source
stringlengths 620
29.3k
| target
stringlengths 12
1.24k
|
|---|---|
What's the schematic to share one crystal with two micros? How do I share a single crystal (not a complete oscillator module) between two micros? Is it OK to just connect everything as normal for the first micro and also directly connect its XO to the XI of the second microcontroller? I expect to place the micros very close together on the board. <Q> I would expect problems trying to drive two oscillators from the same crystal, if for no other reason than that the capacitance of the long PCB traces required to go between the two processors would cause a malfunction. <S> simulate this circuit – <S> Schematic created using CircuitLab <A> That is not quite what you want. <S> When you use a crystal to form an oscillator, you are using an inverter internal the the microcontroller to drive the crystal. <S> If you connect both micros to the same crystal, they will fight and not work. <S> Therefore you will want to pick one of the two micros to serve as the crystal driver (configure as shown in section 6.2.6 of the datasheet) and the other micro to use an external clock input (6.2.1 of data sheet). <S> Then change your schematic connection like this: <S> simulate this circuit – <S> Schematic created using CircuitLab <S> Alternatively, if symmetry is desired you can use an external oscillator as opposed to a crystal. <S> Then both could operate in external clock mode. <A> The problem with that approach is that the connection to the second oscillator input will affect the capacitance on that pin, and alter the frequency slightly as well as reducing the crystal drive. <S> It might not matter, though, and is very unlikely to prevent oscillation. <S> When I've needed to do that for a product I've used a crystal oscillator module for both MCUs. <A> You may build oscilator using cheapo invertor chip (0.1$), or specialized clock generator (slightly more expensive), and feed that clock to both uC. <A> If it's anything like the PICs I've used, you have an XTALout and and XTALin with the actual XTAL between them and caps (maybe around 22pf) to ground. <S> Try connecting the XTALout of one to the XTALin of the other. <A> Make an oscillator out of your crystal and an inverter like BarsMonster suggested, and feed it into the input of a fanout distribution buffer IC . <S> One fanout buffer output goes to the clock input of each MCU. <S> I can't really recommend a specific part without knowing what micros you're using, but you'd want to start with these filters: <S> Type: <S> Fanout Buffer (Distribution) Number of Circuits: 1 Ratio - Input: <S> Output: 1:2
|
What I have done when I needed to share a clock between micros is to have the crystal drive the oscillator on one micro and then use the oscillator output pin (typically CLKOUT or OSCOUT) from that micro to drive the second micro. Simply connection crystal to both would not work (reliably).
|
dc motor and hung (not stable) pic I'm running a dc motor using PIC16 family microprocessor.I connect it to my pc, I can send a command to run the motor in certain speed or stop. It works for small motor (3V, 100 mA). But the moment I use bigger motor (3V, 300 mA), I can only I send 1 command e.g. run fast. After that, the pic refused to process the next command. Even when I turned off the device and turned it on again, the motor still running the last command. It seems like memory hardening or something like that. I'm not sure what happened- Could it be a kickback current from motor caused the instability? I'm using TIP120 as motor driver.- I have put capacitors, diodes around the motor driver. But didn't help. Can someone help me? I appreciate your thoughts. <Q> Federico Russo said it well: "decoupling, decoupling, decoupling". <S> ICs like microcontrollers require smooth power supplies, that is without disturbances. <S> Small negative spikes may cause a reset or cause your software to go bananas. <S> Positive spikes may do the same, and even damage the part. <S> You definitely want to get rid of those disturbances. <S> There are two ways to attack the problem, and the best way is to apply both. <S> First consider the cause of the disturbances. <S> This is often difficult to find out, but in our case it's definitely the motor. <S> Place capacitors between the power supply and ground, close to the motor's connection pins. <S> Have an electrolytic capacitor of \$100\mu F\$ <S> (actual value depends on the motor current, but this is a good start), and place a \$1\mu F\$ ceramic capacitor parallel to it. <S> The latter is needed because the elco is not good at high frequencies, and there the ceramic takes over. <S> This is the first step. <S> It's not just necessary to solve our problem, it also reduces EMI (ElectroMagnetic Interference). <S> There are regulations about the level of EMI you may create. <S> Then we go to the microcontroller. <S> The power supply will not necessarily be clean yet, there may be other noise sources. <S> Here we do the same: place capacitors on the power pins, between \$V_{DD}\$ and ground, as close as possible to the pins. <S> The microcontroller doesn't use high current, so we won't need the \$100\mu F\$ elco. <S> Usually a \$100nF\$ ceramic will do. <S> To calm your nerves :-) <S> you may add a \$1\mu F\$. <S> What else? <S> Some other pins may also be sensitive to noise. <S> Look at the reset pin. <S> It's no good ensuring that your power supply is clean if the microcontroller would reset due to spikes on the reset pin. <S> So also a cap between reset and ground, again as close as possible to the pin. \$100nF\$ is fine. <A> Solderless breadboards are hardly ever appropriate for power electronics of any kind. <S> (and especially not switching!) <S> Compared to a printed circuit board, they have poor parasitic inductance, resistance, capacitance, and noise susceptibility. <S> Don't expect them to conduct more than 20-50mA without causing undesirable voltage drops somewhere in your circuit. <S> If you are desperate to get your circuit working, you can try putting bypass capacitors on the board, but otherwise wouldn't bother trying to diagnose the problem, and instead I'd move all your power circuitry off of the solderless breadboard and onto a PCB or at least a vectorboard. <A> Noise is getting from the motor circuit into the PIC circuit. <S> The most likely way for the noise to get to the PIC is via the power supply as Leon pointed out. <S> Where is the PIC supply coming from? <S> Don't run the motor and the PIC from the same regulated supply. <S> You probably don't need a regulated supply for the motor at all. <S> At the very least, filter the supply to the PIC a bit. <S> Best would be to give the electronics a separate regulator, preferably with a diode and storage cap in front of it. <S> That way the PIC supply will still be stable even if the main supply is glitched to ground for short periods occasionally. <S> However, there is more to noise immunity than just the power supply. <S> The best attack against noise is to avoid making it in the first place. <S> Do you have a snubber accross the motor, or at least a small cap? <S> You don't want too much else that will put strain on the PWM motor driver, but a little to limit the voltage slope from inductive kickback and commutation is useful. <S> Does the PIC have a good bypass cap as close as possible <S> accross its power and ground pins? <S> It certainly should. <S> What is keeping MCLR high? <S> In a high noise environment, MCLR must not be too high impedance else it will pick up noise and randomly reset the processor. <S> Does this PIC have a PGM pin? <S> If so, it needs to be kept from picking up noise just like MCLR. <S> Also disable the PGM feature unless you really need it. <S> What about inductive kickback from the motor? <S> Does the current have a path to go without creating high voltage spikes? <S> These could not only damage the motor driver, but get back into the PIC and cause unpredictable operation. <S> What about the ground? <S> Is the motor current kept off the PIC ground? <S> It should be. <A> Use separate supplies for the PIC and motor and connect the two grounds at only one point. <A> you should check h21e of the transistor, you should use darlington pair or mosfet, since motor load translates to load on pic pin. <S> when load small pic can drive transistor, when load bigger bipolar transistor will require current = <S> 300mA/ <S> h21e to be open. <A> TIP120: <S> Collector−Emitter Saturation Voltage 4 V ? <S> your supply is 3 V :). <S> Put mosfet instead. <S> Your schematic is wrong.
|
Filtering of the PIC supply should help, with a transient voltage suppressor, if you still have problems.
|
Why does a processor overclock faster when cooled with liquid nitrogen? Why does a processor overclock faster when cooled with liquid nitrogen? Also, is the inverse the reason why the processor slows down when it get's hot? I'm specifically interested in Intel processors. <Q> It's not the temperature that makes it run faster or slower. <S> The lower temperature allows more heat power to be removed from the device, which allows more electrical power to be put in without burning it up, which allows it to be clocked faster. <S> Part of the electrical power required is proportional to the clock speed. <A> You want to keep the core within its operational range. <S> If you want to run the core faster which means it consumes more heat/energy <S> then you need to remove more heat to keep it within its operational temperature range. <S> If you can improve the cooling through various methods, fans, liquid, gas, otherwise, then you can add more energy in the form of an increased clock (multiplier). <S> You can damage the part by trying to run it too cold as well as you can by running it too hot. <S> Also there is a physical limit to this <S> you cannot run it infinitely fast infinitely cold. <S> Your limit is likely determined by the silicon itself, there is a limit to the clock multiplier, and you would have to change the reference oscillator to keep increasing the clock rate. <S> Think about the human body, try running a mile when it is 60 degrees outside, then find a place to try running in the same clothes when it is 110 or 120 degrees outside. <S> Which one are you likely to make it through the mile without passing out (failing)? <S> Which temperature will allow you to push yourself harder than normal? <S> Keep your body within the operating temperature range <S> and you can push it harder, at least for a period of time. <S> If you warmed up inside before the run, you might push yourself a little faster if it were 50 degrees outside. <S> But there is a limit, warm up all you want indoors, but in a t-shirt and shorts, you may not make it when it is 30 below outside, 40 below, 50 below. <A> Semiconductors run faster at a lower temperature and at a higher power supply voltage. <S> Higher voltage means more heat generated, which means that it has to be cooled more. <A> Several folks have correctly answered that more cooling allows the CPU to run faster, BECAUSE if it is cooled it allows you to overclock it more. <S> These same folks correctly implied what I am about to say plainly: <S> This was not (until recently) a native talent of CPUs. <S> And it never has been a physical law of semiconductors. <S> A CPU clocked at 1.2 GHz runs at exactly 1.2 GHZ, whether it is cooled to 60F or running at 160F. <S> If you lose cooling (e.g., remove the cooling fan + heat sink), it will run at 1.2GHz until it melts itself into a puddle and can no longer run at any speed. <S> But it will run at exactly 1.2GHz until the very second it dies. <S> I'll bet many on this forum have actually witnessed/experienced this. <S> Some of the new-fangled computers have their own temperature monitoring and control systems which automatically enable/disable overclocking (or otherwise adjust CPU speed) based on CPU temperature. <S> So if the CPU gets too hot, it slows itself down (reducing energy in) rather than burning itself up. <S> (I reckon it does this by selecting a slower clock or by dividing the existing clock down; but I am not an expert on newer CPU innards.) <S> If the CPU cools down, the automatic governer circuitry reverses this process to allow the CPU to run faster. <A> Just to clarify: It's not about just heat dissipation. <S> You can dissipate 1000 W of power using boring water cooling, but it would not allow you to get to the top. <S> The idea is that property of semiconductors change, as well as resistance of interconnect (=copper).Lower resistance - lower RC constant, which is main factor limiting processor speed. <S> If one could cool it down to superconductivity stage, clock would increase even more - but this is unlikely for copper interconnect we see in current CPUs.
|
The more you can cool a chip, the faster you can make it go.
|
how to connect relay and physical switch to control light I am new in this kind of stuff so please excuse me for my lack of knowledge. I have successfully connected blackwidow (wifi + arduino) with a relay to control a 220V light bulb. Now, I want to embed all this circuitry in my room for each and every socket.The problem is that I don't want to remove my physical switches from the wall. I want them to be there and always function normally. How should I connect both elements so that I can control the lighting from both the switches? Arduino hardware Physical switch which has been there for a long time And how can I achieve the following? Turn on the light with the physical switch, then turn off the same by my android which is connected to the blackwidow via local network and vice versa Thanks in advance <Q> If you want to control a light from 2 points you need the following circuit (the top one): Chances are that your current switch is just on/off (SPST or Single Pole Single Throw) instead of the SPDT (Single Pole Double Throw) switch used in the diagram. <S> You really need the SPDT, but you use it just like the other one. <S> The second switch (B) is your relay, which also has to be SPDT. <S> If you want to switch from 3 or more points you need the center circuit. <S> Switch A and B remain the same, but you need switch C (repeat C for every additional switch). <S> The center and bottom circuit are the same, they just show switch C in both their states. <A> Sorry to revive an old topic, but I've been thinking about the problem of wiring like this WRT the home automation not knowing if the light is on or off. <S> I think I may have come up with a solution. <S> 110V Coil relays are now dirt cheap. <S> What if the 5V DC Coil Arduino/ <S> Pi/ <S> Whatever controlled relay was paired with a 110V AC Coil relay. <S> Installation of this is done near the fixture, not the switch. <S> but just before the load you'd split off the hot wire and connect it to the 110V relay's coil. <S> (Obviously attaching the neutral to the other side of the relay's coil.) <S> Then from the N.O. contacts on the 110V relay you would attach it to your Arduino/ <S> Pi/ <S> Whatever as a digital input. <S> And when the power has been turned off (either through the automation or manually) the low voltage circuit will be open. <S> It seems so simple (so I must be missing something). <S> I've done a VERY rough drawing of what I mean. <A> You're out of luck <S> I'm afraid my friend. <S> Which ever way you go about it will require some re-wiring or replacing of the physical switch and its cable. <S> As the other answer shows, you can achieve it with two-way switches and cabling, but as it says you will probably not have this in situ, so you will need to re-wire it and replace the switch with a SPDT type. <S> This will also allow you to get some feedback from the Arduino as to the real status of the light and toggle it on and off properly.
|
The other way to do it would be to replace the wall switch with a push-button type (you can get them that look like a switch but don't stay in the 'on' position), which connects to an input ( not using mains ) on the Arduino to toggle the light relay. This way when the loads are receiving power, the low voltage circuit attached to the 110V relay will be closed. Basically you'd wire the whole thing up exactly the way stevenh has shown (either 3 way or 4 way setup will work)
|
Internal or external oscillator I always use the internal oscillator that pics have as I have never found the need to run anything at higher frequency than 8 MHz (which is the fastest the pics I use tend to be able to go). Are there any reasons, beyond going above 8 MHz, that mean I should use an external oscillator? It just seems like one more thing to go wrong to me but I'd be interested to hear what others do. <Q> As others have said, accurate frequency and frequency stability are reasons to use a external ceramic resonator or crystal. <S> A resonator is several times more accurate than the internal R-C oscillator and good enough for UART communication. <S> A crystal is much more accurate, and necessary if you are doing some other types of communication like CAN, USB, or ethernet. <S> Another reason for a external crystal is choice of frequency. <S> Crystals come in a wide range of frequencies whereas the internal oscillator is usually one frequency with maybe a choice of 4x PLL enabled. <S> Some newer 24 bit core PICs have both a multiplier and divider in the clock chain so you can hit a wide choice of frequencies from the single internal oscillator frequency. <S> There are of course various applications that inherently require accurate frequency or timing other than communications. <S> Time is the property in electronics that we can measure most accurately cheaply, so sometimes the problem is transformed into one of measuring time or producing pulses with accurate timing. <S> Then there are applications which require some long term synchronization with other blocks. <S> A 1% oscillator would be off by over 14 minutes per day if used as the basis for a real time clock. <S> Accurate long term time may also be needed without having to know real time. <S> For example, suppose you want a bunch of low power devices to wake up once every hour to exchange data for a few seconds and then go back to sleep. <S> A 50ppm crystal (very easy to get) will be off no more than 180ms in a hour. <S> A 1% R-C oscillator could be off by 36 seconds though. <S> That would add significant on-time and therefore power requirements to the devices that only needed to communicate for a couple of seconds every hour. <A> Precision. <S> Internal clocks are not precise, can be affected by noise. <S> Temperature independent precision. <S> Typical oscillators can vary wildly. <S> Specialty temperature compensating oscillators can be needed in low or high temp applications, or if temperature varies wildly. <S> Speed. <S> Internal oscillators may not reach the highest speed of the IC. <S> External ones can be needed for that. <S> Voltage. <S> The speed of an internal timer may be dependent on the voltage it is being run at. <S> Multiple clocks are needed. <S> Some applications want to share an oscillator. <S> Special applications where the internal clock may not be easily used. <S> Dividing the internal clock might be harder than throwing a cheap 31 kHz watch crystal at it, for time keeping applications. <S> Off the top of my head, the ATMEGA 328 the arduino uses requires an external crystal at 5V for its max speed. <S> The lily pad version runs at 8 MHz, on the internal oscillator because it's limited to that at 3.3v. <S> The MSP430 Value Line launchpad is limited to 10 MHZ at 3V, 8 at 2.5V. <A> Frequency stability will be higher with an external one. <S> So if you have a application that really depends on the mcu freq, then you may need to use a external one. <S> But most modern mcu: <S> s has a quite stable internal osc, therefore I think this used to be a bigger question a couple of years ago. <S> Also there is more and more ways to trim the internal one, and compensate for temperature drift (etc etc). <S> On the other hand there is other ways to make sure that you are synchronized, in some countries <S> the freq stability in the power net is <S> 50Hz ±0.01Hz and other places like Sweden actually has ±0.001Hz <S> and I have seen projects using this to keep things in sync. <S> And then you are no longer as dependent on the mcu freq and you can use the internal one. <S> But this is a little bit of topic :) <A> Frequency stability is the main one, particularly for serial comms at high speed. <S> But that also brings up the occasional need for a crystal at a seeming odd frequency to get an exact baud rate, because of the limited options that the clock dividers give you. <A> I have actually come across a scenario where 1% was not good enough for UART. <S> If any of you guys have heard of the Teensy++ v1.0 microcontroller dev board, it has a terribly sensitive UART. <S> I had my host baud set at 115200, and it set at 115200 and for the longest time could not figure out why it wasn't reading the data correctly. <S> Turns out my host was sending closer to 114300 baud. <S> ( 115200 - 114300 ) <S> / 115200 = ~0.9% error. <S> I tried it with two different MCU's and they worked fine. <S> Point is: regardless of your application, if greater accuracy of clock frequency is a benefit, you should use an external resonator, crystal, or even oscillator if you chip doesn't have the necessary driving circuitry. <S> P.S. <S> I wonder if anyone has any insights as to what low level design choice they made on the UART hardware that makes it so sensitive? <A> Sometimes to save money, designers use internal ones. <A> It does not make a big difference though. <S> Check the next chart from the PIC16F628A datasheet: <S> You can notice first that the more frequency then the more power is consumed. <S> It can make a difference in low power applications. <S> And about the internal vs external doubt there is, the INTOSC consumes more power than the XT at the same frequency, around 30% more power. <S> If this information is relevant or not, it depends on the application.
|
External crystal crystal oscillators are more accurate than internal clocks, and they should be used when accurate timing is necessary. Another concern related to the oscillator that has not been mentioned is the power consumption.
|
What happens if an AC voltage is applied to a battery? Today I was checking the voltage between charging terminals going to the battery of emergency lantern using a multimeter. The multimeter showed a DC voltage of 9V and an AC voltage of 4V. Could this AC voltage have damaged by old battery? The DC voltage while being measured was not constant and the decimal part was changing in a cyclic pattern. I only saw 3 diodes on the board. Is this a half-wave rectifier and is the multimeter mistakingly showing the ripple of half-wave as an AC voltage. Please see the image of the circuit. <Q> It's fairly common to see a lead-acid battery charged using rectified AC. <S> As long as the charging current isn't beyond the capability of the battery, it will 'work'. <S> If there isn't a series resistor somewhere, or some primary-side limiter, the winding resistance of the transformer could be what's limiting the charging current. <S> A handheld multimeter is sensitive to 60Hz AC, so yes, your DC reading was likely skewed by the low-frequency ripple. <S> Wall-wart adapters have similar failure modes (usually the transformer goes high-impedance). <A> The circuit you show is a full wave rectifier, not half wave. <S> Half of the transformer secondary will be conducting each cycle. <S> However, the third diode on the right doesn't make sense. <S> There is no need for it since there are already diodes in series with each of the two paths current can come from to get there. <S> The meter didn't mistakenly show you AC. <S> The voltage coming out of the rectifier circuit has a AC component to it. <S> The DC reading was jumping around a little probably because of the meter's sampling interval beating against the 2x power line frequency of the AC component. <S> A simple mechanical meter would probably have shown you a steady DC reading. <S> If this voltage is being used to charge a lead-acid battery, then probably there is no problem, assuming the voltage is in the reasonable range. <S> There is nothing explicit limiting the current in your schematic, but the transformer will have some internal resistance. <S> This is probably good enough, especially if the charger is intended for that unit. <A> The extra diode may be a voltage drop mechanism, to limit supplied voltage to the battery. <S> Could be that they had a transformer they needed to use for whatever reason and the output voltage <S> was just a tad too high, or it could have been a design tweak. <A> At best this looks to me like it would produce a pulsed DC current at twice the frequency of the mains (50Hzx2 or 60Hzx2) with the multimeter measuring the 1/2 wave AC at the higher frequency. <S> If that third diode had a breakthrough voltage, meaning that even in the correct direction the lower voltage wasn't passing, this would turn the current into something more square waved...really noisy on an AM radio.
|
If there isn't any explicit current-limiting protection in the charger, it is possible that the charger can become damaged if subjected to long-term overload. The circuit you have drawn is a full-wave rectifier.
|
Combining an Audio Signal and 9 VDC on the same Trace I have a backplane where 4 individual modules plug into it and provide various sound effects. To control which module is powered and to route the input signal to it, there is a 2-pole, 5-way mechanical switch so the power and signal are sent to the desired module. However, the output signal from each module shares the same trace going to the amplifier. This is a problem because when more than 1 module is plugged, in the output signal goes south. Individual modules work great, but with more than one module installed, not so good. I thought that since the other 3 modules were not powered they would have no effect on the active module, well, I was wrong. So I was trying to think of a way to route the input signal and power to the appropriate module and separate the output signals so they are not in parallel. That led me to see if I can use one of the poles of the switch for both power and input signal and the other pole for the output signal. Two other thoughts were to add a diode to each of the output traces or an audio transformer. I just don't know, I already have hardware built so I am stuck trying to make what I have to work. Do you think this will work or have any suggestions about the idea of adding the power to the input signal? <Q> The cheapest solution would be to use a 3 pole 5 position switch and use the 3rd pole to switch the outputs. <S> It is possible to combine the input and DC voltage into one line, but that brings up new complications. <S> The input signal could be coupled to the DC with a capacitor, then extracted by using a second capacitor on the amp end. <S> Problems to overcome with that would be power supply ripple would need to be very small, and you need to deal with impedance matching on both ends. <S> The switch would energize a 2 pole relay that which would switch the input & output. <A> I don't see why 9v power has anything to do with this, other than it's on the switch and backplane. <S> There are two reasons I see as to why the audio signal is going bad: 1. <S> The ESD protection diodes on the output of each card is getting powered down and so clamping the audio to GND. <S> 2. <S> Something else on each card is loading the signal, causing it to have a low level. <S> The solution could be to make a passive audio mixer using some resistors. <S> This is easy, since all you have to do is add a series resistor to the output of each card. <S> The volume of this mixed audio will go down, dependent on the # of cards, but that could be compensated for by increasing the gain downstream. <S> The value of the resistors really depends on what you have downstream, but something in the 500-10,000 ohm range is reasonable. <S> I should mention that if your problem is ESD protection diodes then the passive mixer could introduce some noise-- although for your application that may be acceptable. <S> Again, it depends on your circuit. <A> This thing is already in use, if I understood the question right: http://en.wikipedia.org/wiki/Phantom_power Lines going to audio will need at least a capacitor on the 'hot' line, preferably another one on the ground in case the connections gets reversed. <A> Why do you need to mux the power and inputs at all? <S> Why can't all the modules always be on creating their outputs, then you selectively mix or select the output you want. <S> Your existing 2 pole 5 posisiton switch could still be used but with 1 pole left over.
|
A third solution could be to use individual relays at each amp module location.
|
Can simple microcontrollers read signals from a USB to RS232 converter connected to a USB mouse? Trying to connect as many usb mice to my arduino and just got informed of usb to rs232 converters. Can anyone tell me if it is possible to simply connect a usb mouse to a usb to rs232 converter, connect some wires from the rs232 output plug to the microcontroller and then read the data for all mouse events? <Q> No. <S> The USB mouse needs to be plugged into a USB host. <S> The USB to <S> RS232 controller is not a USB host. <S> It is a USB device, like the mouse, and relies on the host to perform various USB bus management functions. <A> No, but you could plug the mouse into a USB to PS/2 converter which will make the mouse switch into synchronous serial mode. <S> You would need a matching mini-DIN socket fed with 5V but decoding the data should be fairly straightforward. <S> See here for the wiring. <S> UPDATE <S> As an afterthought - remembering that PS/2 mice work in asynchronous mode with a PS/2 to DA9 adapter - I tried cascading USB = <S> > <S> PS2 = <S> > <S> Serial adapters (with external power grafted-in). <S> Needless to say, USB mice don't support legacy async serial mode (well it was worth a shot!). <A> Ditch the Arduino and use an MCU with host mode or USB OTG (On The Go). <S> It won't cost much (I'd use a PIC24FJ256GB110 with USB OTG), but developing the software will be a lot of work, although Microchip has a free USB software stack: http://www.microchip.com/wwwproducts/Devices.aspx?dDocName=en531089 <A> You should pick up the USB host shield . <S> It has a USB host controller chip and comes with supporting software that will let you talk to your mouse. <A> If you are looking to interface a USB mouse, keyboard, or other HID, look at this website: https://www.circuitsathome.com/communicating-arduino-with-hid-devices-part-1 <A> If you can find a USB to rs232 converter that converts USB device input to rs232 output then <S> yes! <S> unfortunatly that sort of devuce costs $100 or more not including the time spent searching for it. <S> you're better off with a cheap 32 bit micro-controller running s USB stack, or a full OS.
|
I just tried this with a couple of new mice and they worked OK.
|
Why do LDO regulators have so big a voltage drop? Why do LDO linear regulators not use MOSFETs as the main component to be able to have minimal dropout=0 (well, depending on current, must be still a few mV)? Or can one expect to build a 0-dropout regulator based on a MOSFET and an opamp? <Q> There are regulators with a drop out voltage close to 0 mV. Check figure 5 on page 6 in <S> TPS73101, Cap-Free, NMOS, <S> 150mA Low Dropout Regulator with Reverse Current Protection . <S> Another example is LTC1844 - 150 mA, Micropower, Low Noise, VLDO Linear Regulator . <S> The problem with regulators at such low drop out voltages is that in those regions they have crappy parameters (line/load regulation and PSRR ). <S> As to the part if it is possible to build such regulator with an op-amp and a discrete MOS device - yes, it is possible. <S> You will have to use PMOS and take care of stability (it is not easy to make a feedback loop stable in such a configuration). <A> If you want a super-low LDO, you need a device with an extremely-low input-to-output saturation voltage (i.e. a FET) and some way of having the control voltage higher than the input. <S> Using a BJT will always limit you to the \$V_{CE}\$ saturation voltage, plus you need sufficient base current to ensure the transistor will be on fully when necessary. <S> Also, the \$V_{BE}\$ voltage has to be taken into account. <S> If the base is 1V below the collector, then the emitter has to be more than 1V + \$V_{BE}\$ lower. <S> If you're using an N-channel FET as the series pass element, you need to get the gate high enough above the source for the FET to conduct fully. <S> Many logic-level FETs need more than a volt. <S> Many FETs with good \$R_{DS(on)}\$ need even higher than that. <S> If you tie the gate to the input voltage, for example, you can expect that the \$V_{GS}\$ threshold voltage will be dropped across the MOSFET, making it a 'lossy' LDO as per your question definition. <S> But then again, if you already have a higher rail available, why not use it as the regulator input and stop worrying about the super-low LDO? <A> Some LDOs use an external MOSFET: http://www.micrel.com/page.do?page=/product-info/products/mic5156.shtml <A> I designed a discrete LDO linear regulator circuit using an n-channel MOSFET to make a negative voltage. <S> This was 22 years ago, and I published it in an electronics magazine set up for charging SLA batteries at 13.8 volt. <S> Thousands were built in one form or another, and I did not have any stability problems. <S> This old simple circuit could be configured with a p-channel FET and lower output voltages and these days the drop would be limited by the low MOSFET on resistance. <S> SMD parts mean that discretes are not a penalty, so I know that really low drop is now possible.
|
A discrete LDO using a FET and a driver able to fully turn on the MOSFET (i.e. higher gate voltage than the input voltage) will allow you to make an LDO which will only have a series \$R_{DS(on)}\$ loss, theoretically.
|
Memory backup capacitors: why a capacitor? Why would an engineer choose a memory backup capacitor instead of any of the types of rechargeable batteries available? Their energy density seems far poorer, and they can't hold their charge very long. The only advantage I can think of is it's slightly simpler to charge? (Just needs a resistor), but then again, the same can be said for trickle charging many batteries. <Q> They're cheaper and (rightly or not) in product design that's often an argument which overshadows a lot of technical arguments. <S> And, like you said, capacitors don't require special charging electronics. <S> Also, batteries contain chemicals for which you have to comply to special regulations. <S> (Sometimes this is the reason why toys and such are delivered without batteries.) <S> edit Another thing <S> : batteries are not suited for wave-soldering. <S> You'll either have to solder them manually afterward (cost), or, like for coin cells, use a battery holder (cost) and add the battery manually (cost). <A> Rechargeable batteries for memory/RTC backup power isn't a good solution today. <S> The batteries will eventually die due to the many charge/discharge cycles that it can be put through. <S> And many rechargeables will self-discharge in a month or two. <S> There are also regulations regarding the metals in these batteries that might come into play. <S> Non-rechargeable lithium batteries can have a lifespan of 10+ years in a RTC/memory backup use case. <S> This is often longer than the expected life of the product it is in. <S> So, it has become common to see these batteries soldered down to a PCB-- because it's more rugged than a battery holder <S> and it is never expected to be replaced. <S> Of course, this makes many people nervous. <S> These batteries also have the regulatory issues common with the rechargeables. <S> Supercaps are a nice alternative. <S> They are not always cheaper than the non-rechargeable lithiums, and don't operate as long on a charge as the rechargeables. <S> But they are maintenance free and don't have the regulatory issues. <S> So, none of the solutions are a perfect mix of cost, regulatory, lifespan, and maintenance free. <S> The designers of a product just have to weigh the pro's and con's and pick the one that they like best. <S> BTW: <S> One product I did used a Supercap as an RTC backup. <S> It would run about 9 months before loosing <S> it's charge. <S> Of course, turning on the unit would recharge that cap in about a minute. <A> Capacitors are nicer for the end user because they don't require any maintenance. <S> They charge up quickly the first time the product gets turned on, instead of taking a day or two to get fully trickle charged. <S> They work for thousands of times more charge cycles and don't leak caustic chemicals onto the circuit board when they stop working. <S> and they'll stop drawing energy when they're full. <S> Batteries generally either need a chip to monitor charging or will tend to get somewhat overcharged if left on trickle charge indefinitely without being discharged, which is detrimental to their lifespan. <S> Batteries also have to be disconnected from the load when they get discharged down to their minimum voltage, which requires extra circuitry, whereas capacitors can be discharged to zero volts. <S> Batteries hold more energy in the same volume/weight, but otherwise are a pain to manage, all the other advantages are with capacitors. <S> A battery makes the most sense in a small/portable product, or one that is so cost-sensitive that it's worth inconveniencing the end user to save a few pennies in design cost.
|
Capacitors are easy to charge, the circuitry around them is simpler, just limit the voltage and current going into them
|
Do I need to duplicate the 4th "substrate" connection when building CMOS gates out of discrete transistors? All CMOS digital integrated circuits I've ever seen connect all the nFET substrates together to GND. In particular, the IC CMOS NAND gate has one nFET that has its substrate connected to GND, but its source pin connected to some other internal node. If I build a NAND gate out of discrete nFET and pFET transistors for educational reasons, do I need to duplicate that substrate connection by using a 4-terminal transistor (with the substrate separately pinned out) to get it to work?Or would that NAND still work just as well with 3-pin discrete transistors, with the substrate "incorrectly" connected to the source pin? Is there something magic about a 4-terminal transistor that has a "source" pin not tied to its substrate, such as the ones inside an IC, that can't be duplicated by an individual discrete 3-terminal transistor? (This question was inspired by some coments at Recomendation for a digital inverter made of discrete components ). <Q> Some Background: <S> In a simple CMOS process, (P-type Wafer, N-Wells) <S> the substrate contact is directly connected to the conductive wafer. <S> This means that the body terminal of all NFETs are basically shorted together. <S> A similar effect happens with the PFETs, although it isn't as absolute. <S> They aren't shorted together to improve performance, but because it is cheaper and easier to manufacture. <S> This brings up a question: If we had to tie the body terminal of all NFET devices together, what voltage would we like it to be? <S> For NFETs, the body-source and body-drain connections normally look like reverse-biased diodes. <S> In order to maintain these diodes reverse-biased, the body voltage must be less than \$V_D \mbox{ or } V_S <S> + 0.6\mbox <S> { Volts}\$. <S> Typically this is done by tying the substrate/body to the most negative voltage present in the system. <S> In digital systems, this is usually ground \$V_{SS}\$. <S> The body terminal of PFETs is typically tied to the most positive voltage, or \$V_{DD}\$ for similar reasons. <S> For 3-terminal FETs, where the source and body have been internally shorted, the internal diodes will never be forward-biased if the source is always at a lower voltage than the drain. <S> If you are stuck with 4-terminal transistors building discrete gates, it will work with the bodies connected to \$V_{SS}\$ and \$V_{DD}\$, and it will also work with the body shorted to the source. <A> In addition to what W5VO has said, the 4-terminal transistors are also useful for analogue electronics on CMOS processes. <S> In such cases, sometimes, the body of the transistor may be connected to some intermediate voltage instead of VDD or VSS. <S> This can be used to modulate the VTH of the transistor using the body effect . <S> This is described here . <A> You can make logic without separating the substrate / bulk / body connection from the source. <S> But if you wanted to experiment and make circuits that are just like CMOS ICs <S> so you need 4-terminal MOSFETs <S> , you can use CD4007UB .
|
The short answer is that you don't need 4 terminal FETs to build CMOS logic.
|
Why do smoke detectors go off when lightning strikes? Just experienced this, I saw lightning outside my window (not hitting our building or anywhere close to it), and immediately after the smoke detector went off for a short while. Can anyone explain what caused this? <Q> Lightning is a nasty thing. <S> Powerful. <S> Very high current at very short rise time. <S> This causes an strong EMP (ElectroMagnetic Pulse) which will be picked up by anything conducting. <S> A 1m free-hanging wire may create a voltage peak between its ends. <S> Even short connections may see spikes. <S> Decoupling doesn't always work as the EMP can enter an IC directly; it doesn't have to come by the (power) wires. <S> So no wonder some products experience a temporary malfunction during a lightning bolt, and high impedance mean more sensitive. <S> If the disturbance remains within the device's voltage range it may behave wrongly without suffering damage. <S> Higher voltage spikes may destroy (parts of) the device. <S> I heard the story of a Dutch family where lightning had struck in the backyard. <S> Every electronic product in the house was fried, from TV and PC to cameras and mobile phones. <S> Se non è vero... <S> And David with his smoke alarm network/antenna <S> , well... :-) <A> Just to add what the other folks have said... <S> The fire alarms in my house are all inter-connected. <S> When one goes off, they all go off. <S> There are wires inside the walls/ceilings that connect them all. <S> Those same wires, because they are long and unshielded, are excellent antennas and would easily pick up the EMI from a lightning strike. <S> Also, the signal on these wires is very simple and the electrical noise generated could easily fool the alarm into thinking that some other alarm went off <S> and so it should too. <S> What the other guys said could also be true ( <S> except the part about CO2 setting off an ionizing detector, it is actually a smoke particle that does it), but if the alarms are interconnected then that would be the weakest link. <S> It's much easier for EMI to get into a 50 foot wire than something that is about an inch long. <A> I think ionization and EMP glitches are missing the point. <S> I find ionization in particular very hard to believe. <S> How is lightning 100s of meters away going to ionize the air in my smoke detectors pretty much at the same instance as the lightning? <S> It's not. <S> I don't believe the EMP theory either. <S> The pickups are sensitive and high impedance, but also shielded from external capacitive pickup. <S> If not, ordinary power line hum and nearby static discharges would set them off, but they don't. <S> What is really going on is that the power got glitched. <S> A lightning strike makes a mess of the power line for a few 10s of milliseconds. <S> Most smoke detectors, including all the ones in my house, sound off for a short time whenever the power goes out. <S> They are fairly sensitive to this, more so than most ordinary appliances. <S> You may notice a small flicker in the lights or a glitch on the TV at the same time (although lightning causes TV and radio glitches by other means too). <S> When we have a pure power glitch not caused by lightning, it is always the smoke detectors that exhibit the symptom first. <S> Most appliances can take a cycle or two of missing power, but the smoke detectors seem to be the most sensitive. <S> I don't know if this is deliberate or just a byproduct of their sensitive detection circuitry. <A> I'm going to assume an ionisation detector, as those are the most common on the market today. <S> Inside a smoke alarm there is a controller IC. <S> This IC registers the current ionised by the americium-241, typically on the order of 100pA. <S> If smoke (\$CO_2\$) enters the chamber the current stops, triggering the alarm. <S> Bill Hammack explains it all . <S> Now what happens when say a nearby lightning strike introduces a high power RF blast to all electronics? <S> That current goes all over the place. <S> It probably swings between several microamps both positive and negative, for a brief period of time. <S> The smoke alarm IC never expected to see this. <S> It doesn't want to see this, as it isn't designed to handle it. <S> So the internal current comparator probably goes a bit crazy and latches the alarm logic which resets every 15 seconds or so. <S> This causes the alarm to trigger. <A> Smoke detectors are not the only devices that can misbehave when lightning strikes. <S> Because smoke detectors are designed to rather go off spuriously than stay silent wrongly, your smoke detector was triggered by the lightning's signal. <S> Often you can hear and see those signals on radios, TVs, oscilloscopes, etc., too.
