than the strip itself, so that the largest part of the power is actually used to heat the contacts, even melt them. <S> Make sure the contact surface is large enough. <A> No need to wind a transformer, since 1-turn output should work. <S> On ebay find a 120V, 200VA toroid which has an obviously open center <S> (some are epoxy-filled!) <S> The secondary will be unused (can be anything.) <S> Pass a metal bar or even copper pipe through the center. <S> That should easily give you 0.1VAC output on the ends of the pipe. <S> If it's a bit too low to drive thin foil, you might need to use two pipes, with outputs connected in series for 0.2VAC output (really it's a 2-turn secondary.) <S> Very useful is a 100watt Variac on the 120VAC input. <S> That lets you adjust the input voltage between 0% and 140%, for net load wattage 0% to 200%. <S> For general reference, here's measured values on the transformer in a typical (Radio Shack) <S> 230watt <S> soldering gun: secondary is 5 turns, the AC output is 1.0V unloaded, falling to 0.6V with a 200W load (soldering tip.) <S> That's 0.1 - 0.2VAC output for each single turn in the secondary. <S> (This is typical for the power transformers I've measured in the past.) <S> Clamp-on meter says the soldering-tip current is over 250 amps. <S> Note that the output current starts dropping as soon as the load conductor heats up. <S> Also: To avoid the problem with soldering to aluminum, I'd also try copper foil, or for higher ohms, some thin brass sheet from a hobby store. <A> This soldering gun is a transformer with lots of turns on the primary and a single turn secondary (built of very thick copper). <S> The secondary circuit includes the heating element. <S> You could try getting one of these (expect to pay a couple tens of bucks for a cheapo one) and wire your heating element in place of the existing one, with big burly thick copper. <S> Another popular option is to use a microwave transformer , because these include an air gap which means in case of a short (which is the case here...) <S> bad things happen less often. <S> There are lots of DIY spot welders on the internet using these transformers, and you should in fact google "DIY spot welder". <A> You could build a voltage-divider network going into the strip of metal. <S> Use a few high-current rheostats as resistors. <S> With this, you could drive it with direct current and be able to get very low voltages. <A> Consider the N-Generator. <S> Using a basic prototype was able to produce less than 4 VDC with several hundred amps. <S> Melted a 8" #8 piece of copper ground wire in seconds. <S> Working on design that will self sustain. <S> I know sounds like fiction but it does work. <A> I have built such AC/DC source by salvaging a transformer, there were more than 4000Amp at 4 Volt DC. <S> I have rewind a 10/16kVa 400V to 220V transformer, the primary winding has been kept as it is, and the secondary has been rebuilt with a copper strip capable for such current. <S> To adjust the output current I have used variac in the primary side. <S> It was to detect any crack inside an workroll of Rolling Mill.
|
You may also try a soldering gun (the kind with a transformer inside) replacing the tip with your stuff.
|
Can I use a subwoofer as a linear actuator? I need an actuator with a few mm of travel and it has to be very responsive to a very low frequency (DC 10 Hz) sine wave generated by a function generator. I'm thinking of getting a cheap subwoofer and hooking it up to my function generator via a power amp. My questions: Can I use a speaker as a pseudo-linear actuator or is the speaker designed to pull back the membrane? Will the voice coil blow when it's used at very low frequencies with enough power to actually displace the membrane by a few mms? Edit: The load is a PCB assembly weighing around 100 grams. I want to move that assembly up and down in a controlled way at different frequencies which I'm generating with a function generator. I also have linear stages but they are slower and noisy and I've thought about a mechanical contraption with a stepper motor and cam shaft. The actuator solution seems more attractive because I can control the motion more easily and I expect the latency to be better. <Q> The actuator used to position a hard-drive would match those attributes. <S> Funnily enough they are called voice-coil actuators! <S> I guess with any solenoid type of actuator that needs to operate at very low frequencies you probably need to design your drive circuit to not overheat the windings even with a DC current. <A> A subwoofer will work just fine. <S> Something like this would work , you'd get about 12mm of approximately linear travel. <S> You'll have to adapt for the response down to DC. <S> I don't mean just the woofer itself which will have some roll off under fs but amplifiers will roll off at a couple hz <S> (cheap amplifier will be higher) and just the interconnects used in the system generally form a HPF at 5-7hz. <S> Basically you'll need some equalization to keep the excursion response flat. <S> Most off the self EQ will not work at these frequencies so you'll probably have to deal with it by just adjusting the magnitude of your input signal, i.e. pre filter the source. <S> Don't send the woofer DC if you can avoid it but when operating in free air it doesn't take much power at all to move cone a few mm, i doubt you'd have heating issues with the voice coil. <S> Most subwoofers you see used as such are in enclosures. <S> Its the job of the enclosure to provide spring resistance to the driver. <S> It takes a lot more power to move the cone with its in a sealed enclosure than it does when its just sitting on the table. <S> A 9V battery will move a subwoofer cone a millimeter or two. <A> A better idea might be a solenoid with a weak spring attached.
|
Woofers have been used to move mirrors for simple laser light displays, but you probably won't get much displacement out of a cheap woofer.
|
How to figure out the pin out of this undocumented LCD? I have a very tiny LCD ripped from a piece of electronic scrap. The LCD is about 1 inch by .25 inches. It does not include a backlight(it had an external LED functioning as the backlight) and is basically just a flat piece of glass. There are 16 wires coming from the board it's on. (it is just a plain "breakout" board for the small SMD connector) The LCD board says "LCD:91-42959-B01" which does not pull up any relevant results with google. I've tried my best to test it with a multimeter, but nothing seems to even be connected. I'm not sure how most simple LCDs work, so what can I do to try and get this to work (hopefully) without ruining it in the process? Also, because this is a flat piece of glass, I do not believe there will be a command type of interface, because as far as I can tell, there is no actual logic connected to it. It's just 16 wires going into clear glass. Also on the breakout board they are bundled into 4 bundles with 4 wires each(with brown, orange, orange, yellow being the colors in each bundle). I have no idea if this matters, but I figured maybe it could help <Q> Driving raw LCDs is a complex business. <S> Microchip have a good appnote on how to create an LCD driver with a PIC microcontroller , the ideas are all transferrable to other devices. <S> When reverse engineering devices, it's best to gain as much information as possible from a still working unit with a logic analzyer . <S> If you have two units, you can then keep one working while you clone the driver hardware/software for your new system. <S> If you are lucky, your LCD may be a chip-on-glass module with a driver chip built in. <S> By sniffing the data transactions with this, it might be possible to guess at the device. <A> If the display was a segmented display, and it has less than 64 segments total, you could try connecting pairs of pins through a 1K resistor to a function generator outputting +/- <S> 5 volts. <S> If you hit a row and a row or a column and a column you may or may not see two or more segments become somewhat dark. <S> If the display is a dot matrix, it probably has a driver or controller. <S> Without a working device to examine it in, I'd regard it as junk. <S> Without that, it might be possible to experiment with the device and figure things out, but if you guess wrong about things you could unknowingly damage the device, rendering further experimentation useless, and consequently waste a lot of time. <A> Very likely the display is driven directly from an MCU with an LCD-drive output. <S> As a quick test, you can try applying about 3V to various pin pairs until you get a segment to change. <S> If that fails, try 12V. <S> You will have to then experiment with the voltage until you get the correct excitation voltage. <S> Then, you'll need to modulate the drive waveform so that the LCD will get a 0-DC average voltage across it - otherwise, it will eventually die.
|
If you had a working device, you could easily determine which pins are for power, and probing the other pins would let you see whether the waveforms look like those of a driver or controller, and what pins seem to be doing what. The resistance of an LCD segment is typically in the meg-ohms range. If you hit a row and a column, you should see one segment become somewhat dark, and possibly some other segments become less dark.
|
Jig for making jumper leads? Does anyone know of a tool that can make jumper wires to fit standard spacings for breadboards? What I have in mind is a jig where you dial in the number of rows you want the jumper to span, insert your chosen single-core wire, and the tool cuts a piece of wire to length, strips the ends and maybe even bends them. I've bought jumper lead kits before, but I'm not a fan of them because they always seem to have one colour per length. I'd prefer to keep my 'traces' the same colour, which is hard to do with such a kit. <Q> What you're looking for is a sort of lead forming tool . <S> The most common such tools don't strip wire, as they're usually used for components such as resistors. <S> The usual hobbyist version I've seen is a plastic device having a series of grooves in 0.1" increments . <S> Perhaps you could use one of those as a cutting-and-stripping guide before bending. <S> Googling around, <S> the CM-01 on this page claims to be able to form insulated jumpers and isn't obviously overkill, but I didn't find a price. <S> I also found high-priced fully-automatic jumper forming machines. <S> Perhaps a little more research starting with these keywords might turn up something useful. <A> Well, you can buy the individual parts to make the connectors, and crimp them onto any wire you like. <S> Pololu sells the parts to do this ( sockets , pins , and housings ). <S> In terms of crimping the connectors, you can go anywhere from a $200 tool for crimping an entire housing in one go, to a $30 single crimper, to an effectively free pliers and patience approach. <S> Alternatively, they also sell pre-crimped connectors , in a random selection of colors. <S> EDIT: <S> Oops, I seem to be thinking about the wrong type of jumper cable. <S> Oh well. <A> I'm four years late to this question - but this "Wire stripping guage" from BreadBoardManiac looks like the closest you can get for the time being. <S> It screws onto a normal pair of wire strippers, and lets you easily measure off how much to strip. <S> https://www.kickstarter.com/projects/252587878/stress-free-great-tool-for-breadboard-wire-strippi/description <S> Permanent product page is here: <S> https://wakutuku.com/shop/breadboardmaniac/bbm-wsg.html <S> Still, I would love to have a one-squeeze device that cuts, strips, and bends. <S> Maybe I'll design one someday if I'm bored. :-)
|
I also found various hand tools which automatically cut and bend, or just bend, to an adjustable width, but that would probably make stripping the short ends difficult.
|
How Do You Prototype with QFP-Only Components? I'm starting to design a circuit that includes a QFP-38 component. I'm accustomed to using DIP components and prototyping on a breadboard, but in this case I won't be able to acquire a DIP version of the component. Is it normal practice for electrical engineers to have a PCB fabricated without prototyping? Or do you always acquire breakout boards for small components at first? (In my case, I would just go ahead and order a QFP-38 to DIP adapter, but it will cost me $20 including shipping. It seems like I may as well just have a PCB made and hope for the best.) <Q> I just have a PCB fabbed. <S> Its incredibly cheap to do so <S> nowadays unless you need exotic materials or construction parameters. <S> Making your own PCB is reasonable for hobby work <S> but when contracting for $150-200+ an hour as most design houses do, having a 4 board 4 day turn done is just cheaper than screwing around with adapters or etching your own copper. <A> I can make a board in 30 minutes. <S> Another option is to use a breakout board for the QFP, like these Schmartboards. <A> I've used dead-bug style prototyping for such parts. <S> The basic idea is that you start out with a copper-clad board which is all copper, then glue down the parts with their pads sticking up, using cyano. <S> This way you can easily solder to the pads and pins of small parts using wirewrap wire or other fine wire and you will have a good solid ground plane anywhere you want it, which is good for noise immunity. <S> Islands of non-ground connections are made by cutting out some small (like 3x3 mm) bits of PCB and gluing those down with the copper facing up, I've even mounted some SMD parts this way. <A> When more is involved it becomes impractical, because you have to hook so many wires from the breakouts to the breadboard that the mess turns it highly probable that you'll make a mistake. <S> What I do sometimes is make a "partial" board, in which I have the high-density components interconnected and hook just their ends (usually one or two SPI ports) to the breadboard. <S> I do the boards myself on the first design round, and have it fabbed when I arrive to the final design - or sometimes when the prototype board is too complex <S> and I'm unable to make it myself. <S> I know two good ways to make PCB at home. <S> One is with toner transfer paper and the other is with photosensitive PCBs. <S> With the first you print your PCB to that special paper using a laser printer, and then press this paper against the board using a heat source (can be an old iron). <S> After that you use water to wash out the paper and the toner is attached to the copper, ready to etch. <S> With the photosensitive board you print the circuit to a transparent sheet, place it on the board and expose it to UV light; then you apply a developer and again it's ready to etch. <S> The first method is cheaper (assuming you already have a laser printer) but I found the limit for vias to be about 0.5 mm. <S> With the second you can have 0.3 mm. <S> Some links: http://www.pulsarprofx.com http://www.injectorall.com <A> Sometimes, such a QFP device will have either a demo board that you can buy with headers broken out, or have a family equivalent device in a smaller/bigger package (which either is available in DIP or have an easy to prototype demo board). <S> This will let you get started in a hurry and validate your design, and let you get on the software while you are still working on designing/building the "real" PCB.
|
I make my own PCBs at home, it's very easy and doesn't cost much. If I want to tiker with one isolated component that has not too much pins (let's say less than 40), I use a breakout board.
|
Using K-type thermocouple to measure human body temperature Is there anything I need to he aware of when using K-type thermocouples? I got the idea by touching the probe. I can get a reading when measuring temperature for 30 seconds using a DMM. A commercial electronic thermometer needs around minutes for the same result. This gave me an idea to make a device which will use K-type thermocouple to quickly measure temperature. On the other hand if it was so simple to measure temperature, why don't the commercial thermometers produce their result faster. I have a feeling that I'm missing something important here. So my question is: Are there any non-obvious problems I could face when using a K-type thermocouple to measure measure human body temperature? <Q> These themocouples have a tiny spot weld between two thin wires to make the active sensor element, which results in a very low thermal mass. <S> Hence they change temperature quickly, which is the same thing as reacting to temperature changes quickly. <S> As soon as you mount them in a larger substrate or coating (eg to provide a sterile insulation, to prevent breakage, to electrically isolate) you increase the thermal mass, so you have to wait longer for stabilisation. <A> Thermocouples are not stable enough to measure human body temperature (you want 0.05C <S> accuracy there). <S> It only gives about 1-2C long-term stability. <S> Platinum PT100/1000 RTD ones are much more suitable and way way more procise. <S> When you have this extra-accuracy, you can measure target temperature while ovserving how it's rising by extrapolating... <S> This would allow some 1-5sec measurement. <S> The only way to be faster - IR sensors. <A> The use of thermocouples is not an easy task if you want to achieve high accurracy (that you need to measure the human body temperature). <S> To get the right temperature that you measure with the thermo couple you have to consider the thermo valtage generated by the plug. <S> I think you would go better with a PT-100 or PT-1000.
|
You also have to measure the temperature of the plug that connects the thermo couple to the DMM because this plug also generates a thermo voltage.
|
Is parallelling inductors a viable solution? I am doing a space constrained board layout, in terms of surface area. So I found an inductor which is \$15\mbox{ }\mu H\$, \$2.3\mbox{ }A_{RMS}\$. I need about \$3.5\mbox{ }A_{RMS}\$, so I was thinking of paralleling two \$8.2\mbox{ }\mu H\$ (\$2.7\mbox{ }A_{RMS}\$) to get \$16.4\mbox{ }\mu H\$ at \$5.4\mbox{ }A_{RMS}\mbox{ (max)}\$, with each inductor on opposite sides of the board. Is this a viable solution? <Q> That is not a viable solution. <S> The equivalent inductance for parallel inductors is $$L_{equiv}=\Bigl( \frac <S> 1 L_1 + <S> \frac 1 L_2 + ... <S> + \frac 1 L_n \Bigr) ^{-1}$$ <S> The equivalent inductance for series inductors is $$L_{equiv}=L_1+L_2+... <S> +L_n$$ <S> For series inductors, the equivalent current rating would basically be equal to the lowest rated inductor in the circuit. <S> For example, for a 2A RMS inductor in series with a 1A RMS inductor, the equivalent current rating would be 1A RMS. <S> For parallel inductors, the (DC) current would be split evenly between them so the total current through the network could be \$I_{total}=nI_{rated}\$ where \$n\$ is the number of inductors paralleled and \$I_{rated}\$ is again the lowest of the current ratings. <S> For \$8.2\mu\text{H}=L\$ you would need eight of those inductors to meet your spec. <S> That's two parallel branches each with four series inductors. <S> This would split the 3.5A RMS current evenly between each branch (1.75A RMS in each) and yield an effective inductance of \$(1/2)(4)(8.2\mu\text{H})=(2)(8.2\mu\text{H})=16.4\mu \text{H}\$. <S> I would guess that this approach would not save board space. <S> Your best bet is probably to find another inductor with a higher current rating. <S> Or as suggested in the comments to Markrages's answer, you could parallel two larger valued inductors. <S> Whichever uses the least space sounds like it would be the best solution. <A> If build a "super-component" out of 4 identical inductors (or 4 identical resistors) in 2 chains of 2 components each, +--X1--X2--+-- <S> | <S> |---+--X3- <S> -X4--+ <S> (assuming negligible mutual inductance, which is true for many common "shielded" inductors),then if each of the four components has identical impedance X, all 4 of them considered as a whole as a single super-component also have the same net impedance X and can handle twice as much current and dissipate 4 times as much power.(This is related to the idea of measuring sheet resistance in "Ohms per square". ) <S> There may have been times <S> ;-) <S> ;-) <S> where I've already bought a bulk bag of exactly the impedance I need, and then discover it can't handle the power. <S> The square arrangement allows me to continue prototyping with the components I have on hand, where each component has exactly the desired net impedance X, while I'm waiting for the "right" component(s) to ship. <S> Sometimes the current rating is limited by thermal considerations -- higher currents will make the wires overheat and something will melt and cause permanent damage or melt the solder. <S> In those cases, the power dissipated is proportional to the surface area. <S> Using N components rather than one big component makes it easier to cool and can save net space.(Sometimes <S> N components in series, each with 1 <S> /N of the desired impedance X, has the least parasitic capacitance). <S> Sometimes the current rating is limited by core saturation -- higher currents will saturate the ferrite core, causing the inductance to drop out of spec. <S> In those cases, the maximum energy (temporarily) stored in the core is proportional to the volume of the core. <S> Using one big component that holds all the necessary volume usually uses less board area than using the same volume of core divided up into a bunch of smaller components. <A> Parallel inductors don't add, that's capacitors. <S> Inductors work like resistors in this respect. <S> But yes, you should not have a problem with such solution. <A> Paralleling will give you 4.1uH @5.4Arms. <S> Series will give you 16.4uH @2.7Arms. <S> You'll need four of those inductors to meet your specs.
|
You are correct that paralleling would allow you to pass more total current, but the effective inductance would be decreased. N components in parallel, each with N times the desired impedance X, uses the least space.
|
Using a heat gun to transfer toner onto DIY PCB? Is it possible/practical to use a heat gun instead of an iron to transfer toner when making your own PCBs? Can you get a uniform heat pattern without for a small board? Is enough heat applied from the heat gun to actually transfer the toner? <Q> From my experience with iron-based toner transfer you really need to apply lots of pressure to get the toner to transfer properly. <S> I tried lots of different techniques (>10 tries) before finding the correct paper, correct/sufficient pressure across the board was important. <S> I'm interested in your findings. <A> This is just a crazy idea I got, but... <S> Build your own little contraption consisting of two sheet of metal, say aluminum, and drill 4 wholes in the corners. <S> Add some nuts and bolts through said wholes to apply pressure, and lastly apply heat. <S> This could be with a big-ass heat gun, or even a (domestic) oven. <S> Report back with the results if you try it. :) <A> Heat gun is a no go.
|
Yep, so for anyone else looking at this, I tried it and it seems like you do need the pressure of the iron. I suggest you try with an iron first and then move on to heat guns.
|
Is it possible for electrolytic capacitor to develop a short in one way only? I've got a 1 F electrolytic capacitor and it seems to have developed a short. I never used it before and its brand new. It reads zero resistance only when I connect positive lead of my DMM to its negative lead and negative lead of my DMM to its positive lead. In the other direction, there's no short. <Q> It sounds like the capacitor is functioning correctly. <S> Large capacitors are normally polarised with positive/negative terminals, so they only work one way which is why you not getting a short in one direction. <S> The reason you appear to have a short in the other direction is because the capacitor is not charged, and your multimeter is charging the capacitor as it reads the resistance. <S> If you held your multimeter on the capacitor for long enough it will become fully charged and your meter would change to read an open circuit. <S> Don't bother trying this with a 1 farad though, it will likely take a long time. <S> So it sounds like everything is good with your capacitor. <A> The design of electrolytic capacitors requires them to be polarized. <S> With an electrolytic, the end with the stripe on the body is negative. <S> When you connect your DMM to it in that way, the current flows correctly. <A> After being degraded for a while the oxide layer will cause short and a violent explosion.
|
AFAIK electrolytic capacitors (like any other capacitors) will work in both polarizations, but DC voltage of the wrong direction will degrade (like in electrolysis) the oxide layer.
|
Is there a simple Geiger counter component? I'm looking for a simple electronic Geiger counter element. In my design I need to detect radiation and did not know what my options were for components to detect radiation. What components exist for my need and what kind of complexity is associated with them? <Q> Here is a technique for detecting gamma radiation using a PIN photodiode. <A> Sparkfun has a couple of things available: one is just the tube, the other is a module that takes care of all the analog bits. <S> Be aware that all the sensor suppliers are pretty backed up right now, so availability may be limited. <A> These are generally only available used or NOS (New, Old Stock) parts on the surplus electronic market. <S> The tubes are often designed to only measure beta and/or gamma particles. <S> GM tubes are powered by a high voltage (low current) power source which makes them modestly complex if you do not have experience working with HV supplies. <S> The only other "sensor" I can imagine would be to adapt <S> the ionizing smoke detector 's ionization chamber , but as it is designed to detect a disruption of alpha particles (emitted from americium-241 ) caused by smoke absorbing the alpha particles <S> , I don't know is such an approach <S> is very feasible to detect all but high levels of ambient radiation (assuming you removed the Am-241). <A> It is getting comercial: http://www.cooking-hacks.com/index.php/documentation/tutorials/geiger-counter-arduino-radiation-sensor-board <A> For some reason, the Electronics Goldmine weekly sales sheet featured radiation detectors a couple of weeks ago. <S> There's a lot of interesting and informative material on the links. <S> Most of the items are cold war salvage. <A> A recent write up of a repair of a geiger counter could be useful to your design process, if not for component selection.
|
The only direct "sensor" of ionizing radiation I know if the Geiger-Müller tube .
|
What's the cheapest way to detect vibration with Arduino? What's the cheapest way to detect vibration with Arduino? What's the vibration sensor that I need? I want to detect PING PONG NET vibration and light on led if touched. <Q> Piezo sensors are cheap, reliable and designed for the purpose you suggest - arduino tutorial <S> In the uk, bitsbox has one for 75p. <S> You can probably find others cheaper, or salvage from electronic toys. <S> EDIT following poster's clarification of use <S> :Sound is vibration through the air, I see no reason why a piezo sensor cannot accomplish what you suggest, but the form factor may not be ideal for fixing to ping pong nets! <S> I think a flex sensor would be more suitable, sewn into the top of the net. <A> A cheap vibration sensor such as SW-18010P consisting of a fine spring wire coiled around a rigid wire in a sealed package <S> may well be sensitive enough. <S> They act as a momentary switch contact when jolted, and cost a few pennies. <A> Might also detect vibrations in the table, as might any sensor. <A> You may be interested in how to create a simple circuit using a piezo element and Arduino Uno that can detect vibration with high sensitivity <S> http://davidhoulding.blogspot.com/2014/11/advanced-high-sensitivity-vibration.html Have fun.
|
An optical or magnetic sensor mounted mid net, bottom edge might be a good way to acomplish this.
|
How do I determine the maximum current for charging a li-ion battery? I have a cell phone that has a 1500 mAh 3.7 V battery. It comes with a 700 mA charger but I've sucessfully used a 1 A charger with no problems. I'm now trying to make my own multi-device charging station and my first problem if figuring out the maximum amount of current I can use at 5V? <Q> If you are charging the battery through the phone then this will have the charge controller circuitry between the 5V charge supply and the battery. <S> You CANNOT/MUST NOT just connect a battery pack to a power supply and expect it to charge without fire and or explosion. <S> The charge controller in the phone will limit the current supplied to the battery pack to be within the limits specified by the battery manufacturer to ensure that the battery is not damaged. <S> Supplying the phone from a 5V source that has a higher current capability will not make the battery charge any faster. <S> If it did then you would run the risk of damaging the connector on the phone or even melting the tracks on the PCB within the phone. <S> Small USB connectors that I have used have a contact rating of up to 1A on the power lines. <S> If you want to build a charging station to charge multiple phones at a time then you need to have a power source that can supply up to the maximum charge current taken by the phone down each of the charging leads. <S> These can all be in parallel but I would place a diode in each of the positive supply lines to prevent the possibility of any current flowing from the battery back to the charger. <A> It's probably 1C, which means 1xcapacity, or 1500mA. <S> However, there is a reason the manufacturer chose 700mA. <S> It's best to keep to that to preserve battery lifetime and for safety. <S> Overcharging a lipo can damage it. <S> It may, in rare circumstances, become "puffy", or even overheat or burn. <S> So be careful! <S> Never charge a lipo battery without a proper charger. <S> They must not be exposed to a charging voltage exceeding 4.2V. <S> They should be charged with a constant current and monitored for voltage. <S> Never connect a lipo directly to a supply. <A> Current is not limited as long as battery does not overheat internally & externally. <S> You can have higher currents with active cooling. <S> Going above 1C is possible (i.e. 1.5A in your case) but it will reduce lifespan of your battery. <S> Under no conditions you should connect unregulated 5V to LiIon - current will be >10A and something would explode (PSU or battery). <S> You need regulated current circuit. <A> It sounds like what you're building is a dock that will accept multiple phones and power them all at the same time. <S> A block of plastic with a micro-USB connector sticking up for each phone? <S> If every phone has a wall adapter that puts out 5V, or has a USB connector that accepts standard USB output (which is 5V at 0.5A), you can add up all the currents and make sure the power adapter for your dock can supply at least that much current at 5V. <S> It's fine if it can supply more current, because the phones will only take as much current as they require. <S> The official USB spec (thus most ports) only offers 0.5A, although some manufacturers give the device more current when connected to the wall adapter, notably the iPad. <S> Making the power adapter unnecessarily large does add some extra risk in the case of a mishap like an accidental short circuit. <S> A fuse on each line might be a good idea. <S> The phone is expecting to see 5V, not 4.3V, and is designed accordingly. <S> http://en.wikipedia.org/wiki/Universal_Serial_Bus#Mobile_device_charger_standards <S> If you are only powering the phones and not the batteries directly (removed from the phones), you don't need to worry about 4.2V, 1C, or how the battery is being charged. <S> The phone will handle that for you.
|
Ian suggested adding a diode between the main power supply and each phone's connector, but I'd be wary of that because 5V - 0.7V = 4.3V, which does not leave a lot of room for the phone's internal battery charger to step down to the 4.2V the battery will require to be fully charged.
|
Is there an I²C device out there for reading a set of PWM inputs? I'm thinking about getting a bunch of Maxbotix Ultrasonic rangefinders . I'd like to use as few pins on my controller as possible. Since I don't think you can have multiple devices on an RS232 serial connection, I'd quite like to interface their PWM output with the controller using I²C. Is there some I²C device I can buy that accepts around 8 PWM inputs, and stores a representation of \$\frac{\text{pulse width}}{\text{period}}\$ in its registers? <Q> Thank you for linking to exactly what you are talking about -- it makes it much easier to get to the data sheets. <S> The question "How can I use more than one MaxSonar®-EZ1™ in the same system?" in the MaxBotix MaxSonar FAQ lists several ways of using a bunch of them at one time (and some pitfalls to avoid).Some of them use 2 pins of the MCU -- the same number of pins as I2C, but a completely different protocol. <S> If I understand the data sheet correctly, one of those 2-pin methods is: wire pin 1 low the MaxSonar to enable chaining Wire an output pin from your MCU to the RX pin of the first MaxSonar in the chain. <S> Wire TX of each MaxSonar to RX of the next in the chain Combine the signals from all the MaxSonar PW pins (AND? <S> OR? <S> NAND? <S> diodes?). <S> Wire that combined signal to a timer input pin on the CPU. <S> Then software on the MCU pulse <S> the RX pin of the first MaxSonar Time the first pulse width to give the distance the first MaxSonar sees Time <S> the second pulse width to give the distance the second MaxSonar sees ... <S> After getting the pulse from every sonar (or timing out at 49 ms per sonar), do something with the data. <S> repeat. <S> Perhaps one of the other methods listed in the FAQ would work just as well or better for you. <A> Will an 8 channel ADC that supports I2C cut it? <S> http://www.linear.com/product/LTC2309 <A> The microchip PIC18F87K22 has 10 capture compare modules that can handle this function and of course you can talk to it through I2C. <S> You could also do this will a really simple FPGA.
|
The Cypress PSoC family could probably also handle measurement of a bunch of PWM inputs.
|
what size of copper wire can act as a 150A fuse? If I would like to use a piece of copper wire as a 150A fuse, what size of wire should I use? It doesn't have to be 150A very accurately, the parameters I need the wire to meet are: it needs to conduct continuously 70A 10V-30V DC without getting red or overheating it needs to conduct 120A sometimes for 15 seconds without melting it needs to surely be melted before the current reaches 200A Easily available wires over here are: 0.5 mm^2, 0.75 mm^2, 1.0 mm^2, 1.5 mm^2, 2.5 mm^2, 4 mm^2, 6 mm^2, 10 mm^2. Will any of these do the job? I can combine a few if the required value is not in the above series. Edit: now I read more into the specifications of the device, and it is in fact rated to be fine with 600A for 5 seconds. Besides that I will be only ever using it continuously with 50A. Is that gap (50A - 600A) large enough to make a fuse out of copper? This table of AWG wire sizes suggests that a 2.5mm^2 copper wire will melt with about 750A in 1s, and with a little under 200A in 10s, so it looks like it will melt in time for the device not to be damaged. The device itself is wired with 10mm^2 insulated wire, and I mostly worry about that wire not to be damaged. Now I only need to find out to what temperature will a short piece of uninsulated 2.5mm^2 copper wire heat-up with 50A continuous current flow, and will it not melt the thing it is secured in at that temperature. By "continuous" use I mean max 1 hour at a time, with full attendance of me, so it will not run unattended like this. More info: The machine came originally with two 50A fuses wired in parallel. In fact I want to use a wire, because the fuses blow so often it starts to get expensive, on average I need 1 fuse per 1 hour of charging, so the fuses cost more then the electricity to power this. I don't know why they blow, because I have an ammeter wired in series with the fuse, and I never seen anything over 60A on the ammeter! The fuses don't blow randomly, I just can see the fuse slowly get red, it stays red for some time, and at one point it just melts. I've been watching this process and I didn't see over 60A while the fuse was melting, I was watching the ammeter all the time in the slow process of the fuse being melted. So if I need 1 fuse per hour, so be it, but I need some cheaper option then 50 cents a fuse if I am to use them at that rate. <Q> Finding a fuse with those characteristics is going to be difficult. <S> I'd use a current shunt with a suitable MCU and circuit breaker and monitor the current taken by the device, shutting off the current if it rises above 120A for 15 seconds. <S> Two fuses in parallel can cause problems unless they are well-matched and the holders are properly designed. <S> Any difference between them can cause one fuse to take more current than the other one, and fail. <A> It sounds to me like like you're about to violate some NEC/NFPA* fire and safety regulations. <S> Replacing a fuse like this with bare copper wire is ILLEGAL period. <S> Machine downtime is always preferable to property damage, injury or loss of life. <S> EDIT <S> * <S> Assuming you're in the USA. <A> Forget trying to design your own fuse, leave that to experts whose livelihood depends on safe and accurate designs. <S> Call a fuse manufacturer's technical department and discuss the parameters that you provided on your load (600A 5 secs maximum, <S> 50A continuous). <S> Fuses protect the load AND the cable feeding the load. <S> Replace your parallel fuse holders with one fuse that is designed for purpose. <A> Fuses don't work well in parallel, especially if they are installed in mechanical holders and not soldered. <S> Holder terminals inevitably oxidize, which increases the resistance and affects current distribution. <S> Now, if there's a single fuse, oxidation doesn't become an issue. <S> Increased resistance results in voltage drop in oxide layer which gets destroyed, restoring the contact. <S> That doesn't happen if you have a second fuse in parallel: that fuse will shunt the high resistance spot and let the oxide on the other fuse grow. <S> Eventually, the oxide layer will completely isolate one of the fuses, forcing all the current through the other fuse and destroying it. <S> So, get a single 100A fuse. <S> Also make sure that each fuse has a significant (and roughly equal) length of wire before the joint: wire resistance will help to equalize the current in both fuses: simulate this circuit – <S> Schematic created using CircuitLab <A> Sounds like you're planning ahead for a tragedy. <S> Kinda <S> like, "If a robber comes through the door should I shoot him, or stab him?" <S> Forget that and just lock the door! <S> :) <S> What you should do with all your planning is simply buy extra fuses and keep them handy. <S> If on the odd chance that you run out of fuses, then kludge something up however the situation calls for it (keeping in mind the safety and legal aspects of it). <S> But at that point, you're not worrying about what amperage the copper wire is going to melt. <A> Fusible links in automotive use are normally made with a short length of wire (25-35 mm) that is two wire gauge sizes smaller than the wire supplying the load. <S> For example, if the wire is 12 AWG, use 14 AWG wire for the fusible link. <S> (Convert American Wire Gauge, AWG, to whatever system of measurement you are using in your part of the world.) <S> This will prevent a fire that destroys the vehicle, but the wiring in the system may still overheat and be permanently damaged. <S> If super-fast response is not required, consider replacing the fuses with an appropriately rated thermal circuit breaker. <S> A thermal-magnetic circuit breaker, such as those used in house wiring, can be used for faster response to overloads. <S> You may find that replacing the fuses two or three times costs more than the circuit breaker and its enclosure.
|
If you can't, try soldering two 50A fuses to the terminals.
|
how to drive 5x5 IR LEDs with 100 mA, using Arduino I am trying to do my first Arduino project, and I would like to drive a 5x5 matrix of IR leds. I would only like to have one switched on at a time. I understand that in an array I can do it with 5 lines for anodes and 5 lines for cathodes. My problem is that I cannot supply 100 mA with an Arduino. What would be the easiest way for me to supply 100 mA for these leds? I think they don't need current limiting, if I could drive them with 1.35 V voltage, it would work perfectly. I've measured a couple of my LEDs and they used exactly 100 mA at 1.35V. I know that I can use a transistor for supplying a higher current power source, but I don't understand that how could I do it with a 5x5 LED matrix. How many transistors do I need? 5 or 10? And if I need 10 transistors, then I understand the 5 which supply 1.35 V, but what are the other 5 doing? How can a transistor supply 0 V? Or, as an ugly trick, I was thinking about using a Mega, and for each line, using 3 pins in parallel . I mean one pin has a current limit of 40 mA, so 3 pins would be enough for 100 mA. That way I wouldn't need to care about an other power source, and transistors. I could just run the whole thing from USB. All I would need would be 5 resistors. Is that a good option? Or, as an alternative, I was thinking about using Rainbowduino in itself. It has current limiting, and in theory I could possibly run the whole thing from 5V USB. But it seems really complicated to just control 5x5 LEDs. I don't understand the whole idea about source driving the anode and current driving the cathode! Isn't it a circuit , I mean we either set the source, or the current? And why does the anode has 500 mA current limit, while the cathode has 120 mA? So while it looks good on paper, I think I'm really not on a level to understand how it works. All I would need is to control 5x5 LEDs, where each of them would require 100 mA @ 1.35 V. What would be the best way you’d recommend me doing it? <Q> First, a safety warning 25 IR LEDs at 100 mA each will be a LOT of IR radiation. <S> If they're close together, your pupils are dilated (indoors, this is often true), and/or you're close to the matrix, you could really hurt your eyes. <S> The image of the matrix will be focused to a tiny area on your retina, and you'll hear/feel a little pop as the blood and fluid there boils. <S> You'll have a permanent blind spot. <S> Not fun. <S> BTW, it's your job to make sure that you understand what you're doing and don't hurt yourself, not mine. <S> I'll help you begin understanding, but I won't be held responsible. <S> My advice: For development (and production if possible), put bright green LEDs in series and physically close to the IR LEDs so that your blink reflex is activated at a bare minimum. <S> Techniques <S> There are many LED driving techniques. <S> None of them hinge on delivering 1.35V; that will change between LED batches, over time, and with temperature. <S> If you're just interested in current limiting, a few transistors will be sufficient. <S> The total transistor count will depend on your choice of a current limiting circuit topology, which is dependent on your heat sinking capabilities, routing area, and other contstraints. <S> There are also voltage/current regulation ICs and linear LED drivers which would simplify your job. <S> If you're interested in power conservation, many buck converters can be configured for current limiting, which would maximize your battery life. <S> You may want to have unlimited sink and limited source (or vice versa). <S> Alternatively, you might have just one current source, and mux it between the various LEDs. <S> Because of the safety issues inherent in this project, I'd recommend the latter option. <A> How did you measure the led current? <S> An led will act as a diode while in forward conduction and will take as much current as you can supply, or will destruct at its current limit. <S> Sounds as if your 1.35 volt supply might have been limited to 100ma. <S> Also if they are IR led's does your circuit need to pulse modulate them? <S> In any case, you could use a current limiter on the high side which could be anything from a single resistor (only one led on at a time) to a current regulator. <S> See LM317 datasheet for how to make a current limiter with only one resistor. <S> You will need 10 switching devices. <S> Five to switch the high side (positive) and five to switch the low side (negative). <A> First, I make a note that you only want one LED on at a time. <S> (I don't have the foggiest why you would do this, however, given that they're IR.) <S> Let's start with a simpler situation: suppose you would only need 1mA. <S> Then you could use a 74HC238 to drive the 5 rows. <S> The 74HC238 is a demultiplexer with 1 output out of 8 high. <S> Use a 74HC138 to drive the 5 columns. <S> It's a similar device but with the active output low. <S> (I'm following my own convention that rows source current, columns sink.) <S> So you would place the LEDs with the anodes connected to the rows, and the cathodes to the columns, and place a current limiting resistor in each row. <S> Now neither the 74HC138 nor the 74HC238 can drive a 100mA LED directly, so we'll need transistors to increase the current. <S> Let's start with the 74HC238. <S> This has an active high output, which will become active low if we drive an NPN transistor with it via a base resistor (the transistor works as an inverter). <S> Likewise, the 74HC138 is active low, but when driving a PNP transistor this will become active high. <S> Again, don't forget the base resistors. <S> So the functions or rows and columns have switched, and the 74HC238 will drive the columns (sinking current), and the 74HC138 will drive the rows (sourcing current). <S> Never mind that the 100mA will give 1.35V forward voltage, do place the series resistors for the LEDs, either on the PNP's collectors or on the NPNs'. <S> The internal resistance of a LED is too low to let it control the current if the voltage would deviate a little, for instance due to temperature changes. <S> Instead of the 5 NPN transistors you could use a <S> ULN2803A <S> if you want. <S> This would save you the base resistors.
|
Use hardware (in addition to software) techniques to limit the number of LEDs lit and/or the power delivered to the matrix to help ensure that only one LED is on at a time.
|
What exactly is the use of PSoC? Hey, yesterday I saw a demo (which was actually meant for my seniors) of a PSoC 5 board by cypress at my college. They demonstrated how to use the capsense built into the board and one of the PSoC chips to turn an LED on and off. This is basic Hello World stuff. Although I thought it was cool and all, I really couldn't figure out in what way could I use those boards. Yes they eliminate all the need for making my own hardware, but how would I use this capability? <Q> Cypress PSoC devices have blocks (PWM modules, counters, timers, UARTs, ADC, DAC, etc.) <S> that can be configured easily by a GUI, which can speed up the development time of a project. <S> (no need to design external circuitry, lay it out, etc.) <S> Also, the PCB real estate reduction is a nice plus (no need for external chips for all of these functions). <S> Since these blocks are actual hardware modules, you also don't need to spend time writing software to emulate these functions. <S> They can be configured to trigger interrupts, so your state machine can easily interact with the blocks. <S> The PSoC 5, for example, has the following blocks: 20-bit sigma-delta ADC, 8-bit IDAC, 8-bit VDAC, 12-bit 1 Msps SAR ADC, PGA, Op-amp, TIA, frequency mixer, comparator, reference, cap-sense block. <S> This sort of hardware is above and beyond what is provided in most microcontrollers. <A> Those development boards are just a way for someone to quickly get up to speed in using the PSoC. <S> The intention is for an engineer to become familiar with that chip and then go and design it into your own custom PCB (and build millions of them and make everybody fist-fulls of money). <S> They also use those boards at college as sort of a "gateway drug". <S> They get you hooked on the PSoC early, so when you go out into the real world you will tend to use them, ship millions, and make everybody fist-fulls of money. <S> Cypress is by no means unique in this. <S> TI, Atmel, STMicro, Freescale, etc. <S> all do this. <S> So, if those boards work for you then great, use them. <S> Otherwise, um, don't. <S> As an aside... <S> I used a PSoC when making the capacitive touch keypad for this paging station . <S> It turned out to be cheaper, more reliable, and better looking to make our own than buy a mechanical keypad. <S> We started out by evaluating one of the Cypress development boards, then quickly made our own PCB. <A> Some years ago we were going to do a project with another company and their design engineer wanted to sell us the idea of the PSoC, which he seemed to think of as the best thing since bread came sliced. <S> My colleagues and I had a look at it and dismissed it. <S> Is that so great? <S> No! <S> If I have configured blocks as timer that's because I need a timer all the time . <S> Other microcontrollers do have timers which are available all the time. <S> And talking about timers. <S> IIRC one building block could be used as an 8-bit timer. <S> For a 32-bit timer you needed 4 blocks, and with that most of those great reconfigurable blocks were used up. <S> Maybe things have changed since, and there may be more resources on recent parts, but at that time PSoC certainly wasn't an added value over other microcontrollers to us. <S> (We were using for example <S> NXP LPC2100 at the time.) <A> The beauty of PSoC is that they have a number of useful analog blocks already baked into the chip, so that in many cases, you can just use the processor with very few (and mostly passive) components to have a working product when other MCU's would require a bunch of external parts. <S> PSoC also comes with a large library of pre-canned solutions (in the form of software plus wiring diagrams) that allows you to quickly mix-and-match the solutions into a completed product. <S> -- BTW, just as an example - one PSoC design I made had a 2-axis accelerometer, a TFT LCD, and 8 capacitive sensing input buttons (four used as an iPod style "jog dial", and 4 other general button presses), an iButton port, and a audio speaker output. <S> The non-passives on that board consisted only of the accelerometer chip, the FET to drive the speaker, and a 5V-to-3.3V level shifter to interface to the TFT (because we couldn't source a 5V TFT display with the features we wanted). <S> The opamp circuitry for the accelerometer, the cap sense circuitry, and the various digital blocks were all contained within the PSoC. <A> An additional advantage of a device like the PSoC is that you can reconfigure the digital and analog blocks during runtime. <S> This lets you get a lot more functionality out of the chip with fewer pins.
|
Cypress sales engineers stress on the idea that you can reconfigure your PSoC in during runtime.
|
Solving DC Operating Point of Common Source Amp My lab partner and I are really stumped. We are given a circuit that has a FET with a \$V_{DD} = -V_{SS} = 15\mbox{ }V\$. We are told \$A_V = 20\$ and \$I_{DS} = 1\mbox{ }mA\$. The DC circuit is straightforward, containing an \$R_D\$ and \$R_S\$ between \$V_{DD}\$ and \$V_{SS}\$, respectively, and the FET drain and source terminals. There is a resistor \$R_G = 1.5\mbox{ }M\Omega\$ on the gate terminal of the transistor, connected to ground on the other end. We are asked what the DC current through this resistor (\$R_G\$?) is. We are then asked what \$V_{OV}\$ is. It seems to imply that we only need values that I have outlined so far, but maybe it would require values from a data sheet too, I'm not sure. Then it says, using \$V_{GS}\$, which I would know how to calculate if I knew \$V_{OV}\$, to calculate \$R_S\$. It also says we don't know \$V_{DS}\$ nor \$R_D\$ yet. Any help would be appreciated. I have tried reading the notes and textbook, but I can't find what I need. I would also appreciate any sort of good sources/guides to this, if you don't want to just give me the answer. I too, would like to understand so I feel less like a cheater, haha. <Q> I will try to help without giving too much away. <S> I just took an analog IC design course last semester, so I'm all psyched up about this stuff. <S> I hope that this does not come off as too academic... <S> In a common source FET amplifier, the input (gate) resistance is really really high, so the gate current is usually assumed to be zero. <S> So the current through $R_G$ should be zero. <S> Also, you should know that the drain current in a FET operating in the saturation region is given by <S> $$I_D = \frac {k <S> '} 2 \frac <S> W L (V_{GS} - V_t)^2$$ <S> Where $k'=\mu_n C_{ox}$ and is related to process parameters. <S> $\mu_n$ is electron mobility (in NMOS) and $C_{ox}$ is gate oxide capacitance. <S> W and L are the width and length of the channel, respectively. <S> So you have an expression relating drain current to gate-source voltage. <S> And $V_{ov}$ as well, since $V_{ov}=V_{GS}-V_t$. <S> You already know the drain current <S> , $I_D=1m\text{A}$. <S> You also should know that in a common source amplifier, the open circuit voltage gain $A_v$ is related to the drain and source resistances, $R_D$ and $R_S$. <S> And you know that $V_{GS}$ should be equal to the voltage across $R_S$. <S> (And you know the current through $R_S$, which is $I_D$. Correction: <S> $V_{GS}$ is not the drop across $R_S$! <S> But you should be able to determine the relationship from the information you have. <S> Without having a circuit drawing in front of me, and without having thought about it too much (please check my work!) <S> , I think that should give you enough info to make these calculations. <S> You may have to leave in a "transconductance" parameter in the expressions for things if you do not know $k'$ or $\frac <S> W L$. Analysis and Design of Analog Integrated Circuits by Paul R. Gray and Robert G. Meyer et. <S> al is my go-to reference for transistor circuits. <S> I have the 5th edition. <S> It's been around forever. <S> You would probably be okay with a 4th edition. <S> I would highly recommend this if you are interested in linear transistor amplifiers (op amps). <A> You're not saying what kind of FET you're using, a JFET or a MOSFET. <S> In the latter case the gate is isolated from the channel and there will flow no current through the resistor. <S> If it's a JFET <S> it's different. <S> Depending on the type the gate is either the N- or the P-part of a diode formed by the gate-channel junction. <S> This diode is reversely polarized, but there will be a leakage current which may generate a small voltage over a large enough resistor. <A> If I understand correctly, there is a resistor connected between the gate of an FET and ground. <S> Did they give you one? <S> For practical purposes, you don't need one, since this is an FET, and the gate resistance will be extremely high, making the current extremely low, essentially 0. <S> (But since this is a homework assignment, it might not be very practical.)
|
In order to figure out the exact (very small) current through this gate resistor, you'd need a datasheet, and to get a datasheet, you'd need a part number.