|
Like every spark a lightning is a strong transmitter on a broad range of frequencies that can induce wrong signals in electronic devices.
|
Resistors in series with Tx and Rx I'm making my own board and using an ATmega 328 with the Arduino bootloader. I have a DIP switch to select either an FTDI chip (for programming) to be connected to the ATMega's Rx and Tx, or a GPS that outputs serial to be connected. I was looking at this schematic for reference: http://arduino.cc/en/uploads/Main/ArduinoNano30Schematic.pdf Why are there 2 resistors on Rx and Tx coming from the ATMega? Do I need those just for the connection to the FTDI chip, or do they need to be there for the GPS too? <Q> One of them is there to prevent damage that could occur if the AVR has RxD programmed as an output, pins on both devices could be damaged if that happened as AVR pins can source and sink quite a lot of current. <A> It doesn't look like there is any good reason for those resistors. <S> Both parts on that schematic appear to run on 5V with a common ground. <S> There should be no need for resistors in the lines between the two chips. <S> If the lines were going off board, then there might be some point to putting resistors in series to protect the on-board parts, but that doesn't seem to be what's happening in that schematic. <S> Keep in mind <S> this is a Arduino schematic. <S> That means there is a good chance whoever designed it doesn't do this professionally. <S> There are a lot of superstitions out there. <S> Just because something is on the net doesn't mean it's done right. <A> This is an old and already answered question, but I didn't find in any of the answers one of the good and possibly one of the most important reasons for the resistors to be there. <S> Although most people use the RX/TX only to connect the Arduino to their PCs for programming the chip and/or perform serial debugging, others use the Arduino's RX/TX pins to communicate it with other serial devices. <S> In this case, the FTDI chip and this other device would conflict and it is very likely that will damage both due to short circuiting. <S> These resistors "separate" the FTDI from the other device when there is one connected to the AVR RX/TX pins, protecting both and allowing them to be wired and connected simultaneously. <S> One thing to remember is that, once another serial device is connected to the RX/TX pins of Arduino, the resistors will mask the logical levels from the FTDI in a similar way that happens with pullup/pulldown resistors, so, the external device will have "preference" over the FTDI communication. <A> It could be done to prevent the other off board device is powering the Atmel when it is powered off. <A> Adding a small-value (100 ohms or so) series resistor on a signal which is going off-board can reduce RF emissions. <S> The resistors on the illustrated schematic don't seem well-placed for that, though. <S> Another use for resistors is as a really cheap mux. <S> If the FTDI chip tries to drive the Arduino's RX pin and nothing on the header tries to, the FTDI chip will "win", but if something on the header tries to drive that pin without a series resistor, the device on the header will "win". <S> That might explain some usefulness for the resistor on the Arduino's RX pin. <S> Not sure what purpose the one on TX, serves, though, unless there's another external connection for the "TX" wire that's wired to the FTDI's RX pin <S> and I'm just not seeing it (if there was such an external connection, it would be possible for the external device to inject data to be sent via the FTDI). <A> I have seen 100 ohm resistors on I2C and UART buses before, they are often for ESD protection. <S> They work in conjunction with the built in clamp diodes in the MCU.
|
I don't think that the other resistor is necessary. Due to current that runs via the internal clamp diodes of the Atmel...
|
Battery directly connected to input of unpowered opamp. Will the input current drain the battery? Is it possible to hardwire battery to opamp input, considering that most of the time normal power on opamp will be off ?Battery positive wire is connected to non-inverting input of opamp workiing as a buffer (inverting input is connected to output). In opamp specs the input current is in picoampere range, but is it true for the unpowered state ? Thank you <Q> Most ICs have ESD protection diodes on their inputs. <S> That means there is a diode from the input to the positive supply and from the negative supply to the input. <S> These are reverse biased in normal operation. <S> Their purpose is to clip out of range voltages that could damage the rest of the IC. <S> That's not a reliable way to power the device, and it could result in latchup and other bad things. <S> Another point is why are you buffering a battery voltage? <S> Most likely the battery can deliver significantly more current than the opamp can. <S> The buffered output might be a little lower impedance if not exceeding the current limit, but in a battery powered device this current will still come from the battery. <S> The buffer may hold the output voltage close to the input voltage, but the drain on that output voltage will lower the battery voltage at least as much if the battery were connected directly. <S> This isn't making a lot of sense. <A> I found legitimate paper with the description of unpowered mode, which I am interested in. <S> http://focus.ti.com/lit/ds/symlink/opa121.pdf Difet opamp OPA121 inputs are specced to withstand high DC voltages (-4.. <S> +15V) <S> in range of interest, even when opamp is not powered. <A> If the input current is in the pA range, it's probably a CMOS opamp. <S> For this class of opamps input bias current is often a few fA typical, a few pA maximum. <S> (for instance National LMC660 ). <S> Since the gates of the MOSFETs are insulated, input bias current won't increase if the device isn't powered. <A> If you check the data sheet for a typical op amp you will probably see that applying power to an input on an unpowered device is a bad idea.
|
When the opamp is unpowered, the positive supply is supposedly at zero, so applying more than one diode drop to any input will cause significant current to flow in that input and also sortof possibly apply power to the device thru the diode.
|
Alternator with torque proportional to angular velocity Im interested in buying or designing an alternator/electrical generator that increases torque as rpms increase - is there an alternator design that increases resistance as rpm/angular velocity increases? The alternator is for an exercise bike and would simulate real life pedal resistances. I intend to collect the energy (however insignificant it is) to do something small like charge my cellphone. <Q> What you are asking about is a generator that presents "viscous friction" to the rotating shaft. <S> That means the torque is proportional to the rotation rate. <S> Yes, this can be done with a electric generator. <S> Torque is proportional to the current, and open circuit voltage is proportional to rotation rate. <S> A resistor is the right load to get torque proportional to rotation rate, at least it will be as close as the generator is efficient. <S> For more fancy control and to harvest the energy, you need what is essentially a switching power supply on the output of the generator. <S> The controller of this power supply has a input that tells it the rotation speed. <S> It can then put any load it wants to on the generator output to get the desired torque. <S> Usually the generator output is bucked to a higher voltage. <S> This voltage is not regulated except that it gets whatever power the generator puts out at the current operating point. <S> Another switcher then keeps this voltage within some range so that the first switcher can continue to dump power onto it as it sees fit. <S> The second switcher usually produces a well regulated lower supply. <S> It can also handle excess power in various ways, including shunting it to a resistor when there is nothing else to do with it. <S> If you want to get really fancy, you could dump any excess power back onto the AC power line. <S> However, that would take considerable electronics, has regulatory and safety issues, and will certainly not be worth it economically. <A> My inclination, if you weren't worried about harnessing energy, would be to use some big MOSFETs to switch current to some resistors or light bulbs and then use a microcontroller to control the MOSFETs. <S> The basic idea would be to pulse-width-modulate the MOSFETs to vary the load based upon what the processor determines is appropriate. <S> To avoid having massive flyback pulses kill the MOSFETs, I would suggest having a number of parallel resistor-plus-MOSFET combinations, plus a resistor which is "always in", and a circuit to turn a MOSFET full on with a moderate-value resistor if the voltage gets too high. <S> If you switch from having 10 ohms of effective resistance to having 25, that will cause an instant momentary 2.5x increase in voltage; the MOSFETs should be chosen to allow for that. <S> I would expect that monitoring and controlling the effective mechanical resistance is going to be easier if one can switch the MOSFETs and resistors in such a way that you never hit a flyback clamping point. <S> If your allowable flyback voltage is 5x <S> the voltage produced by cranking, it may be good to have two MOSFETs and resistors per decade (e.g. values of 1, 3.3, 10, 33, and 100 ohms). <S> To produce effective resistances between 33 and 100 ohms, turn the 100 ohm resistor on and modulate the 33-ohm resistor. <S> This should limit the flyback voltages to 3x the voltage produced by cranking. <A> Because the voltage a DC machine produces (the back EMF) is proportional to it's speed, if you simply hooked up an electrical resistor to it, it would also translate into mechanical resistance on the input shaft. <S> If you want varying resistances, you could alternately connect varying resistors across the motor to create additional load. <S> Bear in mind <S> that if you want to dissipate the energy that a human can put out for a good length of time you will need to absorb several hundred watts of power, which will mean using quite a large motor, and some non-trivial resistors. <A> Come on people does everything here have to be about quantum equations <S> do we forget the fun of the pure mechanical-physics…. <S> A 200 Lb Pearson standing on a 6” pedal will suplay 100 Ft.-Lb. <S> for a 1/4 of a turn hence 2 pedals the force is applied ½ the time. <S> “There’s a formula to calculate the exact torque. <S> Because the force is applied in a circular motion from a single direction like a piston it’s actually little less.” <S> Now assume the main sprocket is ½ the diameter of the pedals that will make the torque 200 <S> Ft-Lb. <S> Ok so connect it to a 3” sprocket in the back this will double it yet againSo <S> the force is 400 Ft-Lb to the generator. <S> So “1 watt ≈ 44.25372896 ft-lb/min” <S> now in a very crude calculation a 200 lb person can generate 10 Watts per minute <S> This is way more than whets necessary for my phone charger. <S> It takes .5V DC, 850 mA <S> Now as far as increasing the load as you pedal faster. <S> It relay does take some electronic tinkering to design a variable load and some feed back to control it.
|
Light bulbs are cheaper than power resistors, and they're designed to dissipate power (given a suitable enclosure), but their resistance varies substantially with temperature, which could complicate your logic.
|
VGA to HDMI converter I'm making a little project purely as a hobby using a small FPGA breakout board.I'm outputting VGA video which works fine, but would like HDMI output instead so I can connect to a television that has that input but no VGA. I don't believe it's practical to generate HDMI directly from my FPGA board (correct me if I'm wrong...) so I was wondering if there was any encoder chip or board that doesn't cost too much that I could use in the project that takes VGA input (and possibly audio) and outputs a HDMI signal. I've not been able to find anything myself. Any ideas? Or is this impractical and I should settle for analog video instead. EDIT: Basically I'm asking if there is any way to generate HDMI that's possible for a small personal FPGA project, the VGA part isn't necessary if there is another way. <Q> HDMI is just "DVI with knobs on" on the video side. <S> As to "can it be done"... <S> my first question is "what FPGA"? <S> Some of them can create HDMI/DVI signals with the IO blocks, others just fundamentally can't. <S> DVI uses TMDS signalling, which is an encoding on top of a Current Mode Logic (CML) differential pair. <S> CML is actively pulled down by a current source for a '0' and floats high with a termination resistor at the far end for a '1'. <S> Then you have to encode and serialise the data. <S> TMDS describes how to encode the data bits, and then you you "just" have to serialise the data bits across the data pairs. <S> The specification can be found here - see section 3: Digital Visual Interface Spec <S> The TFP410 chip data sheet also has a reasonable description of what goes on: TFP410 - TI PanelBus™ DIGITAL TRANSMITTER <A> I just found this VGA to DVI converter <S> ( Hackaday article ). <S> It converts VGA-compliant R/G/B and sync pulses into DVI, which basically uses the same signaling as HDMI. <S> You'll need the ability to generate a reliable pixel clock, as well as to be able to send bits at 10x the pixel clock. <S> Using an FPGA's DCM (digital clock manager) <S> you should be able to accomplish this. <S> I haven't had a chance to test this code myself (am in the process of adopting one of my VGA-based projects to try it) but it has worked fine for others. <S> Edit: I was able to successfully integrate this into several of my VGA demo projects with little to no difficulty. <S> So I can personally attest to the fact that this works. <A> VGA to HDMI is tricky at best. <S> Simply because VGA is analog and HDMI is digital. <S> You'd need to capture each frame of the VGA signal, digitize it, store it in a frame buffer, and output the HDMI stream. <S> While possible, it's not going to be 'simple'. <S> You can buy some external boxes that supposedly do VGA to HDMI, but I don't know how good they are. <S> There's some on ebay. <A> I don't understand all the discussion about specs, chips, development boards and building a converter from scratch. <S> Just go to Amazon and buy one <S> -- there are several, for example: Sewell Hammerhead VGA to HDMI Active Converter 1080p Compact Size OREI XD-600 VGA PC/Laptop to HDMI Video Converter -Upscaler <S> Up to 720P/1080P Converter with Audio Jack <S> HDE VGA w/ Audio to HDMI 1080p <S> Converter Box w/ DC <S> Adapter <S> I gather <S> since this is a hobby project it is a one-off, and not something to be sold. <S> So it would be much simpler just to buy a ready-made box than build one from scratch, and have to get it working properly. <A> Check out this development board from lattice. <S> They have a reference design that includes dvi in and dvi outputs as well as cameralink inputs and outputs. <S> http://www.latticesemi.com/products/developmenthardware/developmentkits/machx02controlkit.cfm <S> $189 <S> It demonstrates using an lvds cameralink deserializer to drive a dvi output chip. <S> Depending on your Fpga board you may be able to output the cameralink serialized video from your board then use this dev board to convert that to a dvi output. <S> Or just use this board for the whole thing. <S> If your budget is larger they have a nice Hd camera to hdmi output dev kit for $400. <S> http://www.latticesemi.com/products/developmenthardware/developmentkits/hdr60videocameradevkit/index.cfm
|
It might be emulatable for a hobby project by using a bidirectional LVDS pair driven low and using the tristate line to drive and release (a bit like doing an open-drain drive).
|
Circuit only works when o'scope probe is connected I am working on a circuit with a PIC mcu and an LCD display on a breadboard. The PIC communicates with the display via I2C. For some reason I can only get the communication to work properly when my o'scope probe is connected to the SDA line of the I2C interface. I know the communicate is working because the display shows my text. If I remove the probe and then restart the firmware the display stays blank. Any ideas what is happening here? I am using 4.7k pull-up resistors on both the data and clock lines of the I2C as recommended by the PIC manual. I also tried swapping the resistors out with 1k's and 10k's but that didn't seem to help. Also, it doesn't matter if the ground clip on my probe is connected to the PCB ground or not (so I don't think it's a grounding issue). Any ideas what is going on here? I know I should really get a real PCB made but I wanted to prove the design was good on the breadboard first. <Q> You didn't say which PIC you are using, but there have been at least two confirmed bugs in the IIC mode of some of the MSSPs. <S> One of them had something to do with sampling the ACK bit on the wrong edge of the clock. <S> The bugs I know about were quite a while ago, so they have probably been fixed in newer PICs. <S> Still, I tend to use all firmware implementation of IIC in master mode. <S> You need the hardware to implement a slave, but in master mode you own the clock <S> so it's easy. <S> The scope probe is adding a little capacitance to the SDA line, which slows down and delays the edges slightly. <S> The slower or delayed edge of SDA is apparently making it work. <S> Try putting a 22-47pF cap on the line and see if that appears to make it always work. <S> I wouldn't want to ship it that way without knowning the cause and understanding <S> that's a reliable fix though. <S> Another possibility is noise getting onto the line from other parts of the system. <S> In that case the cap to ground is making the line a little lower impedance, which attenuates the noise a bit. <S> If it's noise, then adding a cap to SCL will be fine. <S> If there is a race condition somewhere, adding a cap to SCL will probably make it worse. <S> I've seen too many flaky things with IIC over the years so that now I use a firmware-only implementation whenever possible. <S> In firmware I can guarantee that there are never two things happening on any one edge. <A> Each scope probe has some inherent capacitance to it. <S> I2C has a specification for maximum allowable bus capacitance ( 400 pF ). <S> Adding the scope probe (& its capacitance to GND) <S> to your trace may be causing your circuit to fall more closely within a "sweet spot" that the IC is looking for within the specified capacitance. <S> Read the scope probe, most of them usually list their capacitance. <S> Then, put a capacitor to GND of the same (as similar as possible) on the SDA line and power up your circuity without the probe connected, and see if that fixes it. <A> I found that you cannot program a PIC using a PICkit 2 unless you connect the ground and Vdd to the chip (I was young!) <S> but that when the scope ground was connected it provided a ground to the earth pin which connected to my laptop's earth pin, which went through a class-Y cap to join the PSU's ground to earth, connecting USB ground to earth. <S> I found that even with this configuration I would often get programming errors, because it was a very noisy path. <S> But it just about worked, much better than not at all. <A> What load is the scope/probe set for 1Mohm? <S> 50ohm? <S> make sure it is not on the low ohm setting. <S> Are you the i2c master or depndent? <S> One way to debug it might be to sample the input and echo it out another <S> i/ <S> o <S> pin that you are not using, then probe that with the scope. <S> The controller is not that fast <S> but you can get an idea of what the processor core is actually seeing when it samples that input. <A> I had a UART problem at cold temperature where the probe fixed it when it was touching the flow control line. <S> Long story short - it was because the software wasn't setting the pin muxing properly where on a LOW it was basically a weak pull down. <S> At low temperatures the leakage current gets lower so the weak pull down is more like an open. <S> This is bad because the fall times will be impacted and also it could pick up all kinds of noise and cross talk. <S> There's no way to se what's happening because the measurement instruments impacts the result. <S> Adding a 1Mohm resistor helped pull the line LOW. <S> You could try the same troubleshooting method by soldering a 1M to GND, and see if that fixes it, then check the pin muxing in the software.
|
If you haven't got a good ground connection between the display and your microcontroller, the ground clip of the probe can give you a "good enough" ground to allow I2C to work. Something is probably not right in that there is a race condition in there somewhere.
|
Heat transfer in an airtight enclosure (component -> internal ambient -> heat sink on case?) Are the thermal resistance values of heat sinks/etc reasonably reversible? If I wanted to dissipate an internal ambient temperature rise, could I use an internal heat sink on the enclosure wall coupled to a heat sink on the outside to pull heat out of the box? Can I use the same thermal resistances of the heat sinks in reverse to calculate how efficient this will be at extracting heat from the air? (Assuming additional resistances of sink #1 -> wall and wall -> sink #2) I'm developing a product that must be in a sealed enclosure. I'm also trying to have the highest operational temperature range I can (targeting -40 to 85 °C) I'm sinking some heat sources (power bricks, power op amps) directly to the metal enclosure... however I am not currently able to sink the heat from the CPU itself (~1W) directly to the case. I -could- add a wide temperature range heat pipe, but I'd like to avoid it. We're not talking about that much heat, but I just don't have much room near the top of my targeted ambient range (105 °C Tj and 85 °C Ta)... The CPU would require a heat sink even if it wasn't enclosed. EDIT Okay, to clarify I guess I'm asking if a heat sink's thermal resistance is simply a thermal conductance value that I can use both as a number of how much of a temperature rise a certain amount of power put into it will cause AND the exact opposite in how much power it will conduct given a specific temperature differential across it. (So, a 10 °C/watt heat sink would for 1 watt be 10 degrees above ambient, and that if ambient was 10 °C ABOVE the sink it would pull 1 watt of heat out of ambient) ... Say my CPU is putting out 1 watt of power. With a heat sink, the Tjs (die to sink) is 0.5 °C/watt, and the heat sink's Tsa (sink to amibent) is 9.5 °C/watt, for 10 °C/watt total. My CPU is then 10 °C above the immediately nearby internal ambient temperature. My internal ambient is getting about 1 watt of power sunk into it, and that power needs to go somewhere. There is about 3" between my CPU heat sink and the enclosure wall. With a thermal conductivity of air at 0.025 W*(m^-1)/°C, there would be a rise of about 3 °C across those 3" to transfer 1 watt. (right? This seems too low to me. Also, it's ignoring any kind of internal convection effects) Call this 3 °C/watt = Tax (internal ambient transfer). The path is then normally Tac (resistance internal ambient to case) plus TcA (case to external ambient). If I put heat sinks on the inside and outside, I then have Tah (internal ambient to internal heat sink), Thc (internal heat sink to case), TcH (case to external heat sink), THA (external heat sink to external ambient). (although really I also have Tac and TcA still, for the case surface area not covered by heat sinks)) Say Tah is 3 °C/watt, Thc is 0.25 °C/watt, TcH is 0.1 °C/watt, and THA is 1 °C/watt. I have 1 watt of power for Tjs, Tsa, Tax, Tah, and Thc. I additionally have up to 5 watts of power sunk into the case from other sources directly attached to the case. So, I have 6 watts of power for TcH and THA. Can I then say that my external ambient maximum is: Tjs + Tsa + Tax + Tah + Thc + TcH + THA, or 0.5 + 9.5 + 3.0 + 3.0 + 0.25 + 0.6 + 6.0 = 22.85 °C/watt? With Tj of 105, my maximum external ambient would be 82.15 °C? <Q> Yes, thermal resistance/conductance is a two-way street, for passive conductive cooling devices, with equal speed limits posted for both directions. <S> OK, now for the caveats: If your heat sink is rated for forced air flow, and you have no forced air flow, then the heat flow through the heat sink will be different. <S> Your heat sink is rated for certain conditions at both ends. <S> If you meet those conditions (at both ends), except for reversal of temperature conditions, then you can expect equal but opposite heat flow. <S> Say that your 10C/W heat sink is rated 10C/W for conduction of heat from a 1W source to still air, with the contact area with air being fins. <S> Now, you put those fins INSIDE your enclosure, in contact with still air, and you place the outside end of your heat sink in contact with a device (say, a cold plate) that will keep that end of the heat sink 10C cooler than the inside air. <S> In that case you will get 1W of energy flow from the fins of the heat sink to the cooling device (cold plate). <S> You would want to pay attention to such things as: Warm air rises and cooler air falls. <S> Air fins are most effective when hot air can rise from them and allow cooler air to come in contact with the fins. <S> Cooling fins, on the other hand, would be more efficient when placed such that cooled air can fall away from them. <A> How much heat is drained to the environment is determined by two factors: the difference in temperature between the heat source and the environment and the total thermal resistance between those. <S> One of the elements in this thermal resistance chain is the resistance between the air inside your enclosure and the enclosure's wall. <S> If you place a heatsink inside the enclosure this thermal resistance is replaced by the sum of two new resistances: that between air and heatsink and that between heatsink and enclosure wall. <S> Intuition says that the former will be lower because of the larger contact area. <S> And the thermal resistance between two metal object in close contact is also much lower than that between air and an object. <S> Thermal paste between heatsink and enclosure reduce this resistance further. <S> edit <S> OK, I see what you mean by reverse, and the answer is yes. <S> A difference in temperature will cause heat to flow from the higher to the lower temperature. <S> Peltier coolers use this to force heat away from the source of the heat by creating a much lower temperature. <S> If you can't draw no more heat away the temperature difference will decrease and reach an equilibrium state when the temperature is equal everywhere. <A> A fan and heat sink inside the case, coupled to a fan and heat sink outside the case appears to be a viable design. <A> I wonder if a peltier cooler bolted inside the case might help by reducing internal temperature. <A> Thermal resistance is not a fixed value: it depends, amongst other things, on the air temperature. <S> So technically, the heatsink inside the box will not have the same thermal resistance as the heatsink out side the box, unless the temperature inside the box is the same as the temperature outside the box. <S> Which it cannot be, to get any heat flow at all. <S> This is a second order effect. <S> Thermal resistance will depend mostly on air flow.
|
All in all, the total thermal resistance will be reduced by adding a heatsink on the inside, but you can't just calculate it as a mirror image from the one on the outside. I do not have any numbers to offer, but can tell you that there are systems that do exactly what you propose.
|
Selecting voltage rating for capacitors In my project I want to use some ceramic and electrolytic capacitors, I will need the capacitors be at least 10V rated, but what will happen if I use much higher rated capacitors (just to make sure in case of something went bad they don't explode!)? <Q> In general, the voltage rating of a capacitor is the maximum it can take and still stay within specs. <S> Polarized caps, like electrolytics and tantalum, can take any voltage from 0 to the voltage spec value. <S> That said, different things happen to different cap types as their voltage gets near the maximum. <S> With electrolytics, the lifetime goes down. <S> In theory with a reputable manufacturer, the rated lifetime is at max voltage and temperature unless stated otherwise. <S> You could therefore say the lifetime goes up if you operate the cap below its rated max voltage. <S> The two major stressers of electrolytic caps are voltage and temperature. <S> Large currents can also hurt them, but this is due to heating so is really a temperature issue. <S> Ceramics have different properties. <S> Voltage doesn't effect lifetime of SMD multilayer caps much, assuming of course you don't exceed specs. <S> Some ceramics however do not linearly store charge as a function of applied E field. <S> They hold less additional charge for the same voltage increment at high voltage than at low voltage. <S> This means the apparent capacitance goes down with voltage. <S> The cheap ceramics, particularly those with "Y" in their names and a few others exhibit this effect more strongly than others. <S> If you are just bypassing a digital chip, this doesn't matter much. <S> If however the cap is used in a analog filter, then this probably matters and you generally want to stick to ceramics with "X" in their name and look over the datasheet carefully. <S> There are issues with too low a voltage too, especially with electrolytics. <S> They work on a thin oxide layer on the alumimum. <S> This can get degraded when there is no charge accross it. <S> So to finally give you a concrete answer <S> , if you are going to use electrolytic caps try to aim at running them around 3/4 or 2/3 of their rated voltage. <S> It's OK to have occasional spikes up to the maximum, but don't ever exceed it. <S> It's OK for them to be off too, but it's better that they're not completely discharged for years on end. <A> You can always safely use caps rated at higher voltages. <S> The only reasons not to are that they may be larger and more expensive. <S> See also answer to this question . <A> The value might be different from the marked value, if the operating voltage is much lower than the marked voltage, especially for electrolytic capacitors. <A> For ceramics, if 10V is your goal you will probably end up with 50V parts because that's a common voltage. <S> I don't bother to have anything else on hand. <S> For electrolytics, 25V is a common voltage. <S> Should be fine.
|
Unpolarized caps, like ceramics, can take any voltage +- the voltage spec value.
|
Ground as Heat Sink First of all, this might be a newbie question, but I'm just getting into hobby electronics and robotics, so this is all very new to me. Anyways, I'm looking at the data-sheet for an L293 IC, which shows some of the pins as "heat sink and ground". Does this mean I should avoid using this IC on breadboards, or is it something I don't need to worry about? Should I add a heat sink to the IC to further reduce heat issues, or does the pin heat sinks do the job fine? The IC is going act as a H-Bridge and run motors with the following characteristics: Voltage = 12 VNo-Load Current = 3.5 mANormal Load Current = 120 mAStall Current = 360 mA In case you think the IC is a bad choice, I should let you know that electronic components aren't easily accessible where I live, which means I take the closest I can get. <Q> I've moved this comment to an answer as suggested by @stevenvh (and added a little more information). <S> I don't know much about motor drivers <S> but we used a similar IC for a project and drew a lot of current (at 7.4V), maybe even more than your stall current for a majority of the time. <S> We used three of these motor drivers simultaneously. <S> We had them on a breadboard before we moved to a PCB and everything seemed to work fine although the chips did get very warm. <S> We didn't have any heat sink near those ground pins either. <S> I think you'll be okay <S> but if you can add a heat sink I would. <A> Datasheet clearly says, that maximum 'freeair'(breadboard case) <S> power dissipation is ~1.5-2W.At 360mA, and 1.8V maximum voltage drop inside chip, you are dissipating 0.648W per channel. <S> So, if you will have 3 or 4 channels in high state, you are exceeding your maximum power budget. <S> While reaching this is probably not very likely (all channels connected to stall motors), but if you want to be safe, add some heatsink to this chip, and you'll be fine even on breadboard. <A> With regards to what dhsieh said, they got really warm. <S> If you touched them unexpectedly it was hot enough to startle you. <S> Factor in the fact that we were running at or below 7.4V as or battery drained <S> and I would be a little worried about touching them when running at 12V. That said on page 7 of the datasheet <S> they show some application information which indicates a thermal shutdown inside the IC. <S> I'm guessing you'll be fine to try it and the chip will just shutdown <S> if it get's too hot, then you'll know you need a heatsink. <S> But like I said above, be careful touching it. <A> I think that would be that they can be connected to ground and a heatsink. <S> Not that those pins would act as heat dissipation. <S> So if the chip needs a sink, you need to press one on there since those pins will probably not affect the heat at all . <S> Apparently those chips has thick internal copper connections that can lead the heat out from the chipcore so they are a little bit better than the ones that dont have that feature. <S> And normally you would put paste on top of the chip and then press the sink down, but in this case it sounds like it is more important that the sink is connected to those magic pins.
|
Since they save some pins in the centre it looks like they like to "clip on" a heatsink over those pins that press down on the IC.
|
What's so complex in handling batteries that there's battery attendant specialty? Dictionaries say there's "battery attendant" specialty. The name implies that's a specially trained person who somehow services batteries while they are in operation. AFAIK the only kind of batteries that require maintenance is flooded lead-acid and the actual maintenance is adding distilled water. I don't see how this would require a designated specialty. What's so complex in handling batteries that a designated "battery attendant" specialty exists and what does such specialist actually do? <Q> Depending on the technology, when not handled properly, batteries may catch fire or even explode. <S> They also almost always contain hazardous chemicals. <S> I think the battery attendant would be a person to take care of large battery installations. <S> A hospital for instance has backup power from diesel generators, but needs (much!) <S> battery power to bridge the time between a power outage and the moment the generators are up and running. <S> Power to intensive care units and operating rooms in particular should never be interrupted. <S> (I've worked in a hospital which, as a first measure, received power directly from two different power plants.) <A> Also on aircraft most batteries are composed of Ni Cad cells. <S> The must periodically be tore down, cleaned, tested, reassembled and recharged before being placed back in service. <S> Poorly maintained batteries generate hydrogen gas which can build up and with the right (or wrong depending on how you look at it) mix of oxygen and a spark exploded. <S> Battery explosions almost always cause fired usually resulting in the loss of the aircraft and unfortunately who ever is on board at the time. <S> Typically specially trained electricians maintain the batteries. <A> I'm guessing this job description came from the early telephone days. <S> Everything was run from banks of lead-acid batteries with 48V nominal output. <S> There were enough of these installations and they required enough care that I'm not surprised that a specialty evolved to handle them. <A> The " gravity cell " was popular with telegraph networks until the 1950s. <S> It was typically assembled on-site out of a glass jar, a copper electrode,a zinc "crow's foot" electrode, copper sulfate crystals, and water. <S> How many people do you know that could look through the glass walls of a gravity cell,and see at a glance the following typical problems and fix them? <S> mixing caused by too little current draw mixing caused by too much current draw cell completely "drained" of energy -- time to replace the consumables. <S> The crow's foot, of course, ... and something else? <S> water level low -- did you remember the oil? <A> Imagine if you will a whole room filled with lead acid batteries, all linked together with big chunky copper bus-bars. <S> Now imagine if you will someone dropping a spanner and it shorting said bus-bars. <S> That actually happened with a friend in a Vodafone facility... <S> Not his spanner, but someone else who was working in there without the proper training. <S> It was nasty.
|
Keeping such batteries in good condition is literally a matter of life and dead.
|
What is µMAX IC package? While ordering free samples from Maxim I have often seen components packaged as µMAX . What is µMAX? When googling for it all I get are some Maxim components. If I look at google images I get an assortment of SMD and DIP package pictures that bear no resemblance to one another. Is there a comprehensive list of component packages (cross-provider) around anywhere? <Q> It's an 8 pin SMT package, about as wide as an SO-8, but just 3mm long instead of the 5mm of an SO-8. <S> This is achieved by using a 0.65mm pitch instead of 1.27mm. <S> Other manufacturers also go to ever smaller packages, but most often choose for leadless packages like DFN (Dual Flat No Leads). <S> Intersil has this list of packages. <S> NXP has this list . <S> Unfortunately my experience is that recommended footprints for a certain package aren't always consistent between manufacturers. <A> μMAX packages (Maxim uses 8-pin and 10-pin versions of this package) seems (by PCB footprint at least) the same as <S> μSOP aka MSOP <S> aka micro-SOP packages. <S> This is specified on "Package Information" of the datasheets, where there is always 8L (or 10L) uMAX/uSOP specification present. <A> <A> Just download pdf datasheet for component of interest, and in the end there will be images of all packages with exact geometric sizes & pin locations. <A> As far as I can tell, µMAX is Maxim's idiotic name for the standard TSOP package. <S> They certainly fit on the TSOP footprint from the Altium Designer software, anyways. <S> An 8-pin µMAX is pretty much identical to a standard 3*3 mm body TSOP part.
|
Comprehensive list of maxim packages here: http://www.maxim-ic.com/design/packaging/ Whereas most packages are used by several manufacturers, \$\mu\$MAX seems to be a package solely used by Maxim.
|
Faster alternatives to BC847C / BC859C I wonder, are there faster jelly-bean alternatives to BC847/859 (i.e. low-voltage, low-power, SOT23, high-speed transistors)? BC847/859 are rated for 100Mhz, something at 250-500 (and even 1000) would be great :-) When I go to catalogues, they have RF section where everything is too fast(9Ghz+) & too expensive :-D <Q> From Infineon: SMBT2222A : \$f_T\$ > 300MHz, complementary: SMBT2907A <S> SMBT3904 : \$f_T\$ > 300MHz, dual NPN, low power, complementary: <S> SMBT3906 <S> From NXP: BF570 : <S> \$f_T\$ > 490MHz <S> BF840 : \$f_T\$ <S> typical 380 MHz <S> BFS20 : <S> \$f_T\$ <S> > 275 <S> , typical 450 MHz <A> The BC847/859 are not rated for 100 MHz operation, that is the transition frequency at which they deliver unity gain. <S> Devices are usually used at 1/10 of ft. <S> You won't find anything as cheap, the BC847/859 are intended for audio and similar low-frequency applications. <S> I got 500 of them very cheap on Ebay. <S> :) <S> The 2N5179 is still made, and I've also used that, but it's quite expensive. <S> Someone gave me a SPICE model for the BF959, which is useful <S> (On Semi didn't have one). <S> One is readily available for the 2N5179. <A> Good and inexpensive standard parts that have advantages for fast switching applications are the 2N3904/2N3906 complementary transistors. <S> Also available as SMDs (PMBT3904/PMBT3906 are SOT23). <S> The data sheets mention a transition frequency of 250 or 300 MHz: http://www.nxp.com/documents/data_sheet/PMBT3904.pdf <S> http://www.nxp.com/documents/data_sheet/PMBT3906.pdf <S> For switching ("digital") applications, transistors like these are favorable over small-signal RF BJTs, because they tend to offer both speed and some (peak) current. <S> If you look at the peak collector current specification, you'll see that some RF types have a peak collector current in the range of 25...100 mA while the 3904/3906 types have a rating of 200 mA. <S> The 2N2222/2N2907 or SMBT/MMBT2222/2907 types are even rated at 600 mA. <S> If you need a lot of current, the FMMT491/591 might be a reasonable choice, but these aren't exactly everyday cheapo types.
|
If you can find them, the On Semi BF959 is probably the nearest equivalent for high-frequency designs.