|
Is it possible to power a cordless drill from wall socket adapter? I have a dewalt 18V cordless drill. I am wondering if it is feasible to build an adapter that can power the drill from a standard US wall socket? A typical DeWalt drill needs 2.6 amps with no-load. I am assuming this jumps significantly higher under load. Most DeWalt motors have a stall current over 250 amps... Any insight would be greatly appreciated!! <Q> The answer is simple: Don't. <S> If you did, odds are that you'd spend lots of money on high-current step-down transformers or other kind of power supply and run the risk of destroying your cordless drill. <S> Corded drills, new from Home Depot, start at thirty dollars and will have as good or better performance than a $150 cordless. <S> You'd be time and money ahead by just buying one-- and not have the headaches or risk of electrocution. <A> It would be possible to build such and adapter. <S> It would depend on the rating of your drills and how you want to plug in the power. <S> Probably also need a regulator and a few capacitors. <S> You might be able to use the power transformer from a laptop or some other device. <S> But the voltage must match the drill and the ampere requirement must be very similar <S> (more you can fry the drill less <S> you can fry the power supply.) <S> As for how you attach the power supply I would recommend building the interface out of an old battery pack so you don’t have to modify the drill. <S> But it should be possible to add an auxiliary power jack to the drill as well. <A> It's possible but by no means straightforward. <S> An irreversible change is shown here: http://www.instructables.com/id/Convert-a-battery-drill-to-wall-power/ <S> I could have sworn that I saw plug adapters for cordless drills before but for the life of me I can't find them now. <S> I suppose it would be <S> difficult - batteries are high-current beast and fitting a power supply that could manage that sort of current into a manageable size on the drill might be difficult. <A> The aforementioned corded drills work from 110V which means 1/6 of the current. <S> I think it is easier to get a 18V battery with very high discharge current rating than it is to make a cheap high current 18V supply. <S> If you want to do this <S> then a toroidal power transformer + a bridge rectifier should work well. <S> I don't think you will need to stabilize the voltage but a quick inspection of the internal circuity may prove me wrong. <A> I used a 19v dc power supply from a computer with 4amp rating. <S> My drill would run in bursts unless I slowly pulled the trigger to full power. <S> If I put any amount of resistance, like trying to release a drill bit from the chuck, the drill would stall out and give short bursts of life never with any significant torque. <A> This not only possible, but additionally inexpensively and simply http://www.edaboard.com/thread212631.html
|
Either way a cordless drill has a DC motor so it will require a step down transformer and a full wave bridge rectifier.
|
How can I send data to the arduino from the computer? How can I send data to the arduino from the computer? I'm thinking sort of a prompt on the command line and some commands like "led1 on". Then, the arduino would process that and light up a certain led. <Q> One of my favorite features I added to a project years ago was to implement a VT100 terminal . <S> It was a pretty simple and flexible way to allow users to configure our Ethernet-based devices over an RS-232 connection. <S> It had a tabbed interface, status updates, and could also accept commands. <S> You can get surprisingly fancy with this simple approach. <A> Something that you can take advantage of is the Firmata approach. <S> It's basically a way to control ports / read inputs on a microcontroller using serial commands (which sounds like what you're trying to do). <S> There is already a Firmata library for Arduino here . <S> There are also a number of programs that implement the PC end of the conversation. <A> Your plan should work fine. <S> You do not say what computer you have (PC Mac Linux etc.) <S> but obviously you will need a program running on the computer that acts as a terminal emulator and allows you to send the command to the Arduino. <S> On the Arduino side you need to parse or decode the command, so best to keep the commands as simple as possible. <S> For example instead of "Led 1 on" perhaps "L1+" and "L1-" to turn it on and off. <S> Many other people have done this sort of thing with Arduinos <S> so if you Google you should find ots of samples. <A> if (Serial.value == "DATA") { doWork()} <S> When you get correct value, do anythink. <S> Like gMail notifier . <S> hope this helps. <A> The way I do it is with Max and the [serial] object. <S> You can implement a simple terminal using the text editor GUI object. <S> PureData is a free alternative, which has a similar solution, though I'm not sure what the object for serial communication is called. <A> How about Arduino Explorer: http://www.avr-developers.com/arduino_exp.html From the site:"... allows you to examine memory, examine and toggle I/ <S> O pins, explore the mapping of Arduino pin numbers to AVR port names and bit numbers, scan I2C bus, look at interrupt vectors and much more. "
|
You can use Serial.print("DATA") and on your pc build a program that listen on serial port.
|
Using a transformer to step down 230V to 12V My circuit uses main electricity in parts of the circuit. However my AVR and other components only need 5V. So I'm using a step down transformer to generate 12V. It's then regulated to 5V. The transformer won't convert it to dc for me will it? It just steps it down then outputs a certain voltage depending on it's coils. So I would need to convert it to dc myself right? <Q> You are correct. <S> The transformer will only reduce the voltage (and increase the available current), so you need to add additional circuitry to rectify, smooth and regulate your 12\$V_{AC}\$ transformer output to 5\$V_{DC}\$. <S> This is the type of circuit you should be looking to build: <S> The transformer reduces the voltage from mains to 12\$V_{AC}\$ (RMS). <S> The Diode Bridge (known as a bridge rectifier) will convert 12\$V_{AC}\$ to 15\$V_{DC}\$. <S> The voltage is \$\sqrt{2}\$ times minus two diode voltage drops higher than the input voltage because the rectifier output is the peak AC voltage, not the RMS AC voltage. <S> The first capacitor will smooth out the ripples that come from the output of the AC to DC bridge rectifier. <S> The LM7805 regulator will maintain a constant voltage as the load varies. <S> For example if you are switching a light bulb on and off, the current will go up and down, and if you didn't have a regulator then the voltage would drop as the bulb is switched on. <S> The regulator keeps it at the 5\$V_{DC}\$ your microcontroller needs. <S> The final small capacitor filters out any noise or interference on the regulated side of the circuit. <A> The transformer is the first part of the power supply, but you need more to get DC. <S> The varying input voltage of the transformer creates an equally varying magnetic field in the metal core. <S> This magnetic field in turn creates an, again varying, voltage on the output. <S> Both input and output voltage are sine waves. <S> It's called AC (alternating current) because the sign of the output voltage changes continuously, 100 or 120 times per second, depending on the country you live in. <S> You rectify this AC voltage to get rid of the sign changes; one pin will always be positive with respect to the other one. <S> The one thing which remains to be done now is to flatten the curve, get rid of the bumps. <S> This is done by a capacitor. <S> You now have a DC (direct current) voltage which is already usable for a number of situations. <S> This voltage, however, may still show slight variations, which may be unwanted. <S> To get rid of those you follow the capacitor by a voltage regulator. <A> Such devices do convert to DC for you. <S> A few of them use exactly the circuit BG100 shows, but nowadays many of them use a switching regulator technique that requires more parts, but has a lower net cost. <A> The transformer won't convert it to dc for me <S> will it? <S> Correct! <S> AFAIK <S> it just steps it down then outputs a certain voltage depending on <S> it's coils. <S> So I would need to convert it to dc myself right? <S> Correct again! <S> Your transformer will convert \$120V_{AC}\$ to \$12V_{AC}\$. <S> So you'll need to regulate it to \$5V_{DC}\$.
|
If you need 12 VDC, it may be simpler to use an off-the-shelf wall wart with a 12 VDC output.
|
Can I use an inductive load in an ATX-to-bench supply project? I'm doing my second ATX-to-bench supply conversion, and this time, I've decided to follow the "book" and place the correct minimum loads on all the rails instead of merely the 5V rail. This involves fairly small loads on the 3.3V rail and 5V rail (under 2W each), but a fairly sizable load on the 12V rail (12W). I've sourced very cheap cement resistors to provide the loads, but I understand that they are inductive loads, and I would like to know, will that cause a problem? I assume it will not, because many of the loads in a normal computer are already inductive (HDD motors etc), but I would like confirmation from someone who knows a lot more about switching power supply design than I do. <Q> In this case, you're fine. <S> Inductive loads only show their "inductiveness" when the current changes, and since the load is going to have a constant current then you're all right. <S> If you want to be extra careful, you could put a diode between your output and GND (oriented so it's not normally drawing a current). <S> That would protect your power supply during turn on/off from any inductive kickback. <S> Even if your "minimum loads" were not inductive, this would be a good idea just in case you want to connect motors or whatever to your benchtop supply. <A> I would venture that any inductance in the resistors would be swamped by the output inductance of the power supply, so the likelihood that they'd cause any disturbance would be low. <S> I have a hard time believing that the inductance in a small cement wire-wound resistor could store enough energy to do any damage to the output of a power supply, especially one that doesn't have ORing (e.g. an ATX computer supply, which isn't intended to operate in parallel with another supply) <S> - the output capacitors will do their best to absorb any energy kicked back from the inductance. <S> Imagine that you have 1 microhenry of inductance, and 1 A of current. <S> The energy is: \$\; <S> E = \frac{1}{2 <S> } \mbox <S> LI^2 = 0.5 <S> \mu <S> J \$ <S> If that 1 A were to dissipate in 1 microsecond, the voltage would be: \$\; <S> V = L \frac{di}{dt} \mbox = <S> 1V \$ <A> If you examine the typical circuit used in such regulators, you will see that the load is outside the control loop, so an inductive load should not really matter. <S> This is confirmed by the data sheets of switching regulator devices, such as the Simple Switchers made by National Semiconductor; no requirements are mentioned for the type of load.
|
Of course there are already a bunch of inductive loads already in the system, but also the inductive loads only play a part when the supply turns on or off. The risks are minimal.
|
if a standard three-phase 400V AC connection is rectified what DC voltage comes out of it? If a standard (in Europe and a large part of the world except North America and Japan) three-phase 400V AC (three lines having 230V RMS voltage if measured to neutral each) mains supply is rectified with a standard 6-diode rectifier like this: What DC voltage value will come out of the rectifier? How to calculate it having the RMS AC source voltage given? Are there any other ways to wire diodes to get a different voltage (without using any transformers or anything else then just diodes), what are they and what DC voltage will come out then? <Q> If you measure from the + of the load in your figure to the neutral of your AC supply, the peak voltage will be ~325V. <S> If you hook up a load like that, you aren't actually using a full-wave rectifier. <S> The simplest way to get 565V is to start from 400V and apply the standard \$\sqrt{2}\$ scaling from \$V_{rms}\$ to \$V_{p-p}\$. However, starting from 400V is skipping part of the calculation. <S> The more thorough way to derive 565V is to calculate it as: $$(325 \text{V}) <S> * \max_{\theta <S> } \left\{ \sin (\theta + <S> \frac{2 \pi}{3}) <S> - \sin (\theta) <S> \right\}$$ <S> The expression is maximized when \$\theta\$ <S> is \$\frac{5 \pi}{3}\$, and the maximum value is \$325 \sqrt{3} = <S> 563\$. <S> There is a detailed analysis including some java applets here . <A> This configuration is commonly known as a star, or WYE configuration. <S> Its easier to see if you break it down into two halves. <S> Phase to neutral is 230 vrms. <S> Three phases each connected to a diode anode and all three of the cathodes tied together. <S> If measured from neutral to the cathode connections, you would expect to see 230 * 1.414 = 325 vdc. <S> This represents the "peak" voltage of the waveform. <S> Now do the same with the other half of the bridge, which will create a negative voltage of equal value in respect to the neutral. <S> The pulses interweave with each other affectively filing in the gaps of the positive pulses, resulting in 6 pulses creating a smoother dc voltage. <S> The voltage unfiltered would be slightly less than 325 volts. <S> If a filter were added such as a capacitor, the voltage would average close to the calculated value minus the "ripple" value which is always present in a filter. <S> CAUTION: <S> These voltages are lethal and the proper precautions need to be taken to prevent injury or death! <S> The explanation is for illustration purposes only. <S> In real practices this circuit would be built with an isolation transformer and circuit protection, such as fuses. <A> I think Steve and Andy have it pretty well explained <S> but it really helps me to look at the voltage waveforms and <S> see how exactly they add up. <S> Note that the time between peaks ~5.5ms which is a direct result of the three peaks, one from each phase, being offset by 120 degrees and being added together. <S> Three waveforms are plotted:V(v+) is the voltage from node V+ to ground. <S> V(v-) is the voltage from node V- to ground. <S> V(v+,v-) is the voltage across the load resistors. <S> Also, you can right click and view image to see larger versions which are much more legible. <A> Three phase AC through a rectifier produces this waveform: <S> The "DC voltage" out has two possible meanings: average, and RMS. <S> RMS is how much heating energy a load in this configuration will see. <S> The output waveform is a sine wave <S> between 60 and 120 degrees, repeated. <S> Take the RMS of a sine wave between those two angles and we get the RMS of the entire wave. <S> RMS is root-mean-square: take the square root of the mean of the square of the sine wave. <S> \$V_{peak} \sqrt{\frac{\int_{\frac{\pi}{3}}^{\frac{2\pi}{3}}{sin^2\Theta}}{\frac{\pi}{3}}} \$ <S> \$V_{peak} \sqrt{\frac{\frac{\Theta}{2} - <S> \frac{sin2\Theta}{4}\big]_{\frac{\pi}{3}}^{\frac{2\pi}{3}}}{\frac{\pi}{3}}} \$ <S> \$V_{peak} \sqrt{\frac{\frac{\pi}{3} <S> - \frac{\pi}{6} - \frac{sin\frac{4\pi}{3}}{4} + \frac{sin\frac{2\pi}{3}}{4}}{\frac{\pi}{3}}} \$ <S> \$V_{peak} \sqrt{\frac{\frac{\pi}{3} <S> - \frac{\pi}{6} - \frac{sin\frac{4\pi}{3}}{4} + \frac{sin\frac{2\pi}{3}}{4}}{\frac{\pi}{3}}} \$ <S> \$V_{peak} \sqrt{\frac{\frac{\pi}{6} + \frac{\sqrt{3}}{4}}{\frac{\pi}{3}}} \$ <S> \$V_{peak} \sqrt{\frac{1}{2} + \frac{3\sqrt3}{4\pi}} <S> \$ <S> \$.95577 <S> V_{peak}\$ <S> The average is slightly simpler to compute: \$V_{peak} \frac{\int_{\frac{\pi}{3}}^{\frac{2\pi}{3 <S> }} sin\Theta}{\frac{\pi}{3}}\$ \$V_{peak} <S> \frac{-cos\Theta\Big]_{\frac{\pi}{3}}^{\frac{2\pi}{3}}}{\frac{\pi}{3}}\$ <S> \$V_{peak <S> } \frac{cos\frac{\pi}{3} - cos\frac{2\pi}{3}}{\frac{\pi}{3}}\$ \$V_{peak} <S> \frac{2 <S> cos\frac{\pi}{3}}{\frac{\pi}{3}}\$ <S> \$V_{peak} \frac{1}{\frac{\pi}{3}}\$ \$V_{peak} <S> \frac{3}{\pi}\$ <S> \$.955V_{peak} \$ <S> And the peak voltage is, of course, the RMS of the input times the square root of 2.
|
If you measure across the load shown in your figure, the peak voltage will be ~565V; the DC voltage will depend on your load and filtering as others have noted.
|
Is there a more optimized way of making an incrementer than a full adder? I'm designing a very simplistic microprocessor as a project to help learn VHDL. So I'm needing something to increment the 8 bit program counter. I will need to increment it by two. Is there a better (either faster or less logic required for equal speed) design than using an 8 bit full adder? I'm also interested for the same thing for if you only needed to increment the PC by 1. <Q> Hmmm, this all depends on what exactly you are trying to learn. <S> A counter or adder in VHDL is super easy: <S> signal count : <S> std_logic_vector (7 downto 0) := <S> (others=>'0'); . . . <S> process (clk) begin if rising_edge(clk) <S> then if count_enable='1' then count <= count + 1; -- could be +2 also end if; <S> end if; end process; <S> And that's it! <S> The VHDL compiler will normally synthesize a full-adder for this, and them optimize out everything that isn't needed-- ending up with some sort of half-adder. <S> The nice thing about doing it this way is that your code is readable and easily understood and the compiler deals with figuring out the best way to implement it. <S> Now, if you are trying to learn about adders and counters and such then my little code snippet isn't going to help you much. <S> In that case you should implement a half-adder the manual and hard way. <A> You don't actually need a full adder for increment by 1; using half adders where the first input is set to 1, and carry bits are daisy chained to the next bit would do. <S> I'm not sure if there's a still better way to do it, though. <S> Incrementing by two could be done by ignoring the first bit of a number, and using the same method above. <S> NOTE: not a VHDL expert, can't state if this could actually be faster, but should be less logic. <S> Edit: Also, there's an expired patent for a simple binary incrementer that might be of interest: http://www.freepatentsonline.com/3989940.pdf <A> Counting by 2 would mean loading everything but the least significant bit into the counter since that bit never changes. <S> This probably won't be better than a half-adder implementation, but perhaps it is easier to understand :) <A> What specifically you need is series of T-triggers + a bunch of AND gates. <S> You should pass 'toggle' signal to bit n only if all previous bits were 1. <S> This is much faster than full adder, does not require carry look ahead and eats way less transistors. <S> PS. <S> Some time ago I was wondering exactly the same question ;-) <A> It certainly does seem as though there should be a simplification -- after all one of the inputs is limited to a fixed value that is the same every time! <S> Unfortunately, that's only enough to get a small reduction in number of gates, but not much of a reduction in time to complete. <S> That is because you do get simpler adders -- they only have 1 input plus carry, so the base adders have about 1/3 fewer gates. <S> But the delay is determined by the carries, which are not reduced, they still must ripple up the whole chain. <S> So, you don't get much meaningful speedup. <S> And, if you want to go fast, at least half the gates are dealing with the carries, so that 1/3 gain reduces to about 1/6 fewer gates overall. <S> In the end, roughly the same speed, and 85% of the size of a full adder.
|
Maybe you can use a binary counter built from JK flip-flops , one where you can load the bit value into each flip-flop, then just toggle the clock.
|
How do BJT transistors work in a saturated state? This is what I know about NPN BJTs (Bipolar Junction Transistors): The Base-Emitter current is amplified HFE times at Collector-Emitter, so that Ice = Ibe * HFE Vbe is the voltage between Base-Emitter, and, like any diode, is usually around 0,65V. I don't remember about Vec , though. If Vbe is lower than the minimum threshold, then the transistor is open and no current passes through any of its contacts. (okay, maybe a few µA of leak current, but that's not relevant) But I still have some questions: How the transistor works when it is saturated ? Is it possible to have the transistor in open state, under some condition other than having Vbe lower than the threshold? In addition, feel free to point (in answers) any mistakes I made in this question. Related question: I don't care how a transistor works, how do I get one to work? <Q> Saturation simply means that an increase in base current results in no (or very little) increase in collector current. <S> Saturation occurs when both the B-E and C-B junctions are forward biased, it's the low-resistance "On" state of the device. <S> The properties of the transistor in all modes, including saturation, can be predicted from the Ebers-Moll model. <A> Your \$I_{CE}\$ = \$I_{BE} <S> \times h_{FE}\$ isn't quite right. <S> This equation shows what the collector current could be if given sufficient collector voltage. <S> Saturation happens when you don't give it enough voltage. <S> Therefore, in saturation, \$I_{CE} \lt I_{BE} \times h_{FE}\$. <S> Or you could look at it the other way around, which is that you are supplying more base current than needed to handle all the collector current the circuit can provide. <S> Put mathematically, that is \$I_{BE} \gt I_{CE} \mathbin{/} <S> h_{FE}\$. <S> Since the collector of a NPN will act like a current sink and in saturation the external circuit isn't giving it as much current as it could pass, the collector voltage will go as low as it can. <S> A saturated transistor typically has around 200mV C-E, but that also can vary a lot by the design of the transistor and the current. <S> One artifact of saturation is that the transistor will be slow to turn off. <S> There are extra "unused" charges in the base that take a little while to drain out. <S> That's not very scientific and only roughly described the semiconductor physics, but it's a good enough model to keep in your mind as a first order explanation. <S> One interesting thing is that the collector of a saturated transistor is actually below the base voltage. <S> This is used to advantage in Schottky logic. <S> A Schottky diode is integrated into the transistor from base to collector. <S> The on state voltage will be a little higher since the transistor isn't fully saturated. <S> The advantage is that it makes the off transition faster since the transistor is in the "linear" region instead of in saturation. <A> When it's saturated, the collector current is not \$h_{FE}\$ times the base current anymore. <S> It's less, how much, it depends on the rest of the circuit (I'm talking about the simplest model you can think of). <S> In saturation, the \$V_{CE}\$ voltage can be considered more or less constant and you can call it \$V_{CEsat}\$, let's say around <S> \$0.2\mathrm V\$.TYour BJT is saturated when both its BE and BC junctions are active. <S> That limits the \$I_C\$ current to <S> less than \$I_B h_{FE}\$ and pins the \$V_{CE}\$ voltage drop to \$V_{CEsat}\$. <S> Why do you care of having your BJT in open state if there's no current going through it? <S> It's like having your tap open with no water in the pipe :D
|
When the collector gets low when it's nearly in saturation, it steals base current which keeps the transistor just at the edge of saturation.
|
What's the difference between NPN and PNP transistors? Suppose that I know how an NPN transistor works . How different is a PNP transistor? What are the operational differences between a PNP and a NPN? <Q> PNP transistors work the same way as NPNs do but all voltages and currents are reversed. <S> You connect the emitter to the higher potential, source current from the base and the main current flows into the emitter and then exits through the collector. <S> \$V_\rm{BE}\$ will be \$-0.7\,\rm{V}\$ <S> but it's magnitude <S> should be the same in both PNP and NPN if you use complementary parts. <A> Electrons are more mobile than Holes Which means that PNP is not as good as NPN. <S> For Si BJTs the PNP types are behind when it comes to breakdown voltage and really high power. <S> For general purpose devices like BC337 / BC327 things for all intent and purpose are the same but if you wanted to do a off line SMPS it wouldnt be easy or practical at 1KW. <S> For germanium the NPN is supposed to be better but it is not. <S> This is due to manufacturing issues. <S> The AC127 is not nearly as good as the AC128 and the AD161 is not as good as the AD162 <S> and yes these devices were sold as matched pairs. <S> The ratio of electron to hole mobility is a determining factor in how close the PNP will be to the NPN. <S> This is much worse for SiC so one would expect lousey PNP BJTs so they probably wont bother making them. <S> For some reason PNPs have lower noise so they are favoured in diff pair input stages. <S> The abundance of highside driver chips is proof that PNP is not as good as NPN. <A> The only difference lies within the functionality of the transistors. <S> In grounded (common) emitter configuration, when a base current is provided (or to be more practical-when base is connected to 5v supply) of a PNP transistor, no conduction takes place as the majority carriers in n region are electrons whose motion is suppressed and no path is formed b/w emitter and collector. <S> Thus no o/p is obtained at the emitter junction. <S> If base current is removed from the transistor a virtual path is formed b/w the emitter and collector which offers certain resistance to electron flow which is subsequently altered by the base current (or voltage).If in such case, the Vcc is directly connected to the collector and emitter is grounded through a resistance(possibly 10k), then Vcc gets a direct path to appear at the emitter junction. <S> Thus if o/p is taken at emitter in case of PNP, the config is that of an inverter while at collector the transistor works as a simple switch or buffer.(This is exactly the opposite of NPN config.)Due to dearth of certain simulation software, i am unable to present a pictorial view. <S> But i hope this would serve the purpose.
|
NPN and PNP transistors are different.
|
Wiring transformer output in series to get twice the voltage possible? I already have a transformer 2x 6V outputs. I was wondering can I these in series to produce 12V? Similar to how you would wire two 1.5v batteries to produce 3V. <Q> Yes, transformers with this configuration are designed to be used in this way if required. <S> You can wire the secondary coils in series to get twice the voltage. <S> You can wire the secondary coils in parallel to get twice the current. <S> You can use each secondary independently to get two power supplies. <A> If wired in parallel, you will get the same voltage as just using one, but you will double your VA. <S> You MUST use transformers of the same VA and turns ratio. <A> The answers to these questions are not as straightforward as the responses suggest. <S> Connecting transformers in parallel can be dangerous if the output voltage of each is not very close to the same. <S> Transformers exist that are made for the purpose. <S> Even a difference of 1V between the two can cause big problems. <S> The voltage difference causes circulating currents that will quickly overheat windings and, either rapidly or over time, break down the insulation between windings and short them out. <S> Because transformers are often used with currents and voltages that can be hazardous to life, an investigation of the specific considerations applicable to the application is much safer than experimentation.
|
If wired in series, you will get double the voltage but the VA will be equivalent to only one of the transformers.
|
Where to order small batches (say, 100 units) of custom boxes for your projects I'm creating a simple effect for musical instruments and I'm wondering how I will package it when it'll be ready. I mean, I can create some kind of stomb box with my own hands but if I'm ever going to sell it I need to find a way to create some small batch of them. The problem is it's hard (impossible?) to find some manifacturer who cares of building, say, 100 boxes. On the other hand I cannot order thousends of them because I'm not sure I'll sell the out. So has anyone faced/solved this problem in the past?Thank you very much ;) <Q> Looks like <S> Pedal Parts Plus has a variety of pre-drilled cases and will do custom drilling for you. <S> This looks like your best bet, if you ask me. <A> Some enclosures on Digikey have notices that the company will do custom drilling and cutouts on their stock enclosures (in sufficient quantities). <S> This is one such enclosure: notice the 'Customize your Enclosure' link: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=SCR4B-ND <S> Here's a link to a presentation on the custom capabilities of Bud Industries: <S> http://dkc1.digikey.com/us/en/tod/BUD/ModifiedEnclosures_NoAudio/ModifiedEnclosures_NoAudio.html <S> So if you can find an enclosure on Digikey that gets close, some of the suppliers may be able to help you. <A> One option would be to use a service like Ponoko , who do custom laser-cutting on a variety of materials. <S> You can do the assembly and finishing yourself. <A> Like the others said, manufacturers will often mod small quantities. <S> However, for stuff like guitar stompboxes, some builders go the other way and make the hand-modification of the cases a reason to charge extra. <S> I don't have my links anymore, but I have seen a lot of boxes that were just off the shelf aluminum cases <S> then hand painted, stained, corroded, or just polished with high quality hardware then sold for upwards of $300 each with about fifty cents in electronics inside. <S> Some of these are truly works of art! <S> Another poster pointed out that the electronics is usually trivial, the real work is in the enclosure. <A> I would find a local guy with laser cutter - this way you can quickly make cases out of wood or metal(harder to do, obviously). <S> Personally, I love wooden cases, no production items can have them :-) <S> Ordering 100 non-standard plastic cases on a factory is not possible, production preparation will eat your money for such a small series. <A> Also, check out ToolLess - <S> www.toolless.com <S> - they claim to have a tool-less no-mold capability for unique shapes. <S> I haven't personally used them, but heard about them from a friend-of-a-friend, and it sounded pretty impressive. <S> The idea is that they "build up" the case by fusing/joining together sub-panels made up of cut sheets or hogged-out stock.
|
Check out Polycase, PacTec, Hammond, Bud for off-the-shelf enclosures that you can get custom cutouts done either turnkey from the manufacturer, or shopped out to a machine shop.
|
What good resources are there for a high school student to start learning electronics? I have a couple years of experience with computers - I can program, and I know computers really well. But I don't understand them. This leads to a broader scope - I don't understand electronics in general. I have read some internet articles, like howstuffworks.com, that kind of thing. I understand the basics, but about 90% of it is beyond me. Where do I learn? What books can I read? What websites can I go to? I'm especially interested in knowing what books I can read to understand more about electronics. <Q> Considering you're of 'high school age', I think the MIT courses are maybe very difficult (from Mark's post) <S> They generally require a lot of math and previous knowhow from highschool. <S> Or am I only able to pick out the courses that handle circuit theory and maths? <S> You may find this very interesting: <S> http://www.allaboutcircuits.com/vol_1/index.html <S> Start at chapter 1 and read through. <S> You may find some parts that you think is familiar, but read through it anyway. <S> You may already know what a Volt and an Amp is, but it might also be nice to read through some literature so you can conform your previous knowledge. <S> This website is focussed on analog electronics. <S> It's about understanding how electricity and electronic components work. <S> This doesn't have a particular big focus on digital electronics, which you may want to know more about considering you're interested in writing computer programs. <S> As Mark said, Arduino is a great platform to start from for digital electronics. <S> It contains a small microcontroller with an easy to pick up language to start with programming embedded systems. <S> Out of the box you won't be that concerned with configuring registers of microcontrollers. <S> You can get a ton of addons created by 'the community' for it. <S> You can get a ton of stuff from Sparkfun . <A> Get yourself some basic components (either individually or just get a starter kit from a supplier ): <S> Components Wire / alligator clips LEDs, Lamps, buzzers, motors, etc Resistors, Capacitors <S> , Some transistors (TIP120 are a great start for learning about switching) <S> Switches, pressure sensors, etc Breadboard Battery holders (can be scavenged from consumer electronics / toys). <S> Batteries are a good way to start out playing with electronics because there typically isn't enough juice to hurt yourself. <S> Tools And some very basic tools (if you don't already have them) <S> : <S> Multimeter <S> (so you can get an idea what is happening at various points in your circuit). <S> Some alligator clips Note that I left out a soldering iron. <S> You don't need one to start out with (but you'll likely want one if electronics holds your interest). <S> Start Hacking! <S> Try making buzzers buzz, LEDs light (without burning them out). <S> Good luck! <A> MIT's Free Online Courses are a good place to start. <S> Another option is to just grab an Arduino and start hacking from there. <S> They have a pretty large and helpful online following. <S> As for books, hopefully someone else can provide some good starter references, my only experience with intro books are the ones I had to use in college which were... <S> meh... <A> Make - Electronics Discovery . <S> This starts at a very basic level explaining resistors, capacitors and transistors. <S> It is more a small taste in the different areas of electronics. <S> Getting Started in Electronics . <S> This is also an introductory text. <S> It is an older book but many of the principles still apply. <S> The author Forest Mims is well known. <S> He published a few small books which you were able to purchase from radio shack they dealt with different areas of electronics. <S> It would be great if you could get your hands on them. <S> Finally at a more advanced level <S> Art of Electronics <S> This book is considered the standard text on electronics in some circles. <S> It is still considered an introductory text but it is a little more advance. <S> There is some math sometimes but you can still get the general ideas without fully understanding the math. <S> I am currently studying this book. <A> I'm a high school student. <S> A good basic introduction is the MAKE: <S> Electronics book. <S> It teaches you hands-on. <S> The handbook I use for my Electronics class is Teach Yourself Electricity and Electronics by Stan Gibilisco. <S> It's good, if not dry. <S> Another good way to learn it is to become a Ham Radio operator. <S> That's how I learned the meat of my electronics. <A> Fix stuff. <S> Get all kinds of broken electronic things and figure out how to fix them. <S> "Oh boy, it's broken! <S> Life doesn't get any better than this. <S> " <S> Make sure you don't do unsafe things <S> (mains voltage is deadly, for example, and microwave ovens are no toys because they present numerous safety issues when the cover is removed), but otherwise, go and fix whatever you can. <A> When I was in high school my understanding of electricity was very flawed. <S> I eventually worked out my misunderstandings in college, but if I had done it sooner <S> I could have built actually working circuits earlier, instead of just being confused. <S> I think this list of Electricity misconceptions is really helpful to read. <A> If books is what you want, you might want to look at this thread: <S> Basic Electronics Book
|
There are a tremendous number of resources available on the web, and some great starter kits available from many vendors. The best way to learn is by playing with circuits.
|
How can I duplicate Apple's tiny USB power adapter? I'm designing a project that controls 110VAC (plug in, socket out), and also has a microprocessor inside. I'd really like this project to not need a wall wart to get it's 5V, so I want to include a small DC power supply inside. Apple's USB power adapter looks like a great small design that I can fit inside my case. I'm guessing this is a switching power supply but I can't find any teardowns or schematics anywhere. Does anyone know how this thing works? I know I could make my own 110V->5VDC supply from a bridge rectifier and some caps, but I'd like to also be super safe. What's in the Apple brick? <Q> You're in luck: I just posted a detailed teardown and schematic of the iPhone charger. <S> Internally it's a complex flyback quasi-resonant switching power supply controlled by a L6565 . <S> Building your own switching power supply is probably more complexity and danger than you want to include in a project. <S> I'd recommend going with the wall wart - there's a reason most products use one. <S> If you really want a built-in power supply, I'd recommend getting a pre-built OEM one rather than building your own. <S> And if you really want to build your own, I'd recommend a simple linear power supply instead of a switcher. <A> The simplest way would be probably to buy a USB power adapter (probably not from Apple but a cheap replacement) and use this (preferably without removing it from the casing to avoid any electric shock hazards). <S> For mass-market solutions you may check LinkSwitch chips. <S> They are quite cheap and require almost no external components. <S> I am afraid that in small runs you will have a problem with obtaining cheap ferrite power transformers <S> but I would love to be proved wrong on this one. <A> Here are some photos of the Apple iPhone AC adapter, it's got two small boards with components on both sides of each. <S> For the power output it could not be anything other than a high-frequency switching regulator. <S> However, the controller IC is not marked with a part number. <S> I tried to track down who made it, even contacted an EE Times editor to request a teardown article. <S> He asked Portelligent for their teardown <S> but they had to look at the bare IC under a microscope to find the ST logo and <S> even then they weren't able to point to a publicly available part. <S> So either it's something custom for Apple, or Apple got temporary exclusivity before the chip was released generally. <S> http://www.myinnergie.com/mMiniAC/ <S> For your purposes, assuming you don't need several watts to run a little microprocessor, there might be a solution that's much cheaper or less complex. <A> Looks like MAX 611 <S> couldn't make it any easier: <S> http://www.selectronic.fr/includes_selectronic/pdf/Maxim/MAX_611.pdf <S> (sorry about the random site, not turning up a link on Maxim's site). <A> The safety first. <S> But you are describing some rare case, when galvanic insulation is not needed that bad. <S> To provide the power with 3-5 volt and low current you can use UL certified capacitor under 1 uF, diode bridge and zener. <S> but only and only if total current is less than 10-15 mA. <S> With larger current things become difficult because of thermal dissipation and real fire hazard dangers.
|
If you are not experienced with electronics I would suggest against making you own power line switched-mode power supply. You could also easily do a small transformer based supply but it will be quite a lot heavier and larger than the USB "chargers". A few clones of the adapter have appeared on the market so there must be something available by now.
|
Standard way to reverse a motor? Is using a relay and some switches a standard way to reverse a motor or is the preferred method to use a H-Bridge? The circuit for reversing a motor with a relay looks like this: Note that this circuit is just to show the relay and motor connections. The main control would be the coming from a PIC not an op-amp. Me and my teacher are having disagreements about the best way/standard/preferred way to reverse motors, so what would engineers do? <Q> Whenever I have done motor control it has been with an H-Bridge. <S> But the ideal solution depends on several factors: <S> What is the voltage/current required? <S> For <24 volts and <5 amps, an H-Bridge is normally the preferred solution. <S> The reason for this is the MOSFET's that can handle those voltages/currents are easy to get, easy to design with, and fairly inexpensive. <S> Is speed control, dynamic breaking, or regenerative breaking required? <S> If so, an H-Bridge is again usually the best. <S> But in the case where there is no speed control, and the frequency of "direction reversals" is low, and/or the voltage/current is very high <S> then a relay could be a good choice. <S> The main selling point of the relay is its simplicity and ease of use. <S> An H-Bridge will always perform better but it is harder to design, sometimes can be more expensive, and adds complexity that is not always required. <A> How about the best of both, use a relay to switch the motor direction and a power device to control the speed of the motor (It's how early speed controllers worked - it was a lot cheaper to just have a single power stage, and a relay could switch easily - you just need to make sure the motor is going slowly when switching to avoid arcing in the relay) <S> So 1 pin would control the motor and 1 PWM output run a power stage to control the speed. <S> (A power stage could be a simple as a resistor, a transistor and a reverse protection diode) <A> For small currents, (<2A) a single chip H-bridge is economical and more compact. <S> For heavy duty high current motors, the reverse is true, a relay will be cheaper and smaller than a high-power H-bridge. <S> If you don't need electronic speed control, and aren't changing direction zillions of times an hour, a relay is a fine solution, and a heckload easier to debug. <A> You can also get speed control with an H-bridge. <S> They are popular because of that.
|
If you are switching a lot of current, a relay is cheaper (especially in the old days when high power transistors cost a lot more!), a H-bridge has the advantage of PWM speed control, and instant reversal.
|
Solder very close together contacts I have contacts that are so close together I'm either bridging them, or melting the neighboring one. What am I missing? <Q> Flux. <S> You're missing flux, and lots of experience. <S> I regularly solder with a tip that easily covers 10 or more pins at a time, but by carefully controlling how much solder is on the tip and using lots of flux I can get perfect solder joints. <S> When soldering a TQFP or TSSOP package here's what I do: <S> If the pads on the PCB are not super flat, then use some solder braid and suck up some of the solder on the pads to make them flat. <S> Put the part on the PCB and carefully align the part. <S> I use water soluble liquid flux, but others are successful with other liquid fluxes. <S> Clean the iron tip and put a little bit of solder on it. <S> Carefully touch a corner pad ( <S> NOT the pin itself) and let the solder wick up and tack down that pin. <S> Repeat 4 & 5 for 2 to 4 corner pins. <S> Reapply flux as required. <S> Put some solder on the tip and run the tip down one complete side of the chip. <S> Ideally you won't have bridges, but if the last 2 or 3 pins are bridged then you did good. <S> Ignore the bridges. <S> Add more flux <S> (you can never have too much flux). <S> Repeat 7 and 8 for all sides of the chip. <S> Add flux, clean the tip on the sponge, and touch the tip to the bridge. <S> Solder should stick to the tip. <S> Repeat until bridges are gone. <S> In bad cases, careful use of a solder wick can help. <S> But be careful because it is easy to bend the pins and then you're, um, hosed. <S> Clean the flux off. <S> Some flux can be conductive, like the water soluble flux, so make sure it's clean (and dry) before powering up. <S> I should mention that my soldering iron has a straight, but angled tip. <S> It's often called a chisel tip. <S> This tip makes it easy to do everything from 1206's, 0402's, TQFP's, TSSOP's, etc. <S> It just doesn't work for BGA's, QFN's, and some of the weirder things. <A> Use less solder, lots of flux, and practice. <S> Consider pre-soldering the pads on the PCB. <S> Cover the area with some flux, and use your iron to push a bead of solder over the necessary pads. <S> It is surprising (or was to me) how little solder was necessary. <S> After the small amount of extra solder you have added to the pads have cooled, add a little more flux, place your device, and use just your iron to attach. <S> It takes practice, and it seems everybody prefers something a little different. <S> I learned by watching and reading some of the superb sparkfun tutorials and other random youtube videos. <S> I think watching other people in action is the best way to learn. <A> To solder an item like that you need to: <S> Use flux <S> Make sure your iron is clean and tinned <S> Have your soldering iron hot enough that your solder will melt as soon as you touch it to it <S> I think I have everything covered, am I missing anything? <A> Besides flux (to prevent a short in the first place): solder wick! <S> (to remove excess solder after the fact) <A> Clamping a sewing needle to the shaft worked for me, too. <S> You need lots of flux, less solder. <S> If even that fails, and you need to solder some really tricky stuff like BGAs, I have some other tricks in the box, like stove-soldering, etc. <S> Those tricks are likely to break your hardware, so I'm not gonna advice them before suggesting the needle-trick.
|
Use a solder diameter that is rather narrow Heat up your trace and pad and then slide the solder into it You can try to wrap some steel wire around the shaft of your iron, and use that as a makeshift extension. Don't rush this step, because if you mess this one up then nothing else will work correctly. Put a drop or two of liquid flux on the part.
|
is it OK to stack several DC-DC converters with their inputs connected in series to support a larger input voltage? Imagine I have a single-IC switching DC-DC converter which supports input voltage up to 16V, for example a LM2727 . Can I stack the inputs of such converter in series, for example 64 of them, and be able to apply up to 1024V across the combined input? I need lots of stable 3.6V outputs anyway (for charging each cell in a multi-cell battery separately). And on the output side, can I put a 3.6V cell between the outputs of each DC-DC converter and then connect all the batteries in series - will that be fine, or everything will blow-up? <Q> Connecting the chargers in series Connecting several such switchers in series will be a problem since they would have to share the same supply current (all would have to draw the same current all the time). <S> All the chargers have some input resistance which vary according to their immediate needs. <S> If one of the chargers draws less current then: It's input resistance rises so it will get less current. <S> This causes the voltage on this charger to rise. <S> Other chargers get lower a little lower voltage and when they "notice" it <S> they lower their input resistances to get the required current. <S> This effect is highly unstable. <S> Eventually only one charger (unfortunately the one that requires the least power) will get all the input voltage and (most probably) be destroyed. <S> To verify this logic assume that one charger finished charging and does not need any current at all. <S> The same thing happens. <S> The only way out is to dissipate excess (unneeded at the moment) power but <S> that is probably not something you would like to do. <S> :) <S> High input voltage SMPS <S> A single SMPS supply working from 1000V input is also quite difficult to do since most MOSFETs and IGBTs are only rated up to 1200V. A 600V supply should be doable. <S> Output current regulation in multiple-output SMPS <S> You could do a multi-output transformer based switched supply and regulate the output voltages and currents with a magnetic amplifier (this Ferroxcube flyer <S> is a proof that this is used in real circuits) <A> Edit: I completely misread your question. <S> I think this is still applicable, though: <S> So, it sounds like you want to "stack" the inputs/outputs of these converter IC's, which I would describe as more of a parallel combination rather than a series combination, although I do not think that either term really describes this arrangement perfectly. <S> Maybe someone else can provide a concise answer. <S> Edit: I am not actually sure all of the following would apply to the proposed arrangement, since your switches may share the combined 1kV and therefore no one switch would have the full magnitude of the input voltage across it. <S> A final note: 1kV is really high and generally super dangerous. <S> In order to deal with this kind of voltage, all your switches (probably most of your other components as well) would need to be rated to this voltage, as would your PCB, etc. <S> I'm guessing you were exaggerating on that number in order to get your point across. <A> Tamura and CLC are companies making DC/DC converters with rated input ranges of 200-1500Vdc. <S> Both are certified to UL1741 and/or IEC62109-1, which is required per article 690 of the electric code. <S> Both have outputs that are rated as SELV (safe to touch), and are rated for overload and over-temperatures. <S> If you are not an experienced power supply designer and attempt this and stuff goes bad, one or more of several things will happen: <S> your house and/or neighbor's house will burn down member of your family or a first responder incurs shock or burn injury <S> your insurer legally abandons you the local AHJ cites you for code violation, so resultant damage becomes civil and/or criminal issue.
|
From what I know about the various popular DC-DC converter topologies, my intuition tells me that this would fail, although to fully explain why would take a bit of digging and explanation and circuit theory (probably).
|
What is the purpose of solder mask expansion? I see recommendations of 2-4 mils for solder mask expansion. But why is it necessary? <Q> If the solder mask expansion were 0, in theory -- assuming everything aligned perfectly -- the board would work fine. <S> In practice, things never align perfectly. <S> The actual hole punched in the solder mask may be slightly less than what you specified ("shrinkage"), and that hole is always be placed in a slightly different location than what you specified <S> ("movement").If <S> your solder mask expansion is too small, then these misalignments cause the solder mask to partially or completely overlap SMT pads and through-hole pads. <S> If the solder mask completely covers most or all of the pad, the SMT part will be completely disconnected from that pad. <S> Then the board will immediately fail the end-of-line go-nogo test. <S> Many people specifically design the pads of a footprint to comply with IPC's fillet recommendations. <S> If the solder mask even partially covers some of that pad,then the fillet of solder will be smaller than a person looking only at the copper might expect. <S> If the fillet of solder is too small, then the (SMT or through-hole) part will not be mechanically attached as well. <S> Then your customer will notice the problem.(This is much worse than a board failing the end-of-line go-nogo test). <S> Daniel Grillo gives an excellent explanation of what happens if the solder mask is too big. <A> Perhaps this scan of a (rather poorly) commercially made board will illustrate the tolerance issues involved: <A> Solder resist masks are plotted oversize to allow for mask shrinkage, movement and inaccuracies, according to the manual for the Pulsonix PCB software I use. <S> The default value is 5 mil. <A> There are various sources of errors in the plotting and placement of different layers (and holes!) <S> that go into the board. <S> The board house knows their capabilities/yields, and will tell you the allowance that you have to make for that. <S> That is why they have things like minimum annulus (when the drilled holes end up so off center that it might end up not plating to the pads). <S> And the solder mask clearance ensures that the worst-case mask placement offset does not end up covering the pad. <S> For simple prototypes, it's usually not a big deal, and you can just use "default numbers" and then tell the board house to accept the clearance violations. <S> However, for very expensive boards, large runs, et cetra, you should get the exact numbers for all the various clearances to meet the board house's capabilities. <S> Note that board houses can also have different classes of ever increasing tolerances, and charge different amounts of money. <S> Unless your product really requires super tight registrations, use as "sloppy" a clearance number as you can afford. <A> With some designs, you want to make sure that the solder mask is never on top of SMD pads. <S> This is especially critical with QFN or LGA packages where the contacts of the parts don't stick out over the plastic molding or with parts that have a very fine pitch: Even small registration issues would cause the solderable areas of the tiny pads to become even more tiny.
|
After a few thousand cycles of vibration, the solder may eventually crack, and the part will be completely disconnected from that pad or hole.
|
Best pattern for WFI (wait-for-interrupt) on Cortex (ARM) microcontrolers I'm looking into developing battery-powered software using the EFM Gekko controllers (http://energymicro.com/) and would like the controller to be asleep whenever there's nothing useful for it to be doing. The WFI (Wait For Interrupt) instruction is used for this purpose; it will put the processor to sleep until an interrupt occurs. If sleep were engaged by storing something someplace, one could use load-exclusive/store-exclusive operations to do something like: // dont_sleep gets loaded with 2 any time something happens that // should force the main loop to cycle at least once. If an interrupt // occurs that causes it to be reset to 2 during the following statement, // behavior will be as though the interrupt happened after it. store_exclusive(load_exclusive(dont_sleep) >> 1); while(!dont_sleep) { // If interrupt occurs between next statement and store_exclusive, don't sleep load_exclusive(SLEEP_TRIGGER); if (!dont_sleep) store_exclusive(SLEEP_TRIGGER); } If an interrupt were to occur between the load_exclusive and store_exclusive operations,the effect would be to skip the store_exclusive, thus causing the system to run through the loop one more time (to see if the interrupt had set dont_sleep). Unfortunately, the Gekko uses a WFI instruction rather than a write address to trigger sleep mode; writing code like if (!dont_sleep) WFI(); would run the risk that an interrupt could occur between the 'if' and the 'wfi' and set dont_sleep, but the wfi would go ahead and execute anyway. What's the best pattern to prevent that? Set PRIMASK to 1 to prevent interrupts from interrupting the processor just before executing the WFI, and clear it immediately after? Or is there some better trick? EDIT I'm wondering about the Event bit. By the general description, it woulds like it's intended for multi-processor support, but was wondering whether something like the following might work: if (dont_sleep) SEV(); /* Will make following WFE clear event flag but not sleep */ WFE(); Every interrupt that sets don't_sleep should also execute an SEV instruction, so if the interrupt happens after the "if" test, the WFE would clear the event flag but not go to sleep. Does that sound like a good paradigm? <Q> Put it inside a critical section. <S> ISRs won't run, so you don't run the risk of dont_sleep changing before WFI, but they will still wake the processor and the ISRs will execute as soon as the critical section ends. <S> uint8 interruptStatus;interruptStatus = <S> EnterCriticalSection();if (! <S> dont_sleep) <S> WFI();ExitCriticalSection(interruptStatus); Your development environment probably has critical section functions, but it's roughly like this: <S> EnterCriticalSection is: MRS r0, PRIMASK / <S> * Save interrupt state. <S> */CPSID <S> i / <S> * Turn off interrupts. <S> */BX <S> lr / <S> * Return. <S> */ <S> ExitCriticalSection is: MSR PRIMASK, r0 / <S> * Restore interrupt states. <S> */BX <S> lr / <S> * Return. <S> */ <A> Your idea is fine, this is exactly what Linux implements. <S> See here . <S> Useful quote from the above-mentioned discussion thread to clarify why WFI works even with interrupts disabled: <S> If you're intending to idle until the next interrupt, you have to do some preparation. <S> During that preparation, an interrupt may become active. <S> Such an interrupt may be a wake up event that you're looking for. <S> No matter how good your code is, if you don't disable interrupts, you will always have a race between preparing to go to sleep and actually going to sleep, which results in lost wake up events. <S> This is why all ARM CPUs I'm aware of will wake up <S> even if they are masked at the core CPU (CPSR <S> I bit.) <S> Anything else and <S> you should forget using idle mode. <A> I did not fully understand the dont_sleep thing, but one thing you could try is do the "main work" in the PendSV handler, set to the lowest priority. <S> Then just schedule a PendSV from other handlers each time you need something done. <S> See here how to do it <S> (it's for M1 but M3 is not too different). <S> If you enable it, the processor will go to sleep after exiting the last ISR handler, without you having to call WFI. <S> See some examples here . <A> Assuming that: The main thread runs background tasks <S> The interrupts only run high priority tasks and no background tasks <S> The main thread can be interrupted any time (it does not normally masks interrupts) <S> Then the solution is to use PRIMASK to block interrupts between the flag validation and WFI: mask_interrupts();if (! <S> dont_sleep) <S> wfi();unmask_interrupts(); <A> What about Sleep on Exit mode? <S> This automatically goes to sleep any time an IRQ handler exits, so there is not really any "normal mode" running after that's configured. <S> An IRQ happens, it wakes up and runs the handler, and goes back to sleep. <S> No WFI needed.