|
Why are CPUs becoming smaller and smaller? It is a known fact that over time processors (or chips) are becoming smaller and smaller. Intel and AMD are in a race for the smallest standards (45nm, 32nm, 18nm, ..). But why is it so important to have the smallest elements on the smallest chip area? Why not make a 90nm 5x5cm cpu? Why squeeze 6 cores into a 216mm2 area? It will be easier to dissipate the heat from larger area, manufacturing will require less precise (and thus cheaper) technology. I can think of few reasons: less size means more chips could be made on single wafer (but wafers aren't very expensive, right?) smaller sizes are important for mobile gadgets (but everyday PCs still use tower boxes) small size is dictated by light-speed limit, the chip can't be larger than the distance an EM field can travel in 1 cycle (but thats approximately several cm at 3GHz) So, why do chips need to become smaller and smaller? <Q> It's like candy bars. <S> They keep making them smaller at the same price to increase profit. <S> Seriosly though, there are good reasons for smaller chips. <S> The first and foremost is that more chips can be fit onto a wafer. <S> The cost to process a wafer is pretty much fixed, regardless of how many chips result from it. <S> Using less of the expensive wafer is only one part though. <S> Yield is the other. <S> All wafers have imperfections. <S> Think of them as being small but randomly scattered about the wafer, and any IC that hits one of these imperfections is trash. <S> When the wafer is covered by lots of small ICs, only a small fraction of the total are trash. <S> As the IC size goes up the fraction of them that hit a imperfection goes up. <S> As a unreal example that nonetheless points out the issue, consider the case where every wafer has one imperfection and is covered by one IC. <S> The yield would be 0. <S> If it were covered by 100 ICs, the yield would be 99%. <S> There's a lot more to yield than this, and this is greatly oversimplifying the issue, but these two effects do push towards smaller chips being more economical. <S> For really simple ICs, the packaging and testing cost dominates. <S> In those cases, the features size is not so much a driving issue. <S> This is also one reason we have seen a explosion of smaller and cheaper packages lately. <S> Note that extreme small features size is being pushed by very large ICs, like main processors and GPUs. <A> As the process size gets smaller, power usage decreases. <S> The reason for this is that as the transistor gate gets smaller, threshold voltage and gate capacitance (required drive current) gets lower. <S> It should be noted that as Olin pointed out this level of improvement doesn't continue to smaller process sizes as leakage current becomes very important. <S> One of your other points, the speed at which signals can travel around the chip: <S> At 3ghz the wavelength is 10cm, however the 1/10th wavelength is 1cm which is where you need to start considering transmission line effects for digital signals. <S> Additionally remember that in the case of Intel processors some parts of the chip runs at twice the clock speed <S> so 0.5cm becomes the important distance for transmission line effects. <S> NOTE: they may be operating on both clock edges in this case, meaning the clock doesn't run at 6Ghz but some processes going on are moving data that fast and have to consider the effects. <S> Outside transmission line effects, you also have to consider clock synchronization. <S> I don't actually know what the propagation velocity is inside a microprocessor, for unshielded copper wire its like 95% of the speed of light but for coax is like 60% the speed of light. <S> At 6Ghz the clock period is only 167 picoseconds <S> the so high/low time is ~ 84 picoseconds. <S> In vacuum, light can travel 1cm in 33.3 picosends. <S> If the propagation velocity was 50% the speed of light then its more like 66.6 picoseconds to travel 1 cm. <S> This combined with the propagation delays of the transistors and possibly other components means that the time the signal takes to move around even a small die at 3-6Ghz is significant for maintaining proper clock synchronization. <A> The main reason is the first one you mentioned. <S> Wafers (what you call plates) are very expensive, so you want to get the most from them. <S> Earlier wafers were 3 inch in diameter, today's are 12 inch, which not only gives you 16 times as much real estate, obviously, but you get even more dies out of them than that. <S> So it's clear that they would use this technology also for CPUs used in tower PCs, even if it doesn't look like it's necessary there. <S> And don't forget that laptop PCs also have this kind of CPUs, and they are on a budget as far as space is concerned. <S> Speed is also a concern, at 3 GHz signals travel less than 10 cm per clock cycle. <S> As a rule of thumb from 1/10th of that we have to take care of transmission line effects. <S> And that's less than 1 cm. <S> edit Smaller feature size also means less gate capacitance, and this allows for higher speed. <S> Faster switching means less power consumption, since MOSFETs will go faster through their active region. <S> In practice manufacturers take advantage of this to clock faster, so that in the end you won't see much of this power reduction. <A> The CORE reason why CPUs keep getting smaller is simply that, in computing, smaller is more powerful : To a first approximation, computation involves two basic actions: transmitting information from one place to another, and combining strands of information to produce new information. <S> Since we are used to using electronics here, let's call the hardware for these actions 'wires' and 'switches'. <S> For both of these, smaller is better: Wires: <S> Since the speed of transmission on a wire is essentially constant, then if you want to get information from one place (e.g switch) to another, you have to shorten the wire . <S> (you may be able to achieve a faster speed, but eventually you hit the speed of light limit, at which point you are forced back to shortening). <S> Switches: <S> A switch works by information from one or more input wires entering and suffusing the body of the switch, causing its internal state to transform so as to modulate the information on one or more output wires. <S> It simply takes less time to suffuse the body of a smaller switch.
|
For large chips, the cost is all about what fraction of a wafer it uses. Smaller transistor processes allow the use of lower voltages combined with the improvements in construction technique mean that a ~45nm processor can use less than half the power that a 90nm processor uses with similar transistor counts.
|
Zener diode in series? Is it OK to use a zener diode in series? Normally I see them across the output of a power source to limit the voltage to the value of the zener. Putting it in series instead should reduce the voltage by the value of the zener, yes? (so a 5V zener on a 12V supply would give a 7V output). The thing is, I have a -12V supply and I need a -5V (or near, -7 will do - I have added a couple of Si diodes to drop the voltage a little more) supply from it - BUT - I don't want to lose the -12V supply (that is needed for other parts of the circuit), and I don't have any negative regulators, only positive. I have tried it and it appears ok, but will it cause problems if I leave it like that? <Q> A regulator would be the best solution, but a zener (one 'n') is ok, at least if you don't want to draw too much power from the regulated voltage. <S> You don't place the zener directly on the -12V, but use a series resistor to limit the current. <S> It's this series resistor which dictates how much current you can draw. <S> The -12V will still be available. <S> To calculate the resistor value, you have to know how much current your load will draw. <S> Suppose this is 1 mA. <S> Also suppose the zener diode needs 10 mA. <S> That's 11 mA through the resistor. <S> Voltage drop is 12V - 5V = <S> 7V. <S> Then R = 7V / 11 mA = <S> 640\$\Omega\$. edit <S> You may think that in my example I'm exaggerating a bit to have 10mA for a zener if the circuit would require only 1/10th of that. <S> But you'll find that zeners are often specified at much higher currents, like 50mA. Certain newer zeners are specified at much lower currents, these ones only need 50\$\mu\$A. YAE (Yet Another Edit) <S> This diode for instance is specified at 5.1V @ 50mA, but only drops 1V at 10\$\mu\$A. <A> In a couple rare instances, I've used a Zener diode to get a supply voltage down to the input range of a regulator. <S> For example, I had an off-the-shelf transformer that was working fine in a switching power supply design except for the bias power winding, which had too many turns and generated 34 volts after rectification. <S> Since bias current requirements were rather low (a few milliamps), and bias voltage required for the chip was 12-32 volts (it has a built-in linear regulator), the previous engineer used a resistor and 30 volt Zener combination like the one in stevenvh's answer. <S> While the circuit worked, its quiescent current was twice what was expected. <S> We looked at the circuit with the thermal camera (Fluke Ti25--awesome tool if you can afford it) and the Zener was glowing hot. <S> So we changed the circuit to use a series 10 volt Zener, reverse biased, to get the voltage down to about 24 volts, below the chip's maximum. <S> The built-in regulator does the rest. <S> We just had to make sure there was a minimum amount of current through the diode, but that wasn't difficult. <A> Safest you handle the situation by setting the Zener parallel with the load. <S> You calculate the resistor and Zener combination so that when your load takes zero Amps, your zener is able to handle the current without burning. <S> Example your load takes from 0mA to 100mA current. <S> Supply Voltage is <S> 12V. Needed Voltage is 6V. <S> You need a Zener Pd=5W to handle the current when load takes no current. <S> Resistor value is (needed Voltage drop)/(max load current) so in this case 6V/0,1A = 60 ohm. <S> Check the power dissipation of the resistor (PdR). <S> PdR = <S> Iload <S> * Iload <S> * R = 0,1A <S> * 0,1A <S> * 60 ohm = <S> 0.6W. <S> So 6V 5W Zener and 60 ohm 1W resistor are needed exact values without the the fear of overheating. <S> You can use a resistor value of 56 ohm(easier to find from shop). <S> The Zener can handle the extra current.
|
The reason why you don't want to use the zener in series to get the voltage drop is that especially at low currents the reverse voltage may be much lower than the rated value.
|
Can capacitors be replaced with higher capacitance ones? I have a bad capacitor on this LCD power board. I have rev 2.6, and I noticed there is a rev 2.8 board that looks like it didn't get skimped on in parts. Can I install capacitors in my rev 2.6 board as they are used in the 2.8? My board: 9 x 470μF , 1 x 100μF, 1 x 22μF, 1 x 10μF Here's the rev 2.8: 6 x 1000μF , 4 x 470μF, 1 x 100μF, 1 x 22μF 1 x 10μF The differences are circled in blue. My board seems to use lower rated capacitors, is missing one, and the microchip looks slightly different. The pictures are of the entire boards. <Q> The IC that you are pointing out is a 3W (or maybe a 2x 2W) <S> stereo audio power amp. <S> The one in rev 2.6 is a SM7496L, which I am having a hard time finding any info about, but it is essentially the same thing as rev 2.8 just a different manufacture, so it is probably safe to assume that circuit should be the same around them. <S> This means it should be safe to put a larger capacitance value on it. <S> For the group of 3 with one missing, it looks like these are next to something that is probably a voltage regulator with a big heat-sink on it. <S> If this is the case, go read the answers for this question Capacitor Sizes for 7805 Regulator <S> For the other 2 capacitors, these look like they could either be used with the 2 diodes or could also be for the voltage regulator. <A> An important consideration is the 'type' of capacitor you intend to replace with. <S> You should seek out the specifications of the part you wish to replace and make sure that your replacement is comparable in terms of ESR, ripple current rating and rated life. <S> If you inadvertently substitute a general-purpose capacitor (which usually has its ripple specified at 120Hz, and often doesn't specify its ESR) where a low-ESR, high ripple capacitor was (which usually has its ripple specified at 100kHz, and does specify its ESR), you're potentially going to be in for a smelly surprise... <A> If the capacitance is between a transformer and a rectifier you need to take into account the conduction angle at the transformer. <S> More C means smaller angle, means higher charging currents, more stress on the rectifier, so a higher current rating required for it; and a shorter duty cycle also means more energy to be dissipated as magnetism during the off cycle, so more magnetism, so more magnetic hum present to be coupled into the circuit via induction rather than conduction, ... and <S> the transformer can only have been designed with a certain set of these parameters in mind. <S> Small increases may be safe, large ones not. <A> This is the limiting factor of a capacitor due to dielectric breakdown voltages that the manufacturer chose. <S> Varying capacitance gets a little trickier. <S> If the property of capacitance is used for power supply filtering, then it is generally fine to increase the value. <S> I would verify that all the traces are the same and all the components are the same. <S> If so, it looks like they decided that it needed higher values. <S> This could be the reason your cap failed in the first place.
|
You can almost always replace a capacitor with one of a higher voltage.
|
power source affecting my IR receiver I have an arduino with a simple 3-pin IR receiver like you would find at radio shack. The receiver works fine when the power supply to the arduino is 6 AA batteries, but not when the power supply is a wall plug with 7.5V DC. It seems to work for a few seconds, then not at all. I'm using the IRRemote library here: http://www.arcfn.com/2009/08/multi-protocol-infrared-remote-library.html To further help describe the problem, the Arduino is controlling a bunch of LEDs, and when I use PWM to turn down the brightness of these LEDs, the IR receiver ceases to work as well, either when battery powered or wall powered. When the PWM is not being used and the LEDs are full on, it's not a problem. I'm not sure how to debug or fix this. I'm sure that the IRRemote is using a different timer than the PWM, and the power source exhibiting the problem when the PWM is off confirms that that's not the root cause. I suspect that there may be some kind of noise introduced by the wall power supply or the rapid switching of the LEDs. Both the receiver and all of the LEDs are tied to the same ground pin of the Arduino. What can I do to separate or filter the receiver so that it is not affected by the power supply or PWM? <Q> The datasheets of most IR receivers recommend some minimal filtering of the power supply, check for instance http://www.sharpsma.com/webfm_send/1351 <S> p 4 : 47 Ohm / 47uF. <S> Even with such power filtering I have experienced that an IR receiver on one of my boards worked much better when powered from a wall-wart+7805 compared to being powered (directly) from USB. <A> Which Arduino do you have? <S> I took a look at the Uno schematic and it uses MC33269 series regulators. <S> They'd need voltage of at least 6.35 V to guarantee reliable operation. <S> The rectifier needs additional 1.1 V, so minimal input voltage should be 7.45 V, which is pretty close to your input voltage. <S> Make sure to make input voltage measurements when the problems appear. <S> Be sure to check what the supply actually gives with a multimeter, because a small voltage drop may be enough to bring the system out of regulation. <S> Next, check the current consumption and make sure that it is within limits of the on-board regulators. <S> It should be well below 1.6 A. <S> At this point, if you still have a problem, it probably isn't related to the external power supply. <A> It sounds like your wall adapter might be unregulated, or overloaded. <S> Drawing largish pulses of current such as blinking a string of LEDs may cause its output voltage to fluctuate significantly - enough to annoy a sensitive part like an IR receiver anyway. <S> You could have the same problem when running from battery if its internal resistance is high enough. <S> As @stevenvh points out, you probably need a voltage regulator on the IR receiver to give it cleaner power. <S> An IR part that operates from 2.7V was likely designed to run from an LDO that has the 3.3V rail as its input (the output being something like 2.7-3.0V depending on the LDO), although the LDO might be powered from the battery if there's no other regulator in the design.
|
You might even see it if the power wiring is too high resistance, like your battery holder has cheap metal in the spring, wires are too narrow or too long, your board traces for power are too narrow, or some other component in the power path adds too much resistance.
|
Why is high voltage AC more common than high voltage DC? Why is that high voltage AC is more commonly found than High voltage DC? Example my battery powered fly swatter and fluorescent lamp both use high voltage AC. Why can't these devices increase the DC voltage from battery and directly use the High voltage DC? <Q> Another reason for high voltage AC has to do with arcing. <S> If an arc is formed with DC, it's very difficult to extinguish it (you need to disconnect the power source until the air gap de-ionizes). <S> In the case of AC, the arc is extinguished in each cycle. <S> Once your fly is fried, you are not left with a continuous arc. <A> All it takes to make high voltage AC is low voltage AC and a transformer. <S> To make high voltage DC, you have to chop it into (what else) AC, run it through a transformer, and then rectify it back to DC. <S> Quite a bit more hardware is necessary. <S> So, with mass produced products, there's a strong economic bias to use AC high voltage, <S> so that's what you'll see, unless there's a compelling reason that the high voltage needs to be DC. <A> DC to DC converters usually generate some form of AC to do the conversion most efficiently with a switching circuit of some type. <S> If the device can function on AC, then there is no reason to have the performance losses of converting back to DC after stepping up the voltage. <S> This is also why power distribution is via AC. <S> The voltage can be stepped up to a very high voltage, which makes the current drop for the same power. <S> This allows power to be supplied with less losses due to the resistance of the wire. <S> Then it is progressively stepped down until it gets to the 220-240 that most homes are fed (to be used as both 110 and 220 in the US and usually 220 only else where.)
|
The changing current in AC makes it possible to step up and down voltages.
|
Logic to decrement by one I am looking for some way using discrete logic (TTL) to decrement an 8-bit value by one. Basically I want to present it with an 8 bit bus with a binary number on it, and have it give an output of the input minus one on its output: eg: ...I3 I2 I1 I0 | ...Q3 Q2 Q1 Q0 0 0 0 0 | 1 1 1 1 0 0 0 1 | 0 0 0 0 0 0 1 0 | 0 0 0 1 0 0 1 1 | 0 0 1 0.... 1 1 1 1 | 1 1 1 0 I have been googling and drawn a blank for a chip that can do it, and I have been racking my brains of the best way to do it using just normal gates. Does anyone either know of a chip for this, or have some ideas on how I could construct it from discrete gates? <Q> Have you considered adders? <S> You can subtract with an adder. <S> You just need to add -1 to the input value. <S> A pair of 7483 4-bit full adders should handle your 8-bit input. <S> Apply the input X to the 'A' inputs, and apply the binary representation of -1 to the 'B' inputs. <S> Fortunately, the 2's complement representation of -1 is all ones; this means that if you just tie the 'B' inputs all high, you're done. <S> What comes out should be X-1. <A> It's not impossible to do this using discrete gates , but it isn't trivial either. <S> You can always cheat and extend on the schematic of the 74HC83 :-) <S> One other solution is to use a CPLD . <S> You'll need to know some HDL (Hardware Description Language), like VHDL. <S> In VHDL decrementing is just something like output = input - 1 and if you want to you can define input and output as 32-bit values, or even 128-bit. <S> The VHDL synthesizer will translate this into the correct logic. <S> You only need a 256 bytes part, so just any existing device will do. <S> Fill the 256 bytes with a table of 8-bit values, starting with -1 (0b11111111), then 0, 1, and so on. <S> If you supply an 8-bit value to the address inputs you get that value - 1 on the data outputs. <A> Will 255 IDs do? <S> You could implement a 255 step sequence like a linear feedback shift register using only wiring and an XOR IC such as 7486. <S> Consider your 8-wire bus to be the current state of the LFSR. <S> Shift each individual wire over one step and produce the new bit using the XOR or XNOR gates to produce the output bus for your next module. <S> Each module then has a different number in the sequence. <S> I.e.: <S> Taps: <S> * <S> *ID1: 0 0 1 <S> \ <S> \ <S> 0 <S> ^ <S> 1=1ID2: <S> 1 0 0 <S> \ <S> \ <S> 0 <S> ^ <S> 0=0ID3: 0 1 0 <S> \ <S> \ <S> 1 <S> ^0=1ID4: <S> 1 0 1 <S> \ <S> \ <S> 0 <S> ^ <S> 1=1ID5: 1 1 0 <S> \ <S> \ <S> 1 <S> ^0=1ID6 <S> : 1 1 1 \ <S> \ <S> 1 <S> ^ <S> 1=0ID7: 0 1 1 <S> The backslashes represent simple wires, and each ID is generated by a new module. <S> A completely different approach is to equip every module with a simple up counter and a single 1-bit register. <S> Wire the register's output to both the enable for the counter and the next module's register input. <S> The clock will be common to registers and counters. <S> Then clear all registers and counters, and start clocking in the opposite state through the registers; as each register switches, the counters will stop (or start), giving a unique address to each module. <S> This only requires three wires and as many clock ticks as modules. <S> I don't know if either of these fits your problem; what are you using these IDs for? <A> You could maybe chain two 74LS191 or '193 together. <S> Use the "down counter" clock input. <S> Or a single 74LS469 will do what you want, if you can find one.
|
Another way to do this is using a parallel lookup EEPROM (or EPROM, whatever).
|
Why is power supply from electrical utility referenced to earth? In response to the question Basic Training for working with 120V AC zebonaut had replied that "A regular wall outlet is referenced to earth" and this provides the close loop for the circuit to complete. Also it was stated that for safety if an isolation transformer is used the current will have to flow back to the transformer and you won't get shock by touching with only one hand. So why don't the power supply from electrical utility come from an isolation transformer? This way won't there be less risk of electrical shock from electrical appliances ? <Q> You don't want your whole house electrical system arbitrarily floating. <S> First, just a tiny leakage or static electricity could charge it up to high voltage. <S> The whole system would have enough capacitance to ground so that the discharge could cause damage. <S> Then what if any of the three lines were accidentally shorted to ground? <S> You wouldn't know anything happened if one of them got shorted, but suddenly other parts of the system are at lethal voltages. <S> Here in the US, there is a final transformer near your house, often on a utility pole at the street in front of your house. <S> That makes center tapped 220V from whatever the higher voltage feed on the utility pole is. <S> These are left isolated on the pole and all three lines brought into the house. <S> The center tap is then grounded with a thick cable to a copper pipe that goes into the ground, or to a ground stake just for that purpose. <S> This leaves two phases of 110V with opposite polarity. <S> Most circuits are connected between one of the phases and the center, called the "neutral". <S> A few special high power circuits, like for a clothes dryer or electric range, are connected accross both ends and are therefore 220V instead of the usual 110V. <S> Part of the reason for the ground setup it to deal with lightening as best as possible. <S> The system is grounded as closely as possible to the breaker panel in the house to minimize the common voltage on the system due to ground offset. <S> Imagine what would happen if the transformer secondary center tap were grounded at the utility pole instead of your house, and there was a nearby lightening strike. <S> There could easily be multiple kV offset between where the center tap ground and the actual ground potential other things in your house might be connected to, like the concrete floor you're standing on in the basement, water pipes, etc. <S> Even well insulated and properly designed appliances aren't going to protect you from that. <A> There is always several hundreds or thousands picofarad of capacitance between primary and secondary windings. <S> If you will not ground the secondary winding, then on all secondary outputs you will have floating 60Hz AC voltage of about half of primary voltage, say like 5500 volts relative to earth. <S> And, yes, on top of simple first order effects there are dozens of safety, regulatory, cost, environmental and other aspects. <A> Without a ground reference, there would be a very high risk if the high voltage primary of the utility pole shorted to the low voltage (120V or 240V) secondary - thousands of volts sent to your appliances. <S> With a ground reference such a short will blow a breaker on the utility side.
|
Because any real transformer is not an ideal insulator.
|
Want to get a heat gun for SMT - what should I get? I need to get a heat gun so I can do some SMT work. What wattage / heat range should I be looking for? Obviously it needs to get hot enough to melt the solder (preferably lead-free) yet not fry the components. I have seen everything from about 1400W up to 3400W with max temperature ranges anywhere from 350 to 650°C So what wattage / max temperature do I want? <Q> What you are asking for is specific to soldering. <S> It's not called a "heat gun", but a "hot air soldering station". <S> At a minimum, it must have a way to set the output air temperature over the normal soldering range. <S> This should be calibrated somehow, like a degF or degC reading, not just a warm/hot dial. <S> Having air flow rate control is also very useful, and I think most hot air stations have that. <S> Otherwise, I wouldn't worry about power too much. <S> That's its business, as long as it can maintain temperature at whatever flow rate you set. <S> I just looked at the back plate on ours, and it says 270W and has a 5A fuse. <A> I do extensive SMD work/rework using a Weller WMD 1A station using a hot air pencil among other things. <S> You need a temperature control of 300-500 degrees celcius, it depends a bit on what you are going to solder. <S> Power is not the primary selector here. <S> Airflow is more tricky as, like barsmonster wrote, the crucial thing here is that you do NOT want the airspeed velocity to blow away your parts. <S> Thus, the flow per minute is not enough by itself, you need the nozzle diameter of the hot air tool. <S> For the Weller I use, it is about 6 mm wide <S> and you can set the airflow between 1-10 <S> L/min - I usually never go below 50% but at that flow, 100% will blow away stuff. <S> The 150 l/min <S> you quote seems far out crazy, that sounds like a blowdryer :) <S> Or was that in some kind of US units? <S> Note that in practice it is not useful to use hot air for reworking small SMT components. <S> A soldering tweezer is much more useful here (or even a normal soldering pencil with an SMD tip). <S> You need the hot air for desoldering chips, and for soldering chips with a ground pad. <S> You also NEED a microscope or benchmounted magnifying lens, but I'm sure you know this :) <A> Power is not as important as temperature control. <S> Also ability to control air speed is nice (too much air will blow stuff away).
|
You need it to be able to stabilize air temperature at 250-300-350C.Power might be from 200W and up, as long as it is temperature controlled.
|
How to make a digitally controlled power switch? I am new to electronics. I want to make a power switch controller by a digital pin as shown on the following figure. How to do this safely? Update: I forgot to let you know the bulb is just an example for the sake of drawing. The real scenario, it is a female socket into which any equipment male socket is inserted. <Q> Updated after update of the question. <S> Changes are regarding the zero-crossing <S> The best solution for an incandescent lamp is an SSR (Solid State Relay) with zero-crossing detection. <S> (Zero-crossing switching increases the bulb's life.) <S> If you want to switch other (unknown) loads you better pick a random-switching SSR, i.e. without the zero-crossing circuit. <S> If price is an issue you can better build the SSR from discrete components . <S> Below is an example using the MOC3041 as opto-triac. <S> Will cost about 2 euros. <S> Again, the MOC3041 has a zero-crossing circuit. <S> For other loads than an incandescent bulb use the random switching <S> MOC3051 . <S> Not as nice, but easier <S> : an electromechanical relay : The transistor is needed because the microcontroller can't supply the required current for the relay, so you have to amplify this current. <S> About the zero-crossing: you may have noticed that incandescent bulbs always fail when they're switched on. <S> That's because the mains phase can be near its maximum when switching on. <S> When you switch on a zero crossing you avoid these peaks. <A> There's many things you can do to separate the relay from the digital side. <S> You could use a peripheral driver chip. <S> You could set up a darlington pair transistor arrangement. <S> - it's quick and easy. <S> Most driver chips are just glorified darlington pairs anyway. <S> One commonly used driver is the ULN2803 <A> Depends on the current form your digital output. <S> Is it a microcontroller, some circuit, a switch, or something else? <S> If it is a low current output, you have to use some kind of simple transistor, as example, to control the relay. <S> Also check that the relay is rated for 240V. Depending on what controls what, some kind of isolation between the circuits might be good.
|
Combined with the low resistance of a cold bulb this results in a high current peak, which may burn the filament. You could opto-isolate it. An SSR module is the most convenient, but they're not cheap. Personally I'd just use a driver chip
|
Low PPM Tc Resistors for Voltage Divider circuit? I have a voltage divider circuit that cuts an input voltage in half. The half voltage is fed into an ADC. I need this reading to be stable over temperature. The resistors are 100K, but I don't think this matters. Do I need to use resistors that have a low temperature coefficient to have a stable reading? Or, because the resistors are the same value and have approximately the same temperature coefficient, can I use standard resistors? The resistors will track each other over temperature and the midpoint will always be half the input voltage. <Q> Like Olin I was thinking of a resistor array. <S> (Array being a big word for two resistors in a single package.) <S> Firstly, being manufactured on the same process will give you a good matched value, and secondly being in the same package will make their temperatures also matching. <S> At Vishay I found these "High Precision Thin Film Chip Resistor Arrays" : <S> The networks provide 1 ppm/°C TCR tracking, a ratio tolerance as tight as 0.01 % and outstanding stability. <S> edit Linear Technology has the LT5400 resistor array with the following specs: 0.01% Matching 0.2ppm/ <S> ° <S> C Matching Temperature Drift Prices start at USD 3.49 quantity 1000, so that's pretty steep. <S> For a couple of resistors, that is. <A> You should not assume two resistors will track just because they have the same specs. <S> In reality, they will probably track somewhat, especially if they are from the same production batch, but you don't know that and shouldn't rely on that. <S> Either get resistors with low enough absolute temperature coefficient, or get resistors that are specifically matched. <S> Look at manufacturers of more than just jellybean resistors, like Bourns, Vishay, etc. <A> If the voltage range you are measuring at the output of the divider is between about 2 and 20 volts, you could use a "rail splitter" IC like this instead. <S> It has an onboard precision-trimmed and temperature compensated resistor divider. <S> The input has to be within the compliance range of the chip, however, hence the output voltage range restriction. <A> Without giving us an error budget, e.g. max error at 25C, max error due to temperature, it's hard to say, but I'm assuming that max error due to temperature is more stringent than max error at 25C, because the latter can be calibrated out whereas <S> the former cannot. <S> Nowadays they're fairly inexpensive. <S> Digikey sells 10K 0.1% 0603 resistors for 25c apiece <S> that have <S> 25ppm/C max tempco, with the price dropping to 10c apiece at high volumes. <S> At that max tempco, even if one resistor has a +25ppm/C tempco and the other has -25ppm/C tempco, that will affect the output ratio of the divider by 0.0625% of fullscale -- at a 3V supply that's just under 2mV. <S> If you need tighter specs, get 10ppm/C resistors (more expensive: <S> Mouser sells some from Xicon that are 75c apiece , dropping to about 20c apiece at very high volumes) Or use an integrated matched pair of resistors meant for voltage dividers -- those are even more expensive, but you can get 5ppm <S> /C tracking tempco from TT Electronics sold by Digikey at about $2.00 apiece dropping to 65c at very high volumes. <S> Or use a switched-capacitor voltage divider, and filter the heck out of it to get rid of switching noise. <A> You can also look into using a zener diode in series with a PN junction. <S> Zener diodes have a positive temperature coefficient that can largely cancel out the negative coefficient of a PN junction (like one in a BJT). <S> This, of course still assumes that both components are at the same temperature, which may or may not be a valid assumption. <S> EDIT: @stevenh is probably thinking of this: <S> This is for a compensated zener diode, which has a silicon diode in series. <S> Since the temperature coefficient varies with zener voltage, a 5.6V zener diode nicely cancels out the -2 mv/K of a silicon junction, creating the intersection in the above plot. <S> An example of how the coefficient varies with zener voltage is below.
|
You can get multiple resistors in a single package that have matched temperature coefficient to well below their absolute temperature coefficients. I would recommend just using a pair of 0.1% resistors if you can.
|
What does the x mean on this schematic What does the x that's wired to pin 3 and 6 mean? Does it mean it should be wired to gnd? <Q> Some schematic programs use this as a symbol to show that the pin is specifically not connected to anything. <S> However, this looks to be a schematic from Eagle. <S> Eagle requires no such symbol. <S> Output or passive pins are allowed to not be connected without any electrical rules check error. <S> You can make a symbol and footprint for anything you want in Eagle, so only the designer knows for sure what this means. <S> It could mean a test point, but most likely not since there is no component designator. <S> A bunch of test points on the board are pretty useless unless labeled, so I don't think that's it. <S> Most likely whoever created that schematic was used to a different schematic program and felt compelled to put Xs there to indicate not connected. <S> In Eagle, you simply don't connect anything and don't give it a name <S> and it won't be connected anywhere. <A> It looks like the pin is not used. <S> If you're using this schematic to design a PCB, just do not make any traces coming out of those pins. <A> I would mark pins that way as a check that they've been accounted for. <S> On a finished schematic all IC pins should either be connected somewhere or have this marking. <A> The purpose to me says clearly that the pin is left floating. <S> It may be a clear indicator from the designer to himself and others that these pins should be left floating in this design. <S> It may otherwise be unclear whether the pin should be connected (forgotten, removed by accident etc.). <S> This is an explicit sign the pin is left floated. <S> In Eagle you don't need to do anything with unused pins, so this would be the only useful explanation to me. <A> It is an unused/unconnected pin. <S> In this case, it is a No-ERC Directive , and as such, it tells the design compiler to not verify the pin's net against the design rules. <S> As such, you may see similar directives even in situations where there is a valid connection. <S> This is useful in odd cases, such as if you are powering a small, low-power sensor off a MCU output pin. <S> Normally, the schematic checker will complain that you have an output connector, and a power pin on the same net, since this is normally something you do not want to do. <S> In this case, it lets you override the checking, and suppress this error message. <S> Alternatively, it is also useful in a more traditional role <S> if you have a device with an input pin, that has an internal pull-up. <S> Normally, you want to avoid leaving floating inputs, so the design compiler warns you if you do so. <S> In this case, you want to suppress this warning, since there is an internal pull-up, so you tag the pin with a No-ERC directive.
|
Others already said it indicates that the pin is left unconnected. In some EDA suites (I use Altium Designer, this is also likely true in other packages), the "not-connected" symbol has an additonal role. They are probably just indicators for test points.
|
How to securely solder a wire to a vibrating SMD PCB? I have a problem with a "4in1" control unit for a RC helicopter. Due to some problems with helicopter setup, it is currently vibrating heavily which makes things even worse. The 4in1 unit basically has two PCBs with electronic speed control, accelerometer, radio receiver and servo controller. The receiver has short (say 2.5 cm) piece of wire which acts as a 2.4 GHz antenna and is soldered to an SMD pad. It is also glued to the PCB with hotglue. The SMD pas is around 2 mm away from the edge of the board towards which the antenna is supposed to point. My problem is that after a while, the solder joint breaks and so does the glue joint and the antenna falls off. Fortunately, I haven't had any crashes due to that, but I'd rather fly with antenna than without. So I'm looking for tips how to securely solder a wire to a SMD pad on a vibrating board.I already tried just resoldering the wire and gluing it with hotglue, but it seems that it doesn't help much. Here's the photo of the part. The solder ball is where the antenna should be attached. I removed the layer of hotglue that was over it. Here 's full size image, for those who dare click it. <Q> That should be done with a thru-hole pad. <S> You might even want two thru-hole pads to loop the wire thru for better strength. <S> If the wire is vibrating a lot, <S> then maybe a few zig-zags in the wire near the PCB will cushion or decouple the wire motion from the PCB a bit. <A> For better vibration performance use silicone instead of hot glue. <S> Silicone will stay more flexible. <S> Taper <S> the silicone so that it is a big blob at the PCB end and tapers down the wire. <S> This reduces vibration and doesn't focus the flexing. <S> It's standard procedure in mil-spec connector wiring for this very reason. <S> What kind of silicone : not acid cure, it eats the board. <A> I second David's idea of supporting the antenna to minimize vibrations. <S> I would take a 2cm piece of stiff paper the length of the antenna, fold it lengthwise so that both halves are perpendicular, glue it on the housing of your receiver so that it stands up and glue the antenna to the fold. <S> Not for stiffness, but better soldering might help as well. <S> If the wire has been soldered well the solder would flow up against it a mm or so, and not be a nice round blob like you have now. <A> I'm guessing that the tip of the antenna wire is free to vibrate back and forth, until the antenna breaks loose. <S> I suspect the hotglue near the edge of the board is breaking first, and then the solder joint breaks. <S> so it doesn't vibrate as much. <S> Perhaps hotglue a larger piece of cardboard or plastic sheet to the opposite side of the PCB for the antenna to rest on.(Hopefully a piece of plastic a couple inches long won't be too heavy ...) <S> +---------++-*-*- <S> * || | || <S> o=*=*==*||PCB <S> | <S> |+-*-*- <S> * <S> | | card <S> | <S> +---------+ <S> Where = <S> : antennao : soldered end of the antenna* : hotglue dots or superglue dots or tape dots <S> I'm sure you can invent a smaller, more lightweight shape for the card, and cut the card into that shape after the glue hardens.
|
Perhaps you could use hotglue (or superglue, or wire dots, or tape dots ) on some or all of the antenna wire A SMD pad is not a good way to attach a wire.
|
When is it an instrumentation amplifier (In-Amp) and not an operational amplifier (Op-Amp)? I've seen a number of different configurations for instrumentation amplifiers, including 2 opamp versions. This is also one. But it's just a differential amplifier preceded by input buffers. When do you call it an instrumentation amplifier, in other words, what's so special about it that it deserves a separate name? <Q> "An instrumentation amplifier is a precision differential voltage gain device [...]. <S> " <S> One of the important words here is "gain". <S> An OpAmp has infinite gain (in theory) and only gets a defined gain by adding circuitry around it. <S> Usually, when using one OpAmp only, at least one of the inputs loses its extremely high input impedance because external resistors are necessary. <S> If you need two (differential) inputs with both a very high input impedance and a defined gain, you can use the two-OpAmp-InAmp you are talking about or the three-OpAmp-InAmp-configuration your picture shows. <S> There are also readymade IC InAmps by such companies as Linear Technology or Analog Devices. <S> The three-OpAmp-InAmp circuit in in the picture of your question shows that two OpAmps are used as buffers, where they still have a high impedance at their otherwise unconnected non-inverting input pins ("+"). <S> By feeding their outputs into another OpAmp, the upper non-inverting input ("+") becomes an inverting input ("-") because it is connected to the 3rd OpAmp's inverting ("-") input. <S> The lower non-inverting input ("+") remains non-inverting due to its connection with the 3rd OpAmp. <S> Common three-OpAmp-InAmps use a slightly different configuration compared to your picture to set the gain with one resistor only (the external gain resistor in the case of completely integrated InAmps). <S> Please refer to the links I've provided for more details. <S> With the three-OpAmp-InAmp, you get both a very high input impedance at two differential inputs (while you would get only one input with such a high input impedance with a regular OpAmp buffer) and you get a very good rejection of common-mode signals (that is achievable with one OpAmp, too, but at the cost of lowering the input impedance with the resistors you have to use to turn the OpAmp into a difference amplifier). <S> The two-OpAmp-InAmp circuit needs less parts, but at the cost of a not-so-good common mode rejection ratio (CMRR). <S> Here is a link to a very good book about InAmps by Analog's Charles Kitchin and Lew Counts where you can find a more in-depth look onto all these issues. <A> I agree with what Zebonaut said, but here are my criteria for being a "instrumentation amplifier" more concisely: <S> The gain must be finite and known. <S> Sometimes the gain is fixed, like 10x or 100x. <S> Other devices allow a selection of preset gains or you provide a resistor or something that sets the gain. <S> The inputs are differential. <S> The common mode rejection is usually very good, generally much better than you could do with a opamp and discrete parts. <S> The inputs are high impedance. <S> You can make a differential amplifier from a single opamp, but then the inputs are no longer high impedance, and one of them is partially driven with the signal on the other. <S> This is all in one itegrated package if it's sold as a "intrumentation amplifier" chip. <A> An ideal opamp has a infinite input impedance and infinite amplification . <S> Through feedback you can set the amplification to a realistic level, but this is at the expense of the high input impedance. <S> An instrumentation amplifier (inamp) is a difference amplifier which solves this. <S> There are several instrumentation amplifier configurations, this one is probably the simplest to understand: <S> It's a regular differential amplifier with an opamp (the one on the right), with two voltage followers to buffer the inputs, so that they are high impedance. <S> This inamp has an amplification of \$\times\$100 (100k\$\Omega\$/1k\$\Omega\$), and you have to change two resistors to get a different amplification. <S> Other inamp configurations let you set the amplification with a single resistor. <S> Here \$R_{GAIN}\$ sets the amplification: \$V_{OUT} = <S> \left(1 + \dfrac{2 R}{R_{GAIN}}\right) <S> \times <S> (V_2 - V_1) <S> \$ <A> An op amp is a device which attempts to yield an output which will make the difference between two (typically high-impedance) inputs zero. <S> It is possible to use an op amp and some resistors to build a circuit which will produce an output which is proportional to the difference between two NON-HIGH-IMPEDANCE inputs. <S> It is also possible to use an op amp to produce a low-impedance output which is equal to a high-impedance input. <S> An instrumentation amplifier is conceptually similar to a pair of op amps which convert high-impedance input signals to low impedance outputs, and feeding into a third op amp which then produces an output proportional to the difference between those buffered signals. <S> In practice, things are a little more complicated since the difference-measuring circuits require some precisely matched resistors, and there are some tricks to make them less sensitive to component variations, but conceptually an instrumentation amp is a pair of high-impedance buffers which then drive a difference-measurement circuit.
|
An instrumentation amplifier is a device which yields an output proportional to the voltage difference between two high-impedance inputs.
|
Wireless communication between more than two Microchip PIC microcontrollers I would like to create wireless communication between Microchip PICs for my simple project.It's actually an one way communication but there is one server and more than one client (about 2-4, all client needs to the same number at the same time, so they are identical). I have to push through very few bytes. Im absolutely beginner with wireless communication, or almost any hardware communication at all.So please help me, what kind of wireless transceiver should I use ? I have some conditions: It must be relatively cheap . It must be easily available . It would be nice if it easy to use. about 10 meter range (at least) Firstly I found "Serial Bluetooth RF Transceiver Module rs232" but I can't find any info about how to use. (I guess, it's too simple?) And I also don't know is it capable to connect to more clients.And I also saw ZigBee, but I found it a "little" overpowered for my needs. (And also complicated.) So what kind of wireless transceiver do you recommend me? <Q> The Nordic Semi nRF24L01+ is ideal for that sort of thing, low-cost modules are available on Ebay: <S> http://cgi.ebay.co.uk/Arduino-NRF24L01-Wireless-Transceiver-Module-2pcs-/280640828189?pt=LH_DefaultDomain_0&hash=item41577f331d <S> The nRF24L01+ is often used in wireless sensor networks. <S> An MCU is required. <S> I have a suitable design and test software here . <S> It uses the much more expensive Sparkfun module, I've designed a board for the cheaper modules but haven't had one made for testing. <A> You may want to have a look at Digi XBee . <S> They have both point-to-point and point-to-multipoint solutions. <S> They're easy to use as you simply connect them to your microcontroller via its UART connections; the complete IEEE 802.15.4 implementation is transparent. <S> I found the price OK (something like 18 euro for a point-to-point module, IIRC). <S> edit Indoor range up to 30m , <S> but you probably know that this depends very much on the building's construction. <S> Line-of-sight up to 90m. <S> XBee- <S> PRO version: up to 90m and 1.6km resp. <A> You could use something like the XBee, which handles the node linking for you and abstracts the communication for you. <S> Or you could be more ambitious and use a bunch of discrete ISM transceivers (Industrial/Scientific/Medical - refers to the frequency range it works in) and write your own protocol (maybe something along the lines of how I²C works) for the communication. <S> The fact that you have one master and a number of slaves makes it easier to do. <A> This kind of radio is intended for low power and relatively low data rate. <S> Otherwise, your question is too broad to give a meaningful answer. <A> Is line of sight enough, perhaps with a transmitter or reflector in the ceiling? <S> If so, you don't get much easier or cheaper than infrared. <S> You can use a common 38kHz demodulating IC for reception, and send using a 38kHz clock (perhaps off a microcontroller timer) and a digital pin. <S> One example of this is the Lego RCX and Power Functions remotes. <S> Bluetooth is designed for point to point links, not broadcasts like these. <A> I sell RFM70 modules (so I might be biased, beware!). <S> These are cheap, but maybe not that easy to use: <S> 3.3V (but 5V-tolerant data pins), 1.28mm pin grid, Chinese-English datasheet, software interface is a bit complex and the explanation in the datasheet 'could be better'. <S> Range ~ <S> 70 m line-of-sight, but 'within one room' is more realistic. <S> I read somewhere that the chip (RF70) is a lot like the Nordic chip. <S> I am working on a C library with a better explanation of the interface (for now for LPC2148/GCC and 16F887/HiTech-C, which are so different that other chips should be no problem). <S> (update: the library is available from http://www.voti.nl/rfm70 )
|
You might take a look at Microchip's 802.15 radio modules and the MiWi stack that can use them. You can use pretty much anything for the communication - it all depends on how abstracted you want the system to be.
|
Displaying text on a LilyPad? What are the best options to show a line of text on a LilyPad Arduino, probably using a matrix of LEDs? Are there any thin and inexpensive displays? The main problem is that the only things I can find on the Internet are around 8x8, but I need much more than that (about 4x40). Also, the number of pins on the LilyPad seems insufficient, but I might be wrong (I hope so) :) Can anyone help me out? Thanks in advance. <Q> try earth lcd, sparkfun, or crystalfontz. <A> If you want to make a led matrix, a common way is to use 74hc595 . <S> They are shift register with latch to update all output simultaneously and can be daisy chained up to your need. <S> It may only use the SPI pins (miso, clk, cs). <S> some example : <S> http://www.sparkfun.com/products/760 <S> (they are using bitbanging instead of SPI). <S> a simple led array (7x25) <S> I've quickly made a year ago ( pdf ). <A> Most LCDs operate on a standard 8 or 4 bit parallel interface that is easy to work with. <S> There are also serial boards available from various suppliers which will convert from a TTL serial signal to the LCD's parallel interface to allow you to use serial if you're short on pins. <S> They often abstract the LCD protocol for you making it simpler to program. <S> You can get the LCDs in many different styles and sizes, such as character based 2x16 (2 rows of 16 chars), 2x40, 4x16, 4x40, etc. <S> You can also get graphical ones with resolutions like 128x64, 192x64, etc. <S> These are harder to work with but give much greater flexibility.
|
There are a number of lcds with serial interfaces, text and graphical.
|
Replacement for the MC14495-P1 I need to find a device I can use in place of the MC14495-P1 binary to hexadecimal LED driver. This part is sadly very scarce and very expensive these days. The device is a 16 pin DIP with +5V on pin 16 and Gnd on pin 8 (standard logic chip arrangement) The closest I have come so far is a 14 pin PIC16F and a bit of reworking of the PCB, but that's going to be a pain in the sphincter as I have to do about 80 of them. So I am wondering if there is a programmable device from one of the other manufacturers, like Atmel, Maxim, TI, or whoever, that is 16 pins with the power laid out like I need and all the other pins available for IO? I haven't found one as yet, but if anyone has any suggestions I'd be very grateful. Oh, and they have to cost < £1 each. <Q> You might be to put a small PIC on a daughter board with pins sticking down to emulate the 16 pin DIP footprint. <A> According to Octoparts XsMicro may be able to help you get the original. <A> You also have a problem with the drive arrangement. <S> The the device you used is designed to drive the LED segments directly (supply the current) and includes internal current limiting resistors. <S> You don't find many MCU that can directly drive LEDs like that (at least not so many) <S> and I don't know of any with internal current limiting resistors. <S> Fairchild makes a similar part , but the pin out is different and much more "standard", i've never actually seen a 7-seg driver with a pin configuration like that Motorola part, I've seen some differences in the input side but all the segment outputs are always at least on one side of the IC. <S> As such using any other driver i can think of would require PCB rework or an adapter board. <S> Even the PIC would still require additional external circuitry that likely isn't there for the current chip. <S> Unless you can find more of that Motorola chip I think your stuck creating an adapter board for something like the Fairchild IC listed above.
|
In theory a DIP-16 PLD is an option but they are getting harder to find (expensive), would require additional external circuitry (current limiting) and usually the programmers are quite expensive.
|
On Board Unit communication via RS485 I've heard of On Board Units (OBU) can communicate via RS485. My question is, how? For example, I would like to ask the state of some runtime variables from the OBU (what doors are open?, etc.) Or another example, I would like to send some data to the OBU (outside temperature). What protocol must I implement to do these? Are the any sample implementations available somewhere? <Q> One of the more common ones is Modbus. <S> From a hardware stand point you will need an RS-485 level translator attached to the UART module of the microprocessor. <A> I don't know what you mean by "on board units", but it doesn't sound relevant anyway. <S> What you are apparently asking about is the protocol. <S> RS-485 is a electrical standard only. <S> It's a multi-drop differential single-signal bus. <S> How you send bits over it, how they are delimited into bytes, and what the bytes mean is up to the implementation. <S> That said, most implementations use UART-like signalling. <S> That means there is a start bit, 8 (usually) data bits, and one (usually) stop bit. <S> It's very likely <S> your protocol uses that or something close that can be sent and received with a hardware UART and the appropriate bus transceiver. <S> However, that's where common practise ends. <S> There are some officially published protocols above that, but its very common for individual implementations to roll their own. <S> You get this information from the manufacturer. <S> By the way, I wouldn't use RS-485 for new designs. <S> CAN is electrically similar in that it is a differential multi-drop bus, but the standard goes as far as defining whole packets of up to 8 data bytes each, with checksum, ID, and other out of band signalling. <S> Another big advantage of CAN is that the hardware to send and receive whole packets is available in small micrcontrollers. <A> If you are asking about about the protocols used in cars, its not RS485, its a protocol called CAN accessed using a connector standard called ODBII <S> The traffic on the network as I understand <S> it is just ASCII so easy to hack <S> so theres a lot of ODBII reverse engineering going on, have a google!
|
There are a number of protocols that you can support over RS-485.
|
Servo jitters at specific positions, why? I have a Pololu Micro Maestro, and it's great.But when i tell the servos to move to specific positions, sometimes it makes a beeping-like noise. This only happens at specific positions. It's not because I'm trying to make it go farther than it can, it also happens when it is positioned around the middle. This happens when I use a laptop to control the servos via an application written in .NET I have only tried controlling them either via my own application or the maestro control panel. Is there any way to avoid this? I hope you guys understand what I mean.Thanks <Q> The R/C servo control signal is a pulse train. <S> There are two parameters that describe a pulse train: <S> pulse width (PW) and pulse repetition interval (PRI). <S> PW is what controls the servo position. <S> It is critical, and must be stable. <S> PRI is not nearly so critical, at least with R/C servos: they will generally be happy with a PRI anywhere from 20 to 50 ms, and they don't mind if it varies some. <S> (You will also see PRF, for pulse repetition frequency, usually in radar systems. <S> PRI 20-50 ms corresponds to PRF of 50 down to 20 Hz.) <S> Any kind of instability in the pulse width (PW) will cause the servo to "buzz", as it chases the control signal. <S> Such instability can be triggered by the HUGE current draw of a servo in motion, if the power supply AT THE CONTROL CIRCUIT has insufficient surge capacity. <S> (A standard servo in controlled motion can easily draw 250 mA. A "buzzing" servo can draw half-amp spikes. <S> Yes, this is from experience.) <S> Instability in the PRI is generally not a problem for an R/C servo. <S> The very first thing I'd do is hang about a 250 uF capacitor directly across the servo power leads. <S> The second thing I'd do is scope the control signal, triggering on the rising edge, and looking at the problem ranges to see if the timing on the falling edge is varying AT ALL. <S> Anyone planning on playing with servos should start by wiring up a 555 and a one-transistor inverter, generating the servo control signal entirely in hardware, and playing with this. <S> Watching a servo sitting there, buzzing, watching your power supply ammeter going all over the place, registering half-amp spikes, is INSTRUCTIVE. <S> Taming that beast is even more so. <S> Note: This circuit will also demonstrate the servo's tolerance of varying PRI: the simplest 555 pulse generator for this purpose will vary PRI as it varies PW. <A> Is the laptop doing the closed loop control? <S> If so, it's probably due to trying to do real time control on a operating system that isn't real time. <S> I've measured Windows going out to lunch for a few 100ms at a time just because the mouse was moved. <S> If your control iterations are faster than that, such delay will make things unstable. <S> If the laptop is only sending high level commands to a dedicated controller that is doing the actual controlling, then it's hard to guess what's going on. <S> How is the position feedback signal produced? <S> If it's from a pot, it could be getting scratchy. <A> Our experience with model remote control servos and Pololu cards is that some servos have positions that are noisy. <S> Its the servo, not the controller. <S> Swap the servo if you must have better performance. <A> It's a single PIC doing more PWM channels than it has timers for. <S> How well it manages to do this will depend entirely on what the program in the PIC does. <S> For this design, I'd go with the timer to mete out pulse lengths, interrupts to do the output toggling, and software polling for the USB and serial so it cannot interrupt the timing of the servo outputs. <S> Many other combinations will have interference between the timing of servo channels or external updates. <S> It may well be the specific positions you're seeing coincide with the heartbeats on USB. <A> If your servo oscillates with amplitude of one encoder step size then check that feature named like "dead band something" is used correctly. <S> If not enabled, then, your servo always has a chance to hit an uncertain spot near encoder position N and N+1, or N-1. <S> Servo tries to eliminate smallest 1 step error and overshoots one more step, or encoder has sensitivity to magnetic field noises or light and keeps changing its mind between N and N+1. <S> Uncertain reading is unavoidable in any builds no matter how good and expensive are the parts. <S> To fix it, PID setup must ignore position errors less or equal to 1 encoder step. <A> If this is a conventional R/C servo, it uses a potentiometer to sense shaft position and develop the internal error signal that drives the motor. <S> Potentiometers can get dirty. <S> You MIGHT try careful surgery on the servo and spraying the pot down with contact cleaning solution. <S> Or try another servo.
|
Depending on the math of your control law in your PC (.NET -> PC controller), you might have a mathematical instability that shows up in the commands to the servo.
|
What causes "hard working" growling sound in a heavily loaded electric motor? From my experience just any electric motor when under huge load would make recognizable growling sound. This would happen regardless of the motor type - brushed/brushless, AC/DC, small size/huge size - it would just happen. Where does this growling sound come from exactly and how does it form? <Q> Under high load significant force is 'trying' to move magnets & individual wires inside coils - so if there is even tiny loose piece - it start to move back & forth with the frequency = frequency of magnetic field change effectively working as a speaker. <A> Ball Bearings could be an answer. <S> No motor is perfectly symmetric around its axis of rotation. <S> As the load increases the precession about the axis increases and this in turn places greater stress on the bearings, producing sound energy. <A> Transformers hum because of magnetostriction .Electrical <S> current induces a magnetic field whichcause <S> the iron core in most transformers and motors to physically change shape (magnetostriction);when the current returns to zero (typically at mains frequency)it returns to the original shape. <S> Heavier loads cause higher peak current which cause a larger peak change in shape which sounds louder. <A> If you use a doctors stethoscope you can locate the source of a baring failure vibration or bushing wobble. <S> But most often than not if it’s a “Growling” as you describe it its harmonics. <S> I have dealt whit this problem in brad new motor assemblies. <S> The motor balance is set for a max RPM <S> so it’s good. <S> The pulley/gear/sprocket is balanced in the same way. <S> Now put them together <S> and there’s a point in watch <S> this 2 minutely of balances create the harmonic. <S> Manufactures tray to make this happen outside the Min/Max rpm and Max load rage <S> but it’s not <S> 100% <A> You can get the growl out a hard working transformers too. <S> Somewhat like @BarsMostner thinks I think it the whole assembly moving as speaker. <S> The spider of the motor and the shaft are somewhat flexible. <S> We had a weapon system project with a 10kW brushless motor driver that I suggested we cold use a an audio output device. <S> So it could growl (actually quite loudly) when you get close to it. <A> In a heavily loaded motor the case and rotor are oppositely REACTING from either the cogging forces of a brushed motor and/or the AC cycle forces of any AC motor, and even the CURRENT waveform of a PWM supply. <S> When a motor is lightly loaded these currents are less as are the forces. <S> The rotating load also feels the same forces. <S> The same thing happens with gasoline engines and other devices.
|
Various resonances may also come into play if the motor speed changes.
|
Why do we need brushed motors? Looks like brushless motors are more convenient - no brushes means no maintenance required. For example, power tools manufacturers often cite that graphite brushes need to be replaced every 50 hours of continuous use. I'd think that this alone should make brushless motors much more convenient. Still most power tools use brushed motors. What's the reason to use brushed motors instead of brushless motors? <Q> The electronics for this add cost and complexity. <A> Because they are cheaper and lower-tech - no need for tricky semiconductor devices. <S> Also, switching electronics for high-power devices (1-5kW+, 1000V+) is tricky to implement (but nothing modern electronics can't handle). <S> Sometimes companies produce crap just because it's cheaper & gives some long-term income on maintenance, or they just have 50 years old production line deep in China's village which is nearly free to operate. <A> Why do we have brush motors? <S> I can (and have, on multiple occasions) <S> constucted a brushed DC motor with nothing more than cheap wire, thread or tape, a paper clip, a scrap of wood, and thumb tacks. <S> A cheap permanent magnet makes things easier but is not necessary. <S> Also, as Georg partially pointed out, you get higher (power/weight) with brushed motors. <A> Others have covered servicing and cost issues <S> well <S> .There <S> is still a place for the brushed motor when it has a series would field .Remember that BLDC motors have rotating permannet magnets which are not as strong as an electromagnet .This <S> means that the BLDC motor wont make as much torque and its characteristics are more like a shunt wound motor. <S> The BLDC motor can spin much faster without flying apart due to its simple robust rotor so if you gear it down you can get the torque back. <S> Most electric motors are thermally limited so the better cooling characteristics of the BLDC combined with higher RPM give a higher continious horsepower rating. <S> However the peak torque of the series wound brushed motor is greatest .This <S> is why the electric drag people use them and win .I <S> guess that if reduction gearing is not an option then the series brushed motor is useful.
|
A brushed motor can be driven with direct current, while a brushless motor requires the drive current to be electronically commutated. Mainly because they are cheap and easy to construct.
|
What affects IMU (Inertial Measurement Unit) accuracy? Context I need to analyse the performance of a system that uses inertial sensor readings as input. My knowledge of IMUs is very basic, having only read a few introductory texts. My understanding is that IMU accuracy degrades with continuous use due to drift, which results in accumulated error. Questions What other operational factors (i.e. not the specific device/components, but IMUs in general) effect the accuracy of IMUs? I have noticed in my application that the system performance degrades when the object moves, especially when the movement is fast, and then improves once the movement slows, and reaches a halt. What aspect(s) of IMUs causes this? EDIT To be more specific, the sensors give orientation as quaternions (rotational offset from the sensor coordinate system to a global coordinate system). The error fluctuates over time (duration of about 15 seconds), however there is not an incline in error over time. Based on this, I believe IMU drift is not an issue here. What I am interested in understanding is why the estimate error of the object's orientation fluctuates based on the amount and speed of the object's motion. If it helps, I am using Xsen MTx IMUs. <Q> I have noticed some accellerometers and gyros to have a bit of a memory effect. <S> For example, you can place a accellerometer horizontal and null out the reading. <S> I will stay close to zero <S> but of course still drift a little. <S> Now move it sideways and back again to the original position and the reading <S> won't be zero anymore. <S> The stronger and more sudden the movement, the more it seems to change the zero offset. <S> It is this unpredictable drift in the zero value that messes up inertial navigation over a longer time. <S> Sometimes long is only a few seconds with MEMs gyros and accellerometers. <S> I did a sports head tracking device once, and even if you had the person hold still before starting a motion, the position would unusable after a couple of seconds. <S> Fortunately that was long enough, and we designed the algorithm to use things like accelleration and angular rate directly instead of relying much on their integrals. <S> This was with rather cheap MEMs units. <A> Correct axis alignment is one: linear sensors have linear transfer functions between the actual pitch/yaw/roll/3-D acceleration axes and the outputs of each corresponding sensor (or subset thereof, if you're not using a full 6-axis IMU). <S> Ideally the transfer function is the identity matrix: motion in each intended axis affects only the corresponding axis. <S> In real systems, there is cross-coupling that depends on sensor alignment. <S> If you're off by 1 degree, for example, you'll get a small output signal on one axis for motion in another axis. <S> Pendulum or fluid tilt sensors at large angle displacements have linearity issues. <A> Even the best IMU will have some error. <S> Normally those errors add up over time to the point that their output becomes useless. <S> Of course, the better units will have less errors, so it takes longer for the errors to build up-- but they will still build up. <S> Normally, IMU's will have some method to get rid of these errors. <S> They do this by using multiple sensors that complement each other. <S> For example, a GPS has great long-term (long distance or long time period) accuracy but terrible short term accuracy. <S> An accelerometer is the opposite and has terrible long-term accuracy but fairly good short term accuracy. <S> By using both types of sensors you can get the best of both worlds. <S> Of course a GPS somewhat violates the "Internal" part of IMU, but you get the idea. <S> It is impossible to get an IMU that will work forever without external input (GPS, star tracker, etc.) <S> since even with the best of systems there is always a source of error that accumulates over time. <S> But it is impossible to really give you a more detailed answer than "errors build up over time" without looking at the exact set of sensors that's being used-- <S> and how that sensor data is being integrated into the final output. <A> On agricultural precision farming machinery (GPS guided tractors), IMU's are often used to correct for the antenna offset of the tractor. <S> When a tractor tilts, the antenna moves outside of your predetermined driving path. <S> The RTK-DGPS devices can measure up to about 1cm XY (2cm Z) <S> accurate, so you will need to correct for it. <S> Fortunately it is possible to have redundant systems: <S> roll/pitch can be done with inclino and gyro's, where a yaw can be done with gyro's and the driven route. <S> To get a high data rate of accurate angles, you will need to combine different devices. <S> Each sensor has a limited bandwidth. <S> An inclino is an absolute measurement device with low bandwith. <S> It may take a couple of seconds for a stable signal, then again: it is absolute. <S> A gyro can measure angular rate and can be integrated to get an offset, thus you get a different angle. <S> Filtering it with the inclino value, you can get a live angle. <S> However, gyro's have a limit to how much degrees/second they can measure precisely too. <S> If it does go wrong, it should not drift away because its corrected with an inclino.
|
Another factor has to do with nonlinearity, which is very dependent on the particular mechanism used for sensing.
|
Impedance Matching and Large Trace Widths I am currently working on a design in which one of my ICs specifies the use of a 50 ohm trace. The answer to this question, Characteristic impedance of a trace , shows that a 120 mil trace is required to get this impedance. The IC only has room for 18.8 mil traces, and that is assuming no space between traces. So, how can I actually design with that trace impedance kept in mind? Obviously I can decrease the board thickness or increase the copper height, but only to some extent and I would like this to be fabricated for somewhat cheap. How is this usually dealt with? The IC that I am using is the MAX9382 which can operate up to 450 MHz, I will probably be using it around 400-450 MHz. The data that is being used is initially analog, but has to be hard limited to become digital in order to be used with that IC. <Q> You don't have to worry about the impedance of very short PCB traces as part of a longer trace. <S> So you will have a thinner trace directly next to the chip. <S> But if the trace has to go any distance, then you need to adjust the thickness of the trace as it gets away from the chip. <S> You will just "fan out" the trace width away from the chip. <S> That is how I have always seen it done. <S> This is not unlike the connectors of any transmission line. <S> The impedance of a single short element might be a little less, but it is slight when compared to the overall transmission line. <A> Use a 4 layer stackup. <S> Calculating the trace width needed is pointless unless there is a solid ground plane under it, with a 2 layer design you may need to route traces on the other side which then pretty much ruins your impedance if they come anywhere close to your trace. <S> At 450Mhz you really should have <S> solid, continuous, properly decoupled power and ground planes. <S> This will improve noise performance, EMI issues, give you better impedance control, etc. <S> Fabbing a 4 layer board isn't that much more to expensive than a 2 layer. <S> Use a 4 layer like: >----------------Signal <S> 18.3 mil>----------------Ground39 <S> mil>----------------Power8.3 mil>----------------Signal 2 <S> Spacing could change a little based on your copper thickness choice. <S> That will give you something like 10-20mil for your 50ohm trace on Signal 1/2 depending on final dielectric and copper thickness on the Signal layers. <A> Often having overly wide traces can cause issues with the capacitance of the trace. <S> Making the trace thinner will reduce the capacitance. <S> Of course having thinner traces messes up the impedance. <S> On a multilayer PCB this only works when the signal is also on an inner layer-- making it difficult to have the proper impedance AND capacitance on an outer layer. <S> The end result is that it's all a compromise. <S> I usually run those signals on inner layers with optimized PCB stackup's-- but then keep the traces skinny and very short when it has to go to an outer layer to get to a chip. <S> On a 2 layer PCB <S> it's very hard to have the proper impedance on narrow traces-- <S> so I usually don't bother. <A> Can you route adjacent reference trace along with your signals? <S> I've been told that routed triplets, or even quints if you can't fit triplets, etc. <S> can sometimes work in situations like yours if you don't have a close plane to reference to. <S> If you have a diff pair then it might be more like a quad, with adjacent references/returns outside on both sides of the diff pair. <S> The same mentor suggests that a two layer board should be treated as two unrelated boards due to space between the layers, and routed references/returns are the way to go if more layers can't be had. <S> I was wrong about the quad for a diff pair. <S> My notes from the relevant presentations say to use a triplet, with a reference BETWEEN the two signals of the diff pair. <S> Still looking/waiting for impedance calculations this way. <S> I'm told he's looking to find which RF/Microwave book they are in <S> , he has a number of them. <A> First figure out if it's a real requirement. <S> Over what distance must this be kept to? <S> If it's a seriously high speed signal (look at the edge-rate compared to the length of the trace) you may need to perform some simulation. <S> The Howard and Johnson reference that is in the answer to your linked question is a great resource on this sort of thing. <S> If the requirement is real, then figure out much tolerance there is (your board fab can probably only get to +/- <S> 10% of what you ask for, so take that into account). <S> EDIT: <S> Looking at your part you've now posted, you are in "real requirement" territory. <S> 80ps edges are pretty quick! <S> The "knee frequency" at which the harmonic start to drop off rapidly is upwards of 6GHz. <S> Assuming propagation delay is about 66% of speed-of-light, 80ps is 16mm. <S> The rule-of-thumb is that anything longer than 1/4-1/6 of the transition time is going to need to be treated like a transmission line, which means any trace longer than a few mm! <S> I'd hesitate to attempt this on a 2-layer board over any difference without doing some simulation. <S> You'll likely have to go multilayer to get the reference plane closer to the trace which allows thinner traces to meet the impedance specification. <S> (EDIT: As pointed out in the comments, you could do it in 2 layers, but you'll have a really thin board then!) <S> Alternatively, you might be able to build a coplanar waveguide structure on 2 layers which can provide the impedance you are looking for. <S> Or maybe increase the termination resistance, which means changing the trace impedance to match, which means a thinner trace. <S> AppCAD can help you play with parameters for these options. <S> Sounds like fun :)
|
If the PCB stackup is done differently, where the signal layer is closer to the power/gnd plane, then the trace can be thinner while still having the proper impedance. If impedance is critical I will go to at least a 4 layer PCB.
|
Unnecessary pull down resistors on BJT and FET transistors? I commonly see weak pull down resistors at the base of NPN transistors. Many electronic sites even recommend doing such things, usually specifying the value as something like 10x the base current limiting resistor. Bipolar transistors are current driven, so if the base is left floating, I see no need to pull it to ground. Also, I commonly see gate current limiting resistors on FETs. They are voltage driven and there is no need to limit current feeding the gate. Are these two situations examples of people confusing the rules between transistors (which need base limiting resistors) and FETs (which need pull down resistors) or combining the rules or something... or am I missing something here? <Q> The reasons become clear when you are considering not only the ideal behaviour of the transistors but also their parasitic elements. <S> The pull-down resistor at a npn-type BJT's base helps keeping the base "low" whenever the driving element for the base resistor should be unconnected or in a tristate mode. <S> Without this resistor, charge entering the base via the capacitance between collector and base ("Miller capacitance") could remain there and turn on the transistor. <S> There are two common reasons for a series gate resistor in a MOSFET circuit. <S> One is that the resistor limits the driving current and allows for some control of the gate charge current (think of the gate as a capacitor that needs to be dis/charged in order to turn the MOSFET off or on). <S> With a carefully chosen resistor, you are able to get some control over the turn-on or turn-off <S> transition times of the MOSFET. <S> Sometimes, you even use a resistor paralleled by a diode and another resistor to have different charge and discharge currents, i.e. a chance to influence the turn-on-time in a different way than the turn-off time. <S> The second reason for a base resistor is that the trace inductances around the MOSFET form a resonant LC tank with the MOSFET's parasitic capacitances. <S> When all you want is a clean transition of the gate voltage (rectangular waveform), you may get a lot of ringing in reality. <S> The ringing may be so severe that the MOSFET turns on and off a couple of times before settling and finally obeys to what the driver requests. <S> A resistor inside of the LC resonant circuit around the gate driver is able to damp this resonance and the path between driver and gate is the easiest spot to put the resistor. <S> For small-signal circuits, these resistors may not be necessary, but when driving power MOSFETs, you absolutely need them. <A> A series resistor in the gate line of a MOSFET will protect the driver (microcontroller) from ringing effects caused by parasitic inductances. <S> The optimum value for Rg is very application dependant. <S> You want the MOSFET to switch as quickly as possible to minimise switching losses, but not so fast that parasitic inductances and capacitances associated with the pcb layout and any wiring to a load, will cause high di/dt voltage spiking or ringing. <S> This will bypass Rg during turn off thereby speeding up the turn off. <S> Placing a resistor in series with the diode will enable you to control turn off time independantly of turn on. <S> Further reading (for all aspects of mosfet switching). <S> For switching small loads (like 100mA), or when a real MOSFET driver chip is used, the gate resistor is probably not needed. <S> (Note: these links were on the first G results page for "mosfet gate resistor") <A> The series resistor to the MOSFET gate is sometimes required to reduce a current peak when switching due to the gate's capacitance. <S> Logic circuit, esp. <S> microcontrollers allow only for a very low capacitive load. <S> It can also be used to reduce slew-rate (the switching speed). <S> A pull-down on the gate is used to prevent the gate from floating if the controlling <S> I/O is configured as input. <S> In this case the resistor's value can be chosen pretty large (1~10M\$\Omega\$). <S> The base resistor on the BJT is often combined with a pull-up, and this combination is used to set a stable quiescent point . <S> [ our teacher in college, not very good at English and apparently only having seen the word in print pronounced it as "keskent". <S> It took us a while to understand what he meant :-) ] <A> Most transistors have a small amount of collector-base leakage; if there isn't any pull-down, this current will be amplified by the transistor's gain. <S> In situations where leakage isn't a concern, the resistor may be omitted, but if leakage current is a concern, adding the resistor can reduce it.
|
If you find that an optimised value of Rg controls switch on OK but slows the turn off too much, then a fix is to place a diode across Rg with its cathode towards the gate drive circuit.
|
Why doesn't LCR meter measure expected reactance? Measuring an unknown capacitor with a Tenma 72-960 LCR meter , I got 89 nF at both 1 kHz and 120 Hz, which I believe because I measured other known capacitors, too. Then I tried measuring with the resistance function, and it gave me: 180 kΩ at 1 kHz 1.5 MΩ at 120 Hz But the reactance of an 89 nF capacitor is: 1.8 kΩ at 1 kHz 15 kΩ at 120 Hz Also confirmed that in resistance mode, it measures 1 kΩ for a 1 kΩ resistor at both frequencies. Why are the measured values off by exactly ×100? Am I misunderstanding what the LCR meter measures? (Is it magnitude of total impedance \$|Z| = \sqrt{R^2 + X^2}\$ or just the resistive component R in \$Z = R + jX\$?) Update with some more measurements: 10 µF: 9.393 µF @ 1 kHz 9.71 µF @ 120 Hz 185 ohm @ 1 kHz (reactance is 16 ohm) 5.6 kΩ @ 120 Hz (reactance is 133 ohm) 680 nF: 683.5 nF @ 1 kHz 686 nF @ 120 Hz 63.22 kΩ @ 1 kHz (reactance is 234 ohm) cannot measure at 120 Hz (reactance is 1.9 kΩ) So the exact ×100 numbers may just be a fluke. <Q> Well it's plausible but the numbers seem suspicious (I think). <S> This suggests that it is trying to compute the equivalent parallel resistance of the network. <S> The real part of the impedance of a parallel RC network is \$\frac{R}{1+\omega^2R^2C^2}\$ ... and this is frequency dependant. <S> The equivalent parallel resistance in this case is constant with frequency (by definition it is R) <S> and this is what I would expect the meter to display. <S> However in a real capacitor, the equivalent shunt resistance is not formed by a real resistor. <S> It would be interesting to switch to "series mode" if possible and see what the numbers are then. <A> "Then I tried measuring with the resistance function, and it gave me:" When you switch to the resistance mode in an LCR meter, it expects to find a resistance connected across it's terminals. <S> If I'm right, one cannot measure Inductive reactance or capacitive reactance using an LCR meter. <S> Like the manual says, you can measure L,C,R,Q and the dissipation values only. <S> Update: <S> An LCR meter cannot be used to measure (as in display) inductive or capacitive reactance. <S> The reactance value depends on a lot of factors,including the parasitic influences mentiond by others. <S> It internally computes the reactance using an AC voltage and freqency and a suitable circuitry and from it the C or L values are approximated and displayed. <S> Various meters use various circutry to achive this approximation. <S> The series or parallel mode also changes the circutry and approximation model. <S> In practice , use the series mode for large capacitive and small inductive values and the parallel mode for vice-verca,assuming you know aprori the range in which the value of the component falls. <S> I still believe that an LCR meter is just what it is say it is: <S> An L meter, a C meter and an R meter. <S> When it is in the R,mode, it expects a resistor to be connected across it's terminals. <S> Connecting a cap/inductor in the R mode and trying to make sense of the reading might be a waste of time. <S> Feel free to correct me if I might be wrong :) <S> I found a very good link <S> that explains impedance measurement in detail. <A> Just a hunch: it may be interpreting its resistance reading as parasitic resistance (parallel leakage resistance?), rather than reactance. <S> The x100 error is suspicious, though.
|
Your meter's manual refers to measuring capacitance/resistance in "parallel mode" by default, but that this can be changed to "series mode".
|
Designing a simple CW transmitter/receiver I'm interested in building my own radios for use on the air. I'm still rather new to electronics so I'm looking for simpler projects that I can build to learn on the way, and I'm thinking that a CW transmitter and/or receiver would be appropriate (unless someone wants to tell me otherwise?). I've been reading the ARRL Handbook and "Experimental Methods in RF Design" trying to find the information needed to build one of my own but to no avail. While there are circuit diagrams available in those texts, even the simplest ones are too much for me to understand (I did read the accompanying text) and I feel like if I were to just build it and have it work, it wouldn't progress my understanding of it at all and I'm not confident in my ability to build something large without error if I don't have the possibility of testing the smaller subcircuits individually. Looking at the block diagram below, I decided to look into the workings of an oscillator, since it doesn't need any other parts of the radio to function, so it'll be easier to test on its own. I understand how a simple RLC circuit behaves (mathematically), but when it's combined with other components to make a complete oscillator, that's when I get lost. "Experimental Methods in RF Design" mentions the NE602 IC which has an oscillator and a mixer, making it easier to make a complete radio. Another way I could go is using this for the radio and work on the amplifier and antenna, then replacing the IC with discrete components, but I couldn't find any circuits using this IC that are simple enough that I can confidently build it error-free. Any advice on the path I should take (be it one I've mentioned above or not) and useful resources to get me started or get past the barriers I'm facing? Thanks. <Q> To actually use a CW transmitter in the ham bands you would need a license. <S> Further you need to insure that you do not cause harmful interference to other users which could happen because of a problem with the circuit. <S> Why not start with a receiver which does not require a license? <S> You could start with a simple crystal set. <S> Here is a YouTube video showing how to make one: <S> http://www.youtube.com/watch?v=skKmwT0EccE <S> You will learn skills like how to wind a coil, how to tinker with a debug a simple circuit to get it working etc. <S> It will not require much time or money. <S> Once you get it working <S> you can re-use the parts to make a more sophisticated one transistor set. <S> More importantly you will be reusing the skills you acquired getting the first set to work. <A> It's been so long since I used my ham license that I forgot my most recent callsign :-) <S> I would start with the ARRL Radio Amateur's Handbook: it has lots of practical circuits from simple to very complex <S> and you can build a CW transmitter from it in an evening. <S> The first one I built <S> , waaay back in high school was Doug DeMaw's Tuna Tin 2: a two transistor CW transmitter that fit on top of a tuna can. <S> It's fun and you will learn a lot. <S> These days, a lot of what I used transistors for back in the mid 80's - my high school years, can be done much easier using opamps. <A> I've been reading the ARRL Handbook and "Experimental Methods in RF Design" <S> These are the prime sources for building Ham Radio equipment. <S> Let me assure you that there is plenty of info. <S> The issue is that building radio equipment, even "simple" rigs, is actually very complex. <S> I have found out that it is not enough to just throw together some parts according to a schematic. <S> You need to understand the underlying theory and have some practical experience. <S> I recommend to read and reread those books, especially the first few chapters in each. <S> And keep trying to build. <S> With time comes experience. <S> Any advice on the path I should take... <S> You can order a Pixie II kit on eBay. <S> These cost generally $2.75 or so post paid free shipping. <S> Very cheap! <S> If the first one that you build doesn't work (and this is possible) then order a few more and keep trying. <S> There is a nice one that comes with a case/enclosure for $5. <S> Find some other hams who also like to build radio equipment, and make a friend! <S> Actually, we would like to have someone to build with. <S> Your question is seven years old, but this advice still applies. <S> Good luck.
|
If you're into CW, I'd recommend starting with a direct-conversion receiver since they are very easy to build and work pretty well for CW reception.