|
Another thing you could use (maybe together with the previous approach) is the Sleep-on-exit feature.
|
How to switch many (15) low power (40W) lightbulbs from a microcontroller? The power rating on the bulb is the amount of power the PSU has to deliver in order for the bulb to light up the brightest right? Hence if it's a 40w bulb I need 40watts of power for it to light up properly right? If so then my dilema is I have 15 or so light bulbs which need to be switched on and off invididually. The only way I can think of is using a relay but that means 15 relays!! It's turn on and off by a microcontroller which switches the relay on and off. I need to use bulbs because the person am doing it for insists on bulbs <Q> You seem to be asking two different questions here... <S> Do I need 40w of power to light up a 40w bulb? <S> Answer: <S> yes you do, if it's less then the bulb will be dim or not light at all. <S> How do I switch 15 bulbs independently using a microcontroller? <S> Answer: yes, you can either use relays as a nice simple solution. <S> But if you need to switch them quickly, or many times over a long period of time then you should use solid state relays or triacs. <A> Are these large DC (eg automotive) <S> bulbs, or small mains bulbs? <S> If you're fuzzy on what a Watt means, you're not ready to be tackling high voltage circuits. <A> You can use mechanical or solid-state relays (SSR). <S> A mechanical relay offers the advantage that when off there's a complete separation between the bulb and the mains. <S> You'll need a transistor to drive the relay. <S> You can buy SSRs which you can drive directly from your microcontroller, but they're not cheap. <S> No problem, because you can create the same functionality cheap with just a few components. <S> You start with an opto-triac like the MOC3041 (that's for 230V applications, use MOC3031 for 115V). <S> It's like an ordinary opto-coupler but with a triac as the switch instead of a transistor (a triac can switch AC, which a transitor can't). <S> If you don't use this opto-triac your whole circuit will be connected to the mains! <S> The MOC3041 has the additional feature of switching on the zero-crossing of the mains' sine. <S> Incandescent bulbs tend to fail when being switched on, and when the mains voltage is near a maximum. <S> Zero-crossing switching prolongs the bulb's life. <S> The MOC3041 can't drive large currents directly, therefore we use it to drive a power triac. <A> There are no easier or cheaper ways around this. <A> What is the objection to relays? <S> If they are DC bulbs you can use 15 transistors. <S> Or you could build a 3x5 matrix and switch them that way with just 8 transistors and a lot more complexity...
|
You can use either relays or triacs (with optoisolation).
|
Why do real-time clock chips use BCD I have seen dozens of different real-time clock chips on the market, as well as a number of processors with a built-in separately-powered real-time clock module. Nearly all of them not only store time as year-month-day-hours-minutes-seconds, but even the individual fields are stored in BCD rather than binary format. Is there some underlying reason for this? Are there any microprocessor applications that do anything more sophisticated than simply display a clock where the BCD format is more useful than binary, or where year-month-day-hour-minutes-seconds format would be more useful than a straight 47-bit count of oscillator state changes? From what I can tell, it seems RTCC makers add a lot of extra circuitry to make their chips less useful; the only reason I can figure for RTCC modules in processors to behave that way is that the processor vendors use some pre-existing BCD implementation rather than producing their own. <Q> Do all RTCs use BCD encoding? <S> RTCs from Philips/NXP (both standalone and integrated into ARM7 or Cortex-M3 chips) do not use BCD encoding. <S> What's wrong with a BCD RTC? <S> When compared to flat counter the only operations which are more difficult with a split BCD clock are time difference calculations (adding seconds or calculating elapsed time). <S> Time comparisons like: "is current time greater than the alarm time set by the user" are just as easy. <S> What's nice about BCD (and generally split-field) RTCs? <S> Splitting the fields is really nice when you care for the calendar date. <S> Human calendars have funny things like months of different lengths and on top of that leap years. <S> Try to do that in a single counter (you can get a bonus point for using almost no power). <S> Oh and try supporting week days (quite useful in all kinds of devices meant for humans: from alarm clocks to heater controllers) with this. <S> The BCD approach has one additional feature: you get "every second" or "every ten seconds" interrupts for free, without having to do any calculations on times or dates. <S> For the record leap year calculation is a little off in the NXP RTCs since it only cares for the divisible by 4 rule and does not check the division by 100 and 400. <S> If it kept the year counter in BCD this would be trivial and most probably done right. <S> Summary <S> If you want a monotonic clock then use one. <S> You can buy a PIC or AVR with the "RTC counter" (which is just an asynchronous counter with an autonomous 32kHz oscillator). <S> Just keep in mind that simply displaying the date will be difficult. <S> :) <S> When you need to display the time and date and set alarms based on user input of times and dates then use an RTC. <S> And remember that when the user changes the current time and date your RTC based interrupts may be inaccurate. <A> When using clocks in the end you're more likely to be interested in minutes and tens of seconds(towards displaying them) than just the total of seconds, minutes and so on. <S> In case you're not interested in separate digits chances <S> are that you don't care about separate minutes or seconds values either, and that you might as well use a long binary counter like you suggested. <S> It's easier to convert from BCD to binary in software than the other way around. <S> And since BCD counters don't require that much extra real estate over binary counters it makes sense to choose for BCD. <A> I suspect several reasons: Historical - they've been doing it this way for some time now. <S> If you want your new part to replace some other part, then it has to more-or-less work the same. <S> So you keep with the BCD. <S> Application - if someone is using an RTC from a small micro (something in the 8 bit range, like a low-end PIC), then dealing with a large number (such as your 47 bit counter) is a big pain in the neck. <S> Not that hard - Doing the BCD counters isn't that hard, and in fact I think it isn't many more gates than doing them binary. <S> One can imagine a system where you get separate hour, minute, etc counters in binary instead of BCD (thus avoiding the 'breaking down the 47 bit number' issue), but it's not that much easier, and you're going to do some conversions when displaying the thing anyway. <A> I agree with Michael Kohne that there's a lot of historical momentum. <S> Early MCU's also had much less space for code and data (think 128 BYTES of RAM, for example). <S> Since time information is often used for human-interfacing purposes, it made more sense to keep the data closest to the format used to display to/input from humans. <S> Some newer MCU's with more code and data space sometimes implement hardware real time counters -- these devices often keep binary counts of 32kHz ticks. <A> In case anyone is interested, I just looking at ST's 32F series and it seems that while the newer 32L series uses a BCD RTC, the 32F uses a straight 32-bit counter with configurable prescalar and provides a separate battery input for it (hooray!). <S> I would have rather had a longer straight counter without a configurable prescalar <S> (so I could get 1/256sec accuracy but keep time for years without having to worry about wrapping) <S> but if I were to set the prescale for 1/64sec the timer could run two years without overflowing. <S> Not ideal, but not too bad. <S> A little unaesthetic that if someone powers the machine on after it's been off for too long (2.1+ years), the time/date would undetectably slip back by 2.1 years, but hardly a major problem (the counter has an overflow flag, but in many cases that wouldn't be terribly helpful. <S> If the machine was on for two years prior to being powered off, and was powered on three months later, the timer would be expected to overflow; the question would be whether it had overflowed twice, and I don't know of any flag for that. <A> Maxim seems to be doing just what you want with DS1372U . <S> It needs less than 1μA, costs 1.7 USD and is available(!) on DigiKey and Mouser. <S> The only problem is that it does not seem to offer alarms with more than 1 second precision and the lowest output clock rate is $\approx$4kHz.
|
It's MUCH easier to deal with the BCD digits, as you don't have to work at breaking things up.
|
How do I find the maximum "discharge C rating" of a battery? I've been experimenting running my 3A LED strips off battery power and the results are disappointing. I'm only getting a few minutes (< 10) out of a YSD-168 1800mAh lithium ion . How can I find the "C rating" of this battery? Is it something mainly dependent on battery chemistry (as this Dan's Data article I just remembered implies , search for "[space]C[space]" on the page) or something mainly dependent on manufacturing quality/technique? Macho RC battery packs always quote C-ratings, other battery types not so much. And it's difficult to google. Also, what are typical maximum "discharge C ratings" of AA and AAA NiMH batteries? <Q> You're not really looking for the C rating (maximum discharge current in multiples of nominal capacity), you're looking for the adjusted capacity at your nominated discharge current. <S> A 1.8Ah lithium battery can theoretically give 1.8A for 1hour, or 3A for 1.8/3h = 36 minutes. <S> HOWEVER <S> the capacity for a battery is traditionally quoted for a 20 hour discharge. <S> That is, a capacity rating of 1.8Ah means the battery delivered 90mA for 20 hours in testing. <S> The relationship between continuous current and time-to-full-discharge is NOT linear. <S> You have discovered that when discharged at 3A, your 1.8Ah battery is delivering much less capacity (only about a quarter!) <S> than a linear interpolation of the amp-hour rating would suggest. <S> This is not unusual. <S> A battery intended for remote control uses will probably give better performance, as quick discharge is the intended application of these batteries. <S> (I use a 1500mAh 3-cell (11.1v) <S> 25C lithium pack to power a 3A LED bike light, and I get around an hour, which is close to rated capacity given a reasonably efficient buck regulator). <A> This battery pack contains a protection device to protect the battery from overcharging, over-discharging and over-current. <S> You're load seems to trip the over-current protection build into the battery pack. <S> I suggest you do not try to do something about it as it might result in big explosion. <S> NEVER use this battery without it's protection circuit. <S> (the circuit contains some MOSFETS and things like that <S> and they might not be rated for your demand.) <S> Finding the capacity of batteries at different discharge rates is best done with Peukert's law that you can find here: <S> http://en.wikipedia.org/wiki/Peukert%27s_law <S> The wiki page mentions lead acid batteries. <S> But it can also be used for Li-Ion batteries. <S> However, getting the right numbers from a battery manufacturer can be a problem. <S> A more practical explanation can be found here: http://www.bdbatteries.com/peukert.php Regards, Hendrik <A> The best bet is to check datasheet and/or run an experiment :-)And yes, it depends on chimestry AND quality. <S> But for Lithium & NiMh batteries you usually can get more than 1C reliably. <S> The more current you get - the more looses & heat generated. <S> I wouldn't got over 3C in any case. <A> Acoording to Energizer for NiMH batteries, the normal discharge rate is 0.2C. <S> The battery nominal capacity is measured at this rate. <S> At higher discharge rate the actual capacity of the battery will be much less than the nominal. <A> Many RC aircraft batteries can be run at 20 or 30C continuously, but those are specifically designed to do so. <S> I wouldn't imagine a C rating exceeding 1C for a consumer battery, like the CCTV camera. <A> Poorly designed or cheap batteries may allow you to exceed 0.5C, but the battery may become dangerously hot and (both temporarily and permanently in the case of NiMH) <S> lose capacity (MaH). <S> I hope this helps. <A> In the past, most "power" batteries (UPS Systems) used a nominal 8-hour discharge rate for battery amp hour capacity. <S> If one had a 160 AH battery, then only at the 8 hour discharge rate would the exact 160 AH be realized, or 20 amps. <S> if 10 amps was the load, then more than 160 AHs would be available, and if 40 amps was the load, then less than 160 AH would be available. <S> When I was designing my home battery back-up system, I could find nothing about a nominal 8-hour discharge rate, as used in the utility industry, but instead I found a "C" rating. <S> The "C" was similar to my 8-hour rate, but it seemed to change from battery to battery and from manufacturer to manufacturer. <S> It appears that NiMH battery manufacturers have consistently used a "C" factor of 1C for full capacity, but the actual discharge time will be a fraction or multiple of 1C. e.g take the energizer datasheet: http://data.energizer.com/pdfs/nickelmetalhydride_appman.pdf <S> All the charts refer to a typical C/3 or C/5, indicating that 1C is AH capacity but the AH rating is still a factor of time. <S> Energizer uses a C/5 for its full capacity rating, or 1/5 the mAh rating for 5 hours. <S> Any load above the 1/5 mAh rating will reduce capacity and below the 1/5 mAh rating will increase capacity.
|
The better batteries will give rated capacity at several discharge time samples, or even a graph of current vs capacity. You'll find the "C" rating of most standard alkaline and NiMH AA batteries falls between 0.25C and 0.5C.
|
How to discharge smoothing capacitors? I have a simple 12V 10 A power supply with just a transformer and a rectifier. After doing some research and simulations, I've added 3 10 mF capacitors in parallel to smooth out the output. My problem is that after turning the supply off, capacitors remain charged for quite some time. I can get small sparks after shorting the output even 5 minutes after turning the supply off. Right now I only have a single LED connected to the capacitors and it takes more than 10 minutes for it to turn off completely after powering the supply down and the capacitors still aren't fully discharged when it turns off. The most obvious way to solve the problem would be to put a resistor and a switch on the output and connect the resistor to the capacitors after turning off the supply by hand, but I'm hoping to get something a bit smarter and a bit safer. Another point is that I want to use the supply's original case which has very little free volume now that I've added the capacitors, so just putting a ceramic 11 W resistor could be a problem because there would be very little free space around it for safe cooling. <Q> Appropriate bleeder resistors are the usual solution. <S> They aren't usually switched, although they can be. <S> The value depends on the time you require to discharge the capacitors. <S> The formula is $$ V_{t} = <S> V_{0} \, e^{ -t / RC } <S> $$ where \$V_{t}\$ is the voltage at time t and <S> \$V_{0}\$ is the initial voltage at time 0. <S> It's an exponential function, so I'd just assume 1/10 of the initial voltage. <S> It isn't a power function, as someone edited it! <A> What you want is a switch which is open when the circuit is powered, and closed when it is switched off. <S> When closed it should discharge the capacitor over a resistor. <S> You don't want to short the capacitor; they don't like that. <S> Two approaches I can think of <S> (from the top of my head): Use a depletion MOSFET as the switch. <S> Depletion MOSFETs conduct when there's no voltage applied to the gate. <S> Apply a voltage to switch it off. <S> This voltage can not be derived from the capacitor you want to discharge! <S> Otherwise the MOSFET would never be switched off. <S> (You think about this, if you don't get it tell me, and I'll try to explain.) <S> Use an ordinary NPN transitor which you drive from the capacitor's voltage. <S> Pull the transistor's base to ground if the circuit is switched on. <S> Again, the voltage to do this is from a separate power supply. <A> If it's regulated (linear/pulse) you would need to tune it till ripple would be acceptable with much less output capacitor. <S> If you have alot of high-freq noise - you would need to add several ceramic caps. <S> Also, make sure that your inductor at the output is calculated correctly.
|
As long as there's a voltage present, it will discharge. You should find that the power taken by the bleeder resistors is negligible compared to the 120W capability of the supply. Such huge caps seems to be an overkill...
|
Acoustic Measurement of water level Is it possible to measure the amount of water in a rubber water tank by using sound waves? I want to make a non intrusive measuring device that attaches to the outside of a tank, and is able to approximate the amount of water in the tank however I don't know how to calculate what sound of a specific frequency will "sound" like given different mediums (e.g. sound travels faster in water than air). <Q> You could place an ultrasonic transducer (transmitter-receiver pair) in the top of the tank, sending its signal down to the water, and measure the time between sending a pulse and receiving its reflection. <A> Also consider using a capacitive sensor. <S> If you glue two plates side-by-side on the outside of the tank, they will act as a capacitor with the water (or air) in the tank as the dielectric. <S> Because water has a dielectric constant that is 100x that of air, you can easily and accurately relate the capacity of the plates to the water level in the tank. <S> See Measuring Water Level Without Getting Wet for more information. <A> This is not a method I've tested myself, I am just offering a potential solution. <S> If there is more water in the tank, the mic will pick up less volume from the speaker. <S> To determine the amount of water, you just need to calibrate the device by recording the mic level at each water level.
|
You can place a speaker on top of the tank that outputs at a tone at a certain frequency, and place a waterproof mic on the bottom of the tank.
|
Graphical equivalent to HD44780 LCD controller The Hitachi HD44780 LCD controller is an extremely common character-mode LCD controller. Though I believe it is now discontinued, compatible controllers are used in many hobbyist and commercial grade LCD displays that are very widely and cheaply available. Interfacing instructions are also widely available. I would like to know what an equivalent controller might be for graphical LCD displays, such as a 128x64 pixel display. Something widely and cheaply available to the hobbyist, that is reasonable to interface with from a MCU, either parallel or serial. Possibilities include the KS0108 (parallel) and ST7565 (serial), but neither seem to be as popular as the HD44780. <Q> The wonderful thing about standards is that there are so many of them. <S> The closest thing to a standard is, unfortunately, for LCD panels that have controllers but no drivers. <S> IIRC, a typical interface will have signals for phase polarity, frame clock, line clock, data clock and 4 data bits. <S> Every line of pixels one should clock in enough groups of four pixels to fill the width of the display (extra bits will be ignored), driving the data clock high and low for each group. <S> The drive the line clock high and low to strobe the line. <S> The first line of each frame should have the frame clock high, and the phase polarity signal should toggle every frame. <S> The line clock signals, and those derived from them, must be sent at a uniform rate. <S> The precise timing of the data clock signals, however, doesn't matter provided that all the clocks happen for a line happen within the proper window. <S> If you don't have DMA, it may be possible to keep a small display happy and still have time to do something else, but refreshing the display will be a pain. <S> If you do have DMA, however, and can manage a small CPLD to handle a few aspects of the timing, implementing the display that way may be very rewarding. <S> I've done a display panel like that and achieved display-update performance superior to anything I could have done with a conventional display controller. <S> I even achieved 4-level gray-scale by running the display at 100 frames/second <S> and, every three frames, driving the display twice using one buffer and once with another. <A> For graphic LCDs there are lots of chips. <S> For small screens (older cell phones) there are a lot of Philips/NXP chips with I2C interface. <S> They are often used as SOG (silicon on glass) integrated with the LCD. <S> The interface is often a flex cable or a zebra strip. <S> Not very hobbyist friendly, but very cheap in mass production. <S> For 128x64 displays the KS0108 is common. <S> Slighly larger displays can have for instance a SED or a T6963 controller. <S> There must be many others I don't know. <S> Recent microcontrollers can have an integrated GLCD controller, so they can use controller-less LCD. <S> Those are cheaper, and the screen updates can be faster. <S> ============================= <S> (added 2016-06-24) For a small and simple BW display the Nokia 5510 (or is it 5110?) <S> style 84 <S> *48 pixel displays are very cheap, but the quality is often bad, and there seem to be at least 2 versions of the controller. <A> For smaller LCD sizes they often include the drivers, memory, and controller, so a single chip can do everything. <S> They are popular enough that they are being cloned by other manufacturers, so the command sets for some chips very closely relate to one of these two. <A> A common upgrade to the HD44780 <S> I've seen is the ST7920 . <S> This display controller will drive a 128x64 monochrome LCD. <S> (It's frequently sold as a "12864 LCD" module, with a pinout compatible with typical HD44780 displays.) <S> By default, it uses a built-in 16x8 ASCII font -- yielding a 16x4 text display with extremely large, clear letters -- and can be controlled using a HD44780-compatible command set. <S> In this mode, it is largely a drop-in replacement for a HD44780 display. <S> However, it adds a number of new features: <S> It can be driven using a SPI-compatible mode, as well as the standard 8-bit and 4-bit parallel modes. <S> The command set is identical in all modes. <S> Along with the aforementioned 16x8 ASCII font, it also contains a sizable 16x16 Chinese font, which can be accessed by writing high-ASCII pairs to the display (compatible with either GB or Big5 encoding, depending on the model). <S> While most of these characters are only useful to Chinese users, there are a number of general-purpose characters and symbols available. <S> It contains a number of "extended" commands which can be accessed using encodings that are unused by the HD44780. <S> Among these is a command to place the display into a graphical mode, at which point you can write a bitmap directly to the display. <A> I found this BV4512 controller via Google. <S> It looks to be reasonably easy to control via I2C. <A> If you're interested in using an HD44780 display, I have a short PIC18 hd44780 tutorial here . <S> As for a graphic LCD, check out this Graphic 128x64 LCD at SparkFun: http://www.sparkfun.com/products/710 . <S> It seems to be pretty popular and uses the KS0108B parallel interface.
|
For character LCDs the HD44780 (or one of its equivalents) is everywhere. The SED controllers and the KS controllers are the most common graphical LCD controllers.