|
TSSOP pads lifting when soldernig I am just soldering my first ever TSSOP-16, on my first perfect toner-transfer board, and when I am tinning the TSSOP pads some of them are lifting up. Which of these is most likely responsible? The soldering iron is too hot (it's an uncontrollable 30w) The board is inferior (it was cheap stuff from Maplin) I have Parkinson's (Apologies to anyone who has a relative / friend with Parkinson's - no offense intended) Something I haven't thought of? All of the above It's not the end of the world - the pads have remained pretty much in position and I have successfully soldered in the TSSOP chip, but it would be nice to know why it happened so I can stop it happening the next time. <Q> My first thought is inferior board. <S> It never happened to boards we purchased from a professional PCB shop. <S> The iron is also a suspect. <S> 30W is not a particularly high value, but the fact that it isn't temperature controlled means that it becomes very hot, no matter how low-power it is. <S> The iron's power determines how fast heat can be drained from the iron without dropping temperature, and SMT pads don't draw much heat, so the iron will be and remain too hot. <A> If you are planning to do more SMD work I certainly would recommend buying a tempearture controlled iron. <S> Mine costed ~ $40, so it probably won't empty your wallet. <S> How long did you touch any individual pad? <S> When I am doing semi-mass production a single pad takes less than a second. <S> But that is on a pre-tinned (or rather pre-golded) board, with solder mask (which might act as additional glue). <S> If you make your own PCBs it might help to make the pads as large as possible, preferrably extending beyond the area you intend to solder. <S> If you are into more SMD work you might consider using solder paste and a solder oven. <S> temperature controlling a small toaster oven might be a nice project. <A> I tack down two leads on opposite corners, whilst holding the chip in position, using my smallest tip cartridge in my Metcal system, apply plenty of jelly flux, and drag-solder the leads with a mini-hoof cartridge. <S> The Metcal system was bought second-hand and was quite cheap. <S> They have exceptional temperature control, I never have problems with pads lifting on my home-made PCBs and commercial ones. <S> The Metcal system I use comprises an old STSS power unit with the later MX-500 handpiece and cartridges. <S> It cost me £120. <A> If you're tinning the board with your soldering iron, yourproblem is that the small trace is sticking to the iron. <S> The surface tension of the solder is a large force for small features,and the copper-to-board glue has a small area. <S> Large beats small. <S> Adhesion might be better with a larger pad, but the usual solution is to use a bit of adhesive to hold the component down to the board, then flux and solder the joints (the component would have to lift, too, in this case). <S> In hand-solder application, I've also held the component down with a bamboo stick while applying heat.
|
A temperature controlled iron is a must for decent soldering, both SMT and PTH. I have no experience with Maplin, inferior board quality might be an issue.
|
Current sensing power supply for hobby servos I'm trying to design a power supply for 8 hobby servos which senses the current consumed by each of them. Size and cost are the main drivers, much more so than absolute accuracy. I'm designing for 1A per channel, although it might be nice to have a couple of 2A channels. I have an MCU with enough single-ended analog inputs. If the best way to do this is differentially, which I suppose might happen if the sense resistor is on the high side of the load (is there a reason to put it there), then I might need more pins and would probably add something like a MAX11609 or MAX11611 I2C ADC . Apart from knowing whether it needs to be single-ended or differential, I've got the ADC part under control. My questions are: Where do I put the sense resistor? How do I do sufficient low-pass filtering that I can sample these current measurements slowly (say 10Hz) and smooth out whatever is going on at frequencies higher than that? Does it make sense to look at something like a current-mirror FET? I can't seem to find any with current ratings remotely as low as 1-2A, they all seem to be rated for 40A+ and sized accordingly. It would be nice because then there would be over-current protection, but I can certainly live without it if its going to be a 300% price difference and a 100% size difference. How do I amplify the signal? Does it make sense to amplify before or after filtering, and before or after multiplexing? Is there something I'm not thinking of which is going to cause a major problem. ;) <Q> An alternative approach might be an IC like MIC2545A , Programmable Current Limit High-Side Switch. <S> To address your question points: The device replaces the current sense resistor, goes on the high side , and also provides a current sense flag output, current limiting, and a logic level MOSFET switch with a 50 milliohm RdsOn, in an 8-pin package (6 actual pins). <S> An optional series RC, parallel to the current limit set resistor, provides the smoothing out / low pass you need - see datasheet. <S> The steady-state current limit is set by a single current limit set resistor, to a range of 500mA to 3 amperes, although with only a +/-20% accuracy in your desired range. <S> Also, as the IC is a current limit switch, your requirement for limiting is met. <S> Not required, the device does this internally. <S> Yes: This may not be the optimal part for your specific need, there may be similar, cheaper devices that meet your specification. <S> Experts in this community would suggest better alternatives, I trust - an n-way array version might exist, so you can address your 8 motor requirement with 8/n of the ICs. <S> As both through-hole DIP and 8-pin SOIC packages exist, you have both prototyping and final product options. <A> Your options are high-side or low-side. <S> High-side is the best, because it doesn't modify your ground level, but is also the more difficult to measure. <S> Choose a value which doesn't decrease the power supply too much, yet gives you a decent reading. <S> A few hundred millivolts voltage drop may be a good value. <S> An RC filter helps, and you can take advantage of the microcontroller to average over the latest N samples <S> (This is effectively a FIR filter) <S> Current mirror would be nice to create at the mirror side a larger voltage (larger sense resistor), but I don't think you'll need it. <S> Depending on the accuracy you may or may not need amplification. <S> If you do need it a non-inverting opamp amplifier is a solution. <S> You need an RRIO <S> (Rail-to-Rail I/ <S> O) type. <S> \$\frac{V_{OUT}}{V_{IN}} <S> = \frac{R_1 <S> + <S> R_2}{R_1}\$ <S> No, unless I forgot about it myself as well :-) <A> Putting the sense resistor in the high-side is generally recommended. <S> If it's in the low-side, the micro will operate at a different ground than the load (lower by the voltage drop across the sense resistor). <S> There are pre-fab devices like TI's INA213 that contain a high-precision, low-offset differential amplifier with fixed gain and are intended for high-side sensing. <S> You can easily add a low-pass filter to the output of this device (plus scaling, if necessary) and bring that signal to your ADC. <S> It's a matter of accuracy and linearity vs. cost - the INA213 is a bit pricey but is extremely accurate and very small. <S> If your power supply has a remote inhibit, you can use the micro to cut the power if one of the rails goes over-current. <S> You may also wish to consider self-resetting PTC polyfuses in series with each servo (again on the high-side) which will give per-motor protection. <A> Depends on what you do with question 3 and amount of wires per motor. <S> I'd suggest to sample it 20x of your position loop bandwith. <S> Normally its >=2Khz. <S> The current sense, amplifier and torque PID normally are insulated from rest of curcuit. <S> Having high power side insulated, gives you more freedom to choose where to put current sense. <S> For 3-4 wires motors, you will need only 2 sensors per motor and some triginometry calc to derive the 3rd, 4th value. <S> Dont filter ouput current or use very rudimentary 800Hz filters. <S> For 1-2A current it is not important. <S> Have amplifier individual per motor. <S> If you multiplex amplifier, then you will have much worse frequency bandwidth. <S> The major rule is that current PID bandwithd should be 10 times higher than position PID bandwidth. <S> Also you dont need to have more than 800 Hz for current loop. <S> so 80Hz is normal limit for PID. <S> Choosing bandwidth is sort of frequency planning to decouple 2 cascades of servo, to avoid chaotic oscillations. <S> You possibly did not think that motor servo is actually a PID position servo cascaded to Current/Torque servo. <S> So the current should not only be measured, but be tightly looped throught local current amplifier controlled by its own PID fed from current sensor.
|
You could also build discrete differential op-amp circuits (use a rail-to-rail part with low input offset for accuracy). I would go for low-side, i.e. resistor between ground and servo V-.
|
Limiting current from a 5v power supply Possible Duplicate: Choosing power supply, how to get the voltage and current ratings? I have a 5V / 1A regulated supply. I'm thinking of using it to power a PCB which asks for 5V and 0.25A. In this case, I am thinking the power supply could be too powerful and damage the board.... My best guess is to put a resistor in parallel with the load, so 3/4 of the current goes through the resistor, and the rest goes through the board. Any ideas? <Q> The power supply will definitely not damage your PCB. <S> The PCB will draw only as much current as it needs. <S> You could even use a 5V/10A regulated supply and it will give only as much current as the PCB asks for (in your case 0.25A). <S> So there's no need for resistors. <A> The current rating of a power supply is the maximum it can deliver if the load demands it . <S> A power supply can not dictate both the voltage and the current. <S> In this case the supply will keep the voltage at 5V and the load will draw whatever it needs. <S> The 1A rating means the load can draw up to 1A before the supply might not be able to keep its output at 5V. <S> That all said, there are some supplies that are designed to require a minimum current to function. <S> Those that work like that usually require something like 10% of the maximum rated current. <S> In this example, it would mean the supply is only promising to keep the output at 5V if you draw from 100mA to 1A. <S> What happens above or below that depends on the supply. <S> The only way to know if your supply has a minimum current requirement is to check its datasheet. <S> If your board draws 250mA, then almost certainly the supply will be able to maintain the specified voltage, but always check the datasheet anyway. <S> In no case can the supply somehow force 1A thru your board at 5V if the board only wants to draw 250mA. <A> Well, talking about ideas, this is a bad one. :-) <S> You'll lose 3.75W for nothing. <S> It's the load that determines this . <S> If the load needs 250mA it doesn't matter if the power supply can deliver a thousand amperes, there will only flow 250mA. <A> Under normal operating conditions you will only pull 0.25A from your power supply.
|
The regulator won't supply 1A if the load only requires 0.25A. Use a fuse if you are worried about an overcurrent condition damaging your board.
|
Can a water drop do a "fake" touch on a capacitive touchscreen? If a capacitive touchscreen is hit by water drops, like rain, will it report a touch? I'm trying to evaluate what is the most appropriate touchscreen solution to use in an outdoor kiosk. <Q> Capacitive sensors react to the polarization of a conductor or dielectric that touches (or is close enough to) its surface, so the size or connection to the water should be taken into account. <S> An isolated drop might not affect it, while a stream of water will. <S> I have a large trackpad for my computer, and an app that visualizes its input. <S> I put a fairly large drop onto the trackpad, and it wasn't registered. <S> When I touched the drop, it opened a path for the electric field into my body and that activated the sensor and registered, etc. <S> So what I'm saying is that drops of water on a touch surface won't affect it by themselves, though your results might vary depending on the implementation. <S> Still, water on the surface might mess things up since it could make streams, or connect with someone's finger and cause jitter. <S> Hope that helps. <A> Water drop on capacitive touchscreen is usually recognized as a touch. <S> But, there are certain ways to avoid the fault output. <S> Try to search with "digisensor" in YouTube. <S> Answer is simple and clear. <S> If we have a high resolution capacitive sensor, then we can distinguish water from human finger touch. <S> Generalized answer to avoid water drop is not easy because it is strongly dependent upon surrounding electrical pattern. <S> (Precisely speaking, any electric field absoption). <A> My phone (SE Xperia X10) goes nuts if there's water on the screen. <S> Even a sweaty finger can make it mis-register touches. <S> If it's a worry I'd suggest a resistive touchscreen. <A> The short answer is if you want to use pro cap (projected capacitance) touch, you need something that has been specifically designed to not produce ghosts touches in the presence of moisture. <S> This is considered a premium solution and they do exist <S> but they tend to be prohibitively expensive. <S> Search for "rugged" or "ruggedized" touch screens and you will find them. <S> I have worked in the touch screen industry for 4 years (N-trig) and another 4 at Synaptics doing Clickpads and touch pads. <S> I am now at my 3rd capacitive touch company (an Austin TX based startup). <S> Some of the stuff posted here is pseudo intellectual technobabble. <S> Examples: "Capacitive sensors react to the polarization of a conductor or dielectric that touches (or is close enough to) its surface... <S> So what I'm saying is that drops of water on a touch surface won't affect it by themselves" <S> All of this is wrong. <S> We call this mutual capacitance or trans capacitance (short for transmission line). <S> This is the most common kind of capacitive sensing and it has zero to do with polarization of anything. <A> I can confirm that the drool of my nine month old son also makes my Samsung Galaxy S go bananas. <S> I seem to recall (might not be the case though) that when it has happened, the false touches seem to register further out on the display than i actually touched. <S> Can be totally wrong there though. <S> Also, if my memory serves me right, it never actually registered any touches when i didn't actually touched it, just that they registered wrong. <S> Take my word with caution though. <S> Next week when my son gets back from a trip, i can do some bug testing if you'd like ;) <A> Capacitive controllers can support water rejection though algorithms or sensing methodology. <S> Generally each vendor has "secret sauce" to make it work <S> but they expose a method to leverage that sauce. <S> The ultimate challenge will be what volume of water do you want to allow to be present and still support touch operation. <S> Resistive doesn't have this issue since the water does not have any interaction with their sensing method, BUT resistive has <S> it's own challenges in the kiosk environment. <S> Resistive is dependent on physical displacement of two layers to create a contact and therefore an electrical circuit which will wear out. <S> Additionally depending on the type of resistive interface you may have to handle calibration and sensor drift. <S> All solutions have their trade offs so make sure you weigh which is more important to you.
|
And yes, drops of water on the sensor, if they produce signal above threshold, will cause ghost touches (which is the industry standard term for "false touches, etc.). Capacitive digitizers react to changes in the amount of signal that is capacitively coupled from the transmit side (i.e. rows) to the receive side (i.e. columns) of the very tiny capacitor in the sensor.
|
Why does the distance between the plates of a capacitor affect its capacitance? Why does the capacitance of a capacitor increase when its plates are closer in distance to each other? <Q> Intuitive approach: if the distance wouldn't be a factor then you would be able to place the plates at an infinite distance apart and still have the same capacitance. <S> That doesn't make sense. <S> You would expect a zero capacitance then. <S> If the capacitor is charged to a certain voltage the two plates hold charge carriers of opposite charge. <S> If the distance becomes too large the charges don't feel each other's presence anymore; the electric field is too weak. <A> FIG 1 to 4: Capacitor: <S> It is obvious that as the distance between plates decreases, their ability to hold charges increases. <S> fig.1 <S> = <S> If there is unlimited distance between plates, even a single charge would repel further charges to enter the plate. <S> fig.2 <S> = <S> if distance bet plates decreases, they can hold more charges due to attraction from the opposite charged plate. <S> fig.4 = <S> with minimum distance between the plates, the max attraction between them enables both to hold max amount of charges. <S> As Capacitance C = q/V <S> , C varies with q if V remains the same (connected to a fixed potential elec source). <S> So, with decreased distance q increases, and so C increases. <S> Remember, that for any parallel plate capacitor V is not affected by distance, because:V = W/q (work done per unit charge in bringing it from on plate to the other) and W = <S> F x d and F = q <S> x E <S> so, V = F x d /q = <S> q <S> x E x d/q V = <S> E x dSo, if d (distance) bet plates increases, E (electric field strength) would drecrese andV <S> would remain the same. <A> Capacitance is charge per EMF. <S> Specifically Farads are Coulombs per volt. <S> As you move the plates closer at the same applied voltage, the E field between them (Volts per meter) increases (Volts is the same, meters gets smaller). <S> This stronger E field can hold more charges on the plates. <S> Remember that the charges on the plates would otherwise repell each other. <S> It takes a E field to keep them there, and the stronger the E field the more charges it can keep there. <S> The higher charge at the same voltage means higher capacitance (more Coulombs at the same Volts). <A> To get technical, you want to look at Coulomb's law . <S> This states that "The magnitude of the Electrostatics force of interaction between two point charges is directly proportional to the scalar multiplication of the magnitudes of charges and inversely proportional to the square of the distances between them." <S> - Wikipedia <S> The formula for this is: \$F = k_e \frac{q_1 q_2}{r^2}\$ <S> Where \$F\$ is the electrostatic force between two charges, \$k_e\$ is a 'proportionality constant' <S> (eg the dielelectric constant in a capacitor), and \$r\$ is the distance between the two charges \$q_1\$ and \$q_2\$. <S> There are other forms of the equation - such as this one specifically for an electric field: <S> \$E = \frac{1}{4\pi\epsilon_0}\frac{q}{r^2}\$ <S> Which tells us the force at a distance \$r\$ from the single point charge \$q\$. <S> If you want to start getting really technical <S> then you need to start reading up on quantum mechanics and the interactions between particles and the energies involved in it. <S> When two particles (say electrons in this case) <S> interact they send quantum particles between them (photons). <S> These, like the rats in the basement, require energy to move. <S> The greater the distance the higher the energy. <S> The higher the energy taken to move the photons the lower the charge left between the two plates. <S> That's a very simplistic view of it and there is one helluva lot more detail in there to be discovered - such things as Quantum Tunneling, Leptons, Fermions, Bosons, etc. <S> It's fascinating reading if you have the time. <S> I'd recommend Steven Hawking's A Brief History of Time as a good starting point. <S> Follow that up with F. David Peat's Superstrings and the Search for the Theory of Everything <S> and you won't go far wrong. <S> While both these books are getting a bit long in the tooth now and the theories are all still evolving, they give good insights into the workings of the universe at a subatomic level.
|
Opposite charges attract each other, creating an electric field, and the attraction is stronger the closer they are.
|
Can you run an x86 class processor ramless? Modern x86 processors have at least 512K of L2 cache. There are applications which would fit entirely into this amount of memory. Can you run these chips with no RAM attached? If so, is there a way to do it that eliminates the writeback timing penalty when the CPU attempts to maintain RAM coherency? I don't have a specific application in mind, it's just idle curiosity. I'm certain that somewhere there is a niche application where this would be useful though. <Q> Yes you can. <S> By faking reads from consecutive (non-existant) physical memory locations, you set the tags in the cache. <S> Then you switch off further filling of the cachelines and enter writeback, thus confining reads/writes to the cache and it will behave as a normal RAM. <S> Some of the bios-replacement projects do this because then you can spend much more code on the chipset and chipset memory controller setup, so you can write it in C for example. <S> This practice is widely used for embedded-class CPUs as well for handling bootloaders. <S> The methods to turn the cache into a RAM-like mode vary a bit. <S> For a brief low-level introduction you can check out this presentation. <S> Note that as others have pointed out you still need to load the boot-code from somewhere obviously. <A> When the CPU comes out of reset, cache is turned off. <S> The BIOS is what initially configures and clears out the cache. <A> I don't know how accurate this is, but these are my thoughts: <S> I don't think there is any way you can get programmatic access to the cache. <S> You cannot guarantee from one instruction to the next what would be in the cache, and where it would be located, so you cannot use it reliably as RAM even if you could access it directly. <S> You could run an x86 without RAM, but you wouldn't be able to get it to do very much useful. <S> You'd be restricted to purely using the internal registers for data storage.
|
So no, you can't run it RAM-less because there is no RAM to boot the thing up in the first place.
|
Extremely Long Prototyping Board for LED Strip I'm interested in building very long led lighting strip. Does anyone know where to get a long, thin prototyping/matrix board(if they even exist)? I can't think of a good approach for mounting/soldering two dozen leds(each with its own resistor) in parallel, spaced 2" apart. Thanks! <Q> However, with 24 LEDs spanning 4 feet, I'd want something more permanent. <S> I think the cardboard trick with maybe a wooden dowel glued to the back for stiffness would do it. <A> All Flex claims they can make flexible printed circuits 40 feet long. <A> 24 LEDs 2 inches apart, that's 46 inches (+ 1 solder pad :-)). <S> The largest PCBs I've seen (minicomputer in the 80s) were 25"x25", and I don't believe that they're being made bigger than that. <S> You'll have to construct the strip from several shorter pieces. <S> Unless you wanted to leave them bare this isn't necessarily a problem, since the seams would be hidden from view. <A> Prototype Boards - Perforated <S> (alas, it's just insulation with holes in it: no copper) <S> Injectorall Electronics B3429D: 17 inch x 5.75 inch (431.8mm x 146.05mm) <S> Drilled Copper Clad Breadboard <S> (completely covered with copper; you'll have to cut/etch a slot down the length to separate the LED terminals) <S> Vector Electronics 169P44WEC1: 17" x 4.5" (431.6mm x 114.3mm) <S> Prototype Boards - Perforated; FR4 Fiberglass and Copper Clad, Single Sided, 1 oz. <S> (completely covered with copper; you'll have to cut/etch a slot down the length to separate the LED terminals) <S> I'm pretty sure they're all available from Mouser, Newark, or Digikey.
|
If you want them 2 inches apart, you could punch the leads thru cardboard and solder them together with the resistors on the back. If you're asking about off the shelf true prototyping boards, many of the common "solderless breadboards" can snap together to make any length and width you want. The longest off-the-shelf protoboards I've seen are 17 inches long: Keystone Electronics 3407: 17 inch x 2.73 inch (431.8mm x 69.3mm)
|
Difference between regulator's tab and pin? Take the classic LM1117 linear regulator http://cache.national.com/ds/LM/LM1117.pdf Pin 2 is Vout. The tab is also Vout. From a component layout perspective, it would be nice if I could connect the output capacitor to the tab instead of the pin. What is the functional difference between the tab and the pin? Does one have less current carrying capability than the other? Can I leave the pin floating and instead connect the tab to the output capacitor? Must I hook the output capacitor to the pin? Can I just short the pin to the tab? <Q> I think the explanation behind this is interesting as well. <S> Basically, the exposed tab and the middle leg is generally the same piece of metal . <S> Basically the tab and middle leg are formed from a single piece of bent metal. <S> It may even have a (marginally) lower inductance. <S> Image borrowed from ESP <A> There is no functional difference between the tab and the pin. <S> On a lot of SMT versions of these kind of devices the middle pin is cut off and not used - just the tab is used: Tabbed packages are often used for SMT parts which, despite their small size, have to dissipate up to a couple of Watts. <S> The tab will drain much more heat to the PCB's copper than a pin would. <A> There's no important electrical difference, but of course the pins are solder-plated and all have similar thermal response to soldering temperature. <S> Making a solder joint to the tab will be more challenging than soldering to the pins.
|
Therefore, the tab literally is the middle pin, and you can use it as such.
|
Why some through-hole component leads are ferromagnetic? I was surprised when saw a magnet picking some random 0.125 W resistors leads. I was always thinking that leads are made of copper alloy. Are there some subtle reasons to make them using metals other than tin plated copper or copper alloys ? Thank you <Q> Main (and probably only) reason: copper is expensive . <S> I've commented before that cost can outweigh other factors in design and production, and that goes also for components. <S> Every milli-cent counts. <S> I guess that for very low value resistors (< 0.01\$\Omega\$) copper may be used. <S> Copper is also used for certain power devices because it conducts heat better. <S> These 500\$\mu \Omega\$ examples from Isabellenhütte illustrate both: <A> Leadwires are attached to provide both electrical and mechanicalsupport, and must survive the component fabrication process beforethey are assembled. <S> Glass/metal sealing for diodes and some kinds of resistors makesKovar and similar alloys <S> (iron/nickel based) a natural choiceof lead material, because it adheres to glass and doesn't cause stresses on cooling. <S> Even when (for power diodes like 1N4001)the leads are heavy copper for cooling, <S> the button that seals againstthe glass is a magnetic material welded to the copper. <S> Nickel, also ferromagnetic, is frequently employed on surfacemount components, because a thin layer of nickel will hold solder,where a thin layer of copper might dissolve into the solderinstead. <S> Copper isn't suitable for thin items that must be fired at hightemperatures (it oxidizes), and has chemical incompatibility with some materials (only a few high-tech ICs have copper near the silicon parts). <A> With resistors, there are two technical reasons I can think of: <S> Heat: <S> Copper is not only a very good electrical conductor, it is also a very good conductor for heat. <S> Sometimes, you want to run a large and leaded resistor's body at 155 °C (or even 175 °C, if its specification allows), but for safety reasons and regulations <S> (UL, mostly), you may be limited to 130 °C at the solder joint (for standard FR4 and solder). <S> With some sort of steel alloy, the heat remains around the resistor, and can be radiated or convected into the air around it, while not so much heat is transferred into the board. <S> Mechanical Strength: <S> Sometimes, resistors are not mounted flat onto the board. <S> (Sometimes, it is because of constraints of pick-and-place machinery, sometimes it is done to save space, sometimes, you want the hot resistor at a distance from the board - see "Heat"). <S> When the board is exposed to mechanical shocks or vibrations, the resistor's leads last longer when they are made from some stronger material like steel. <S> The process of forming the leads automatically may also be a bit rough on the leads, and steel may be the better choice. <S> Both reasons are especially true for NTC inrush limiting resistors. <S> You want them to run as hot as you can ( <S> e.g. 175 °C) <S> because then, the resistance is low and the losses are low. <S> At the same time, you need power to make and keep the resistor hot, and in order to save power and keep the solder joints from becoming hotter that (e.g.) 130 °C, you don't want the heat to go away from the resistor's body into the board . <S> And typically, an NTC's leads are bent to keep the body <S> 5...10 mm above the board, requiring some mechanical strength of the leads. <S> All of these reasons favor steel. <S> Then, of course, cost may also be an issue that favors steel alloy. <A> The only reason I ever learned is this: <S> The leads were made of steel so a person would be less likely to destroy the component while manually soldering, since steel is about 20x less thermally conductive than copper. <S> (Manual soldering was the only assembly technique for years.) <S> Typically it was semiconductors and some capacitors that had this kind of lead - and still do. <S> The other answers have a valid point about strength, though: in leaded resistors sometimes the little caps inside the epoxy that connect the copper leads to the thin-film resistive element are made of steel - I assume for strength. <S> I have never seen a resistor lead made of steel though.
|
The alloy used (no, I don't know what it is) may have a higher resistance than copper, but over the whole the difference will be negligible. Instead, the leads are bent to allow for upright mounting or to keep the resistor at a certain distance from the board, or both. In the old days, some components were easily damaged by heat.
|
How to measure electrical noise? Another question I asked about Power Supply Noise pushed me to ask this question. The background is as follows: I have a design with PD (Photodetector) and Opamp at medium frequency (couple hundred KHz). I am trying to test my design when it come back. I am looking at power supply noise, opamp supply noise and opamp output noise. I realized, there is more to it than just touching the probe. People talk about power loops, having special cables etc. I will spin another board in two weeks and I wanted to ask, if you are designing your board now, what type of test points or elements you place on your board so that you can measure the noise accurately. We are talking about <20mV type signals. Bonus question: The output of the opamp is connected to an ADC of the processor. Could I simply run the ADC and plot that to gain a deeper understanding of the noise compared to connecting my cheap scope? <Q> Noise is difficult to measure, and the amplitude you see on your your scope is only a first indication of the level. <S> Do you want to measure absolute noise levels, or just comparative? <S> In the latter case the scope could be a good instrument, but at the given levels the average $500 scope will have so much noise itself that any measurement becomes in fact meaningless. <S> You need a high quality scope + ditto probes to do this. <S> The difficulty with measuring noise is that it has a wide bandwidth continuous energy spectrum (the continuous spectrum makes it hard to separate noise from signal, notch filters may work). <S> Ideally you measure the noise energy through RMS-to-DC conversion . <S> This is not for the faint-of-heart, as your RMS-to-DC converter has to be very sensitive due to the low levels, and wideband. <S> And of course be low-noise itself! <S> Liquid nitrogen helps :-). <S> In any case, absolute signal-to-noise ratio is not as easy as reading amplitudes. <A> Typically you design some parts on the first spins of the board to help measure stuff like this. <S> Be generous with footprints for bypass caps and filters, and put for example SMA coaxial contacts at key places in the signal chain, but put them with a removable SMD 0-ohm resistor at the T-junction <S> so the stub doesn't have to affect the signal chain if not used. <S> For low-frequency signals you could hook this directly to the scope, but the SMA coaxes have a good feature in that some of the probes with the ground-lead built into the tip can be stuck into the center position of the coax connector and the ground lead will or can be made to touch the shielding.. note though that for best results with a probe it need to be active, and then the probe itself will cost 3000 dollars <S> :/ <S> With photodiodes and transimpedance amplifiers you have the problem that for a typical setup (you don't specify your parameters..) <S> you have only a couple of microamps and a transimpedance gain of several hundred thousand. <S> Inserting elements at the PD-side of the opamp, and indeed just having PCB traces in the vicinity might disrupt your precision and noiselevels. <S> The solder resist coating has a non-infinite resistance for example. <S> Therefore if you can control the gain of the ADC and know you have designed a very good ADC circuit (this you can test separately, with separate SMA coax inputs to it for example) you can use that as a probe replacement yes if the sample rate is high enough. <S> This is a good solution and postpones the need for the more expensive scope (you can rent these if you really need it sometime though). <A> Stevenvh is correct that noise can be very hard to measure, but I would like to point out a different point of view. <S> The only time that noise really matters to you is when it affects your readings. <S> This means that you can just take an ADC of your input, pass it off to a computer, and then do some math on it. <S> I am not entirely up to speed on your project, but I assume that you "know" what your signal is that you are receiving. <S> Most systems tend to convert this to a dB scale. <S> This won’t tell you anything about what part of the system the noise is coming from, but it will allow you to have an idea as to how well your system will function overall.
|
You can come up with an SNR figure by calculating the average power of your signal and then to get your noise, just subtract the known signal from your ADC signal, find the power of the resulting noise, and then divide the two.
|
Do Peltier and heater make a good combo for thermostat? I want to achieve 0.05C stability in closed space to fit 24 bit ADC and voltage reference. Is it a good idea to use Peltier and heater together ? <Q> The Peltier can be used as a heater as well, just invert the power connection. <S> But this kind of precision is very hard to obtain, no matter how you heat or cool down. <S> The reason is the thermal inertia of your system: switching on the Peltier if the temperature rises won't result in an immediate temperature decrease . <S> So you'll have to model <S> the thermal environment and use a PID regulator which takes the thermal inertia into account. <S> To get 0.05°C precision the environment has to be extremely stable , which means very good isolation from the outer world. <S> It's not impossible, but anything but easy. <S> Temperature stability is a requirement for meaningful 24-bit ADC, but you'll also have to take care of offsets and (thermal) <S> noise . <S> Offset case in point: ADCs usually have low impedance input. <S> A 5 cm trace to the input may give an error of several bits! <S> 1 LSB in 5V is 300nV. <A> First, a Peltier device is a heater. <S> It is also a cooler. <S> Which it is at any one time depends on the direction of the current flow. <S> Controlling anything to .05C is not trivial. <S> You will need a sensor with somewhat more resolution and repeatability than that. <S> Get your wallet out. <S> Once you have a suitable sensor, Peltier device, and good insulation, the rest is up to the control loop. <S> That's a whole other subject, but there is nothing special about the control loop once you have the right feedback sensor. <A> I was thinking about similar problem, and there are few suggestions: 1) <S> If you have small, well-isolated stabilized area, you don't need bulky peltier for that. <S> You can desolder usual peltier array, and use just several elements for it (they usually have about 100 modules inside) <S> 2) I personally decided to end up on heating-only on above-ambient temperature (40C) - <S> this is much easier/cheaper and does not need much power if isolation is good enough, as this might be completely sealed system in a vacuum for example, while peltier always conduct heat as it needs to be connected to outer world.
|
Yes, a Peltier device would be appropriate if you think the ambient temperature could be both above and below the desired temperature.
|
What is the simplest way to make an oscillating signal? Imagine you have a black box with 5VDC and ground inputs and you have to create one output that is an oscillating signal. What is the simplest circuit that can do so? Can you create a tank circuit with an inductor and capacitor? The output signal will be detected by a PIC. The frequency is not important but should be rather low (between 10 and 500Hz). The PIC will not measure the frequency but only detect if the oscillating signal is present or not present based on whether this "box" is connected or not. That means the signal can be sin, square, saw-tooth, whatever, the shape doesn't matter. Bonus points for the cheapest, lowest component count and lowest real estate solution! <Q> Lowest component count I can think of: <S> The 74HC1G14 is the single gate version of the 74HC14 in SOT-23 package. <S> OK, I lied. <S> You can do it with less. <S> Take a microcontroller with an internal oscillator and write this incredibly complicated program to output a square wave. <S> Number of components: 1. <S> Board space: 6 mm\$^2\$. <S> If you drop the frequency restriction you can use a LED: f ~ 374740572500000 Hz. <S> ;-) <S> Also out-of-spec is the Schmitt-trigger inverter with the output connected to the input. <S> That's also a 1-component solution. <S> Should oscillate at a few MHz. <A> You want low component count? <S> How about this: You apply power. <S> The relay activates. <S> The contacts open. <S> The relay deactivates. <S> The contacts close. <S> The relay activates ... <S> It's also good as a buzzer, and for generating nice flyback voltages. <S> Be warned - the flyback voltages could kill a µC. <S> But hey - it's one single component - you can't get less than that without getting all quantum... <A> One part: an ATtiny 13. <S> Yes you'd have to program it to output a square wave, but it's a mere 8 pin device <S> , you can run it on low voltage, and easily hit the frequencies you mention. <A> You can use a pkg of 6 gates for lowest cost (as they are so common) or for minimum size use some of the tiny single gate packages. <S> You could also use an opamp or comparator for the same purpose. <S> A unijunction forms a relaxation oscillator with a very few parts. <S> Neon and cap and resistor if getting desperate. <S> Esaki / Tunnel diode and R !!! :-) <S> . <S> BUT if you want something smaller and cheaper and electronic that arguably satisfies the letter and spirit of you requirement even though it does not look like an oscillator in isolation, and that requires a single 0402 <S> packaged 1 cent component then - Software driven sawtooth oscillation cycle. <S> Exponential charge of a capacitor using weak pullups, preceded by discharge of capacitor. <S> With care this gives minimal cost, minimal area, no power drain except when testing (and hardly then), no EMI etc when not in use. <S> PIC pin to Capacitor. <S> Cap other lead to ground. <S> Enable weak pullups. <S> Make pin output. <S> Set low. <S> Set pin to input. <S> Measure time taken to go high as cap is charged by weak pullups. <S> Repeat several times if desired to check value. <S> Can be multicycle oscillator or single cycle. <S> Needs: <S> One capacitor, relatively low value. <S> Can be 0402 if desired (breathing hazard :-) ) <S> Pin can even be used for other purposes if desired if cap not too large. <S> Weak pullups vary in current sourcing by ? <S> 2:1 ratio. <S> The above can be calibrated by adding one more cap on board with cap >> stray and pin capacitance. <S> Cycling this cap tells you how strong the pullup is. <S> Adding offboard cap in parallel increases charge time. <S> Similar can be done with an ADC pin. <S> ADC version has advantage of part charge cycle response. <S> By looking for the shape of the exponential charge curve you can tell how much capacitance is present in << 1 RC cycle. <S> An external pullup R can be added in each case if desired. <A> I would use a 555 timer IC, in astable mode. <S> Two resistors and two capacitors. <S> Five components. <S> $0.50. <S> This is not as clever as the other answers. <S> But it will work. <S> And 10Hz or 500Hz is easily attainable. <S> And other engineers will see it and immediately understand. <S> And you can easily tune it with a pot or by swapping components. <S> This is the engineering solution. <S> I give myself 10/10 and no bonus points. <S> If what you really want is an obfuscated magical trick that depends on temperature, trace inductance, ritual animal sacrifice etc. <S> then by all means use one of the analog hacks.
|
If I wanted a "real" standalone oscillator then something like Steven's Schmitt trigger gate (which I also mentioned in the monostable query) is a practical cheap and flexible electronic solution.
|
Good 5-pin connector under moderate stress I'm fixing up some electronics at my marine ecology lab. One of the gadgets we have is a portable RFID tag reader. It runs off of a PIC board, and has an antenna that attaches to it. The antenna looks like a beach metal detector… a 4'long PVC pipe containing the antenna reader, with a 2' diameter PVC hoop (the actual antenna) at the end. There are 5 wires between the PIC and the antenna reader - 2 power, 3 data. Since the antenna reader already decodes the antenna, I'm assuming I don't have to worry too much about RF. It is advantageous to have an interconnect between the PVC contraption and the PIC board for easy storage and transportation. Previously, the lab has been using MOLEX connectors for this, and have been epoxying the crap out of it. This hasn't worked so well for them, and they have to resolder the connections every couple of months. I need a 5-pin connector that can withstand a reasonable amount of stress(almost anything has most strain relief than Molex anyway). It doesn't have to be waterproof. The first obvious solution for me was DB9. However, the PIC board already has a DB9 connection (serial for uploading the data onto a computer), and I would rather avoid any disasters that might occur from somebody accidentally swapping connections. <Q> Have you considered something like a good ol' fashioned DIN plug? <S> They're old hat, but that also means they're proven technology. <S> How would that sort of connector fit in with the scheme of things? <A> You can still use the DB9 connector , but male instead of female (or vice versa). <S> Or a 5-pin DIN plug , like Matt suggested, but I would take one with a bayonet lock . <S> The bayonet lock works like a BNC connector. <S> :-)), <S> but that's not as user-friendly. <A> I work in the test department at my company, and we use CPC connectors for many of the harnesses that are frequently handled by workers. <S> They're a bit bulky, but that's OK for our application. <S> They're very similar to the DIN plugs you mentioned, and look something like this: <S> They're available in both panel mount and freestanding versions. <S> Note that you can get both male and female plugs for either side, so you can create cables that can't be connected with a little accidental cleverness. <S> The only drawback is that they're designed for more connections than you've got; the best you're likely to find is a 9-pin version.
|
Just like there's also a TNC connector, there are DIN plugs with a threaded lock (not dreadlock! They do have good strain relief, though we occasionally epoxy (with potting compound) connectors that get plugged and unplugged dozens of times each day, seven days a week.