|
How to build an ultra-low power time counter? Inspired by this question I would like to know how low power you could go with a counter + 32 kHz oscillator (possibly made by yourself). I found a nice oscillator circuit on a BJT reportedly drawing less than 1.2 µA from 3V. Unfortunately the counter and/or prescaler parts are a little more tricky. I do not believe you can make low power flip-flops from discrete transistors but most normal logic ICs are not very efficient either (standard counter ICs all draw around 80 µA at room temperature). I do not want a counter that is integrated into a microcontroller (PIC or AVR or ARM). <Q> There are several issues here... <S> The oscillator circuit on a BJT doesn't quite spit out +3.3v logic levels. <S> Fortunately, you want to use a lower voltage to get lower power consumption. <S> +1.8v logic levels would be compatible with that <S> oscillator-- <S> but then you'd need +1.8 and +3.3v power rails (and probably loose all benefits in your voltage conversion inefficiency. <S> Dynamic power consumption (the power used when things switch) is mostly from charging and discharging the parasitic caps on the various signal lines. <S> The way to reduce that is to use shorter, thinner wires. <S> And by shorter & thinner <S> I mean don't use wires and instead use a chip. <S> Building this from a collection of chips and transistors instead of one chip that does everything will drive your power consumption up. <S> You said no microcontroller, but honestly that's the best way to do this. <S> TI has an ultra low power MSP430 that would run at 32.768 KHz at less than 1.5 uW. As you've already seen, this type of performance is really hard to beat. <S> After a microcontroller, my next choice would be a Xilinx Coolrunner-II CPLD-- <S> but I doubt that meets your requirements either. <S> To summarize, an MCU will give you the lowest total power consumption. <S> Otherwise your best choice is to use standard logic parts and suffer with something in the several hundred micro-watt range. <S> Making something out of discrete transistors isn't going to be better. <A> Something better than HCMOS: <S> I'll presume you have a 32.768 kHz clock and want prescalers/counters to get 1 Hz from it. <S> The AUP series is not as extended as the HCMOS but <S> the 74AUP2G80 may be all we need. <S> It contains two D-flip-flops with inverted Q outputs, so that we can make a :2 divider. <S> There's no Q output but that doesn't matter. <S> To divide by 2\$^{15}\$ <S> we'll need 8 devices, that's a supply current of 4 µA <S> maximum (0.5 µA per device). <S> Dynamic power dissipation depends on the supply voltage, and since we want low power we choose the lowest possible: 0.8 V. Then dynamic power dissipation is given by <S> C\$_{PD}\$ is power dissipation capacitance and is 1.8 pF <S> at 0.8 V. <S> Then we have <S> \$ P_D <S> = 1.8 <S> \times <S> 0.8^ <S> 2 <S> \times <S> 0.032768 \times <S> 1 <S> + (2 \times 0.6) <S> \times 0.8^2 <S> \times <S> 0.016384 = 0.050 <S> \mu W \$ <S> The two times 0.6 pF is the load of the D input of the current FF plus that of the D input of the next stage. <S> The power of the next stage is half this one's, since only the frequencies are halved, the rest is the same. <S> So that's 0.025 µW for the 16.384:2 stage, and so on. <S> The sum for 15 stages is 0.101 <S> µF. <S> (Binary people don't need a calculator for this: N + N/2 + N/4 + N/8 <S> + ... = 2N) . <S> Add the static power of 4 µA \$\times\$ 0.8 V and we have a total of 3.3 µW , which is almost 2 orders of magnitude better than the 192 µA with the HCMOS devices (see my other answer). <A> For the oscillator I'd use an RTC like NXP PCF8563 , which provides a buffered 32kHz output and only consumes about 300nA @ 2V. <S> For the counter/divider I looked at HCMOS. <S> According to NXP's HCMOS family specification quiescent current for a flip-flop is 4uA maximum, but that's @ Vdd=6V, <S> so it should be lower at 3V.If <S> you use 74HC93 s (4 FF per device <S> , you'll need 4 of those to get 1Hz) <S> total quiescent current is 64uA maximum, which agrees with the 80uA you mentioned. <S> The good news is that at this low frequency dynamic power is far less than that. <S> Minimum supply voltage for HCMOS is 2V, so working at this voltage should also reduce the current. <S> The PCF8563 even operates at Vdd=1V (since you don't need the interface). <A> Many impressive solutions offered to date. <S> I could offer a 100 nano watt RTC solution but it does not meet the requirement to be CPU free. <S> Even the major watch makers such as Casio use CPU cores in their ASIC watch chips. <S> Once dynamic power is negligible with 32KHz, the drain is controlled essentially by transistor leakage and supply voltage, so the best solutions will be Vdd=1V or less. <S> Microchip: <S> nanoWatt XLP in all non-DSP PIC's 800 nA Real-Time Clock and Calendar <S> SiliconGate: <S> SGC22300 RTC - Real Time Clock Nano Power Series <S> I also have research articles with schematics of 0.3uW clocks only, but not commercial ICs. <S> If interested; I can supply more info; <S> If I may be so bold as to inquire; why stipulate restrictions on implementation if the requirement is for lowest power? <A> Technically, the method to make the most power efficient (and fastest) circuit (anything) would be a Full Custom IC using the tiniest technology you can find. <S> You'd have to design your own IC layout and have a foundry build a wafer for you. <S> There are universities that have agreements with some foundries, maybe you could research your local schools for information on that. <S> If they are willing to help you then it would be free. <A> If you think about simple ICs, you can get decent power consumption on CMOS chips. <S> You may take single inverter in a chip, and a counter. <S> Run it at some 1.8V, and hopefully it will eat very little :-)
|
NXP's Advanced Ultra-low Power (AUP) series. Static Free's Electric is a great free software to start building your own custom ICs.
|
Creating a 12V @ 3.5A out of 15 volt 1A I have a huge led matrix which will draw 3.1Amps of current. However I only have a 15v 1A psu> Would it be possible to use a transistor to amplify the current after I've lowered the voltage? Could you recommend such a transistor. <Q> 15V <S> * 1A is 15 watts. <S> 12V <S> * 3.5A is 42 watts. <S> You cannot create power out of nothing. <A> A transistor does not create current (and power) out of thin air. <S> Transistors can amplify voltage or current but only if you provide them with an adequate supply. <S> A transistor is not magic at all. <S> It works like a valve: it can pass more or less current (water) <S> but if there is none available in the hose (supply) it will not be able to "create" it. <A> If you only have 15W available, then the best switched mode power supply will get you between 1.1 and 1.25 amps at 12V, because switched mode power supplies can be 90% or more efficient. <S> But not over 100% efficient, so for 12V at 3.5A you would need an initial supply of 15V at 2.8A at least. <S> Alternatively, your 15W supply could be converted to a 3.5A at just less than 4V supply. <S> A transistor on its own will not perform this conversion but will be an important part of a switched mode power supply .
|
The amplification part comes from the ability to take small current or voltage and use it to control the transistor (valve) which can pass or block much higher currents and voltages.
|
How to implement SRAM sense amplifier? I am designing a simple programmable LED screen system as an exercise, and I need a non-standard type of SRAM (16x5) so I am designing the memory circuits. I have looked everywhere (Digikey, TI, National Semiconductor, etc.) for a chip that is just a sense amplifier for SRAM and I haven't found anything. I called TI and they said they didn't make anything like that, but I may be able to use a comparator for that. My question is, is it possible to wire up, say, a LM339 as a sense amplifer? And if so, how? Thanks <Q> To answer your question, yes you can use a LM339 as a sense amp, and all you need to do is wire <S> the + input to the non-inverting bit-line, and the - input to the inverting bit-line. <S> BUT... Since we're talking about reading a SRAM, the basic sequence of events during a READ is: Pre-charge both bit-lines to VDD. <S> Assert the desired word-line <S> The selected bits will now pull one of each complimentary bit-line pair towards VSS <S> The sense amplifier will sense a difference between the bit-line voltages (one will stay at VDD, the other will be falling towards VSS) <S> If you look at the schematic for a 4T or 6T bit-cell, then you can see that the falling bit-line will eventually reach VSS. <S> In typical IC designs, where density is very important, the bit-cells have very small transistors. <S> Additionally, they are typically built into large arrays, which have very large bit-line capacitances. <S> This results in a very slow (relatively) discharge of the bit-line by the bit-cell. <S> This is the motivation for the speed aspect of a sense amplifier. <S> It probably doesn't apply to you because you can use large transistors (since they're discrete, or in an array), and your array is small. <S> The isolation aspect of the sense amp is also very important. <S> If you don't use a sense amp, your bit-line will have an unknown load based on whatever you hook up. <S> Also, any noise on the signal will be broadcast to all your bit-cells. <S> Some isolation is always a good idea here. <S> I don't think you're going to see a major benefit from using a comparator to sense the bit-lines. <S> Instead, I would recommend that you use a CMOS inverter or buffer. <S> It will be simpler, just as fast, and denser. <A> Unless you're trying to use the smallest transistors you can get away with to hold data, I'm not clear why you would need sense amplifiers. <S> Wouldn't the data latches and passgates themselves provide enough strength to drive a normal logic input? <S> Actually, I would expect that the simplest way to get a 16x5 RAM would be to simply use a 32Kx8 SRAM and ignore the upper part. <S> Otherwise depending upon your you might be able to use something like a dual 64-bit shift register chip which has taps every 16 bits. <S> Combining that with a 74LS374 or equivalent along with suitable clocking logic you should be able to store 80 bits fairly nicely and access them five at a time. <A> If you are not shooting for top speed(>1Mhz), you don't need any sense amplifiers. <S> Standard 2T trigger (plus some Ts for accessing the data) would give you close to 0 and VCC voltages. <A> (a) Are you using an off-the-shelf SRAM IC chip? <S> If so, it already includes the sense amplifier -- the bitlines cannot be accessed from the outside, even if you wanted to. <S> (b) Are you not using an SRAM IC chip? <S> I.e., are you storing a few bits in a D latch or a shift register IC chip or are you building something out of 3 000 discrete transistors ? <S> If so, then you don't need a sense amplifier. <S> As supercat mentioned, the output transistors of those devices is much larger than the tiny transistors inside a chip, and so that output can directly feed a digital input -- at worst you might need some digital inverter/tristate/line-driver/buffer to give you better fanout. <S> Do you actually have something -- such as, say, magnetic core memory -- that needs a sense amplifier? <S> While I doubt that is actually the case, if so, the LM339 may work fine for you -- please tell use what you are actually building.
|
Something as simple as a CMOS buffer/inverter will work just fine.
|
Disable data leads of USB cable I've got a Sansa Clip that recharges via USB. Problem is that it "connects" to the PC whenever I plug it in. I learned a trick where you pull the cable out of the PC partially such that the data leads are no longer connected but the power leads are. However I am looking for a more permanent solution. How can I modify the cable itself such that only the power leads remain and it never transfers data? (I have an extra cable, so it's OK.) <Q> <A> my first attempt at solving your problem would be to get some kapton (or other non-conductive) tape and just stick it on top of the traces in your USB plug. <A> how about covering the two middle pins on the type A side of the cable with electrical tape? <S> simple, reversible, and if you get it wrong, its a matter of removing the tape ;p <A> The pinout of USB is the following: <S> pin1 (usually red) is Vcc pin2 (usually white) <S> is D- pin3 (usually green) <S> is D+ pin4 (usually black) <S> is Gnd <S> What you want is an cable where pin2 and pin3 are disconnected. <S> Just cut the cable and only reconnect pin1 (red) and pin4 (black). <S> Edit: the following page says: simple USB charger should short the 2 data lines together. <S> The device will then not attempt to transmit or receive data, but can draw up to 1.8A, if the supply can provide it. <S> So if simply disconnecting pin2 and pin3 does not work you might try to short them. <A> I just cut the green wire. <S> No problems. <S> I'd like to get a click switch to turn it on and off.
|
Much as I like hacking things, I think the simple answer here is a dedicated charger with a USB port.
|
Why is three-phase offset by 120 degrees? For three phase electricity the wave is offset by 120 degrees(2\$\pi/3\$ Rad). Why aren't the phases closer together? Is it because it will affect the frequency of the phases? How was this 120 degrees chosen? <Q> When there's 120° between phases the sum of the voltages at any time will be zero. <S> This means that with a balanced load no current flows in the return line (neutral). <S> Also, if each phase is 230V with respect to the neutral (star operation), then there will be 230V \$\times\$ \$\sqrt{3}\$ = <S> 400V between any two phases (triangle or delta operation), <S> and they're also equally spaced, i.e. at 120° angles. <S> (images from http://www.electrician2.com/electa1/electa3htm.htm ) <A> If you had phases 'closer together' as you suggest, there wouldn't be any real advantage over single phase power. <A> In principle, any power generator has a rotor with magents and coil on the periphery, one rotation of rotor is one cycle of 360 degrees. <S> Suppose the generator has one magnet and one coil,then as the magnet/rotor rotates one turn,the voltage generated in the coil gradually rises and reaches peak(max) <S> when the coil comes closest to the magnet and reduces gradually as the magnet moves away. <S> Suppose we connect the bulb then the flicker rate is clearly visble. <S> This is called 360 deg, single phase AC. <S> Now, suppose the generator has two magnets and two coils placed equidistantly, then the flicker rate is increased, it is 2-phase , <S> 360/2=180 degrees AC. <S> Say generator has 3 magnets and 3 coils placed equidistantly, then the flicker rate is much increased; it is 3 phase with 360/3=120 degrees AC. <S> if we have 4 magnets and 4 coils placed equidistantly then the flicker rate is much more increased (not visible), then it is 4-phase with 360/4=90 degrees, 4-phase AC. <S> In practice, 3-phase is much more suitable for design. <A> By separating the phases by 120° one keeps the voltage peaks (for instance) evenly spaced. <S> For example, 60 Hz has peaks every 16.66 msec, so phase A, B, <S> C peaks would come one third of that time apart, in this pattern: A-5.55ms-B-5.55ms-C-5.55ms-A. <S> If one separated phases A & C from B by, say 100° then phases C and A would be separated by 160°, and the pattern of peaks would be A-4.63ms-B-4.63ms-C-7.40ms-A. <S> Such a stuttering set of phases (with, say, 100°, 100°, <S> 160° separation) would entail many inefficient, unnecessary consequences, not least of which would be designing an AC motor which could effectively use the staggering impulses of such syncopated voltage peaks. <A> Most of the electric energy is made by AC generators. <S> 2/3 of electric energy is used by AC electric motors (electric energy in - mechanical energy out), they are built very similar to the electric generators (mechanical energy in - electric energy out). <S> In order to create a rotation in AC electric motors you need to have equally spaced windings in the stator fed by equally spaced magnetic fields; equally spaced magnetic fields are created by equally spaced currents (this answers your question of the 120 degrees for the 3 phase system). <S> The reason of using 3 phases instead of 2, 6 or 12 <S> it's because it's the most efficient system (having 2 would mean more power losses during transmission, having 6 phases would mean to transport the energy with 6 wires instead of 3). <A> Also keep in mind that the phase to phase voltage would drop tremendously with more phases. <S> You would only be able to use it phase to ground if you added more phases. <S> With a regular wye transformer, we can still have equipment be on 208 volts and 240 single phase. <S> Add more phases it would be a lot more harder to add 3 phase equipment or more.
|
Being 120 degrees apart makes the phases balanced such that power transfer at any instant is a constant.
|
Are there any generic "how to use this part" guides for basic discretes? Coming from a programming background, I'm used to user guides telling me how to use something, rather than how it's implemented. In the world of electronics, when it comes to discretes like, say, a MOSFET, all I can usually find is inventor, history, an overview of the theory and some implementation details. It's quite hard to find answers to basic questions, like "does a fully turned-on MOSFET conduct more current from source to drain or from drain to source?" Am I just looking in the wrong places? Or is the best thing to do actually understand the underlying implementation details so that I can deduce the answer from that? It always feels like a bit of an overkill trying to learn how to make MOSFETs when all I want is find out how to use them in a most conventional way. <Q> <A> I don't think there is any such guide. <S> At least if there is one, it will be limited to a few basic uses of the device. <S> The difference is that if you are using, e.g., a programming library, all you really need to understand is the API. <S> The library is designed to do a few well-constrained things. <S> Now, a component like a transistor is also designed to do a single thing: a bipolar transistor controls the output (Collector) current by varying the base current. <S> However, that single "feature" has huge amounts of leverage and hides a lot of complexity. <S> The same transistor can act like a switch: give it plenty of base current and it will saturate and conduct a lot of current. <S> It's a proportional valve: give it varying, small amounts of base current, and you will get a varying, larger amount of current conducted at collector-emitter. <S> It's a temperature sensor: give it a small, fixed amount of base current and the voltage from base-emitter will vary depending on the temperature. <S> As a result, the current at the collector-emitter junction will also vary according to temperature. <S> My point is that the only useful "user guide" to a transistor requires knowing its fundamental properties. <S> You may not need to remember the complete Ebers-Moll model, but you do need to understand the fundamental properties if you're going to use the device to its fullest. <S> Failing that, if you have a specific question, just ask. <S> Lots of smart people here: someone will be able to help you. <A> I never liked the Art of Electronics myself; it's encyclopaedic for sure, but getting knowledge out of it has always been rough for me. <S> When I was learning this stuff, Forrest M. Mims III and Robert Grossblatt were two authors whose works I devoured. <S> Mims' Getting Started In Electronics is geared toward younger people, but it's very approachable and easy to understand.
|
Get a good book, like The Art of Electronics, if you want things explained properly.
|
Can a resistor on its own do anything This is a newbie question. Can someone confirm: a resistor on its own is useless. True or false? I mean, I can't do anything with a resistor alone, or is there something that can be done with it? Sorry, I know the question is too newbie. <Q> Almost everything we have in electronics is useless on its own. <S> What good is a light bulb by itself? <S> With no power source you can't light it. <S> Even with a power source, it is still not all that helpful with out a switch. <S> Even the more complex things, like a microcontroller, are pretty useless on there own. <S> Even if you have a power source there isn't much purpose of them unless they are connected to other things. <S> As for focusing on the resistor itself, there are things that it can be useful for when combined with a power source. <S> For example, creating heat. <S> If your goal is to create heat with a very simple component, using a resistor with a power source is a very simple way of doing it. <A> A resistor on its own may be useless, but that doesn't mean that it doesn't do <S> anything!Resistors produce thermal noise . <S> This noise occurs because of the random movement of charge carriers (mainly electrons). <S> It depends on temperature (the higher the temperature, the higher the noise) and on the resistance value (again, the higher the resistance, the higher the noise). <S> This noise, however, is too small to be useful; you can't power a circuit from a resistor, and the variable nature of the voltage doesn't help either. <S> While not useful as a power source, this thermal noise often plays a role in circuits. <S> It may be amplified with the signal and deteriorate it. <S> But even in a circuit the resistor generates this noise voltage all by itself. <A> Taking the question literally (as a lot of the replies seem to), "Can a resistor on its own do anything". <S> Yes, it can be confused with an inductor (and vice versa), but only if you don't examine it closely : <S> ~) - No power source required. <A> Yes, a resistor by itself is useless. <S> It plays an important part in circuits though. <S> Resistors limit current in your circuit. <S> For example, if you want to light up an LED, you typically can't just connect an LED to a battery because the current is not limited. <S> With a resistor in series with the LED, you can control the current going through the LED. <S> Also, different values of resistors will give you different currents, which also gives you different LED brightnesses. <A> Resistors by themselves are perfectly useful in debugging situations. <S> Ever not have a jumper wire handy <S> so you can beep out a serial cable, or use it as a test lead? <S> I plug a resistor into the cable or breadboard and that becomes the means for connecting my multimeter. <S> Yes, I read on the same end of the wire, not from the other lead. <S> :) <S> I know this was a bit of a silly answer <S> and it doesn't answer the OP's question in an EE/circuit analysis sense, but in all practicality, I have used resistors like this very often. <A> A resistor, on its own, can (perhaps) exist. <S> "Is it useless?" <S> (true or false) is a different question. <S> I find it pleasing ( <S> which is useful to me) to find them strewn about, but that requires me to exist as well. <S> I suppose my answer, in that case, would have to be false (because if I am, the resistor isn't alone). <S> Likewise, the further question (can something be done with it?) <S> would probably be no if there was no one to do that something. <S> If you were looking for more literal non-circuit applications, I've occasionally used them to tie things together (as staples), leave a trace (I wuz here), fill boxes (the one marked "resistors"), and engage in philosophic discussions <S> (Is resistance really futile?). <S> If you want to talk resistor only circuits (excusing the supply, connecting wires, and meters), they can be quite useful for draining batteries, heating the room, testing voltages / currents under load, and even as decoration (for instance, Stackpole's RMCF2512JT33K0 - Don't be fooled by its digikey image, its quite large and has "333" prominently display on its face). <A> <A> You have to define what is usefull and what is useless first. <S> Resistor on it's own (assuming it is connected to a power supply) can only dissipate power and heat up. <S> If that is what you whant then it is usefull on its own. <A> Several resistors can add and average . <S> They can also divide . <S> In a sense, they also multiply, but are reduced in scale.
|
It doesn't do exactly nothing: it will generate heat, depending on the current flowing through it -- this is why resistors have a wattage rating.
|
Looking for a *fun* way to introduce basic electronics to my kids We recently came across this video: http://blog.makezine.com/archive/2011/04/the-awesome-button.html and both my six year old and 13 year old thought it was pretty cool. I am a computer programmer, but never did much with electronics, so I am looking for some suggestions on how to introduce electronics to myself and my kids in a fun lab/project oriented way. Its a little challenging b/c of the age gap between my kids, but I'm ok if the younger doesn't understand everything. As long as he can participate in the lab and be involved in what we are doing, he will pick things up along the way. Likewise, I wouldn't want something so simple that my older son is bored. A few preferences: I would prefer to stay away from robotics initially, maybe we can graduate to that, but I am looking for something not quite as involved. I would prefer that we can bring a programming component into the mix relatively soon, like in the example I linked to above, but that's not a requirement. That also is something we can graduate into. I would prefer not to shell out a whole lot of money initially. Maybe a couple hundred dollars at most, but hopefully less. I'm ok with learning stuff from a book or videos but my main goal is to give us projects to work on together. This is mostly about trying to find something to do with Dad time that is fun for us all, learning electronics is kind of a bonus. The labs or projects are my focus. So, can you suggest books, kits, websites, or other resources that would facilitate what I have outlined above? update @AndrejaKo: I know basically nothing about electronics. I understand that it involves things like resistors and logic gates, etc. but as to what those things are and how they are used, I know nothing. I have a multi-meter I bought at Lowes, but that is about it. If I have to spend the money to get some basic tools & components, thats ok. I just want to keep the cost down until I found out if this is really something they we are interested in doing together or if its going to die after a few projects. solution I began trying to put together an order for the electronic components for the All About Electronics experiments section as well as the first chapter of Make: Electronics . I had a couple hours into it, wasn't sure I was ordering the right components, and was somewhat discouraged when I happened across a blog that is walking through the Make: Electronics book that mentioned MakerShed's component packs for the Make: Electronics book. That sealed the deal for me! I went ahead and ordered the first component pack and the book (which was only $10 if you buy the component pack). We are going to start out with this book. If we make it past chapter two, I will go ahead and order the second component pack. The price has to beat RadioShack and even if its more expensive than ordering the components individually, I'm way sold on the time its saving me. <Q> I would stay away from anything so complicated that you only can use it as a black box, like in the video a USB interface. <S> One possible project is an electronic die. <S> You can do this with a few HCMOS ICs, like a 7-segment LED driver. <S> It has the advantage that you don't have to introduce transistors yet <S> (transistors are Really Complicated Devices!!) <S> and they see something happen. <S> edit <S> (re OP's comment) <S> HCMOS is a family of digital ICs, it comprises basic logical functions like gates (AND, OR, NOR, inverter,...), but also more functional blocks, like counters and the 7-segment driver I mentioned. <S> For the die I was thinking of an oscillator you can start-stop, which makes the counter loop from 1 to 6. <S> (Note to chief engineer: that's a presetable counter, a normal counter would start counting at 0.) <S> The 7-segment driver converts its binary input to a pattern for the 7-segments digit display. <S> Alternatively, you can create a die-like LED pattern and use logic ICs to determine which LEDs have to light up for which counter value. <S> (chief engineer: you can use a Johnson counter, like the 74HC4017. <S> Count from zero to 5.) <S> For instance, the center LED only lights up when you roll a 1, 3 or 5. <S> Then you OR outputs 0, 2, and 4 of the counter (remember, zero-based) <S> The advantage of working with logic ICs is that you don't need to explain about electrons right away, and that they can get acquainted with voltage levels first. <S> If you create the oscillator with a NAND gate like the 74HC132, you can explain its working with the water model : voltage = water level, capacitor = water tank, resistor = thin water tube. <S> For a concrete example see this googled document <A> I recommend the Snap Circuits sets for basic electronics. <A> You're probably looking for a book like these: <S> Make: <S> Electronics (Learning by Discovery) Practical Electronics for Inventors <S> Physical Computing: <S> Sensing and Controlling the Physical World with Computers <S> Making Things Talk: <S> Practical Methods for Connecting Physical Objects <S> Which are introductory, but practical, books on electronics. <A> At that sort of age you can definitely get them soldering and even designing circuits - there is a regular show that comes to the Edinburgh festival that my kids went to and got to do the full build (not design) of solar powered audio circuits and stylophone equivalents, including the soldering at age 4. <S> They can certainly cope mentally - it's just like building with lego or meccano - but be aware it will probably take one instance of dropping a soldering iron on a hand or leg for them to really believe you that it hurts a lot :-) <S> I would also recommend kits such as these ones from sciencekits.com as they do help you pick up the basics without really needing much in the way of tools. <A> What about dcaclab ? <S> students can learn online, and its fun too :)
|
Instead I would use very basic components, preferably things which generate a sound or display something, these will appeal to your youngest too.
|
Is it safe to assume that all pins in a standard VGA cable are wired? I am currently in the process of designing a custom arcade-stylejoystick console for use with emulators on my PC. My design separateseach joystick console from a single keyboard-encoder "hub" such that Imay add additional joysticks or upgrade the encoder circuit withoutmodifying existing joystick consoles. The connection to the hub may spanseveral feet, much like existing joysticks connect to their gamesystems. While designing, I became puzzled as to which cable and connectors to use forconnecting a console to the hub. I need at least 12 signal wires foreach console (joystick directions, buttons, and ground) and so Ithought of using a standard HD-15 VGA cable and terminals. I am nowdubious that this will work in general as I've read that some VGAcables don't wire all the pins. Is the use of HD-15 cables suitable for my design and usage? It isdefinitely attractive as it is inexpensive and generally available. What other cable and connector options are available to me andsuitable for this project? Jumper cables are appealing as theyconnect without soldering, but they don't appear to be available inlong lengths or as resistant to tangling as a VGA cable. <Q> Don't assume. <S> Most VGA cables use some sort of coax or individually shielded conductors for the Red, Green, Blue, and sync signals. <S> The shields for these signals, which have their own pins, are commonly tied together inside the cable. <A> As stated earlier, some cables may not connect everything. <S> Instead, try using a 25-pin parallel port cable instead, pins 1-17 are guaranteed to be connected. <S> I've opened up a few VGA splitter/switchers before. <S> I was surprised to see that those switchers only switch 4 wires. <S> I assume they were 3 color components and 1 ground. <A> You should not assume it, get out an ohm meter and check it out. <A> Not directly the answer you were looking for, but some consoles, like the SNES, use a simple shift register system to pass data with fewer connectors. <S> Something like this may be easier than a N-conductor cable to each controller. <S> Info on pinning, etc: <S> http://hackaday.com/2011/01/30/snes-arcade-controller/ <S> http://pinouts.ru/Game/snescontroller_pinout.shtml <A> The old standard PC joystick pinout seems suitable, and all-connected: http://pinouts.ru/Inputs/GameportPC_pinout.shtml <S> Presumably you can buy premade cables for this. <S> If you chose to match the PC pinout you'd also have the advantage of being able to test with pre-existing joysticks.
|
Additionally, some older VGA cables might not connect the "unused" pins-- which are used on more modern devices. From my experiences so far, all 25 wires seem to be connected individually.
|
Tips for multiple voltage regulators? I've got a project I want to work on, which involves an LCD, MAX232, and MSP430. The LCD requires 5V for logic and 3V for the backlight. The MAX232 will work from 3VDC to 5.5VDC. The MSP430G2231 will handle 1.8VDC to 3.6VDC. The main power source is going to be the 12V accessory power circuit in a car. It seems logical to regulate the ~12VDC to 5VDC, and then use the regulated 5V to create the 3.3V needed for the MSP430. I was thinking about trying 3.3V on the backlight -- the specs state 3V typical, but don't specify a min or max value. I plan to see if I can just measure Vf of the backlight and add a current-limiting resistor to the circuit so I can safely use the 3.3V. Anyhow, since I need two voltage regulators, does anyone have any recommendations for an ideal configuration? It looks like the LCD uses a max of 1.5mA @ 5V, and the backlight takes a max of 45mA. The MAX232 clone uses a max of 1mA @ 3.3V with no load (I have no idea what it would be under load), and the MSP430 looks like it uses about 4.2mA @ 3.3V and 16MHz. I think these are worst-case conditions. Would you use a switching regulator to go from 12V to 5V, and then a linear regulator to go from 5V to 3.3V? Or one of the really low-count options as in the 3V Tips 'n Tricks from Microchip like a single zener or multiple diodes? I picked up some MC34063s to play with, but am open to anything else with a lower / smaller part count. LCD specs I used: MSP430 specs: MAX232 specs: I would sure appreciate any comments, like recommended approaches, or if I have flaws in my interpretation of the specs. <Q> If power requirements are low (tens of millamps) <S> I tend to use linear regulators. <S> It doesn't matter whether you go from 12V to 3.3V immediately, or via 5V, the efficiency is the same (and rather bad). <S> If you want to use SMPSs you'll have to look at the efficiency. <S> This is in general lower for lower output voltages, and also if the Vin/Vout ratio gets higher. <S> Let's say you go from 12V to 5V at 90%, from 12V to 3.3V at 80% and from 5V to 3.3V at 95%. <S> So going from 12V to 3.3V in 1 step will be 80% efficient, if you do this in 2 steps (via 5V) your overall efficiency will be 90% * 95% = 85.5%. <S> You'll have to make the calculation for your specific regulator. <A> Use +5V for the backlight. <S> Select a resistor to drop 2V at 30 mA. (67 ohms). <S> The backlight will get a nicely regulated current. <S> For the 3.3V, use a three-pin LDO. <S> They are not expensive. <S> Be safe when connecting the LCD's output to the MSP430 input. <S> Use a resistor divider or such. <S> TI appnote . <A> One red LED will drop you from 5 to 3.3 for applications < 20mA. <S> Its super-cheap AND tells you if the power is on.
|
Make sure, however, if you use a regulator to go from 5V to 3.3V to use an LDO; the voltage difference is too small for standard regulators.
|
Stability of USB +5V output So i am designing a device that is supposed to connect to the PC via USB and draw power from the VBUS (+5v) of the USB. I am wondering if I need to put a voltage regulator in between the +5v from the USB and the microcontroller on my device. I've tested this device on my own laptop and it works fine. But I'm not sure if I can make any assumption regarding the stability of the VBUS on other computers. I looked at some Arduino designs and they don't seem to place any voltage regulator when it is powered through the USB. So is it safe to assume that the +5v coming from a PC USB connection is fairly stable? <Q> The USB specification states that the 5v is supposed to be +-5% under load, which translates to 4.75v to 5.25v. <S> A simple power connection will provide up to 100 ma. <S> (Up to 500 ma can be drawn, but you have to negotiate for it.). <S> Most IC's designed for 5v <S> should accommodate the +-5% variation also. <A> The 5 V VBUS is definitely not stable, and will usually vary with the computer's load. <S> In USB audio products, this creates all kinds of noise, presumably from audio frequency fluctuations caused by video cards, hard drives, etc. <S> With your micro, it probably doesn't matter, but if you're doing any kind of analog to digital conversion, I'd also recommend a regulator. <S> The USB spec lists a worst-case drop of 4.375 V at the device, after a bus-powered hub, with transient drops to 4.07 V. Keep in mind that not every USB host follows the USB spec, so it could be even worse. <S> Wikipedia summarizes : <S> It is specified that devices' configuration and low-power functions must operate down to 4.40 V at the hub port by USB 2.0 and that devices' configuration, low-power, and high-power functions must operate down to 4.00 V at the device port by USB 3.0. <A> Provided you are using a computer as a source, then yes you can assume that the 5V is stable. <S> You cannot assume this if you use a mains power supply adaptor with a USB output, like the one that comes with an Apple iPhone. <S> (I know that's a bad way of saying it <S> but it kinda makes it easier to understand!).
|
Also, to use a regulator effectively, you should supply the regulator with a slightly higher voltage than the regulator output, otherwise as soon as you put a load on it there won't be enough 'spare voltage' for it to draw from to maintain +5V. For these applications, a regulator is necessary.
|
Lightning/surge protection for a Arduino monitored thermistor circuit I am considering whether to replace the commercial differential temperature controller on my solar water heater with a Arduino based controller of my own design. I know just enough to be dangerous about such things. First question: will the classic 5V/10Kohm thermistor voltage splitter circuit blow something during a lightning storm? The thermistor is located 60 ft away on the roof mounted solar panel. The cable is shielded and grounded. What's needed - surge protector on the thermistor circuit, some RC connection to the thermistor leads, reduce the base resistor for more current flow to the thermistor .... Second question: does the Arduino Atmega microprocessor auto-reboot and resume software execution after a loss of power? Put another way, does the reset button have to be pressed after a loss of power? <Q> Answer to second question: AVRs have a BOD (Brown-out detector) whose purpose is to detect short power interruptions, and reset the controller when they occur. <S> In the datasheets, however, you'll find this statement: If the Brown-out Detector is not needed in the application, this module should be turned off. <S> The reason Atmel gives is that the BOD will consume power, even during sleep. <S> I find this odd: the BOD is a major factor in your device's reliability. <S> If it spends long periods of time in low power modes, and a dip occurs in the power supply it may lock up and require a hardware reset. <S> In practice unplugging it for a few seconds. <S> Not something I would like to tell my customers. <S> BTW, Atmel publishes an appnote <S> "AVR180: External Brown-out Protection". <S> I'm not sure what's the rationale behind this. <S> Does it mean the on-chip BOD isn't reliable?? <A> The second question is easy to answer. <S> The ATmega is a microcontroller, which is hard-wired to reboot and resume after a loss of power. <S> In fact, that's what the reset button actually does on some boards. <S> Every time you apply power, the controller reads the content at 0x00 (usually a jump instruction), and begins executing code. <S> The first question, not so much. <S> Lightning strikes are pretty serious events, and (especially without a schematic), it's hard to say what will happen. <S> I'd suggest that you first provide some isolation for your circuitry. <S> An easier method would be to make the temp sensor completely independent. <S> A little MSP430 + MRF24J40 system could run for months on a couple batteries and cost less than $10, transmitting the current temperature every couple minutes. <S> Then, when lighting strikes, there won't be an easy path to ground through the sensing wire, which means that lightning is likely to strike elsewhere instead. <S> The easiest method (also the least likely to survive a strike) would be to place a zener diode across the thermistor. <S> You'll want to be careful about compensating your measurements for leakage currents through the zener, though. <S> If you can't accept the possibility that the temp sensor would be destroyed by a lightning strike (which is an interesting requirement to design for), you should research transient voltage suppression diodes and be prepared for some much higher system costs. <A> You might want to look into GDTs. <S> Gas Discharge Tubes. <S> These are often used in telecom to buffer sensitive circuits from lightning strikes. <S> The resistance when under their rated voltages (varies from 50v to over 200v) is many megaohms. <S> When the voltage reaches a higher level, the device will move into a glow range (think neon lamp). <S> This is good for small spikes. <S> When it gets hit with REAL voltage, like 40 kV from a strike, it converts into an arc phase, where the resistance is very small and lines are shorted together, protecting the sensitive components. <S> You still need something to handle the low danger voltages of a couple hundred, but after that the GDT takes over. <S> None of these will protect you from a direct strike to the board. <S> Hopefully you have grounding path so a roof hit will mainly be taken to ground and all you are protecting is incidental voltage spikes and not a true lighting current path. <A> Thanks for the input. <S> After studying this a little more, I think a Metal Oxide Varistor would give some level of protection. <S> I wonder what's in my commercial differential temperature controller to deal with this possibility. <S> It's beyond my ability to reverse engineer.
|
Many voltage regulators have an enable pin, and it's very easy to wire it up in such a way that the reset button actually cuts power to the board. A little optoisolator is likely to provide the isolation you need, but you'll need to provide power on the high-voltage side. But a GDT across your thermistor might be the thing.
|
how can I make a large transformer less loud? I have a large 2600W 230V -> 28V transformer, and it makes a constant quite loud bzzzzzz sound while it works. Is there any way to make it more silent (other then buying an equivalent SMPS which I don't have money for now)? I am looking for some quick and cheap ways to make it less loud. Wrapping the whole thing in a large towel might not be the best idea as it will probably prevent the transformer from cooling properly. Why does it make that annoying sound anyway? <Q> The sound is a result of the system creating large magnetic fields as a function of how it operates. <S> These fields are fluctuating, probably at 50Hz based on your input voltage <S> but maybe 60Hz if your in the US. <S> These magnetic fields push and pull on components in the system, since perfect dampening isn't possible these components vibrate. <S> Some of what you hear isn't at 50/60Hz though, it is also harmonics of that frequency and interactions with other components. <S> You definitely shouldn't do anything that will limit air flow or otherwise create a situation where the transformer could overheat. <S> You can poke around a bit (carefully) and see if you can find a particular components or panel that is vibrating and try to use something to dampen the vibration. <S> High Temp Dampening material may be an option. <S> That being said, transformers of that power are basically impossible to silence. <S> The transformer core or the windings are probably vibrating and causing most of the noise. <S> At high enough power levels the iron core itself actually stretches and contracts as well causing vibration. <S> Some designs try to isolate the core to prevent vibrations transferring to other components but its not something you can likely fix in an after-market fashion. <A> The explanation I've usually heard is 'loose laminations'; at least one of the laminations has room to vibrate with the time-varying magnetic field. <S> Not sure what can be done about it. <S> Some transformers have bolts through them to hold the laminations together, and it might be that tightening them would mitigate the buzzing somewhat. <S> If there's an obvious gap somewhere, there might be some way to get some epoxy or something into the gap that would limit the freedom of movement of the adjacent laminations. <A> You might not be able to do much about the noise directly but you can always redirect the noise such that you hear less of it. <S> Lower frequency noises are more difficult to absorb, they tend to travel right through the material. <S> Thus wrapping a towel around it won't do much. <S> They are however easily reflected by a dense hard material. <S> Build a sturdy sound wall around one side <S> and that should redirect the noise somewhere else while keeping good airflow to the transformer. <A> Is the transformer mounted? <S> If you temporarily unmount it, is it still noisy? <A> I've used polyurethane on I large welder transformer. <S> Typically they are coated a the factory with something similar. <S> It just somtimes dries out with time. <S> You can also put shims made from paper or wood between the coils and iron core. <S> After I coated mine I used it for an hour to warm it up and the vibration helped wick the poly into the laminations. <A> Can't really help you with your problem, but I might be able to give you some insight into the cause: Magnetostriction <S> -- sounds cool, huh? <A> Get you some epoxy spray paint, open up the enclosure to get to the core and windings, and paint the heck out of it, pour it on thick, it don't have to be pretty, in fact you want runs, you want the paint to run in between everything the core laminates and windings. <S> Put a fan on it for over night to dry it. <A> can, in many instances, reduce emitted noise. <S> Professional transformer makers will immerse an entire power transformer in epoxy. <S> Yes, too much can reduce the heat dissipation, but to be sensible, if the transformer is running cool but noise, warm and quiet might be better. <A> From the factory, the transformer may or may not have been "conformal-dipped", meaning it was placed into a vat of liquid similar to polyurethane (but specifically designed for high temperatures and electronics work), then let to sit and dry. <S> This helps "glue" everything together and prevent microscopic motion we perceive as "hum. <S> " The better transformers are dipped, then a vacuum pulled on it, releasing all the air bubbles trapped inside, allowing conformal into every nook and cranny. <S> That works better, but obviously is more difficult and expensive. <S> An alternative not mentioned yet would be mounting the transformer with rubber mounts into a sealed metal enclosure, filling it with a compatible transformer oil (or mineral oil), and using a liquid-cooling system. <S> Many such transformers exist in electrical substations all over the world and are quieter and can handle more power due to liquid cooling.
|
It could be that what it is mounted to is acting as a sounding board -- you might be able to use rubber insulators or otherwise modify the mount to reduce noisy generation. If the transformer is so noisy that it cannot be effectively used, flooding the windings with epoxy paint
|
Do 4-bit CPUs still outsell 32-bit CPUs in unit volume? Way back in 2002, Jim Turley mentioned that about 14% of all CPUs sold were 4-bit CPUs, while about 8% of all CPUs sold were 32-bit CPUs.(Most people I know were surprised that any 4-bit CPUs were still being made, much less that they were doing so well). It's 2011 now -- Do 4-bit CPUs still outsell 32-bit CPUs and 64-bit CPUs combined in unit volume?Where could I go to look up the latest numbers for sales by unit volume?What are the top websites and magazines to learn more about modern 4-bit CPUs and their development tools? <Q> The reason that 4 bit CPUs beat out the high end processors at the time was primarily due to their use in watches and clocks. <S> They are used in many other very low cost, very high volume applications. <S> I can't imagine that 4 bit CPUs are going up at the same rate. <S> 64 bit CPUs are still largely related to computers, and are probably still low volume enough that they have not yet eclipsed 4 bit CPU volume. <A> I doubt it. <S> 32-bit has gained enormously in market share the last 5 years. <S> More and more designs that in the past would have been done in 8-bit are now being done by 32-bitters, mainly ARM. <S> Due to technological progression a 32-bit RISC controller doesn't need more real estate than an 8-bit CISC. <S> That translates in lower cost. <S> Same for 4-bitters. <S> Most 4-bit controllers are older designs being produced in an older process (larger feature size). <A> That 2002 statistic might have been a bit surprising, but it wasn't some two bit factoid from USAToday. <S> The author was once an editor for the Microprocessor Report. <S> If you think back to the technology and prices of the mid-1990's (when many design decisions affecting volumes in the early 2000's were made), a lot of embedded applications were simple and well-suited to 4-bit and 8-bit processors. <S> (Heck, it's still true today for most of the "invisible" processors in use - like the LCD thermostat, or the microwave, or the smart dimming dome light in your car.) <S> The problem with 16-bit and 32-bit processors back then was that it was needlessly more expensive to provide memory for them. <S> RAM was not cheap back then. <S> And wider RAM was much more expensive for the same capacity. <S> (In fact, the early PC's were quasi-16-bit machines. <S> They had an 8-bit external memory bus.) <S> Fast forward to this decade, and one key change is that the newer embedded processors have plenty of embedded RAM on-board; thanks in large part to improved semiconductor processes. <S> Without the external RAM penalty, it's just as easy to grab a 32-bit processors for a new design. <S> And the volumes are there that you don't pay much more for the 32-bit. <S> In fact, for bang-per-buck, the older 8-bit processors are awful for new designs. <S> And, I can't even imagine anyone today would even bother with datasheets for a 4-bit processor. <S> So, have 4-bit processors died? <S> Given that even soft core processors are 8-bits, I'd say yes. <S> The fun question today is what the split is between 8-bits, 32-bits, and 64-bits. <S> Circling back to the original question - I used to see "processor yearbooks" that detailed processor offerings from different manufacturers - and they were broken into groups by processor bit size, and whether they were MCU or CPU's. <S> I haven't seen one of those things lately -- I think there's far more players in the market today, many of them from Asia-centric companies that have little or no sales presence in the U.S. <S> Additionally, "processors" may be hidden inside FPGA's so that it would be hard to count them. <A> The size of the chip has less to do with its word-length than it has to do with its architecture. <S> For example, an 8-bit CISC-like machine would probably suck up more real-estate than a 32-bit RISC-like machine. <S> As others have commented, 4-bit CPUs are still being sold and still find application in all kinds of things from air-conditioning units through to microwave ovens and maybe some toasters. <S> However, they will be slowly phased out as the cost of 32-bit machines are now as low as the smaller machines (due to economies of scale). <S> So, if you are working on a new design, it would be a good idea to build it on a newer chip. <S> The slight difference in cost today, may well disappear in a few years time.
|
A 4 bit CPU, running in BCD mode (ie, each 4 bits was one 0-9 digit), is very optimal for clock and timing applications, and in volume they are nearly as cheap as transistors. I don't know whether they still beat 32 bit CPUs, however the trend has been going up steeply for 32 bit processors due to cell phones and portable computing devices.
|
Are there any Wifi/Networked Lights available to purchase? Summary I'm trying to find a light that can be connected straight into a network via WIFI or LAN Cable (cat5, etc..). Details We have a server which handles automated builds and deployment for our website development. If the automated build (aka our continuous integration) failed, we would like to be told. So - traditionally, people have been using lights or monitors to visually display the status of their builds. green - last build worked 100% fine.yellow - in progressred - last build failed. So, I was hoping to find a light (bulb/lava lamp/whatever) that can be connected to the network and I can hack up a program to then tell that light to display as red/yellow/green, depending on the state of the build. So - are there any lights that can be connected to a network, preferably via WIFI but via cable is fine. I'm assuming they will also need a power plug to get power .. unless maybe the lan cable can supply power, if the light doesn't require too much? <Q> I assume that you're ready to write code to make such a device. <S> Even if there was something as simple as a networkable LED, you'd have to write custom code to parse the build output file so you could choose a color to display. <S> And since you want this for a build server, I'm already assuming that you can write code. <S> :) <S> The first product that came to mind was a thing I remember from a couple of years ago -- it looked like a rabbit and the ears would go up or down, and change color. <S> I think it was supposed to report the general mood of the internet, at least for things that interest you, like stocks and sports. <S> Stocks up <S> == good == <S> green == ears up. <S> Favorite team loses = <S> = bad == <S> red == ears down. <S> But I couldn't find that product anymore, so <S> You can read about it here . <S> That said, why don't you use the free network LED, which is basically a monitor hooked up to a cheap PC that meta refreshes the results page from your integration server? <S> EDIT -- <S> so my joke about a "free network LED" wasn't well received. <S> :) <S> If you want to keep development of this to a minimum, then two good options are Arduino / Netduino and mbed. <S> Arduino + ethernet shield make it super simple to get your embedded device on the network and is outlined here . <S> mbed is another candidate and looks pretty easy to get hooked up to your network. <S> There are examples like this and this for getting the networking going. <A> I think you want the Ambient Orb . <S> Unfortunately, I think they're not being manufactured any more. <A> I've searched long and hard for this before and the best I've come up with are networked power switches from this manufacturer . <S> They have an embedded web server, so you could easily hack together a plugin for your CI server to switch lava lamps on and off after a build. <S> Note they're Ethernet only and very expensive as from what I can tell these type of products are intended for data centre operations. <S> (Upon further googling, I also find this , which is wifi and cheaper, but apparently on backorder. <S> They say they have similar products available now though.)
|
the next most logical, hackable, open "network LED" is the Chumby / Chumby One. Might find one on Ebay.
|
why cannot generators having different frequencies supply in the same transmission line? I am thinking of the issue that the TEPCO electric transmission networks cannot be taken over by other electric companies because of frequency difference. TEPCO is the electric company having the nuclear crisis. Why cannot generators having different frequencies supply in the same transmission line? Edit I think most electric equipments can work at either 50 OR 60 Hz. So what is the problem? <Q> Transformers designed to work with 60Hz will have lower ratings when working at 50Hz (Important for power distribution and individual devices) <S> The speed (RPM) of AC motors is tied directly to the power line frequency. <S> AC Wall clocks wouldn't work :) <S> See the plot of both 50Hz and 60Hz superimposed from Wolfram <S> The blue plot shows the difference. <S> Where the function is large, the generators are fighting eachother (Big Boom!) <A> A voltage phase mismatch from one generator and the power grid would mean a voltage difference between the grid and the generator during different parts of the phase(s). <S> This voltage difference means current would flow to/from the grid and the generator, causing resistive heating in the lines. <S> Also, I don't know all the mechanical and electrical consequences of trying to run a spinning turbine in reverse, but I don't think it would be good. <A> All power line components should work fine at both 50 and 60Hz. <S> On the other hand the whole electrical grid has to be precisely synchronized <S> so there is no easy way to suddenly connect several circuits together. <S> There is no difference between connecting: <S> two out-of-phase generators 24V DC to 12V DC a supply directly to the ground (a short circuit). <A> It's a lot more than just the transformers. <S> It's turbines, control systems, meters, loads... <S> TEPCO's sole generating source isn't Fukushima - there's additional nuclear in Niigata as well as fossil fuel generation. <S> You can't mix frequencies unless the whole system is changed. <S> Also according to Wikipedia, there are several HVDC stations <S> that interconnect between the 50Hz and 60Hz grids through an intermediary conversion to HVDC. <S> Inefficient, sure, but at least there is some interconnection. <A> Loads notwithstanding, if you just connect a 50Hz source and a 60Hz source together, they will alternate between being in-phase and out-of-phase 10 times per second. <S> So half the time, things would be fine, but the other half the time, the sources would be trying to burn one another out. <S> This might not be such a big deal except that when the sources can supple megawatts, there would definitely be smoke.