|
Does using 220V-110V transformer matter? An electronic device manufacturer, said that using it with transformer is not recommended. I don't understand why is that. There are products that work with both voltages, and I don't understand how the output of a transformer will be different from an "original" output. Is that really true? If so, why is it? The device is a singer sewing machine , which I want to use in the USA and in Europe, with different power outlets. Even if regular transformer wouldn't work, I'll be glad to hear about other solutions. <Q> RE: <S> why output of transformer is different ? <S> The difference is caused by non-ideality of transformers. <S> Most noticable non-ideality is the leakage inductance . <S> For switching load with some cable or filter capacitance like thryristor commutated load, there can be unpredictable runaway mode with circuit resonating at high frequencies. <S> Another non-idealities of tranformer is non-linearity, caused by saturation, which can clip the original sinusoidal waveform coming from source, asymmetric rectifier, which can magnetize core and saturate it even worse, simple raw peak rush currents, which can overload transformer. <S> Or for very simple case, it can be manufactures disclaimer to save user from misuse of little 90VA "autotransformer" when they plug 1000W hair dryier and cause all kinds of trouble. <S> What is the device in question ? <S> Looks like it is something special <A> If the transformer is 1:1, usually called a "isolation transformer" because that's why you'd use it, then it shouldn't matter as long as the transformer is capable of delivering the required power. <S> If the transformer does something other than produce the same output voltage as input voltage, then it could matter. <S> For example, 220V power tends to be 50 Hz with most 110V power being 60Hz. <S> For example, some clocks use the line frequency to keep time, but there can be other electrical reasons too. <A> Yes, it matters. <S> That transform will put out half the input voltage at twice the current. <S> Some devices can take a wide range, but many others can't. <S> And @Olin's point about 50/60 Hz is important too.
|
Some devices might not work well at the wrong line voltage frequency.
|
safe to use an unshielded jack for 10Mbps Ethernet? I have a hardware device which includes Ethernet connectivity. I have a surface-mount Ethernet jack, with top entry (as opposed to right-angle). Part number is RIA Connect AJS03B8813 , if that helps. No integrated magnetics, this is fine, I have a discrete part. All is well and good. Except that I'm having trouble finding that part, so I need to find a replacement. However, it seems that no one makes a jack that has all of the following features: top-entry (i.e., part sits flat on board and cable comes "down" into it, perpendicular to the board -- in other words, NOT right-angle), surface-mount, and shielded. I can find shielded right-angles aplenty, and shielded top-mounts, but they're through-hole (not an option, the other side of the board is densely populated.) All I can find are unshielded parts, such as Molex 0855135013 Molex 95503-6891 (yes I know this has support posts but that's OK, I have those holes on the board) So, my question is: are there any potential problems using this unshielded jack in a 10Mbps Ethernet application? I can't see the shielding on the jack mattering that much, since most clients don't use shielded cat5 anyway. Any input would be appreciated. Thanks! EDIT: thanks for all the replies. I'll just go ahead with the unshielded part. I know that most cables aren't shielded, so I figured it was probably pointless, just wanted to get some other opinions. <Q> Unless the application is in a very noisy environment shielding is not required, or really even recommended. <S> Shielding may be needed in very noisy environments such as some industrial settings <S> but it comes with its own set of problems. <S> Special attention needs to be paid to ensure that the shield is properly grounded, preferably only on one end of the cable. <S> If both ends are grounded then both systems need to be able to deal with the ground potential offset that will almost certainly exist. <S> In more complicated network topologies this can be a serious issue. <S> In short, if you don't REALLY need shielding, don't use it. <A> About the only purpose of a shielded ethernet jack would be if it mates up with the chassis in such a way to minimize the electrical hole thru the chassis for the ethernet cable to pass thru. <S> If your connector shield is only connected to the board ground and not to anything else it touches off the board, then it's just a waste. <S> I haven't used a shielded ethernet connector yet. <S> The much bigger issue is common mode radiation from the cable. <S> This is where I've seen naive designs fail FCC testing. <S> Some think this is what shielded connectors are for, but they are pretty useless for that. <S> Fortunately the solution is usually pretty easy. <S> This is inductance in series with the common mode signals both wires of a pair pick up from your board, but little or no inductance in series with the differential mode signal (the actual ethernet data). <S> You don't need much capacitance (20-50 pF) to ground on each line to attenutate the common mode high frequencies, well less than the ethernet capacitance spec or a relatively small length of cable. <S> One thing to watch out for though is what that does to the isolation voltage. <S> Ethernet is transformer coupled, in part to avoid ground loops. <S> I forget the spec, but it's 100s of Volts at least. <S> If you do put capacitors there, you either give up some of that isolation voltage, or the capacitors will need to be rated for high voltage and therefore large. <A> Even for 1Gb <S> it's not a necessity (for Ethernet itself), for unshielded 10Mbit <S> I can bet 1 million $ :-) <S> The only slight worry is If you have sensitive millivolt-range unshielded <S> signaling(like <S> sensors reading or audio inputs) on the board, you may get slight interference.
|
Shielding just the jack without the cable also being shielded is rather pointless. Take a good look at the ethernet transformer datasheet, and you will probably see extra windings in series with the outboard leads. These form a balun, or common mode choke.
|
What micro controllers or microprocessors are capable of high resolution video but are not BGA? I am looking for any chip that I can use to display video at a high resolution (at least 720p). There is no need for I know of the Blackfin processors, but they are all BGA.The PIC24FJ256DA210 is capable of only 640 x 480 sadly. I am guessing that FPGA's are the last option that I have in this situation, but am more curious about anything that would require me to do less work than having to make the entire video system on an FPGA. <Q> Nearly all CPUs are not directly capable of producing video -- some other piece of hardware (typically using a DMA engine) <S> fetches data from main memory or video memory and displays it. <S> Some of the few CPUs that are capable of directly producing video include: As far as I know, the only 32-bit CPU currently manufactured in a DIP package is the Parallax Propeller. <S> The Propeller, unlike most CPUs, has built-in hardware to support VGA video generation. <A> Maybe look at the xcore from xmos.com. <S> Bridges the gap between processor and fpga. <A> Here is a demo of someone playing video directly from an xmos chip. <S> http://www.youtube.com/watch?v=vmKrLcJGlmI <S> Here is another video of video output with an xmos. <S> http://www.youtube.com/watch?v=A5eU8pHpy-c&feature=player_embedded <S> It doesn't look high def, but from the notes he states that he is using only 2 of the available 4 hardware threads--and that's on the bottom of the line xmos chip <S> (it does 400 MIPS).
|
I see that several different Propeller boards support1024x768 bitmap graphics http://www.parallax.com/tabid/252/Default.aspx Chuck Moore's F21 , i21 , and MuP21 CPUs, unlike most CPUs, have built-in hardware to support VGA graphics.
|
Where can I get CiA 302, CiA 304 and CiA 305? Where do I find the following CANopen documents? CiA 302 CiA 304 CiA 305 <Q> May or may not be : 305 http://www.elmomc.com/support/manuals/MAN-CAN305IG.pdf 304 online http://www.softing.com/home/en/industrial-automation/products/can-bus/more-can-open/framework-dsp-304.php?navanchor=3010660 <S> Maybe here - free signin <S> http://www.softing.com/home/en/industrial-automation/downloads/drivers-demos.php?T=4&G=2&P=1 <A> The CAN in Automation website has all the CANopen sepcifications. <S> http://www.can-cia.org/ <S> However the specifications you require are in the members only section, and to access that you need to pay an annual subscription, which I think is about £3k <A> I can't seem to see Cia302, or Cia304, <S> but Cia305 is here: http://www.can-cia.org/index.php?id=915&no_cache=1 <A> Update for current status as some of the comments didnt match up with my research. <S> You can get any of the docs that are in public status by registering ( <S> eg full 301 spec) for free. <S> You can buy a series of docs for a one time fee that are in Draft Standard status (depends on series, ~512 euro for all 3xx DS), link here <S> If you want DSP specs, you have to become a paying member (i.e. 402, 305, etc), list of all docs here . <S> Membership prices are annual and based off company size, but its currently ~3800 euro for a company less than 1k people. <S> One additional thing I found recently, some of the specs that are not public are standardized through IEC (ex DSP402 --> IEC 61800-7-201 ). <S> So one additional resource for finding specs that may make more sense than becoming a full fledge CiA member if you just need a single spec.
|
can-cia.org will supply specifications by email to non-members, membership allows you to directly download them from the site.
|
Choosing the right type of wire? I am going to have a something thats runs on 120vac pulling 1-2amps is there a chart or something for choosing the right gauge of wire? The wire will run from an inverter in the front of my car(by the battery) to the back (near the taillights). I need the smallest diameter wire possible. <Q> The maximum current a wire can carry depends on several factor: wire <S> diameter , obviously (Kortuk may take exception on me using the word obvious again, but this time it really is; anyone older than 10 will understand that thinner wires will allow less current through them) wire material . <S> Often copper, but in some applications where weight is paramount aluminium may be used. <S> Think overhead high-voltage power lines. <S> insulation . <S> The energy dissipated in the wire will cause a temperature rise, but the heat will also be given to the environment. <S> The type of insulation determines the amount of thermal isolation. <S> It also determines the maximum temperature allowable. <S> Think melting of the insulation. <S> edit: insulation will also determine maximum voltage. <S> The insulation may breakdown when too thin in combination with a high dielectrical constant. <S> This document gives the maximum current for copper wire as a function of diameter and temperature rise in free air at 30°C ambient temperature. <A> There are charts that give properties for different gauges of wires. <S> However, you haven't said what your criterion is <S> so there is no way to give a single answer. <S> One spec might be how much voltage you are willing to let the wire drop. <S> At your 2A current you can compute a maximum resistance from that. <S> With the Ω/foot from a chart, you can compute the maximum length of each guage of wire your system can tolerate. <S> There are charts for temperature rise as a function of current for various wires. <S> For house wiring there are legal requirements too. <S> These are usually based on temperature rise. <S> Each gauge is specified for the maximum fuse or circuit breaker that must be on that circuit. <A> As Olin notes, if this is for house wiring it will need to meet regulatory limits. <S> If safety is a concern you may be concerned with eg temperature rise. <S> "Smallest diameter possible" may be to fit in a duct etc - you need to say WHY for this to be able to be answered well. <S> Here is a table that will both help and confuse. <S> Note that at lower current levels there are two standards that vary widely. <S> See also the note <S> re contacting your local electrician re legality. <S> http://www.powerstream.com/Wire_Size.htm <S> See also http://www.interfacebus.com/Copper_Wire_AWG_SIze.html <S> which only gives the very conservative rating and suggests you'd want about 18 gauge (likely to be unnecessary but YMMV). <S> As a general guide: http://en.wikipedia.org/wiki/Standard_wire_gauge <S> And http://www.bulkwire.com/wiregauge.asp
|
Or, your criterion might be maximum temperature rise of the wire. All that said, if smallest diameter is really your aim, then something like 26 gauge AWG is liable to be "fairly safe".
|
transistor as low-leakage diode I found this application in AD's OP77 datasheet : Apparently they (ab)use a 2N930 transistor's B-C's junction as a low-leakage diode. Is there any reason why you would choose a transistor over a real low-leakage diode for this? <Q> The reason is a recovery time. <S> Diode can have 3 microseconds, BJT is just a 5pF <S> * N <S> ohm ~ dozen of nanoseconds. <S> The diode is better than BJT in slower range, it has pA current, when BJT has nA. Physics of fast recovery for BJT can be explained by very low volumetric capacitance (low charge) of base, because BJT base is very thin by design. <S> I guess low leakage diodes have much longer distance across junction (falling doping gradients or even gaps), when carriers must approach region by thermal diffusion, tonneling, which takes time. <S> The diode gap has very large volume (comparing to BJT base volume) to fill with carriers during recovery time. <A> The reason is all about leakage current. <S> The B-C junction outperforms all low leakage diodes by more than a factor of 10. <S> A highly used low leakage diode, the BAV116, is rated to 5nA of rev leakage current. <S> The B-C junction of a 2N3904 is well below 30pA at room temperature. <S> Using the B-E junction is even lower leakage, usually around 5pA. <S> You can't touch this with standard low leakage diodes. <S> There are very pricey FET based diodes that are lower, but they are rare. <S> The B-C junction advantage is that its reverse voltage is that of the transistor. <S> Using B-E the junction will "Zener" at 6-7 volts. <S> Still quite useful in low voltage circuits or even as an asymmetric clamp. <S> Just pay attention to maximum forward current through your selected junction. <S> In the example circuit the B-C junction was used because of the range of voltage. <S> It increased the holdup time of the sample capacitor by reducing rev leakage back to the op-amp output. <S> I would think the leakage of the reset mosfet will dominate this circuit as the input bias of the AD820 is typically 2pA. <A> Thermal characteristics? <S> BJT can be heat sinked quite easily with a clip-on, or case mounted for mechanical thermal coupling. <S> If the op amp OP77 are also chosen as circular case, then a dual-mounted heatsink is simple to mechanically fit.
|
I have used 2N3904s for low leakage diodes for years with excellent results.
|
How to pack an arduino project in a box I made a project with Arduino using a breadboard. Everything is working, but nothing is soldered in place. So, for example, I cannot put it outside, otherwise after the first wind, everything will break :-) I would like to pack it in a box and actually used as a gadget. Can anyone suggest ways to do this? For example, are there any breadboard + screws that I could use, without having to solder my Arduino? <Q> I've seen Pelican cases being popular for this kind of thing. <S> There are some videos of people using them: http://www.youtube.com/watch?v=9f8nwwCbfRA and http://www.youtube.com/watch?v=3MGRx9KDDg0 Otherwise, to make a sturdy Arduino circuit without a breadboard, I'd suggest using Perfboard to solder the components in. <S> Check out http://www.instructables.com/id/Perfboard-Hackduino-Arduino-compatible-circuit/?ALLSTEPS for a great tutorial on doing this. <S> It is more work than a breadboard, but not too much and worth it for a very solid Arduino device. <A> I know this is an older question, but I thought I'd throw my 2 cents in anyway. <S> You could find a project enclosure at a local hobby shop, or gut some old electronic gadget you have laying around similar to the way I did. <S> You could always surround any opening(s) with a bead of RTV silicone; this should make it fairly water/weatherproof, and if you need to change the circuit, you could just scrape the RTV off and start again. <A> When I've wanted to use a patch board for a demonstration unit that I use elsewhere than just on a test bench <S> and I want it to be "modestly robust", <S> even though the wires are plugged in and not soldered, I have on occasion added a sheet of foam on top of the wiring and components and then sandwiched the whole assembly between two thin sheets of wood or plastic, with long screws pulling the sheets together and compressing the foam enough to hold the wiring and components in place. <S> This is very "unprofessional" but surprisingly effective. <S> Remember to account for lack of cooling due to foam "insulation". <S> Obviously this does not confer "waterproofness" (not much anyway :-) ) <S> - you still need to mount the "unit" in a suitable enclosure. <S> Even when a proototype HAS been soldered I have used a simila system to reduce the effects of people playing with the circuitry during testing. <S> To soldered board add a sheet pf suitable plastic and retain with screws at each corner and maybe through several centre points if large. <S> Works wonders. <A> You may want to have a look at this http://daddiest.com/arduino-project-case/ <S> I made use of an abandoned accessory box to make an Arduino project box. <S> It is dirt cheap. <S> I can contain any under-development project and reopen it at any time. <S> Just like saving and opening a software project.
|
I, personally, wanted a nice enclosure for my Arduino based project, so I gutted an old cable modem I had laying around, mounted some standoffs with gobs of epoxy, and it has turned out great so far.
|
Efficiency of a power supply I need some info on efficiency of power supplies. Imagin a power supply with 91% efficiency that can provide a maximum of 800watts. Does the rated 91% efficeincy is for all loads that going out from this source? I mean if it is providing 500W, the efficiency is still 91% or less or higher? is there a formula between load amount and efficiency? Thanks. <Q> There is no one formula. <S> This is individual to particular power supplies. <S> Good power supply datasheets will show a graph of efficiency as a function of current. <S> Sometimes you might get a minimum efficiency spec over a current range (assuming fixed voltage supply). <S> Sometimes you only get the efficiency at the maximum output power, which is usually the condition for maximum dissipation in the supply. <S> There is much choice of switching frequency, modulation technique (frequency, pulse width, both), synchronous rectification <S> yes/no, type of magnetic core, etc. <S> There are many other tradeoffs possible between size, weight, cost, transient response, output ripple, idle power, efficiency, and more. <A> This is something you have to look up in the vendor's data sheet. <S> Usually the efficiency is best at the rated power <S> (800 W in your case), degrades a bit at light-load (500 W) and is 0 % at no-load (still some demand for household, but not output at all). <S> See fig. <S> 9-1 of this data sheet for a typical curve: http://www.pulspower.com/pdf/cs5_241.pdf <A> The efficiency will vary with output power, because the supply will always consume some minimum energy itself. <S> There should be a graph provided with the power supply which shows you the efficiency against power output. <S> What is quoted as a single figure in the headline is best case, usually at maximum power output. <A> There's no simple relationship between a regulator's load and its efficiency, so no formula. <S> In any case it can't be constant: with lower loads regulator's losses become more important, and at zero load the efficiency is also zero, no matter how well the regulator is designed. <S> The graph is constructed from measurements, rather than from a formula. <S> This kind of graph is typical for a switcher; the design is optimized for a certain load. <S> Other types of regulators may have monotonic graphs. <A> You're going to find lots of variation in power supply efficiency as a function of both input voltage and output loading. <S> If the manufacturer only specs efficiency at one input/output condition, don't expect that number to hold anywhere else. <S> There are some newer directives out there such as 80 PLUS <S> which, if a product is compliant, provide a standardized measurement methodology as well as efficiency minimums for various loads. <S> 80 PLUS Gold for server power supplies guarantees 88% efficiency at 20% load, 92% efficiency at 50% load, and 88% efficiency at 100% load at 230V / 60Hz input.
|
The reason there is no one standard formula is that there is no one standard power supply topology, especially when you look at the details. A good datasheet should give you a graph of efficiency versus load, like this one for an application of an arbitrary Linear switcher (i.c. LT1912 ).
|
Why is SMPS controller IC failing? I'm back to diagnosing power supply failures, in an LLC resonant SMPS based on the L6598 , from a 400 V PFC supply with ostensibly 400 W output. I'm new to SMPS, and learning this all as I go, and the manufacturers don't know what they're doing, either; they just copied the design and modified it for more bigger power. The circuit is basically like Figure 3 in AN-9425 , but the OPOUT overcurrent detection triggers EN1 (latched) instead of EN2 (soft-start reset). The failures happen during overload, when the supply is making audible screaming noises and the output voltage has dropped (because it's switching below the resonant frequency, right?) Sometimes it can withstand a prolonged overload (without triggering EN1), sometimes it can't. Sudden transient overloads are usually what kill it. Not sure if it's a progressive damage thing that eventually results in failure or a random spike that exceeds a limit. Theoretically, it should handle overloads by shutting off momentarily or otherwise protecting itself, not by exploding. Theoretically, I could change the overcurrent protection so that it detects when it's being overloaded and triggers the soft-start EN2 protection mode, shifting the frequency up high to drop the output voltage instead of dropping below the resonant point. I tried modifying the circuit to match the app notes so OPOUT drives EN2, but that just made it blow up even more quickly when I overloaded it. The L6598 controller IC always fails, the STW20NM60 MOSFETs fail sometimes (high side short across all pins, low side short from D to S), so I think the IC is failing first and taking the MOSFETs with it? The MOSFETs don't have built-in gate protection, so I added Zeners externally and they haven't failed since (yet). New ICs measure open circuit from any pin to any other, but after I remove a failed IC from circuit, some of the high pins will have kΩ between them (12 Vsupply, 14 Vout, 15 high side driver, 16 bootstrap cap). Today I was manually triggering the EN2 pin, which worked fine at low loads, changing to a high switching frequency and output voltage dropping. At about 200 W out, it worked fine when I enabled EN2, but when I let go of EN2 it would shut down (overcurrent protection triggering EN1 is the only way I know of) and then another time failed. After it failed, the low side driver would still switch at fmin (no output voltage to drive the feedback loop), but the high side driver just had spikes and weird shapes on it, not a square wave. The MOSFETs measure normally with an ohmmeter and didn't explode. So some kind of spike is killing the high side section of the IC and I need to prevent that. I certainly don't see any pins going higher than the 618 V absolute max rating. I do see tiny spikes going down to -3 V before the rising edge on Vout and VHVG, which are rated for -1 V. I don't know if that really matters. I'm not sure about the ratings for Vout or VHVG, or how I would even measure those (differential probe?), or what conditions would cause them to go beyond the stated limits. They seem to track each other, as would be expected if the high side driver is floating relative to Vout . What am I missing? Update: There are huge bursts of oscillation(?) ±25 V on the gate driver while overloading, 100s of ns after the rising or falling edges, which coincide with the audible noise. These could easily break through the gates and destroy the FETs, so they are the prime suspect. They originally had ferrite beads on all the FET pins, which did not prevent this, nor does removing them. I've added Zener protection to the gates, though, which should prevent them from being destructive, but I'd like the eliminate the bursts themselves first. <Q> I have a suspicion that your problem might have to do with the bootstrap circuit, as detailed in section 5.3 of the data sheet. <S> Some evidence leads to this conclusion: <S> You are likeley using larger MOSFETs than in the original design ( <S> because your overall power is greater), thus you are dealing with a higher gate charge, resulting in a higher current that must be delivered by the integrated bootstrap circuit. <S> Adding an external, fast and high-voltage bootstrap diode might help. <S> Note that the calculation in Eq. 9 of sect. <S> 5.3 uses the typical on-resistance of the integrated bootstrap circuit. <S> It's a better idea to use the max. <S> value from table 4 which is twice as high. <S> Once the circuit goes into overload, you say that the frequency drops. <S> During these prolonged on-times, the voltage across the external bootstrap capacitor might become too low to keep the high side MOSFET saturated, causing it to have a higher on-resistance, excessive losses and thermal overstress. <S> However, in this case, it would be the high-side MOSFET that fails first. <S> Check the voltage across the bootstrap capacitor. <S> A bigger capacitor might help, but this will likely put more stress on the integrated bootstrap driver or the external diode that might be necessary anyway. <S> Another possibility could be that you are exceeding the maximum slew rates for the high-side driver. <S> This might cause it to do weird things. <S> Concerning the negative spikes, an external protective clamping diode at each driver's output might help (K=Vout,HS; A=GND,LS and K= <S> Vout,HS; A=GND,HS). <S> Something as simple as a 1N4148 might be enough. <A> I think that the burst is the coil ringing at 55 Mhz self resonance frequency. <S> The workaround is to put 2x 470pF caps in parallel to power transistors , to convert coil energy of resonator to lesser peak voltages during ringing oscillations. <S> (This solution I see in similar datasheets from same manufacturer). <S> Independent test is to calculate energy of coil for given current and inductance and convert to voltage on parasitic capacitance (maintaining the same energy) and see if it exceeds the max voltage of IC. <S> V=SQRT(E*2/C), where E= <S> L*I*I/2, if V > max then fail Why the burst is delayed by 366 ns is a mystery. <S> Cable to load, should be about 50 meters to reflect back the transient to make it that late. <A> These sorts of failure modes are always fun to debug. <S> Here are some thoughts: <S> It's very unlikely that the IC dies first - <S> it's much more likely that the MOSFET is being poorly controlled, fails, and the HV comes through the device into the control IC and is killed. <S> (Either that or the converter is getting into ZCS mode and the MOSFET is under heavy stress - either way, <S> dead MOSFET = dead controller; the opposite often does not hold true) <S> A well-designed power supply should never scream when overloaded. <S> Screaming usually implies extremely high, rapidly changing current passing through a magnetic component which tends to support the theory that proper control is being lost and the converter is going bezerk before self-immolating. <S> The common thread in what you've described is that at a certain point when overloaded, the power supply blows up. <S> I'd strongly consider setting the peak current limit lower so that the power supply never 'screams' (voltage divider on OPIN+) and see if this helps. <S> If not, there's a fundamental issue with the implemented protection mode (nothing to do with the overload itself). <S> You were able to blow it both in heavy overload ( <S> > 400W) and at lighter load by playing with the enable circuit. <S> This rules out the PFC stage dropping out as a cause, which helps. <S> This also rules out transformer saturation. <S> It's starting to appear that the supply simply doesn't like to be inhibited or otherwise constrained. <S> Perhaps the range of frequencies is too broad? <S> Measuring tiny spikes on the high-side is notoriously difficult without a good scope and a high-bandwidth differential probe. <S> The driver circuitry in the IC is only capable of 250mA (both high-side and low-side). <S> To me, that's an awfully low amount of current for high-voltage MOSFET drive (high-voltage MOSFETs tend to have large gate charge requirements) - I suspect that turn-ons are going to be slow. <S> What MOSFETs are they using? <S> The theory that Vout is floating up to Vin is unlikely, unless everything on the Vout rail is also blowing up when the primary goes.
|
It could be that the bootstrap circuit dissipates more power than it was designed for and fails, sometimes resulting in a bad output driver for the high side MOSFET, sometimes taking the MOSFET with it.
|
Protection against automotive power supply hazards I'm looking for a way to protect a small circuit which is to be used inside of a car or truck (12V or 24V power system). The circuit consumes about 12-15W. I use an isolated DC/DC converter module which can regulate 9-36V down to 3.3V. I'm looking for recommended circuits or a controller IC that can take care of the usual hazards: Load Dump Spikes Reverse Voltage OV/UV Protection General noise on the power lines. ... Anything I might have missed. Currently I have my eye on the LTC4365 from Linear Technologies. I've thought about using it together with a bi-directional TVS, clamping the voltage to 32V and protecting everything with a fast blowing fuse. Would this be a proper solution or did I miss something here? <Q> Load-dump ... is a killer - your TVS has to turn a huge amount of energy into heat without going pop. <S> ISO7637 for a 12V system has a spike peaking at up to ~90V with a rise time of 5-10ms lasting up to 400ms from a source resistance as low as 0.5ohms. <S> That's several hundred Joules of energy in less than half a second! <S> Not all of that has to go into the suppressor - only the excess above the clamping voltage (but still ~60V <S> in your case) <S> On the bright side, load-dumps are pretty rare, so if it's a one-off and you don't mind the small risk, you could ignore it. <S> Fast transient spikes <S> These can reach 200V when the wipers switch off for example - provide a (high-voltage-rated) capacitive route for those to ground right near the input. <S> Longish-term over-voltage Automotive electronics is often specified to survive 24V for several minutes (for when a car is jump-started off a 24V truck) and 48V for up to a minute (IIRC) as sometimes 2 truck batteries are used to provide a quick boost charge to get a car moving in extremis! <S> Your spike suppressor may pop under those conditions. <S> Dropouts Battery dropouts can also be significant, there's a test in the industry which involves a series of pulses battery voltage falling to 0V - you need to have enough internal capacitance to keep your supply rails up when that happens. <S> Real-world requirements specification <S> If you want an example of how gory this can get, Ford's electromagnetic compatibility (EMC), which includes transient testing, is available on the web: <S> Component EMC Specifications EMC-CS-2009 Search through it for "transient" and "dropout" to see what series-production designs are supposed to live up to! <A> You seem to have answered your own question. <S> The LTC4365 <S> is probably a good solution. <S> The datasheet says no TVS is needed, but I still would use one. <S> Have the LTC4365 followed by a buffer capacitor to handle dips in the battery voltage. <S> If the battery is also used for a starter motor <S> it's probably unavoidable that the voltage drops, especially when you're consuming 15W (that's 4.5A at 3.3V). <S> (The fuse doesn't offer extra protection over the LTC4365 other than limiting the damage in case of a component failure). <S> Any particular reason why you want to use an isolated DC-DC converter? <S> They're usually not needed for battery operation. <A> If you've already got a isolating DC-DC converter that can handle up to 36V in, it doesn't sound like you need much more. <S> I don't understand what you think the LTC4365 will do for you. <S> Your converter can already handle 36V on its own, which is actually a little more than the 34V <S> the LTC4365 is rated for. <A> For a brute force protection device: ST : <S> RBO040 MANY thousands of devices in police and other emergency vehicle applications with this part at the connector to the +12V line. <S> Not fancy but will save your circuit from most transient events.
|
If the capacitor has a rather large value you may want to use a slower fuse , otherwise it may blow when switching on.
|
RS232 and RS485 over same pins I have 2 pins and primary function should be A/B RS485, but is possible that I also put RS232 (Rx/Tx) chip on same bus and then control which one I want to use from microcontroller. I was thinking about MAX232 and MAX481 and enabling and disabling them trough power on/off controlled by microcontroller driving transistor. Is there any single device that does this? <Q> Exar has the SP331 programmable <S> RS-232/RS-485 transceiver which may suit your needs. <A> Since you like parts from Maxim now, take a look at the Maxim MAX3160 and friends . <S> They provide transceivers for both RS-232 and RS-485 protocols, allowing the protocol to be chosen at run time by driving a logic level input. <S> We are using the MAX3161 in a couple of projects which provide only a three contact terminal strip for field wiring the serial cable ( Rx / A+ , Tx / B- , Gnd ). <S> If the port is incorrectly configured the device is still safe because, unlike most RS-422 receivers, it tolerates the full range of RS-232 voltages on the pins regardless of which protocol is currently selected. <S> This part is a little more expensive than a pair of individual protocol drivers, but makes up for that in saved board area and saved complexity. <A> You should not connect ordinary RS485 drivers (e.g MAX481) onto the same wires as RS232, because the negative RS232 voltages can easily exceed the -8V abs-max for the MAX481. <S> Or use some kind of switching, of course. <A> Powering a device down may cause its ESD diodes to start conducting current from the data lines, allowing some power to pass through into Vcc. <S> You should use an RS232 transceiver with an ENABLE line. <S> With RS485 <S> it is simpler, any chip with separate DE and RE (driver and receiver enable) will do.
|
If you can't use the suggested combined part, you'll need to check the specs of the RS485 part you do use very carefully for compatibility with RS232.
|
How do I know where I need decoupling capacitors? I'm building a motor driver shield for the Arduino. Here's the schematics (please forgive the messy layout): And here's the PCB layout: I'm testing it out by breadboarding it while running a simple 'run for 5 seconds then reverse' procedure on the MCU, and I'm experiencing some odd issues with decoupling capacitors. If I leave them off, the motor stutters rather than moving smoothly, presumably due to the motor's power rail dropping. I put a 100uf electrolytic capacitor across it (shown on the board), and it started running smoothly. Now, however, the motor reverses seemingly at random, I presume as the MCU is being reset for one reason or another. Following the 'more capacitance is better' mantra, I installed the second electrolytic cap, between the 5v and ground rails. Now, weirdly, we're back to step 1: the motor stutters. Finally, I added a third, 0.1uf ceramic cap across the motor terminals. Suddenly, everything is fine: The motor runs smoothly, and it reverses when it's supposed to. Adafruit's shield seems to solve the issue by just throwing lots of caps, both electrolytic and ceramic, at the problem. Unfortunately, as you can see, I have quite limited PCB space, so I can't afford to do that. I could take the cargo-cult approach and say "it's working now, great" and stop, but I'd rather understand what it is that caused each of these symptoms, and what I should do to ensure they don't happen on the real board. The first answer to this question answers most of my obvious queries about capacitance, but I have a couple of remaining ones: Is the 100u and 0.1u cap between the 5v (logic) rail and ground that Adafruit's shield has necessary? Removing it on my breadboard seems to have no effect. Do I need the 0.1u ceramic caps across both the motor rail on the H-Bridge input and directly across the motor terminals? Edit: I've updated the schematic and PCB layout with the proposed positions of caps, based on the advice of those kind enough to answer my question. <Q> It's not because removing caps doesn't seem to have any effect that you shouldn't use them. <S> It may work now, but not in an hour or so. <S> The principles are Place caps on the source of the disturbance, so that it doesn't conduct to susceptible components, or radiates via the wires Place caps on susceptible components. <S> Use a bigger and a smaller capacitor. <S> The bigger one will absorb most of the disturbance's energy, but isn't very good at high frequencies, where the smaller one takes over. <S> Is it worthwhile to decouple the same disturbance source twice, over different components? <S> Most likely yes, you can't decouple too much. <S> I remember a colleague's design, where half of the components (about 200) were decoupling capacitors. <A> All digital ICs should have decoupling caps between their power and ground pins. <S> These should be ceramic and physically as close as possible to the IC. <S> Decoupling caps deal with short term current spikes the IC draws. <S> They must therefore be high frequency. <S> A 100µF electrolytic cap is pretty useless for decoupling. <S> 1µF to 100nF ceramic is good. <S> As for the cap on the motor, the idea is good <S> but I think 100nF is too large. <S> That could cause excessive or unnecessary current to flow in the H bridge every time it switches. <S> If you're only reversing motor direction occasionally, then this isn't a big deal. <S> If you're using the H bridge to modulate the apparent motor drive with PWM, then you should lower the cap. <S> Something like 1nF should still cut down the noise the motor is making while not getting in the way of switching. <A> I'd follow approach with 1 goal: minimize the area for each AC current loop, coupled to power lines (Vcc and ground). <S> Make list of all power nets: <S> say, total 2 entires on list : V+, ground. <S> Add 1 capacitor per each pair <S> In layout move capacitor closer to ports to reduce the area of loop, formed by 2 leads of capacitor and 2 ports of the pair. <S> Effectively this will shorten the current path to minimize voltage drop, spikes, caused by line inductance and will isolate max(dI/dt) <S> current loops each from other magnetically.
|
You want to minimize the loop length from power pin to cap to ground pin thru the IC and back to the power pin. Your disturbance source may not be the only one, and you can't always rely on the others being properly decoupled Large capacitance is not necessary, and since large capacitors usually have poor high frequency response, they are worse. Identify ports, connected to this nets, carrying power: 2 or more per IC package, active element Split list of this ports (total say ~15) into pairs, by relevance to package with possible duplicates on ground ports or V+ ports.
|
Why are more powerful tools using higher voltage? AFAIK each of the world leading cordless power tools manufacturers produces several "product lines" of cordless power tools with different voltages. For example, Bosch currently produces tools with Li-Ion batteries with 10,7V, 14,4V, 18V and 36V output and the higher the voltage the more powerful a tool is. Now those tools are powered by batteries that are assembled from cells with lower voltage (something like 3,7 volts for Li-Ion cells I guess) and cells are connected in sequence until the target voltage is reached. They could instead connect cells in parallel. They would have the same voltage, but higher current and that would again yield higher power. Why do they choose higher voltage over same voltage and higher current to get higher power in electric tools? <Q> Matt already explained that using a higher voltage you'll have a lower current for the same power rating . <S> This means thinner and less heavy wires , which means savings (copper is expensive). <S> You may have to pay attention to better insulation , but that doesn't outweigh the advantage mentioned. <S> It's also much easier to place cells in series than parallel . <S> When placed in parallel the voltages have to be exactly equal, otherwise you'll have high currents running from one cell to the other, causing big power losses and reducing the cell's life . <A> TLDR: <S> I disagree with copper efficiency argument (above/below) <S> The only reason is back EMF of motors at high rpm. <S> No matter how much current the batteries can supply, their current translates to torque for motor, but not velocity. <S> At top velocity, theoretical lossless motor has back emf exactly equal to the voltage of supply and consumes current approaching zero while having zero torque. <S> Higher power tools perhaps have higher rpm, rotational velocity figures, and they need higher voltage. <A> For exactly the same reason the power companies transmit power around the country at many hundreds of thousands of volts instead of just doing it all at 110/230v. <S> A lower current means smaller components and thinner wires, thus making it cheaper and more efficient. <S> For example, take a DC motor. <S> For a 12V motor to develop the same amount of power as a 24V motor it would have to draw twice as much current. <S> This would mean that the windings in the motor would have to be made up of thicker wire. <S> This would increase both size and cost. <S> The electricity companies transfer power at a high voltage so the current is low so they can use small diameter cables. <S> It's all the same principle. <A> The higher is the voltage, the lower is the current, at same power. <S> When you suck lot of current from a battery, you'll create a poorly changed area inside. <S> That leads to a worse performance.
|
A higher voltage means a lower current for the same amount of power. Higher voltage is needed for higher rpm.
|
What should I put on my almost empty PCB layer? I have a 3"x3" 4-layer PCB where my stack up is: Signal 1Ground5v PowerSignal 2 My Signal 1 layer has a few traces that carry 500 MHz on them, some high resolution ADCs, and microcontroller/usb circuitry. I have SMA connectors that are carrying the 500 MHz on to the board. Currently this will just be "open air" sitting on a test bench, but long term into will end up in a case in which everything will be contained internally. My Signal 2 layer has almost nothing on it, specifically it has the following: MCLR from programmer connector that is 0.1" long SPI Data and Clock line that are both about 0.1" long Negative voltage (for powering 2 op-amps) trace that is about 2" long I feel like it is somewhat of a waste to have so much unused PCB. I am considering the following options: Fill the layer with my negative voltage rail Fill the layer with ground Leave the layer empty Is there any benefit of one of the options over the other? What is usually done in these situations? Some Additional Details The system will be pulling a peak of 300 mA off of the 5v rail. While the -5v rail will only have about 2 mA load. <Q> What is typically done in the industry in these cases, assuming that ground-fill practices as shown in other answers does not provide significant benefit, is something called thieving . <S> Thieving consists of covering large expanses of unused outer layers with a pattern of shapes, usually diamonds or squares, disconnected from one another. <S> Thse shapes are kept away from other features, such as holes, board edges, or traces. <S> The sole purpose of thieving is to improve manufacturability by ensuring a constant PCB thickness given any particular area on the board, say, half a square inch. <S> Without thieving, the rollers that are used to laminate the layers together will not exert as much force on the copper-starved areas, which could lead to delamination (looks like light spots inside the board). <A> Turn the negative power trace into a pour (small impact <S> but you have the space). <S> BUT...use lots of stiching vias to tie it to the internal ground plane. <S> Make sure there is no orphaned copper. <S> Your goal is to "sandwich" the DC power plane on all side possible by grounds stitched together, this will minimize the RF impedance to ground from the 5V supply to keep it clean. <S> Every 3/4 inch or so in grid. <S> If the board is small, the decoupling on the ICs is likely enough. <S> A pour on the top side isn't a bad idea either although it depends on the layout. <S> Here is an example of what I mean by using lots of vias to couple the pour to the ground plane: <A> Is much of your board HF (tens of MHz, or even 100MHz)? <S> If not, maybe you can get rid of the inner layers , and place components on both sides , so that you can route the power nets in the free space you get this way. <S> Two-sided component placement is a lot cheaper than a 4-layer board. <S> edit Since you seem to have VHF on it <S> I would populate it with decoupling caps , and pour a second ground plane, or power plane. <S> If properly decoupled, for HF <S> all power planes are at the same potential , so it doesn't really matter which net you pour here. <S> I also place as many test points as possible on the bottom of my boards, including pads for in-circuit programming (see also this answer of mine). <A> --2. <S> In this case should be <S> the best - as distance between Signal2 & power is small, it will act like a distributed capacitor and will help with stability & EMI performance. <S> --3. <S> No benefits. <S> --1. <S> Too much hassle for single negative V user. <S> But personally I like just 2-sided boards, probably you may use few 0-ohm jumpers and it might end up cheaper to manufacture.