|
Equipment designed to work with one frequency power will not necessarily work with another frequency. The combination of 50Hz and 60Hz would cause generators to fight eachother
|
Is it safe to drive a MOSFET from an output pin of a microcontroller? I have used commonly-available BJTs such as the 2N2222 and 2N3904 as switches by operating them in "saturation mode" from my MCU. I believe, however, that for these sorts of applications, a MOSFET is a more appropriate device. I have a few questions, however. 1) Does a MOSFET have a "saturation mode" like the BJT does? Is this "saturation" achieved by simply providing a high enough voltage on the base that the MOSFET is completely "on"? 2) Is it safe to drive the MOSFET directly from the MCU? I understand that the gate of the MOSFET behaves like a capacitor, and therefore draws some current while "charging", and then none thereafter. Is this charging current high enough to damage the MCU pin? By placing a resistor in series with the gate, I can protect the pin, but this will slow down the switch, possibly resulting in high heat dissipation by the MOSFET? 3) What is a common "hobbyist" MOSFET suitable for various low-power situation? I.E., what's the MOSFET equivalent to a 2N2222 or 2N3904? <Q> Many power MOSFETs require a high gate voltage for high-current loads, to ensure that they are fully turned on. <S> There are some with logic-level inputs, though. <S> The data sheets can be misleading, they often give the gate voltage for 250 mA current on the front page, and you find that they need 12V for 5A, say. <S> It's a good idea to put a resistor to ground on the gate if a MOSFET is driven by an MCU output. <S> MCU pins are usually inputs on reset, and this could cause the gate to float momentarily, perhaps turning the device on, until the program starts running. <S> You won't damage the MCU output by connecting it directly to a MOSFET gate. <S> The BS170 and 2N7000 are roughly equivalent to the BJTs you mentioned. <S> The Zetex ZVN4206ASTZ has a maximum drain current of 600 mA. <S> I don't think that you will find a small MOSFET that can be driven from 3.3V, though. <A> Note that "logic level" does not seem to be an exactly standardized term, and it won't necessarily show up as a parameter in the parametric search at the vendor sites, nor will it necessarily show up in the data sheet. <S> However, you will find that logic-level MOSFETs often have an "L" in the part number, ex: IR540 (non logic level) vs. IRL540 (logic level). <S> The big thing is to look in the data sheet and check the VGS(threshold) value and look at the graph that shows current flow vs VGS. <S> If the VGS(threshold) is like 1.8V or 2.1V or so, and the "knee of the curve" on the graph is at around 5 volts, you basically have a logic-level MOSFET. <S> For an example of what the specs on a logic-level MOSFET look like, check out this datasheet: http://www.futurlec.com/Transistors/IRL540N.shtml Figure 3 is the graph I was referring to. <S> All of that said, I see that a lot of people still recommend using an opto-isolator between the micro-controller and the MOSFET, just to be extra safe. <A> Re: saturation: yes, <S> but it's confusingly not called saturation (which actually corresponds to the linear region in bipolar transistors). <S> Instead, look at the datasheets and the rated on-resistance Rdson, which is specified at a certain gate-source voltage for each part. <S> MOSFETs are usually specified at one or more of the following: 10V, 4.5V, 3.3V, 2.5V. <S> I'd put two resistors into the circuit: one from gate to ground, as Leon has mentioned (actually I'd put it from the MCU output to ground), and another between the MCU output and the gate, to protect the MCU in case <S> the MOSFET has a fault. <S> More discussion on this blog entry . <S> As for what MOSFET to use, there really isn't a parallel to the 2N3904/2N2222. <S> 2N7000 is probably the most common & cheapest FET out there. <S> For other jellybean FETs, I'd look at Fairchild FDV301N,FDV302P, <S> FDV303N,FDV304P. <S> For the next step up (higher power level), I'd look at IRF510 (100V), or IRFZ14 <S> (60V), both in TO-220, though these are basic FETs spec'd at 10V gate-source. <S> Logic-level FETs (IRL510, IRLZ14) have Rdson specified at 4.5V gate-source. <A> In answer to question 3, I found the Fairchild FQP30N06L is ideal for driving a high power device from a MCU at logic levels. <S> It's not cheap (0.84 GPB) but great for lazy n00bs like me. <S> I'm using them for supplying 12V RGB LED light strips. <S> Some stats: Vdss <S> Drain-Source Voltage: 60 VId Drain Current: <S> Continuous (TC = 25°C) 32 <S> A Continuous (TC = 100°C) 22.6 AVgss Gate-Source Voltage: ± 20 VVgs(th) <S> Gate Threshold Voltage: <S> 1.0--2.5 V <S> Therefore, Raspberry Pi's 3.3v is above the 2.5V upper Gate Threshold, which will ensure that the drain is fully open. <A> driving power mosfet is driving capacitor 1-8nf. <S> it is slow with resistor and not possible without. <S> mosfet require fast, powerful switch in front of it: bjt
|
It is safe - in general - and it will work if you select a "logic level" MOSFET.
|
When is a MOSFET more appropriate as a switch than a BJT? In my experimentation, I've used only BJTs as switches (for turning on and off things like LEDs and such) for my MCU outputs. I've been repeatedly told, however, that N-channel enhancement-mode MOSFETs are a better choice for switches (see here and here , for examples), but I'm not sure I understand why. I do know that a MOSFET wastes no current on the gate, where a BJT's base does, but this is not an issue for me, as I'm not running on batteries. A MOSFET also requires no resistor in series with the gate, but generally DOES require a pulldown resistor so the gate doesn't float when the MCU is rebooted (right?). No reduction in parts count, then. There doesn't seem to be a great surplus of logic-level MOSFETs that can switch the current that cheap BJTs can (~600-800mA for a 2N2222, for example), and the ones that do exist (TN0702, for example) are hard to find and significantly more expensive. When is a MOSFET more appropriate than a BJT? Why am I continually being told that I should be using MOSFETs? <Q> BJT's waste some current whenever they're switched on, regardless of whether the load is drawing anything. <S> In a battery-powered device, using a BJT to power something whose load is highly variable but is often low will end up wasting a lot of energy. <S> If a BJT is used to power something with a predictable current draw, though (like a LED), this problem isn't as bad; one can simply set the base-emitter current to be a small fraction of the LED current. <A> A good N-channel MOSFET will have a very low \$R_{ds(on)}\$ (drain-source equivalent resistance) when properly biased, which means that it behaves very much like an actual switch when turned on. <S> You will find that the voltage across the MOSFET when on will be lower than the \$V_{ce(sat)}\$(collector-emitter saturation voltage) of a BJT. <S> A 2N2222 has \$V_{ce(sat)}\$ from \$ 0.4V - 1V \$ depending on biasing current. <S> A VN2222 MOSFET has a maximum \$R_{ds(on)}\$ of \$ 1.25 \Omega\$. <S> You can see that the VN2222 will dissipate much less across the drain-source. <S> Also, as previously explained, the MOSFET is a transconductance device - voltage on the gate allows current through the device. <S> Since the gate is high-impedance to the source, you do not require constant gate current to bias the device on - you need only overcome the inherent capacitance to get the gate charged up then the gate consumption becomes miniscule. <A> BJT's are more suitable in some situations because they are often cheaper. <S> I can buy TO92 BJT's for 0.8p each but MOSFET's don't start until 2p each - it might not sound like much but it can make a big difference if you're dealing with a cost sensitive product with many of these. <A> BJTs are much more suitable than MOSFETs for driving low-power LEDs and similar devices from MCUs. <A> FET devices having almost no input current (gate current) are the best choice for the LEDs driven by the micro-controller as micro-controller doesn't need to provide much current through its die, keeping itself cool (less heat dissipation on chip) while LED current is almost all driven through the external FET channel. <S> Yes, it is also true that the Ron of the typical FET devices are very low keeping low voltage drop across FET which is advantageous for low power application. <S> However, there is some disadvantage when it comes to noise immunity at the gate of the MOSFET, which may not be the case for the BJTs. <S> Any potential (noise) applied at gate of the MOSFET will make the channel conduct upto some extent. <S> It is not highly (but still adequate) to use the Mosfet to drive the relay coils with low Vt (threshold). <S> In that case, if your Microcontroller is driving the FET, you might want to get a FET with higher Vt (threshold). <A> MOSFETs are more robust for high current requirements. <S> For example 15A rated Mosfet can pass 60A (f.e. IRL530) of current for a short period. <S> 15A rated BJT can pass 20A pulses only. <S> Also Mosfets have better thermal junction to case resistance even if it has smaller die.
|
MOSFETs are better for high-power applications because they can switch faster than BJTs, enabling them to use smaller inductors in switch-mode supplies, which increases efficiency.
|
Replacement for Queues in RTOS For Inter-task communication or to share data between two tasks of RTOS, We use Queues.But Problem with Queues is that they are slow.... They copy data in Buffer then Mutex Handling and then Data Transfer. It's irritatingly slow if you have to transfer large data.Another problem is if same queue is accessed by Multiple tasks. Then Picture becomes like this:- First Wait to get access to The Queue then Queue internal Mutex Handling then Data Transfer. This increases overhead on the system. What could be the Efficient replacement for Queues? (I guess this question is Independant of RTOS we use. Most of the RTOS handle Queues in this way only) <Q> One easy way is to put a pointer to the data on the queue and consume the data using the pointer. <S> Note that you're trading safety for performance this way as you have to make sure that: the buffer remains valid until the consumer has consumed the data someone deallocates the buffer <S> If you're not using dynamically allocated memory you don't have to deallocate it, but you still have to make sure that the memory area is not reused before the data has been consumed. <A> Queues operate that way because that is a thread-safe transaction model for inter-task communication. <S> You risk data corruption and/or ownership issues in any less-stringent scheme. <S> Are you copying the data into a buffer in memory then passing a pointer with the queue elements, or trying to pass all the data in the queue elements themselves? <S> If you're not passing pointers then you'll get an increase in performance doing that instead of passing one byte at a time through queue elements. <A> Lock-free queues can be implemented for the single-producer/single-consumer case, and often you can architect your software to minimize the number of multiple-producer or multiple-consumer queues. <S> A lock-free queue can be constructed like so: <S> Allocate an array of the elements to be communicated, and also two integers, call them Head and Tail. <S> Head is an index into the array, where the next item will be added. <S> Tail is an index into the array, where the next item is available to be removed. <S> The producer task reads H and T to determine if there is room to add an item; writes the item in at the H index, then updates H. <S> The consumer tasks reads H and T to determine if there is data available, reads data from index T, then updates T. Basically it's a ring buffer accessed by two tasks, and the order of operations (insert, then update H; remove, then update T) ensures that data corruption doesn't occur. <S> If you have a situation with multiple producers and a single consumer, or a single producer and multiple consumers, you effectively have a resource limitation of some kind, and there's nothing else for it but to use synchronization, since the performance limiter is more likely to be the lone producer/consumer than an OS overhead with the locking mechanism. <S> But if you have multiple producers AND consumers, it's worth spending the time (in design-space) to see whether you can't get a more coordinated communication mechanism; in a case like this, serializing everything through a single queue definitely makes the efficiency of the queue the central determinant of performance. <A> One can get efficient operation in a lock-free multi-producer single-consumer queue if the queue itself holds items that are small enough to work with a load-store-exclusive, compare-exchange, or similar primitive, and one can use a reserved value or reserved values for an empty queue slots. <S> When writing to the queue, the writer does a compare-exchange to try to store his data into the next empty slot; if that fails, the writer tries the following slot. <S> Although the queue maintains a pointer to the next empty slot, the pointer value is "advisory". <S> Note that if a system uses compare-exchange rather than load-store-exclusive, it may be necessary to have a 'family' of different 'empty slot' values. <S> Otherwise, if between the time the writer finds an empty queue slot and attempts to write to it, another writer writes the slot and the reader reads it, the first writer would unknowingly put his data in a spot where the reader wouldn't see it. <S> This problem does not occur in systems that use load-store-exclusive, since the store-exclusive would detect that the data had been written even though it was written back to the old value. <A> You can access queues more efficiently by writing on top of the queue Normally most of the RTOS does give the support of adding to the front of the queue which doesn't require acquiring of mutex. <S> But make sure you use adding to front of queue as minimal as possible where you just want to execute the data faster. <S> Normally queue structures have max size limit so you may not put all the data in queue hence passing the pointer is always easy. <S> cheers!! <A> Queues are not inherently slow. <S> The implementation of them may be. <S> If you're blindly copying data and using a synchronous queue, you're going to see a performance hit. <S> As other posters have indicated, there are lock-free alternatives. <S> The single-producer/single-consumer case is straightforward; for multiple producers and consumers, the lock-free queue algorithm by Michael and Scott (those are their last names) is the standard, and is used as the basis for Java's ConcurrentLinkedQueue .
|
It's possible to optimize out the need for queues in certain cases, but they provide concurrency guarantees that usually provide huge simplification benefits to systems by allowing you to decouple tasks.
|
which kinds of electrical equipment need a pure sine wave inverter to work correctly? I am planning to purchase a good inverter, and pure sine wave inverters cost about 3 times as much as modified sine wave ones of the same power. I am thinking about a 2000W continuous/4000W peak inverter 12/24VDC -> 230VAC. Which commonly used equipment requires a pure sine wave, and what kind of equipment will be equally fine with a modified sine wave? I am especially wondering about compressor and absorption refrigerators, dehumidifiers, electric drills, fans, microwave ovens, common power tools and kitchen tools. Purely electronic, computer and lighting equipment will not be used through this inverter, as these will have dedicated low-voltage DC-DC converters straight from the battery. Are pure sine wave inverters any more efficient then much cheaper modified sine wave ones, in terms of input battery consumption per 230VAC output power? Do there exist inverters which can accept a wide range of input DC voltage, for example will work on all of: 12V, 24V and 48V DC? <Q> Some people claim that certain loads "may" not work as well, or "may" be damaged , with anything other than a pure sine wave. <S> Since the power coming out of my wall sockets is significantly different from a pure sinewave, I suspect these sincere and well-meaning people are merely repeating propaganda from the manufacturer of a pure sinewave inverter manufacturer,much like people repeat urban legends. <S> There is only one kind of device that I know <S> doesn't work as well with a non-sine-wave inverter: <S> devices that use a "capacitive power supply"-- see How efficient is a capacitive power supply? <S> for details. <S> Since a "capacitive power supply" has a power factor near 0, it is questionable whether any "capacitive power supply" meets EU-mandated power factor laws, such as EN61000-3-2. <S> I suspect that all products -- as long as they were designed after EU-mandated power factor laws went into effect -- should work just as well on "modified sine wave" as with "pure sine wave" inverters. <S> Of course, I can't possibly know about every product ever designed -- if there is any specific product (that was designed after those EU-mandated power factor laws went into effect) that can't work or doesn't work as well with "modified sine wave" power, I would be interested. <S> If anyone can explain why it doesn't work, in enough detail that I can try to avoid that kind of failure, I would be fascinated and grateful. <A> The modified sine wave inverters generally cause more power loss in your products' power supplies. <S> So the inverter itself may not be any more efficient, but the equipment running on a pure sine wave inverter will most likely run more efficiently. <S> This is especially true for inductive loads, such as all the equipment you listed. <S> I'd guess that most of your equipment with linear AC/DC power supplies will work, but perhaps not as efficiently or with reduced performance and increased heat. <S> Other products that use switching supplies may not work. <S> I'd recommend ponying up the extra money for the pure sine wave inverter for anything other than an emergency power supply you need to use temporarily when power is lost, for example. <S> Especially as you are looking to drive inductive loads. <S> I'm not an expert on inverters, but <S> I know there are plenty that run on 12V designed for the RV community. <A> Generally speaking, anything with an inductive or motor load, pure sine wave is better. <S> A load that first rectifies the input (PC power supply, phone charger etc) a modified output would be sufficient <A> Most equipment is likely to work but quite possibly at a lower efficiency. <S> Lower efficiency can lead to overheating. <S> There may also be problems with interference. <S> Any EMI filter capacitors will experience higher currents than normal. <S> If those capacitors were already marginal then the increase can lead to failure. <S> Rectifier circuits will likely experience higher peak currents. <S> Again if those components were maginal to start with then the increase can be the difference between survival and failure. <A> ALL switching based supplies that run with an AC input can be run on DC directly, gotta love bridge rectifies and EU laws... <S> I've run everything from motors to "sensitive" medical equipment on modified sine, the only difference between the two is that modified sine wastes more power in the equipment being run. <S> But seeing as how I've run most industrial equipment on modified sine with no problems for days on end <S> I think it's just marketing bull. <A> Some kinds of power meters may also not work correctly (or they may be damaged) if they are connected to a modified sine wave source. <S> See an example of a burnt wattmeter here.
|
Many radios, for example, will sound worse when run off a modified sine wave converter.
|
Crafting a digital watch I've been wondering. How hard is it to craft a (digital) watch from scratch? I'm currently thinking that (with a lot of assumptions) you can buy/import quartz chips or something like that from overseas, and then after putting them on a case, stick them to some circuits, batteries, some LCD display etc. So yeah, what other considerations could there be, besides manufacturing (cases, straps etc)? disclaimer: I'm a complete newbie when it comes to electronics, so you probably cringed just by reading that. I'm just curious about this because after trying to shop for a good watch, it seems that there's almost too many choices, and I can't help but wonder how hard is it to get to the watchmaking business. <Q> The quartz crystal should be the least of your problems. <S> You can find them in miniature dimensions (2mm diameter x 6mm length) at any distributor. <S> You may have some trouble to find the LCD, though. <S> Microcontrollers in small packages are readily available, but they usually don't have enough I/ <S> O pins to control the LCD + some buttons. <S> In common watches you'll see that they solved this by bonding the die (with sufficient I/Os) directly to the PCB. <S> Standard components are never used for the battery holder and the push buttons, because they wouldn't fit. <S> Instead, custom made metal metal strips are used. <S> Frankly I think this will be a difficult to do. <S> If you really want to build a clock I'd suggest to make it a table top clock. <A> Not at all difficult - the most expensive part would be making a nice case. <S> For minimum cost you'd be constrained to a small selection of off-the-shelf LCDs (which also constrains choice of case), but if you;re serious you can get a batch of small custom-design LCDs for something like $1000-2000 for several hundred units including tooling. <A> What about the TI Chronos? <S> 1 <S> I've considered buying one for its large display, which would be very nerdy but also very practical with my poor eyes. <S> It's relatively inexpensive, too.
|
The main difference between you and the watchmakers business is that you want to build 1 (one) watch, while they build them in large quantities, which allows them to use custom components at a low cost.
|
Why are there teardrops on PCB pads? What's the use of the teardrop shapes around some PCB pads? <Q> There are primarily two reasons to use teardrops: <S> It avoids a pocket (where the trace meets the pad) that could collect acid from the PCB etching process which would later do bad things. <S> That being said, in professionally made PCB's teardrops are rarely needed. <S> It's almost more of an aesthetic thing than a solution to a real problem. <S> I've done many boards with and without teardrops <S> and I have yet to notice a difference. <S> In my opinion, they are more trouble than they are worth. <A> It prevents drill breakout where the trace joins the via or through hole. <S> Sometimes it's not necessary do it because the manufacturer can do it for you. <S> Where I work, our repair department recommend us do it because it increase the strengthening of connections between pads and tracks. <A> They are mainly used on single-sided boards, to increase reliability. <S> I often use them on my home-made PCBs. <A> If you look at the picture the Ben Jackson posted in answer to a question about solder mask expansion , you'll see that the drilled holes can be significantly off-center from the pad centers. <S> In extreme cases, you could actually end up with holes that leaves no or very little annulus to make the connection to outgoing traces. <S> The tear drop ensures that there's enough pad material left over to have a solid connection to the pad. <A> The simple answer is for Strain Relief. <S> Like a plug with a thin cable, it needs a graduated support to prevent the shear forces of thermal expansion when desoldering a component or when a human interface force is involved ( eg mic jack or a heavy part on a board.) <S> Where you do not have strain or any thermal or mechanical stress in the application, it is unnecessary. <S> But I have seen dozens of times when this would have prevented a track failure that was hard to see with the naked eye. <S> One example was a $2K Mac vertical monitor with the flyback transformer on the main board... <S> It created a micro-sheer crack impossible to see, but whenever the CRT went blank a slap to the side of the case fixed it. <S> Which lasted for a few days, until the secretary cried.... <S> help.. <S> so I came to the rescue with a soldering iron. <S> and said to myself... <S> I wish the board designer knew about tear drop pads. <A> I thought it had to do with IPS-6012C for rigid board. <S> Specifically related to section 3.4.2 Annular Ring Breakout. <S> For example, Class 2 allow for an annular ring to break out (be missing) for 90 degrees. <S> This is due to drill misalignment. <S> On large vias or plated through holes, this is not likely to happen in modern fab facilities. <S> When dealing with 10 mil finished drill holes and 20 mil pads this is much more probable. <S> If the breakout happens on the side of the trace exiting the pad, Class 2 states that it can cut the trace down to 50um and still be acceptable. <S> If there was a teardrop (or fillet), the misdrill would not cut into the trace. <A> A teardrop via make it easy to remove embedded resistors when using ECL logic. <S> This was done by Harris Computer Systems about 1987 to make it easier to repair memory PCA's that cost upwards of $15,000 (or more) in their h1000 and h1200 systems. <S> To remove the embedded resistor that was connected to the via all one needed was a pin vice and then drill a hole at the point of the teardrop. <S> Then add a wire to a spare embedded resistor or add an external resistor. <S> I remember building the light box to see the teardrop via.
|
It reduces mechanical & thermal stress resulting in less hairline cracks in the trace. The teardrop is to help the drilling process.
|
Adding a small protoboard-like area to a PCB I'm currently laying out a PCB in eagle for a circuit that will form the basis for some experimentation. Instead of adding a row of headers so that I can plug it into a breadboard, I figured I would try and layout a small protoboard directly onto the PCB itself. I have room to spare on the board and the resulting creations will be somewhat more resilient. This PCB is a good example of what I'm shooting for. I have considered using vias, but I seem to recall they are usually sealed off with a chemical to prevent solder from bonding. This is obviously not what I'm going for. Adding hundreds of single-pad components to my schematic does not seem very appealing either. What is the best way of achieving this using Eagle? EDIT: Thanks for your help everyone. Here's the design and here's the finished product . <Q> You could use the 1x25 package in the SparkFun Eagle library as a starting point. <S> It gives you 25 pins spaced 0.1" apart (used for single row headers). <S> Modify to taste. <A> At some point, with any EDA tool, you're probably going to have to create a custom part for something; so you might as well dive in and create a proto-area part with an array of pads the way you want. <S> You need to make the schematic component for that as well and place it on your schematic. <S> This is a good technique also for design-specific holes, especially if the holes have to align to some externally defined dimensions. <S> Having a pseudo-component in your schematic to call out those features would make those features "official" parts of your design. <A> There is a thread on the Eagle Support forums that deals with this issue in detail. <S> One of the replies includes a link to a library of prototyping components . <A> EAGLE trains you to add schematic parts for everything that appears on the PCB, but in this particular case, you really want to just add the holes, traces and silkscreen elements directly to the PCB. <S> Think of it more as technical drawing than building a circuit. <S> I did this in the spare space in the upper left corner of my PIMETA v2 board: <S> The holes are 40 mil drill with 70 mil pads, and the traces are 40 mil. <S> I highly recommend adding silkscreen outlines, as you see above. <S> This makes clear which pads are connected to which. <S> That was particularly helpful on this board, since the traces were on the bottom, but even if they were on top, I'd add the outlines. <S> The contrast of silk on solder mask is simply a lot better than for the copper under the mask. <S> Most of the pattern is intended to support DIP chips. <S> The bits at the leftmost edge deviate from that partly due to lack of space but also to support an optional switch. <S> (That's the translucent yellow overlay you see.) <S> It's not important, here, to discuss what those switch pads are good for. <S> The point is that you may not want to make your prototyping area completely generic. <S> Another example where I deviated from generic patterns is that some of the pads connect to the board's existing power and ground rails: V+, V-, B+ and IG. <S> Doing that is one of the major advantages of having a prototyping areas on a special-purpose PCB, as opposed to using generic off-the-shelf protoboard: it means that things built up in the prototyping area can be run directly off the board's existing power supply and you don't have to run hookup wires across the board to get back to power points elsewhere. <S> I recommend that you do the same. <A> In most good PCB programs you can get rid of the solder mask over a via by expanding the hole cut in the mask. <S> This thread seems to be useful. <A> I'm not acquainted with Eagle, but I guess it has copy/paste like any EDA program. <S> Set your grid to 0.1", place a free pad, copy and paste. <S> Select both pads, copy/paste. <S> Select all four pads, copy/paste. <S> You can do the whole area in less than a minute. <S> Vias are a bad choice because 1) they're too small, or you'd have to make a custom one, and 2) the pad of a via usually has a solder mask, which you would have to remove.
|
You should think through the scenarios of how the prototyping area will be used, and if there are special features you can add that will make it more useful than a generic protoboard pattern, do so. What you are talking about with regards to the vias is solder mask capping them.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.