|
If the board is larger, you can also use the space to sprinkle decoupling capacitors around on this layer tying +5V to ground. Do a ground pour on the rest of the layer...
|
Part for VHF/FM booster? I'm needing to create a booster and the easiest schematic I've found is this: http://electronics-diy.com/500mW_FM_VHF_Transmitter_Amplifier_Booster.php however, they do not document what part they use for the amp. Does anyone know what amp they may use? Or a similarly easy way to create booster inexpensively? <Q> The MAX2650 LNA (Low Noise Amplifier) is not the amplifier they are using, but it may be suitable, depending on the gain you want. <S> The MAX2650 has a gain of 18.3dB at 900MHz. <S> If you need higher gain you can cascade a couple of them. <S> (Monolithic Microwave IC), like Leon says. <S> MMICs often consist of just a couple of transistors. <S> This is a typical MMIC package: Notice that it has two ground pins on opposite sides for improved PCB layout, but that also means that the power supply has to share a pin with the output. <S> The MAX2650 comes in a SOT-143 package and has separate pins for these. <A> It's an MMIC. <S> At VHF virtually anything should do, such as the $1.25 <S> MCL MAR-1 <S> + . <S> I've constructed similar amplifiers by mounting the parts directly on a scrap of PCB material. <A> As Leon says - MMIC is the name you want. <S> Go to Digikey and search for MMIC and go from there. <S> For the same pinout style you could use NLB310 data sheet In stock at Digikey for $US3.44 in ones. <S> That's a 10 GHz part !!! <S> You can get functionally similar ones for under $US1 with lower bandwidths. <S> That's known as a Micro-X 4 package. <S> Digikey has about 5 different parts available in 1's in stock. <S> Surprisingly, I find that the one I listed above is the cheapest Micro-X 4 available in 1's and in stock. <S> If you are willing to go to a different package, same principle otherwise, there are many more available. <S> Cheapest is BGU7031 from NXP. 1 GHz. <S> Datasheet in a SOT363-g pkg <S> Pricing 48 cents in ones. <S> 14.5 cents at 10k <S> They say 1.2 Features and benefits <S> Internally biased Flat gain between 40 MHz and 1 GHz Noise figure of 4.5 dB <S> High linearity with an IP3O of 29 dBm 75 Ω input and output impedance <S> ESD protection > 2 kV Human Body Model (HBM) on all pins 1.3 Applications <S> Terrestrial and cable Set-Top Boxes (STB) <S> Silicon and “Can” tuners Personal and Digital Video Recorders (PVR and DVR) <S> Home networking and in-house signal distribution
|
The one in the schematic you linked to is an MMIC
|
Electrical and electronic engineering magazines What electrical and electronic magazines can be recommended in order get up-to-date information about the latest technologies? I would like magazines listing from hardware hacking to proper professional design. <Q> Ones that I read frequently: Nuts & Volts , great monthly with lots of microcontroller projects <S> Servo , sister publication to N&V with focus on robotics Circuit Cellar , more professional than N&V, lots of great hardware info Elektor , just started distribution of English translation in US Everyday Practical Electronics , British magazine full of construction projects <A> This one's easy, Make magazine http://makezine.com/magazine/ <A> Everything by UnwiredBen and Vineeth are great, but dont forget online websites for the freshest hacks. <S> My particular favorite <S> Hack A Day is a great resource <A> Of the "free" ones, there are: EDN EETimes Electronics Product News <S> I will give you a warning, however. <S> If you subscribe to any of these, they WILL send you lots of spam emails. <S> Do not, under any circumstances, give them your normal email address! <A> I always like what they have to say in Circuit Cellular . <S> I'm not sure how much "breaking" technology they have since they're not trying to sell you on new products like many other free EE magazines. <S> They definitely always have articles that seem relevant, immediately implementable in any designs I do, and more technical than other magazines targeted at hobbyists. <A> http://hackaday.com/ <A> Some mention the EEvblog. <S> I'm sorry to say, but the EEVblog with it's rambling <S> is all nothing but plain bla bla and nothing to learn. <S> You better not copy he's schematics if you want to build something that work. <S> He doesn't even know how to make an emitter follower that doesn't oscillate. <S> (For the record, it's not the op-amp that causes this circuit to oscillate. <S> It's the emitter follower itself <S> who's oscillating. <S> It would even oscillate without feedback.) <S> If you wanne seem him going down? <S> Check out this one. <S> http://www.youtube.com/watch?v=b7UQVZaqxg0 <S> Base stopper resistors? <S> Huh? <S> Anyone. <S> And apart from that. <S> He sounds like a cat. <S> Regards <A> PCB Design and Fab CircuiTree <S> PCB Design 007 (and its related online mags) <S> SMT/EMS007 ECN (Electronic Components News Power Electronics Technology <S> ESD (Embedded Systems Design) (under EE Times) Evaluation Engineering <S> Nasa <S> Tech Briefs <S> (publications on left side) RTC (Embedded stuff) <S> Element 14 <S> Technology First Journal(I have a few issues but can't find a subscribe link, forget how I got it) <A> Also electronic, free, good electronic design website. <S> Mailout material offered. <S> As for "spam". <S> Most reputable sites have a working opt out policy and the sort of mail they send is in ant case liable to be of high interest and relevance to electrical engineers and fellow travellers. <S> I'm happy to have such sent to my main mailbox, where I can filter into folders etc as desired. <S> More later ... <S> There are groups who manage dozens of magazines - electronic only and real-paper. <A> There was a question relating to Blogs and Podcasts that somebody asked 6 months ago in a similar vein, some of the resources there might be of interest: Electronics Blogs and Podcasts <A> These are the ones I enjoy reading: Elektor Circuit Cellar <A> EE Power (www.eepower.com). <S> A digital publication in power electronics focusing on technical articles, market insights, and design trends from industry-leading electrical engineers. <S> Offers full content RSS feeds and Apple News integration.
|
Nuts and Volts Everyday Practical Electronics Magazine Servo Magazine Silicon Chip LEDs Magazine {electronic, free, good} electronic design update
|
Getting started with PCB designing I'm a hobbyist EE in college, and have been building breadboard circuits and perfboard circuits for the past year or so. I want to move on to doing PCB design. I wanted to start with something small so I designed a small amplifier circuit for a microphone that I want to make into a module on a PCB. Here is the file for what I designed in Fritzing (I went with this software because it's user friendly): http://www.mediafire.com/?2b2as7iibys68cu Here is an image if you don't have the program: Is this a good design? How can I improve it? The general schematic I followed was this (in case you wanted to know): (Credit @Olin Lathrop) What advice can you give me on getting started? Any resources you can recommend?What software would you recommend? Ideal would be free and easy to learn.What type of class would you suggest I take in college to get deeper into this? <Q> I'll just comment on the design: <S> Replace C5 with a 100nF ceramic capacitor and place it close to the power supply pin of the MCP6022. <S> Put the designators on the PCB-Design, not values. <S> Make it fareasier to understand the layout. <S> Avoid 90 <S> ° trace bends, they can cause problems when etching theboard. <S> They're also bad for high-speed stuff <S> (at least that's thecommon opinion on the matter). <S> Use two 45° bends instead. <S> Consider flooding one side of the board with a GND-Plane. <S> Use wide short traces for power supply connections. <S> Use one side of the board for mostly vertical traces and the otherside for horizontal traces. <S> Take more care of component placement. <S> Place them in a way wherethey are easier to route. <S> Component placement is 70% of the job. <S> Place them BEFORE starting to route a single trace (Won't alwayswork out). <S> Just use the ratsnest (the lines which indicate connections <S> whichare not routed yet) as a rough guideline. <S> Do not see a trace which is already routed as something which is fixed. <S> Ifits in the way <S> or you don't like they <S> way it looks, rip it up and try again. <S> When in doubt, start from scratch, try not to rescue something which can'tbe rescued anymore. <S> Rule of thumb: Create something which pleases the eye. <S> Others will have an easier time to understand it and sometimes it will even work better. <S> While the second one more about EMC compliance it helps to understand WHY these things should be done in a certain way. <A> Eagle CAD is popular package with hobbyists. <S> There is a free version available (limited to 2 layers and small boards for non-commercial use). <S> Eagle has many advanced features beyond what Fritzing does, but it's bewildering to learn for a beginner. <S> I found this video series helpful in getting started. <S> http://www.youtube.com/watch?v=qG0O9LKH-_E <A> Corners are usually chamfered so they don't cause problems during etching. <S> Your circuit design has filter caps. <S> These are usually unpolarized disks and the lead spacing <S> might be more than 0.1", depending on voltage for your part. <S> The silkscreen shows polarized (electrolytic?) caps. <S> If you do go that route, then maybe flipping two of them around so that the polarization direction is consistent would help prevent build errors. <S> Masta79 covered everything else pretty well after I got started, so I'll just send dittos there. <S> Especially 7. <S> Put all the parts on the board. <S> Flip and shuffle them until the rats nest doesn't look too dense anywhere, then start routing traces.
|
There are two Books i can highly recommend for learning Electronics/PCB Design: The Circuit Designer's Companion and EMC for Product Designers .
|
How can I sense the motor's current? I need to drive a DC motor @ 24V, 6A with a MOSFET. How can I sense the current that the motor is drawing with a microcontroller? I have to know when the motor is stalled. <Q> You place a small sense resistor (typically < 100m\$\Omega\$ for the voltage and current involved) in series with the motor and measure the voltage drop. <S> There are two methods: high-side and low-side , depending on the position of the sense resistor. <S> Low side is easiest, as the voltage drop you want to measure is directly related to ground, but it lifts the low side of the motor's voltage a few tens of millivolts above ground too, and not everybody likes that. <S> If it's no more than these few tens of mV it shouldn't be a problem though, and you can use an opamp to amplify the voltage in a simple non-inverting amplifier configuration. <S> A 10m\$\Omega\$ resistance will give you a 60mV drop, which is acceptable, and at the same time high enough to measure properly. <S> You don't necessarily need a physical component for this; a 1cm PCB trace 0.5mm wide has a 10m\$\Omega\$ resistance . <S> Make sure to select an RRIO <S> (Rail-to-Rail I/O) opamp. <S> For high-side measurement you have to use a difference amplifier to measure the voltage drop. <S> There are special ICs for that, some of which have the shunt resistor integrated, for maximum accuracy. <S> But you can also construct your own difference amplifier with an opamp. <S> If you just want to detect a stall you probably don't need the A/D converter but can use a simple comparator . <S> Be sure to filter the measured voltage with a capacitor. <S> A (not very thorough) search turned up the SiLabs Si8540 high-side sensor, available from Mouser from USD 0.65 quantity one. <S> edit <S> The Zetex/Diodes ZXCT1009 is comparable, but only needs 3 pins of its SOT23 package. <S> Further reading: <S> Linear Technology Current Sense Circuit Collection <S> (warning: heavy product plugging!) <S> Collection of documents on current-sense amplifiers by Maxim <A> People who think the only way to measure DC current is to use a shunt resistor may be surprised to learn that a variety of current sense techniques exist. <S> Hall effect sensors are nice for measuring large high-side DC currents. <S> Some have analog out, eating up one of the analog inputs on your microcontroller. <S> Others have an integrated internal ADC, with digital pins that directly connect to your microcontroller. <S> A few also have an integrated power FET driver, and are smart enough to unconditionally turn off the FET when it measures over-current. <S> It makes the rest of the system much simpler to use a "smart switch" that automatically turns itself off when the motor stalls. <S> The Allegro Hall effect sensor chips look nice. <S> The IR intelligent power switches <S> look nice. <S> Related: <S> Best shunt resistor for power meter application? <S> and High-bandwidth current measurement <A> Put a small (\$<0.1\Omega\$) resistor in series with the motor. <S> The microcontroller can measure the voltage drop across it (you may want to amplify it through an op-amp) using ADC. <S> This is something I have wanted to do myself for a while, and I understand the theory - just haven't worked out how to measure the voltage difference yet <A> As Andrew Kohlsmith corrected me here's the edit: For DC, the only way to sense current is by a Shunt Resistor . <S> This method is derived by the Ohm Law: <S> \$I=\dfrac{V}{R}\$ <S> Where <S> 'I' stands for current and will be the only variable solved by the µC. <S> In the same way, 'V' stands for Voltage, which will be measured by an ADC (Analog-Digital Converter) inside the µC. Finally, 'R' stands for the resistor you must know for calculating the ecuation. <S> There are two ways of designing the shunt resistor: <S> Using a resistor conected in series with the motor. <S> Which value has to be known and you'll have to consider the power dissipated. <S> For example: if you use a resistor of \$1\Omega\$ and you want to sense a current around 6A, the power dissipated by that resistor would be 36W. <S> So I sugest you to use values around \$10m\Omega\$. Using the board trace in a PCB to fabricate a Shunt Resistor. <S> As [1] says, depending on the following parameters in the formula, you'll get a resistance value: <S> \$R=\rho <S> \times \dfrac{L}{t \times w}\times(1+Tc \times (T-25))\$ Length (L) <S> Thickness (t) <S> Width (w) <S> Resistivity (ρ). <S> For Cu, \$\rho=1.7*10^{-6}\Omega\$-cm Temperature (T) <S> Tc = <S> 3.9 \$10^ <S> -3\Omega/\Omega <S> / <S> C\$ <S> (I don't know what does it stands for, ideas?). <S> Some, ommit the temperature product part [2]. <S> There are plenty of webs you can use for generating the aproximate resistor of the PCB trace, for example in [3] and [4]. <S> Anyway, I would measure the value with a multimeter with a \$m\Omega\$ capability. <S> If you want more information, check [5]. <S> On the other hand, the only way to measure the voltage of that Resistor is using an Instrumental Amplifier, just as Stevenvh suggests. <S> [1] AN894 - Motor Control Sensor Feedback Circuits by Microchip. <S> [2] AP144 - Calculating PCB Track Resistance by Polar Instruments. <S> [3] Trace Resistance Calculator by EEWeb. <S> [4] PCB Thermal Copper Area by The CircuitCalculator.com Blog. <S> [5] Contructing your Power Supply - Layout Considerations by Robert Kollman [TI].
|
As current, voltage, and resistance are all related (Ohms law), you can measure current by measuring the voltage drop across a known resistance and calculating it: \$I=\frac{V}{R}\$ In many cases, I don't really need to know exactly what the current is, I just want to keep things from getting permanently damaged when the motor stalls out.
|
Are 4xAA DIY Gadget Chargers Safe? I've found some Do-it-yourself guides to making a 4-AA battery gadget charger, which are basically four AA batteries connected to a cut up usb cable: http://robotification.com/2007/09/15/diy-usb-travel-charger/ http://www.instructables.com/id/Build-a-4-x-AA-USB-Altoids-Battery/ Such a device uses the AA batteries to recharge your gadget. This looks like a useful project, but I was wondering, is this safe to use long-term, for gadgets like smartphones/android devices/mp3 players/tablets/etc.? Assuming it is connected with the right polarity, and lower-voltage rechargeable batteries are always used, it should always be around 5 volts, which is standard for all usb, but isn't there supposed to be a circuit to prevent overcharging? Is the worst case here simply not charging, or bricking the device? Also, what would keep this from reversing? If the AAs went dead, would plugging in your phone actually drain the phone's battery to the rechargeable AA's? Update - A 5v regulator with red battery connected to input, red usb connected to output, both blacks connected to ground, doesn't seem to be charging when the battery is dead. I'm wondering why this recommends connecting green and white to ground. Wouldn't this make voltage return through -, d+, and d-, which is not recommended? <Q> The use of 4 x AA Alkaline would usually be safe BUT does exceed the USB spec and damage may occur in some cases. <S> I have seen IC's in this role with max operating voltages of 5.5V (which is ludicrous) - you'd hope designers had more sense, but it can't be guaranteed. <S> While some devices may use converters between charge input and battery, many don't (probably most). <S> A LiIon battery has max charging voltage of 4.2V <S> so a 5V nominal USB input will usually meet this need with enough headroom for a linear regulator. <S> An Alkaline cell can be nearly 1.6V when fully charged - about 1.55V is common or 6.2V for 4, and up to 6.4V may be seen. <S> There is not much energy in this initial high voltage "tail" and voltage falls to 1.5V or below very quickly. <S> So, you should be safe, but YMMV, alas. <S> A solution would be to use an LDO (low dropout voltage) regulator OR a clamp regulator which takes the peak energy out of the battery or a series diode to drop 0.4 to 0.8V (Schotky / Silicon). <S> LDO is best solution <S> but you want as little drop as possible. <S> Clamp to drain peak battery voltage is unusual but viable. <S> A zener could be used but is too inexact. <S> An eg <S> TL431 clamp regulator in a TO92 or other largish package (to get OK dissipation capability) ould do. <S> A TL431 plus a transistor would be safer. <S> Series diode is cheap and easy but prevents full battery use. <S> Say minimum usable battery voltage is 4.6V (may be higher). <S> At 1.15V/cell there is still some battery capacity left. <S> Adding a Schottky diode increases minimum battery voltage to 4.6 + 0.4 = 5V or 1.25 V/cell. <S> Some capacity wasted. <S> At the top end a 0.4V drop diode results in Vbattmax of say (1.55V x 4 - 0.4) = <S> 5.8V or 1.45V/cell. <S> "Almost <S> certainly safe". <S> At 4.6V, V per cell is 1.15V where NimH still has modest energy left. <S> At top end Vmax = say 1.35V, maybe 1.4V for short periods at start. <S> 4 x 1.4V = <S> 5.6V. <S> Very probably safe. <A> The charging intelligence is in the phone that's being charged (or even its battery). <S> When you charge via USB it sees just a fixed 5V, current limited at 500mA, so on that side there's no control over the charging. <S> The only thing that might make you frown is that the 4 AAs don't give the 5V a USB port would. <S> Most chargers can work with voltages to at least 6V, so you should be safe. <S> Your phone's battery won't drain to the discharged AAs; they're not directly connected, or simply over a resistor. <A> Here's a writeup on its build process from Lady Ada: <S> http://learn.adafruit.com/minty-boost <S> It's using a MAX756 , which only works down to 0.7V, but that's more than low enough to call 2 AA batteries completely drained. <A> The protection is the in equipment that is being charged.. <S> Take a look at this IC ( <S> STC4054 <S> http://www.st.com/stonline/books/pdf/docs/12666.pdf ) from ST. <S> This is a common charger IC that guides the charging process.
|
There's a DC-DC converter in between, which may actually charge the phone's battery from a voltage which is lower than the battery's. Using NimH works but is more marginal at bottom end and safer at top end. Take a look at the Mintyboost from Adafriut, it has a proper regulator that allows it to output a constant 5V and drain the battery completely.
|
HUGE capacitor recommended in datasheet for Audio Amp I'm using the TPA3111 audio amplifier in a circuit of mine and I'm working on adding the appropriate capacitors. However, on page 18/19 of the datasheet TI recommends to use a 220mF capacitor on the power line along with a 220uF capacitor on each of the PVcc pins. Is this entirely necessary? If you take a look at the evaluation board TI offers for this chip they don't do any of that. The same goes for the application case in the datasheet. I'm already using quite a few capacitors according to the sample application and eval board. The largest so far are 100uF electrolytics and then there are bunch of smaller ceramic caps. Any idea on what dangers I might have if I do not include the additional 220mF or 220uF caps? Thanks. <Q> The Example schematic only has 100uF <S> + 0.1uF <S> + 1000pF across the rails. <S> Have you considered the "mF" may be a typo? <S> In one place, the datasheet says typically 0.1 mF <S> to <S> 1 uF . <S> I wonder if they meant to type n, and accidentally hit m. <S> Also, I copied the u symbol out of the PDF, and it got printed as m when it was pasted. <S> Cut&Paste may be at fault here <S> , it certainly seems to be used within TI's various datasheets. <S> Also, millifarads have come to be an almost unused unit. <S> It's generally Farads -> microFarads -> <S> (nanoFarads - somewhat uncommon) -> picoFarads. <S> Furthermore, looking at the TPA3111 Evaluation Kit is informative: The device is bypassed with two 100uF electrolytics (along with 0.1uF and 1000pF ceramics). <S> Also, looking at similar parts from the same line is informative. <S> The TPA3110 <S> (15W vs TPA3111's 10W) merely says a larger aluminum electrolytic capacitor of 220 uF or greater placed near the audio power amplifier is recommended . <S> It's worth noting that the same datasheet's example schematics only use two 100uF caps for bypassing. <S> The same note as in the TPA3111 is present in the TPA3112 datasheet. <S> It's also worth noting that the TPA3110 and TPA3113 have identical "Power Supply Decoupling" paragraphs, despite the fact that one is half the power of the other (15W vs 6W), which further inclines me to think typo. <S> The 25W <S> TPA3123 only recommends 470uF of bulk capacitance. <S> The 100W <S> TAS5121 only recommends 1000uF. <S> Edit: We will see if it is a typo: "Below is what you submitted to tis-doc-errors@list.ti.com on Tuesday, July 26, 2011 at 04:13:23; E-mail: <S> tis-doc-errors@list.ti.com Lit Number: <S> SLOS618BB <S> Part Number: TPA3111D1 Error Page <S> No: 19 Error Description: Please see this thread: <S> HUGE capacitor recommended in datasheet for Audio Amp " <A> It all depends on how hard you intend to drive the amplifier. <S> If your goal is to get close to the full 10W over the full audio range then you may need substantial capacitance on the power rail. <S> Class D amplifiers have very spiky current draw and the gain of the Class D amplifiers is directly proportional to the rail voltage. <S> Any dip in the rail will cause poor frequency response, usually on the low end. <S> For your application this may not be important. <S> Additionally the part you've chosen is a half-bridge design which can suffer from "bus-pumping" issues which the can cause currents flowing back into the supply to push the rail voltage up to levels dangerous to the circuit. <S> A common way to deal with this is to use very large decoupling capacitors to "soak up" the pumping. <S> Generally this is more of an issue with high power amplifiers switching under something like 100Khz. <S> However the designer for the datasheet may have added large capacitance as a safety net. <S> So i would say in summary, unless your pushing this device to its limits, you don't need as much capacitance as they discuss in the datasheet. <A> Audio amplifiers are notorious for massive current spikes, especially for lower frequencies. <S> Think about how much more current is consumed when someone bangs the bass drum <S> and it has to move <S> pump extra current into the speaker to replicate that massive one-way movement of the speaker cone. <S> The only problem you should see is possibly some distortion when you have higher volumes at lower frequencies, like with drums, and there may be a noticeable drop in power to the rest of the circuitry (lights may dim, etc when you bang that big ol' bass drum) at times.
|
The big capacitors are there to provide that extra boost of current when it's needed.
|
Is there an IC for a 32:5 Encoder? I need a part similar to a 74HC148 8:3 encoder, but which takes 32 input lines and encodes them to a 5 bit value. I can guarantee that only one of the 32 will be high at a time, so I don't even need the priority feature although it would be a plus. If possible I would like to stay in the 74HC series, or at least compatible chips. I could implement it with 5 16-input OR gates, but that would require something like 10 chips to implement the entire encoder. It feels like there must be a better way. Is there a single chip made to do this? Google and digikey searches failed me. Is there a better way to implement it than with a whole bunch of OR gates? I was trying to find a way to combine the 8:3 encoders, but a clever working solution escapes me. <Q> Page 7 of the very datasheet you linked shows how to cascade the 74HC148 for additional inputs. <S> This datasheet has a better diagram showing n-device expansion: <S> I don't know of any 16/32 input encoders, if you really must do this in a single IC then a simple FPGA / CPLD is your best option. <A> You don't need a 4GB VHDL development system for this, you can write your code in Abel , which is a lot easier to start with. <S> edit <S> Oh, that's right <S> , you're the one with the homebrew CPU! <S> Well, I guess you'll have to stick to 74HCxx then :-). <S> You won't need 16-input gates (don't you mean <S> 32-inputs? <S> Well they don't exist anyway), I would take the outputs of four 74HC148s and feed them to a fifth. <S> Further reading Xilinx appnote: Using ABEL with Xilinx CPLDs <A> There is not one that I know of, but this type of thing is relatively easy in a CPLD. <A> Yes, there are single chips for this called microcontrollers. <S> You need one with at least 37 digital <S> I/O lines but not much processing power. <S> I don't know (without looking) <S> whether there is one with enough I/O lines in a 44 pin package, as these are often really 40 pin packages with a few redundant pins for the TQFP. <S> Some newer ones that don't come in DIP anymore might give you all 44 pins to do useful things. <S> Of course there are 64 pin parts that definitely have enough digital <S> I/O pins. <S> Something like a PIC <S> 16F1526 should do it.
|
I don't think a priority encoder 32 bits wide exists, but it should be easy to implement in a CPLD .
|
Sourcing parts for Really, Really low frequency filters I'm doing a project at work that has a couple of esoteric requirements, particularly for a 0.1 Hz (0.07 Hz really, due to part availability limitations) Highpass filter (in the data acquisition system). Right now, I'm using a 22uF film cap, and 100K resistor, and the whole affair works quite well. However, the film cap is enormous (1.240" L x 0.532" W), and the resultant PCB is really a very large (there are many channels). I really don't want to go too much higher for the R in the filter, since it's going into an op-amp. With the existing system (OP27, need the really low 1/f knee), +-10nA bias current, you get \$10*10^{-9}A*100,000\Omega = 0.001V,\$ or 1mV of offset due to bias current. WIMA used to manufacture some compact 22uF 16V film caps , but they have EOLed them without a replacement. Unfortunately, the application is a bit extreme. The caps need to be able to handle extremely low temperatures, and hard vacuum, which I think means electrolytics are out. Does anyone manufacture large, low voltage film caps (voltages in question are +-5V, nothing major)? Alternatively, does anyone know how electrolytics fare in vacuum? <Q> I'd forget analogue techniques and use DSP. <S> At 0.1 Hz virtually any MCU could be used, but I'd use a dsPIC <S> as I have the MDS dsPIC filter design utility. <S> It actually writes the code for me. <S> It'll be cheaper, smaller, and operate in a vacuum without any problems. <A> Why does it have to be a film cap? <S> Why not a ceramic? <S> I'm no expert on vacuum, <S> but I think they should be able to handle that fine. <S> According to my calculation, you only need 16 µF with 100 kΩ to get a 100 mHz rolloff. <S> In any case, a couple of 10 µF 20V ceramics in parallel with good dielectrics should work. <S> Using them over a small part of their voltage range keeps the capacitance reasonably constant. <A> Instead of using an RC highpass filter, why not use an RL highpass filter - but instead of using a real inductor use an inductance gyrator. <S> You can use active components (and a couple much smaller capacitors) to simulate a massive inductor connected to ground to give you your low frequency cutoff point, and it will save you a lot of board space and the other problems of using a large capacitor. <S> Here are some notes on gyrators . <S> Edit: <S> Here's a gyrator RL filter design for a cutoff frequency of 0.1 Hz, using 2 opamps, resistors and a 0.1uF capacitor to simulate a 1000 H inductor. <S> The gyrator design is based on the one here by Jim Thompson. <A> Many EKG and EEG machines have "a very small AC signal of interest, modulated onto a very large, (slowly) variable" unwanted near-DC signal -- the "baseline wander". <S> Since the heartbeat can drop as low as 40 Hz, we typically want a linear-phase highpass to cut off everything below around 0.5 Hz. <S> a b <S> Perhaps you could use the same techniques they use for their highpass filter: <S> Servo loop: <S> Instead of passing the signal through the capacitor of a high-pass passive RC filter, they use an active filter that integrates the DC component and subtracts it from the signal ("servo loop").A active lowpass filter somehow tweaks the main signal chain to produce a highpass effect. <S> My understanding is that this approach can be scaled up to extremely high resistances -- say, 10 MOhm and 1 uF to get a roughly 0.015 <S> Hz high-pass corner frequency -- without the noise that such high resistance values normally cause. <S> digital filtering: some people say that baseline wandering is easier to filter out in software than in hardware. <S> a b c <S> (See the "Highpass Filter Simulation" page -- how do I link directly to that page?) <S> The INA322 datasheet in Fig. <S> 9 <S> "Simplified ECG Circuit for Medical Applications"uses <S> a servo loop driving the REF input to produce a highpass effect. <S> Figure 37 of the INA333 datasheet has another servo loop. <S> Figure 69 of the AD8420 datasheet has another servo loop: <S> 0.5 Hz high-pass. <S> Figure 70 of the AD8295 datasheet has another servo loop. <S> Figure 5 of " Getting the most out of your instrumentation amplifier design " has another servo loop. <S> The ECG prototype from Matthew Shieh has another servo loop. <A> EPCOS have Metallized Polypropylene Film (MKP/MFP) series capacitors listed on their website . <S> Digikey have these capacitors up to 110µF!!
|
The Imac Engineering people claim they have a hipass corner frequency of 0.03 Hz.
|
How is using a transformer for isolation safer than directly connecting to the power grid? How is using a 1:1 transformer safer than using the mains straight off? Is it because you can limit the current coming from the transformer whereas straight from the mains its not current limited? I fail to see how its "safer" when playing with electricity is dangerous. Could someone please explain why it is considered safer to be isolated by a transformer. <Q> Without a transformer the live wire is live relative to ground. <S> If you are at "ground" potential then touching the live wire makes you part of the return path. <S> {This image taken from an excellent discussion here With a transformer the output voltage is not referenced to ground - see diagram (a) below. <S> There is no "return path" so you could ( stupidly ) safely touch the "live " conductor and ground and not received a shock. <S> From <S> The Electricians <S> Guide <S> I say " stupidly " as, while this arrangement is safer <S> it is not safe unconditionally. <S> This is because, if there is leakage or hard connection from the other side of the transformer to ground then there may still be a return path - as shown in (b) above. <S> In the diagram the return path is shown as either capacitive or direct. <S> If the coupling is capacitive then you may feel a "tickle" or somewhat mild "bite" from the live conductor. <S> If the other conductor is grounded then you are back to the original transformlerless situation. <S> (Capacitive coupling may occur when an appliance body is connected to a conductor but there is no direct connection from body to ground. <S> The body to ground proximity forms a capacitor.) <S> Murphy / circumstance will work to defeat this isoation. <S> This is why, ideally, an isolating transformer should be used to protect only one item of equipment at a time. <S> With one item a fault in the equipment will propbably not produce a dangerous situation. <S> The transformer has done its job. <S> BUT with N items of equipment - if one has a fault from neutral to case or is wired wrongly this may defeat the transformer such that a second faulty device may then present a hazard to the user. <S> In figure (b) above, the first faulty device provides the link at bottom and the second provides the link at top. <S> Similarly: <A> Mains is referenced to earth ground: the same potential that you're standing on, that your workbench is at, that nearly anything you touch (that's not isolated) is at. <S> It develops voltage with respect to those items. <S> After an isolation transformer, the earth ground reference is gone. <S> If you touch the line side after the transformer, suddenly that becomes the reference for the whole system, and the neutral side (which you're not touching, and which would normally be safe) is a dangerous sinusoid with respect to earth ground. <S> The important thing is that the point you touch is still safe. <S> However, touching both sides of the transformer simultaneously is just as dangerous as not using the transformer at all. <S> Use one hand when working with high voltages. <A> It is because using an isolation transformer means there is no connection between the earth and the live/neutral wires. <S> You could handle one of the phase wires (as long as you don't handle the other) with no harm as it doesn't matter if you are earthed.
|
SO a transformer makes things safer by providing isolation relative to ground.
|
Best way to get +5V or +12V or 0V from arduino pin I am working on an eeprom programmer as an Arduino shield. There are two pins on my EEPROM that require a 12V signal as part of the programming process. One of them is an address pin that I am also driving by a signal through a 595 shift register, and the other is the output enable pin which is connected to a regular Arduino pin for 5V and 0V operation. I would like to have a pin on the arduino for each of the 12V pins that will set it to the higher voltage without damaging any other components. I was thinking I could use a simple transistor setup like Figure 1 on this page, but I'm really not sure how well that would work when connected to both sources at once. I would need diodes to protect my 5V components from getting damaged by the 12V signal, but I suspect that circuit may work out. I will simplify things by ensuring I that Vin is a solid 12V supply. Here's my initial design: My main problem is that I don't understand the circuit elements in play here very well. Will this explode? I'm still a bit confused by simple transistor circuits, and not sure what specific parts to use. I feel like what I am trying to do is not that hard, but my circuit skills are not good enough to know the best way to accomplish this. Thanks. <Q> Here is a circuit that should work: <S> Your basic idea of using a high side PNP switch was fine. <S> The problem with it is that this exposes the high voltage to the micro pin. <S> In this circuit, Q2 is the high side switch that turns the 12V output on or off. <S> Q1 switches the high side switch, thereby isolating the micro from the 12V on the base of Q2. <S> When the base of Q1 is held at 5V, the emitter will be 4.3V, so there will be 1mA thru R1. <S> Most of this also flows thru the collector of Q1, which thereby acts like a switchable 1mA current sink. <S> Most of this can only come from the base of Q2. <S> Figuring the two transistors each have a gain of at least 50, then the output is good for at least 45mA for Q2 to stay in saturation. <S> The purpose of R3 is only to make sure Q2 is really off when Q1 is off. <A> While this is an old post, I recently had this same problem when I was making an EEPROM burner for retrocomputing (where I needed to write some 27C512-style EPROMs). <S> The currently accepted answer has the correct idea (which I used), but it is missing the logic level (5V) control and protection, and pull-down on the output side. <S> The following circuit adds these, and worked for me: <S> The microcontroller pin connecting to <S> HVCtl controls the "high" voltage ( VPP , e.g., 12V), <S> LVCtl controls the logic level voltage (e.g., 5V), and when both are low, R4 pulls the pin low. <S> The Schottky diode protects the <S> LVCtl <S> pin on the microcontroller from VPP , while having such a low voltage drop that it doesn't affect recognition of the logic level. <S> That is, the output ( OUT ) is at VPP <S> whenever <S> HVCtl is high, otherwise it is at the state of <S> LVCtl . <S> The diode and transistors may be substituted for other parts, these are just the ones that I used (because I had them around). <S> (edit: In retrospect, I would suggest a lower value resistor for R2 to support higher programming currents, e.g., 10K or 4.7K. <S> However, the 22K shown works in practice even for vintage EPROMs with up to 30 mA or so of programming current.) <A> From previous questions I gather that you don't want a universal programmer, but rather one which can program the device you need for your homebrew CPU, CMIIW. <S> I would pick another EEPROM device. <S> 12V programming voltage is really Fred Flintstone! <S> Modern devices are programmable at 5V. <S> The SST39SF010A <S> can write at 5V , and for the money you get twice 64KB , so if you wish you can load two programs in it and switch between them by toggling A16. <S> For future enhancements there are pin-compatible Flash devices with 256KB and 512KB (same datasheet). <S> And it's available in DIL package. <S> Who needs 12V?!
|
I already suggested to have a look at Flash memory instead of EEPROM.
|
PIC UART Interrupt Not Triggering I have PIC16F628A that I am trying to have read from UART. Without interrupts, it reads the first 3 bytes fine but hits an OERR. To combat this, I thought an interrupt would be good and load any received bytes into a buffer variable that could be read in later (ring buffer of array type char). But the interrupt is not triggering and I've run out of ideas. CMCON = 0x07; //16F627/8 spcial function reg (RAx is port)CCP1CON = 0b00000000; //Capt/Comp/PWM offOPTION_REG = 0b00000000;T1CON = 0;INTCON = 0;PIR1 = 0;GIE = 0;PIE1 = 0;BRGH = 1; /* high baud rate */SPBRG = 19200; /* set the baud rate */SYNC = 0; //AsyncTXEN = 0; //Disable transmitTXIE = 0; //Disable transmit interruptRCIE = 1; //Enable Receive interruptSPEN = 1; //Enable serial pinsCREN = 1; //Enable continuous receiveSREN = 0;TX9 = ninebits?1:0; /* 8- or 9-bit transmission */RX9 = ninebits?1:0; /* 8- or 9-bit reception */PEIE = 1; //Enable external interruptGIE = 1; //Enable global interrupt I have simplified my interrupt to turning on a light: extern interrupt isr(void){ RB5 = 1;} But it's not triggering. The project is reading a barcode scanner over serial and processing the barcode. Can anyone offer some assistance? EDIT Ok since you don't seem to understand. I'm going to post the actual routines: void initialize(){ CMCON = 0x07; //16F627/8 spcial function reg (RAx is port) CCP1CON = 0b00000000; //Capt/Comp/PWM off OPTION_REG = 0b00000000; T1CON = 0; INTCON = 0; PIR1 = 0; GIE = 0; PEIE = 0; PIE1 = 0; sci_Init(BAUDRATE ,SCI_EIGHT);// Baud set and Bit set TMR0 = 1000; T0IE = 0; PEIE = 1; //Enable external interrupt GIE = 1; //Enable global interrupt //Set inputs to input SetButtons(); //Set relays to output SetRelays(); TRISB5 = 0; LEDStatus = 0;}unsigned char sci_Init(unsigned long int baud, unsigned char ninebits){ int X; unsigned long tmp; /* calculate and set baud rate register */ /* for asynchronous mode */ tmp = 16UL * baud; X = (int)(FOSC/tmp) - 1; if((X>255) || (X<0)) { tmp = 64UL * baud; X = (int)(FOSC/tmp) - 1; if((X>255) || (X<0)) { return 1; /* panic - baud rate unobtainable */ } else BRGH = 0; /* low baud rate */ } else BRGH = 1; /* high baud rate */ SPBRG = X; /* set the baud rate */ SYNC = 0; //Async TXEN = 0; //Disable transmit TXIE = 0; //Disable transmit interrupt RCIE = 1; //Enable Receive interrupt SPEN = 1; //Enable serial pins CREN = 1; //Enable continuous receive SREN = 0; TX9 = ninebits?1:0; /* 8- or 9-bit transmission */ RX9 = ninebits?1:0; /* 8- or 9-bit reception */ rxBuffIndex = 0; rxBuffRead = 0; return 1;}void sci_LoadBuffer(void){ rxBuffer[rxBuffIndex] = RCREG; rxBuffIndex = ++rxBuffIndex % MAXBUFFER;}unsigned char sci_ReadBuffer(){ unsigned char byte; do { byte = rxBuffer[rxBuffRead]; }while( byte == 0 ); //Block until valid data rxBuffer[rxBuffRead] = 0; rxBuffRead = (++rxBuffRead) % MAXBUFFER; return byte;}void interrupt isr(void){ if(RCIF) sci_LoadBuffer(); LEDStatus = 1;} I know that's not EVERYTHING but that should be enough to diagnose why the interrupts aren't triggering. THAT'S ALL I NEED! Triggering the interrupts. I'm using MPLab with Hi-Tech C Compiler. Which from the manual automatically saves state and restores it when entering/exiting the interrupt. <Q> TRISB1 needs to be set to 1 to configure RB1 (RX) as an input. <S> I'm not sure what the default is, so it may be ok. <S> You need to clear the receive interrupt flag (RCIF) by reading the receive register (RCREG). <S> In addition, since the receive register is doubled-buffered, you may need to read it more than once. <S> So your interrupt routine needs to look more like this: extern interrupt isr(void){ while (RCIF) { char ch; RB5 = 1; ch = RCREG; // normally would go into an array and increment a counter }} <S> I don't know if that is your only problem, since you indicate you are not getting into the interrupt routine at all. <S> But the above is the correct way to read the characters out of the receive buffer. <S> ====================================== <S> EDIT: <S> Don't know if this will help or not, but in this post , before enabling interrupts, the code clears out the FIFO first. <S> (Their code also clears out the RCIF flag, but since it is readonly on your chip, that isn't needed.) <S> ch = RCREG; // <S> clear FIFO <S> ch = <S> RCREG;ch = <S> RCREG;// then enable interruupts ... <A> Ok, two things. <S> First, within your interrupt routine, you usually have to clear the IF flag RCIF to enable the interrupt to fire again. <S> That's not the reason the interrupt's not firing at all though. <S> It also terminates the function with a "return from interrupt" command instead of a simple "return" command. <S> What it doesn't do is link the function into the interrupt vector. <S> There is usually only a small amount of space around the interrupt vector area, so it is normal to place a goto at the address of the interrupt vector which calls the name of your interrupt routine. <S> Depending on your compiler there are a number of ways of doing this. <S> I suggest you read the manual for your compiler about interrupt vectors and the sample code for it. <A> I didn't read all of your long post, but happend to notice this on skimming: extern interrupt isr(void){ RB5 = 1;} This is definitely wrong. <S> I don't know what the interrupt condition is, but you're not clearing it. <S> The processor will get hung on the first interrupt because it will re-enter the interrupt routine immediately after it completes because the interrupt condition is still active. <A> Have you tried looking at the RS-232 signal on a scope to see if the baud rate is correct? <S> Try transmitting a few characters from the PIC to make sure the UART has the right baud rate setting.
|
The problem with your code is that you are defining a function as an interrupt - which is fine - that causes the compiler to push things to the stack for you automatically, and pop them off after the routine is over.
|
difference between COM port and SATA can somebody please explain why can't I say that SATA is a COM port, SATA is a serial port and COM port is a general name for serial ports, am I wrong? thanks in advance <Q> COM ports are the common name for EIA-232 ports (aka RS-232), i.e. ports where only the lower levels of communication are defined: physical (mechanical and electrical) and data link. <S> For SATA these are different, and SATA defines also network and transport layers, which means packaging and error detection/correction. <S> You can run a protocol on CMO ports which do this too, but they're not part of the COM specification itself. <A> The "COM" port is the name given to an RS-232 port in DOS (and continued into windows). <S> It is specific to that type of port on that operating system. <S> As far as I am aware no other operating system uses the name "COM" port. <S> Unix & Linux refer to them as "Serial TTY ports". <S> COM is the original, yet still common, name of the serial port interface on IBM PC-compatible computers. <S> It might not only refer to physical ports, but also to virtual ports, such as ports created by bluetooth or USB-to-Serial adapters. <S> - Wikipedia SATA, while yes it is a "serial protocol" is not a COM port and has nothing to do with COM or RS-232. <S> Line signaling levels are completely different, the data contained within the bit-stream is formed differently, etc. <S> Serial ATA (SATA or Serial Advanced Technology Attachment) is a computer bus interface for connecting host bus adapters to mass storage devices such as hard disk drives and optical drives. <S> Serial ATA was designed to replace the older ATA (AT Attachment) standard (also known as EIDE), offering several advantages over the older parallel ATA (PATA) interface: reduced cable-bulk and cost (7 conductors versus 40), native hot swapping, faster data transfer through higher signalling rates, and more efficient transfer through an (optional) I/O queuing protocol. <S> - Wikipedia Trying to compare the two is like trying to compare a parallel printer port with a parallel IDE connector. <S> While they are both parallel, they are completely different in the way they operate and the signals that are passed through them. <S> While you're about it, you may as well say that USB is the same as COM and SATA, and why not include DVI in that as well? Or Ethernet? <S> They're all "serial protocols" in that they transmit one bit after another, but they are all so different to each other in so many ways that they just cannot be compared to each other in a sensible manner. <A> It's a name specifically for PC serial ports. <S> Even though serial data transfer takes place in both COM and SATA connections, the two are completely incompatible, so it's just incorrect to call a SATA connection a COM port.
|
The thing is, "COM port" is not a general name for a serial port.
|
Reading empty (new) AT24C16 I2C EEPROM - 0xFF or 0x00? What byte vaule should I see when reading empty EEPROM chip via I2C? <Q> An erased EEPROM normally reads as 0xFF, but it may also contain seemingly random data. <S> We've had this often in microcontroller's Flash, which sometimes has code programmed in it for production testing. <S> This may also be done with EEPROM. <S> Start with erasing the device if you want to be sure. <A> The way an EEPROM works, the act of erasing sets every bit to 1. <S> When you write to it the bits that shouldn't be 1 are set to 0. <S> It's all to do with the movement of free electrons in the semiconductor layers. <S> Therefore, an erased 'empty' EEPROM will always read 0xFF for every memory address. <S> However, you cannot always guarantee the current state of an EEPROM when it arrives from the supplier / manufacturer. <S> It may have had some data written to it in order to test it, and not erased afterwards. <S> You'd expect a good manufacturer to always finish the testing with an erase, but this isn't something that can be 100% guaranteed. <S> Therefore, you should erase the EEPROM before first use. <S> That way you don't have to worry about what state it is in when first inserted into the circuit. <A> Every eeprom I've ever encountered reads 0xff in its factory state.
|
If the EEPROM is to be used to store data that is generated by the system it is placed in, then you could have an 'identifying' byte in a known location - check to see if that byte is there, and if not erase the EEPROM and write the identifying byte.
|
ARM controllers in small packages Are there ARM controllers for small applications (like Cortex M0) available in small packages with maximum say 20 pins? I have the impression that in this area they don't quite are a threat for the usual suspects, like PIC and AVR. <Q> Usually, because it also depends on the technology; QFP technology for instance is cheaper than CSP <S> (Chip Scale Package). <S> I presume this WLP (Wafer Level Package) for the LPC1102UK is the smallest ARM package to date, body is 2.17 x 2.32 x 0.6 mm, with 16 bumps. <S> That's damn small, yet it costs almost USD 5.00 quantity one (Digi-Key). <S> Even at 3000 pieces the price is still over USD 2.00. <S> (Remember, this is a Cortex M0, the lowest end ARM.) <S> From recent limited research I found that there are few Cortex M devices in very small packages, I haven't found anything like a SOT23-8, for example. <S> Apart from the TI LM3S101 in a Fred Flintstone Package (aka SOIC-28) most packages seem to be QFP and QFN , and more of the former than the latter. <S> This is somewhat surprising, since the PCB assembly technology for both is the same, both can be inspected using flying probe , for example (which isn't possible on CSPs). <S> Yet the QFN needs much less space than an equivalent QFP. <S> The explanation is demand , of course. <S> Apparently most customers don't need the smaller space of the QFN (yet). <S> Some manufacturers are pretty flexible about packaging, and may be prepared to introduce a new package for an existing device if you buy, say, 100k devices per year. <S> This has more administrative than technical implications. <S> So while ARM is widespread most customers will either need smaller quantities or don't really need the new package. <S> Still I expect ARMs to become available in smaller packages, like less than 20 pins. <S> Especially for Cortex M0 this is going to be needed to successfully take 8-bitters the wind out of the sails. <S> While SOT23 may not be an option I see many possibilities in QFN and particularly DFN. <S> Unlike DIL DFN is not limited to a specific width. <S> This table shows how many variants there are available from just 1 manufacturer . <S> So there is always a solution for a specific number of pins and die size. <S> Small controllers like the LPC1102 would easily fit in a 3 x 3mm QFN-16, for instance, but apparently (and unfortunately?) <S> this hasn't happened yet. <A> NXP LPC1102 16 pins <S> http://www.nxp.com/documents/data_sheet/LPC1102.pdf <S> There are also several 32 pin M0 and M3 parts in NXP's range <S> However for very small apps, 8 bit MCUs often still have advantages, even if cost is similar, e.g. lower density packages, wider supply voltage, onboard eeprom, lower power consumption. <A> The LPC810 comes in a DIP8 package. <A> The Kinetis KL03 chip-scale package (CSP) <S> MCU is the next world's smallest ARM Powered <S> ® MCU designed to support the latest innovation in smart, small devices. <S> Available in the ultra-small 1.6 x 2.0 mm² <S> wafer-level CSP, the Kinetis KL03 CSP (MKL03Z32CAF4R) reduces even more board space while integrating even more rich MCU features than previously seen in the market.
|
The smallest ARM microcontroller to date (March 2014) is the Freescale Kinetis KL03 micocontroller , based on the 32-bit ARM Cortex-M0+ core : Smaller packages, more specifically packages with less pins, are usually less expensive .
|
Is FCC-accredited testing required again if circuit board is placed in a different package? I am considering enhancing a commercially-available device so that it can be installed outdoors. The enhancement would be to seal the product inside a resin case to make it weather-resistant. The result will be something that looks like a brick with wires sticking out for 12v DC power and Ethernet data. Once inside the sealed case, the product, along with its FCC sticker, would be invisible. The resin case would have my branding on it and not that of the sealed device. Can I just remove the FCC sticker on the original device, before encasing it, and apply it to the exterior of the new sealed case, or do I need to go through the FCC conformance testing as if it were a new totally different product? The sticker would have the branding of the sealed-device manufacturer on it. The device is an unintentional radiator. I have the option of either sealing the entire device, as it is sold, or removing it's circuit board from it's original enclosure, only sealing that, and discarding it's original plastic non-shielded enclosure. Would it make any difference between any of these two scenarios, from the point of view of FCC conformance testing requirements? Thanks for your attention. <Q> I am going through this too, and likely late for the original question, but thought I'd share. <S> FCC requirements are for anything marketed or sold that emits electromagnetic radiation. <S> If it's a computer or gear that isn't supposed to transmit then it is, as mentioned, an "unintentional radiator"--so it's now in that ballpark of rules. <S> If the device is for commercial use, not in homes, then it falls under the Class A rules (less stringent). <S> If it is for residential use you have to follow the rules for Class B. <S> If you are encasing the whole product inside something else then you don't have to go through FCC validation and can use the existing approval. <S> If it is a Class B device with an FCC statement on its label then you can put that same info on the outside of your box. <S> If you don't want to use the full product and put parts in your own resin case then you have to go through the FCC processes. <S> If you market and sell and then get caught you could end up with severe penalties, along with being stopped from marketing your product. <S> Start at: http://transition.fcc.gov/oet/ea <S> Note the "Application Information" and "Equipment Authorization Procedures," including: VerificationDeclaration of Conformity (DOC)Certification <S> To get an idea of testing go to this page: http://transition.fcc.gov/oet/ea/fccid/ and plug in the FCC ID from any product you have handy that has an FCC ID (or you search the Web for one) <S> and then you can see the report that was generated--including a description of the lab setup, the tests, and the report data. <S> We will work with a lab for the testing but also work on our own testing ability so that we can do at least part of it on our own (so that we don't go to a lab with something that won't pass). <S> Good luck. <A> If you are only putting something in a box and not re-selling it, there is, in my understanding, no requirement to do any further testing or make any further declarations. <S> The problem will be if you indend to re-sell the device, as it will now not be the same as originally tested. <S> If all you are doing it putting it in a box, it is highly unlikely that the emissions characteristics will change. <S> However, you can no longer use the original FCC sticker as that only applies to the original device that you put in the box. <S> You will need a new sticker that identifies you as the responsible organization. <S> You don't necessarily have to re-test, you can self declare if you can make the case that the emissions will not have changed. <A> If this is a unintentional radiator and you are using it completely within its specified limits, then I think you're OK. <S> Permanently keeping the cables plugged in shouldn't matter because the device was originally tested with cables hanging off. <S> If the device is specified for maximum cable lengths, make sure you don't exceed them. <S> If you open it, then you're on your own. <A> See Olin's answer. <S> Everything electronic must be within the conducted and unintentional radiator limits. <S> Many people pay a lab to do this testing to make sure their device is ok. <S> Unless your box somehow acts as a waveguide and focus some low level emissions, you should be ok. <S> You can get a "modular" approval for an intentional radiator and include this is a device. <S> Then on the label you say "Contains FCC ID: foo". <S> Modules have additional rules to me such as they must have a shield. <S> You can't take a device with an FCC ID that doesn't have a modular approval and stick it in something and use "Contains FCC ID: foo". <S> If you ship something without a module (e.g. an embedded board with a WiFi USB stick) <S> the WiFi stick has to be visible from the outside and user serviceable.
|
Unless putting this device in a box is outside its intended use, I don't expect any problems. Devices with an FCC ID are intentional radiators and have further testing and must be certified (that's when they get the FCCID). There isn't a certification for this: you must comply.
|
Difference between Zero-Crossing Optoisolator and Regular Optoisolator I am trying to find any documentation on what the purpose is of a Zero-Crossing Circuit TRIAC Optoisolator. The datasheets don't explain the concept well enough. If you do answer, please include references or explain how you found out. Thank you! <Q> Zero-crossing is typically used for incandescent bulbs. <S> You may have noticed that when incandescent bulbs fail they always fail when they're switched on. <S> That's because the mains phase can be near its maximum when switching on. <S> Combined with the low resistance of a cold bulb this results in a high current peak, which may burn the filament. <S> When you switch on a zero crossing you avoid these peaks. <S> How I found out? <S> I've known this since my time in college. <S> It simply makes sense. <A> Wikipedia has a little information on the theory of Zero Crossing. <S> Basically, the triac will hold off its switching until the alternating waveform of the load signal crosses the 'zero' or midpoint of the waveform. <S> This helps alleviate sudden voltage spikes when the switched load jumps from 0V to say 100V in one instant. <S> By ensuring that the load is only switched when it's waveform is crossing the mid-point, the voltage increase will be a smooth rise from 0 to maximum. <A> According to my research, a zero crossing detector is used specifically for protecting a connected load. <S> It does not exactly matter whether it's with a triac or directly with mains. <S> As mentioned in one of the replies above, a load could most vulnerable at the introduction of a sudden peak level of the AC phase during a power switch ON, a zero crossing sensor ensures that the load is always switched ON at the first zero crossing of the applied AC phase, thus safeguarding the load from the dangerous peak levels. <S> With triacs a zero crossing could be useful for protecting the loads as well as for reducing RF interference. <S> I have tried to explain the concept comprehensively in the following article How to Make a Zero Crossing Detector Circuit
|
Zero Crossing refers to the load voltage at which the triac will switch.
|
Whats inside a resistor? How does a resistor "resist" current/potential? I know it's an elementary question, but I'm sure others are wondering too. <Q> Just as it happens I'm reading this application note by Vishay titled " Basics of Linear Fixed Resistors ", which explains the construction of PTH and SMT fixed resistors. <S> Most resistors have a resistive layer on the surface of a non-conductive carrier, either carbon or metal-based or thick film. <S> The resistance value is obtained by laser-cutting lines in the film. <A> The simplest is just a wirewound resistor. <S> Metals are not perfect conductors, so a long thin piece of wire will have a predictable resistance to it. <S> Cram it inside a little component by wrapping it into a helix. <S> There are many other types, too, using thin films, etc: <S> As for what causes resistance, a simple explanation is that it's analogous to friction. <S> It converts electrical energy into heat energy, the same way friction converts mechanical energy into heat energy. <S> For a more detailed description, you need to get into physics . <S> In metals, like a wire-wound resistor, "the thermal motion of ions is the primary source of scattering of electrons (due to destructive interference of free electron waves on non-correlating potentials of ions), and is thus the prime cause of metal resistance" <A> Here is a link which I suppose, will definitely quench your thirst if nothing else has yet. <S> http://www.ecawa.asn.au/home/jfuller/electronics/resistors.htm <S> The track may have been machined, or 'burnt' away with a laser beam.
|
Inside a carbon resistor is a 'ceramic 'core' on which is deposited a spiral carbon 'track'.
|
Awesome modems with Arduino compatibility? I'm looking for a modem that can hook up to the input port on a portable ham radio and an ardunio board serial connection. Not really interested in APRS, it's too high level for my needs (long distance remote control). Just need to be able to send 10 characters maybe 3 times a second (>1200 baud) low power if possible distance of like 40-50m max Seems like that will all be taken care of by the radio, just wondering if there is a good hardware modem for this kinda of thing. <Q> As you probably know, often one has a choice of implementing a function is either hardware or software. <S> I understand that this is a hardware rather than a software forum, but you may want to consider the fact that this can be done completely in software. <S> Here is full documentation on the project: https://sites.google.com/site/ki4mcw/Home/arduino-tnc <S> In response to the comment I will add some background. <S> Generally modems, either telephone or radio, use tones of various frequencies to transmit data. <S> On the transmitting end one needs to generate tones, and the receiver needs to detect them. <S> Timing and accurate frequency are important to successful operation. <S> To generate tones of a particular frequency, one can use purely analog circuits employing precision inductors and capacitors. <S> Think of touch-tone telephones from the 1960s. <S> Alternatively one can use a digital to analog converter (DAC) and a micro-controller to generate accurate tones, generally at far less cost. <S> On the receiving end one can also use a purely analog tone detector, once again with precision inductors and capacitors, to detect tones of a preset frequency. <S> A key drawback of this is that a separate tuned circuit is required for each tone frequency to be detected. <S> The digital alternative is to use a analog to digital converter (ADC) to convert the incoming signal to digital format and pass it to a micro-controller. <S> In the processor one generally runs a fast fourier transform (FFT) to detect the tone frequencies contained in the signal. <S> A further refinement is that many micro-controllers designed for digital signal processing will have on-board ADCs and DACs. <A> It includes V.21/Bell 103, V.22, V.22bis, V.23, V.32, V.32bis modems. <A> Frequency shift keying is a pretty simple means of digital signaling that can work with either a single sideband or FM transceiver. <S> You can implement it with a few ops amps or at low data rates (75 baud?) <S> possibly even in software... <S> maybe or maybe not on an atmega, but certainly on other small chips. <S> An advantage over on/off keying is that both symbols are positively detected - you can basically say "which is louder, the high frequency or the low frequency" More primitively after a limiter you can put both on the sloping skirt of a filter to get some frequency-driven change in voltage and then threshold that voltage, ie, use an FM slope detector. <S> DTMF is commonly used to control VHF/UHF FM repeaters, and single-chip detectors and generators are available. <S> This might work on HF SSB under good conditions (and is presumably used on 10m FM) <S> but I don't know if it's allowed on HF in general. <S> Sub-audible tones have also been long used to open squelch on VHF FM land mobile and amateur service.
|
Microchip has this free dsPIC soft modem library.
|
What would be the best way to design a real time clock for the MSP430? Basically that. The way I am doing it now is with the TimerA set to 1 second interrupts. But I think that it's very annoying. Are there any other ways to do it? I want to basically set timers on that clock, like, shutdown until 40 seconds have passed... <Q> There are MSP430 devices with a low-power oscillator that use a standard 32.768 kHz watch crystal and are intended specifically for that sort of application. <S> A typical one is the MSP430F1101. <A> It's not perfectly clear to whether you want a real-time clock or a stopwatch <S> (the 40 seconds you mention). <S> You could use an RTC IC (Real-Time Clock), like the NXP PCF8563 . <S> This one is available in several packages, including both the old DIL and a very small DFN . <S> But you probably don't need a separate RTC IC. <S> It's typically used because it consumes very little power , and the rest of the circuit can power down while the RTC keeps running of a battery or supercap. <S> The MSP430, however, is also a low-power device, and it has a low-frequency oscillator which can run on the same 32.768kHz crystal you would use for the RTC. <S> In one project I had the MSP430 running continuously on a 32kHz crystal, and yet it consumed less than 5\$\mu\$A. <S> That's more than an RTC (the PCF8563 only needs 250nA), but it will be acceptable for many applications. <S> What's your problem with the 1s interrupt? <S> anyhow, whether generated internally or coming from an external RTC. <S> Upon interrupt you can perform required updates of seconds, minutes and hour counters, and wait for the next interrupt. <S> You even could work with 10ths of a second , although with the 32.768kHz this will have a minor deviation. <S> You would have four tenths of a second or 3277 clock ticks, followed by one tenth of 3278 ticks, to get exactly 1/2 second, so repeat this pattern a second time to complete one second. <A> Usually a real-time clock is done by either having a real-time clock module built in to the microcontroller, or to use an external real-time clock module such as the DS1305 from Maxim. <S> These can typically be programmed to trigger an interrupt at a pre-defined time, and run off their own 32.768KHz crystal. <S> They often have their own battery-backup circuit, and some even contain charge circuits too.
|
If you want to make a real-time clock you'll need a time-cue
|
Sensor reading is chaotic when Arduino is powered by laptop unplugged I have a Sharp proximity sensor connected to my Arduino (analog input). This Arduino is connected and powered through USB by a laptop. When the laptop is plugged in, everything works fine (I getconsistent readings). When the laptop is running on batteries, the values I'm reading are"chaotics". I mean that they vary a lot. I also have an Arduino motor shield, which is powered by an external battery, but that's just for the motors (sensors are powered by the Arduino). The GND of the battery and of the Arduino are connected though, maybe that's causing the problem? Or is too much power required from the USB? Well, do you have any idea to understand that and fix it? Additional informations: Sensors reading are "smoothened" using a low pass filter (see my previous question about it ). (if you need some more information, that will show up here) <Q> Sounds like a ground loop. <S> The laptop's DC output (-) terminal is connected to the earth pin of the plug inside the adapter. <S> The problem is usually resolved by eliminating all other ground connections. <S> Check if your Arduino is connected to any other device which is grounded, for example an oscilloscope. <A> Try adding a LARGE capacitor on the power supply rail - at least 1,000uF. <S> 10,000 uF <S> if available. <S> If that makes no difference then it's liable to be a radiated noise issue. <S> If adding the capacitor helps its liable to be a conducted noise issue. <S> Motors can be very nasty - even on their own supply. <S> If motors are driven in one direction only then reverse diodes across the motor are essentially essential. <S> If the motors are bidirectionally driven then some sort of snubbing is essentially essential. <S> (Are motors mono or bidirectionally driven?) <A> Based on the report that it works when the circuit is powered by its own battery, it really sounds like the laptop's USB VBUS supply is noisy when operating off batteries. <S> It should be possible to filter this supply using both a series inductor and a cap - this will be more effective than a capacitor alone. <S> You can even do multiple stages of such filters. <S> But it may be easier to just use a battery if that works for you.
|
Maybe the VBUS comes from a badly implemented switching regulator or DC/DC converter or one with some failed components in it, that is under more stress when operating off batteries than when off the likely higher charger voltage.
|
What is the purpose of the Arduino Uno power pins section? I have an Arduino Uno, and from what I understand, the digital I/O pins on the side of the uno each have 5V and can be controlled from the Arduino software. I am pretty sure from reading the documentation and testing it out with a 9V battery (Please correct me if I am wrong) that the VIM and ground pins on the opposite side of the uno are used for powering the uno from an external, non-USB power source. What I find a bit hard to understand from the uno's documentation, is the purpose of the 5V and 3.3V pins. I tried hooking up a hobby motor to the 5V and ground pins, and it spun - so am I correct in deducing that the 5V and 3.3V pins are used to power external components that don't need to be controlled from the Arduino software? <Q> Yes, you are correct that it can be used to power external components which use a 5V connection. <S> The 3V3 pin is the output of the on-board 3.3V regulator. <S> Same as above as for powering components from it. <S> The VIN pin is slightly more complicated. <S> If you are not powering it from USB but rather from an external power supply, that supply is directly available on VIN. <S> However, the ATmega328 is still powered from 5V which is available on the 5V pin after being passed through the regulator. <S> So the VIN pin is unregulated (unless your external supply is regulated) and should probably not be used to power external components. <S> Unfortunately, I believe all the pins on the arduinos are only rated for 40mA. <S> So while your power supply might be able to provide more, if you take it from the power pins you should not draw more than that. <A> If you have a battery pack, instead of constructing a plug you can just wire it straight to VIN and ground. <A> One use for the 5V pin is powering anything that draws more than 40mA. <S> I have a motor that draws more than that. <S> To control it safely using the Arduino, I have to use a digital pin to control a transistor, as a discrete transistor can switch much more current than the digital pins can handle. <S> The number of Vcc (aka 5V) and Gnd pins represent how much current the board can take, in increments of 200mA. <S> So the Uno, with 1 Vcc and 2 Gnd, can push out 200mA total and is capable of safely sinking 400mA into Gnd. <S> It should also be noted that the 5V (and 3.3V) is coming out of a regulator. <S> Whatever power goes in to the board gets regulated first, so those pins are guaranteed to be at their listed voltage. <S> The Vin pin has 2 uses. <S> If you supply voltage, 5V or more, to Vin and connect one of the Gnd pins, that will power the board. <S> If you are using an external power source through the DC jack, then Vin provides direct, unregulated access to that source while Vcc give a regulated 5V. <A> The Vin pin is connected directly to the DC jack on the board, which is connected to the input of the on board 5V regulator. <S> The 5V pin is the common 5V volt rail shared by the USB input and the output of the regulator. <S> One use for the Vin pin is to power your Arduino using a 9V battery.
|
The 5V pin is the output of the on-board 5V regulator. The VIN pin can also used to power the arduino.
|
LCD : what is the material inside does anyone know what is exact name(s) of liquid crystal material used in modern LCD's and how big should be electric field to make it change polarization? Update for the second bounty: My idea is that I want to make LCD at home. I can get everything except these twisted nematic crystals, and I haven't found any papers or information on how they should be used (like required electric field, preventing electrolysis, required thinkness), or under which names I can buy them. <Q> These operate by twisting between 0° and 90° thus polarizing the light as it passes through. <S> A second polarizing filter either allows or blocks the polarized light, thus making the black areas you see on the LCD. <S> A Twisted Nematic crystal typically requires a potential of around 1V to twist fully through 90°. <A> Some Web search fun turns up: 4'-Propoxy-[1,1'-Biphenyl]-4-carbonitrile (C16H15NO), which is a powder Trans-1-(bromomethyl)-4-propylcyclohexane (C10H19Br), which is a liquid <A> This site (in German) has info, and kits, for building your own LCDs: http://fluessigkristalle.com/ <A> Another common chemical to use in LCD screens is MBBA , or N-(4-Methoxybenzylidene)-p-n-butylaniline. <S> If you do experiment with this substance, please heed these warnings. <S> From this MSDS , you can see that while the material is poisonus if you ingest/inhale it (or get it in an open wound or your eyes), it just causes some skin irritation if you get it on you. <S> If you do happen to get the material inside of your body, then you should seek medical assistance (as it will begin to make methemoglobin in your blood). <S> However, if you get just a bit on your finger (externally only), wash your hands with soap and water and you should be fine.
|
The crystal substance is known as " Twisted Nematic " crystal.
|
How to avoid latches during synthesis I want to design a block of combinational logic using VHDL, but occasionally the synthesized result contains an unintentional latch. What coding guidelines do I need to follow in order to avoid the synthesizer from inferring latches? Example : in small segment of code, should I use if-else statements? <Q> To avoid latches, you need to make sure all of your outputs are assigned at all possible branches of the code. <S> for example, if a = '1' then b(0) <= '1';else b(1 downto 0) <= <S> " <S> 00";end if; would generate a latch, because in the first condition, the value of b(1) is not specified, so the compiler decided you wanted to keep the previous value of b(1) there. <S> One way to write this that would not generate a latch is: if a = '1' then <S> b <S> <= prev_b; b(0) <= '1';else b(1 downto 0) <= <S> " <S> 00";end <S> if;... <S> if rising_edge (clk) prev_b <= b;end <S> if; <S> Here you explicitly state that b should retain it's old value, and then overwrite b(0) with the new value. <S> Another way is to give b a default value, as in @TomiJ's answer. <S> If you post the code you are getting a latch on, we could help you find the specific reason. <A> None of the outputs can be dependent on any of the outputs from "last time" the process ran. <S> Otherwise you infer a latch because the next time the process is scheduled it has to keep the value of the signal which didn't get a new value last time around. <S> I prefer to keep purely combinational logic as continuous assignments, and use processes for clocked logic, then I don't get latches. <A> Four rules to avoid latches: Don't read from the signals to which you write. <S> Have a correct sensitivity list <S> (all signals that you read should be in the sensitivity list) <S> Make sure that all signals to which your write are assigned in every path. <S> (for example: in each branch of an if-else-statement) <S> For processes which use variable, make sure every variable is initialized a default value before reading it (in another variable or signal ). <S> Additionally, if you have several combinational processes, make sure you don't create a loop. <S> Several coding styles can help you stick to these rules, for example the style in @TomiJ's answer. <S> As @Martin Thompson points out, it may be better to avoid combinational logic all together. <S> Put everything in a clocked process instead. <A> As has been pointed out by @fbo and @Martin Thompson, you need to ensure every signal driven by the process is assigned some value in every branch of the process, and that value must not be dependent on the previous state of any of the outputs of the process. <S> The easiest way to ensure this is to assign some default value to each output at the very beginning of the process, for example (co-opting fbo's example): COMBO: <S> process(a)begin <S> b <= <S> (others => '0'); -- Assign default value to b <S> if a = '1' <S> then b(0 <S> ) <= '1'; else b(1 downto 0) <= <S> "00"; end if;end process COMBO;
|
If you are using processes for combinational logic (and I advise against it for just this reason) then make sure every path through the process assigns something to every signal that the process drives.
|
How do I look at all the footprints in a KiCAD library? In the KiCAD PCBnew layout editor,is there some way to see what a footprint looks like before I place it?Or am I forced to select a footprint that has a plausible-sounding (ASCII text) name,place it, and then afterwards delete it and start over ... until I find the right one? How do I look at all the footprints in a footprint librarywithout laboriously selecting one, placing it,going "That's not what I wanted", through the whole list? I would be happy with any one of a quick way generate .pdf of all the footprints in the library a quick way to generate a "dummy PCB" with every footprint in the library somehow show each footprints one by one, with at most a single arrow-key to get to the next one (like the KiCAD EESchema schematic editor, when I place a schematic symbol I can see what it looks like before I place it.) <Q> In recent versions of KiCad there is a Footprint Library Browser, allowing exactly what you asked for. <S> It can be opened from inside PcbNew via the menu option "View >> Library Browser". <A> footprints that come with KiCAD <S> In the KiCAD <S> PCBnew layout editor,when I can choose <S> the menu option "Place <S> > <S> > Module",then <S> right-click on the PCB and choose " <S> Footprint Documentation",it opens up a big "footprints.pdf" file. <S> That file has one page for each of the libraries that come with KiCAD,showing all the footprints in that library. <S> That is perfectly adequate, except it only shows the footprints that come with KiCAD,not any of the footprints that I've downloaded. <S> Is there some way to generate a similar ".pdf" file for footprints in the libraries I've downloaded (or created myself)? <S> footprints in libraries you've downloaded or created Install the <S> library(see <S> How do I install a KiCad schematic or footprint library? ). <S> window(or click on one name, then use the arrow keys). <S> That shows you one footprint at a time. <S> There may be a better way. <A> PCB-only workflows are a bit difficult. <S> If you have a schematic in the schematic editor ( eeschema ), you can use the footprint association tool ( cvpcb ) to assign footprints to components; the association tool has a preview window that shows the currently selected footprint.
|
Start upCVpcb,hit the "view selected footprint" button,then click on the footprint names in the CVpcb
|
I want to build a DC powered computer, what are the gotchas, where are the parts? Google builds its servers with a Battery in the box so that they don't need a UPS. I'm wanting to do the same, but I'm wondering what all the gotchas might be for doing this, and where I'll find the components to put it all together. I'm wanting to figure out how to eliminate the DC -> AC -> DC conversion (and all its losses) and just run the system off a DC power supply with a battery in place of the current ATX power supply. I'm thinking a standard ATX power supply to give me the 12 volts for normal operation, then a kick over onto a gel cell battery when the power goes off, eliminating the UPS. I'm looking for ideas on how to do this economically using off the shelf parts and whatever amount of soldering needed to make it work. <Q> The simplest way would be to replace the ATX PSU with a 12V DC (Automotive) PSU. <S> Then you can run the system off a 12V DC power supply / charger, and a standard car (or leisure) battery. <A> There aren't any real gotchas on the computer end - but you will end up with a computer powered by 12 (or 24 or whatever) volts DC, which you will have to provide via an outside power supply. <S> Which becomes another failure point. <S> I'd only bother with this if you have a real reason to do <S> so - like you've already got battery-backed DC infrastructure in your site or something. <S> If you just want a computer that's got battery in the box, buy a laptop. <A> There are a lot of issues. <S> You can't take off ATX power supply and put 12V DC PSU instead. <S> It's because PC utilizes a lot of different voltages. <S> Such as 3.3V, <S> 5V. Look at ATX standard. <S> Of course, you can add to 12V DC PSU another, smaller size DC-to-DC convertors to 3.3, 5 and other voltages you will need. <S> The charging voltage of battery is above 12V. <S> Usually it is in range from 13 to 15V. <S> You need to check datasheet for battery. <S> So you can't use ATX power supply to charge it. <S> Or you need to rework it to add needed voltages. <S> Think about how you will charge battery. <S> The another good idea may be take laptop parts (as they can work from 12V supply) and build on their base your custom PC. <S> Con is high price for details. <S> Maybe it is good idea not to use ATX PC, but take some monoblock PC ("thin client"). <S> They often have their own power supply. <S> Often they gives 12-19V output, because it is good value. <S> And on motherboard this voltage is converted by internal converters to voltages needed by ICs. <S> It is good idea. <S> But better is to buy ATX PSU with integrated battery. <S> They are used in industrial systems and costs big money. <S> But they worth it if you really need such thung. <S> The really good advice: please, write your situation and what you want as comprehensive as you can. <A> Since the computer will spend most of its time powered from the mains, it makes sense to optimize efficiency for this scenario in order to save money on electricity costs. <S> Powering a PC from 12V (not an ATX supply) only makes sense if the cost savings on the UPS outweigh the extra electric bill caused by inefficiencies introduced by double conversion (mains to 12V, <S> then 12V to lower voltages). <S> However, in a modern PC, most of the power is drawn from the 12V rail, so it might not be as bad. <S> Besides a laptop, there are other simple solutions: PicoPSU . <S> This is a small switching converter which generates voltages needed by a PC from a 12V source. <S> It is intended to power small, low-power mothermoards, like Micro-ITX and the like. <S> Just supply some 12V, and you're done. <S> It is expensive, though. <S> Intel NUC and similar micro-PCs <S> Many of these small PCs run from an external wall wart DC adapter. <S> Some will run on 12VDC. <S> Also, they draw very little power, so this will extend your battery life. <S> However, you also need a screen that will run from 12VDC...
|
If you just want a computer that's got battery in the box, buy a laptop
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.