I was reading this article and it got me thinking: http://www.silentpcreview.com/modul...ns&file=index&req=viewarticle&artid=19&page=1 I'm searching the net for info on how I could measure power use on the three main voltages from a PSU (3.3, 5 and 12V). Since most PSUs connect all wires to the same output, in theory it should be relatively easy to get an ammeter (sp? measures amps) in line with each output. That could go directly to an LCD or moving coil meter, or to the analogue inputs on a PIC. A simple conversion factor will give the output in watts, although since most devices and PSUs give their current use in amps, rather than overall use in watts. Anyone got any thoughts on this? Some schematics would be nice Main thing is to do it in a way that does not mess up the voltages, and above all is safe (150W through an ammeter sounds like fun ) MoJo
His way is to measure the current on the psu mains input, and taking the odd volt away from 110 or 230 across the measurement resistor (1R) makes no difference to the voltages finally produced. Start introducing extra resistance into the final regulated supplies and I think you'll also introduce noise and poor regulation. Keep it clean.
Okay, been doing some more research and I have found two methods that seem to be ideal. Firstly, shunts. They give you a small voltage output that is proportional to the current flowing through them. The output current and voltage drop are in the range of tens of mV, for example RS carry some that will do up to 300A(!) at 75mV (0.075V). An acceptable drop, although you can get even lower ones. A 30A shunt which would be enough for the 5V line on most PSUs is about £8. The second method, and one which looks most promising, is hall effect sensors. Basically, their output is a function of the input voltage and the strength of the magnetic field around them. They can be used to measure current that way, and there does not seem to be any associated voltage drop! More research needed... Zap? MoJo
Using a shunt on the ammeters is the simplest. The downside is you will need to extend the power wires out to the front pannel where the meters are then back to the mobo. This will produce more voltage drop. I haven't used a hall efffect sensor yet. The theory would be to get ferrite rod and wrap as many turns of the power wires around it as possable and attach the hall sensor to the end of the rod.
Sounds a little crazy just to measure voltage lines for an LCD. A PIC comes with multiple ADC channels and is easier to drive an LCD than other circuits. The ADC can only handle 0-5V, so use zener diodes to cut out the unnecesary voltage. Use a 3.3 zener on the 5V line, and get a signal around 1.7V and add the 3.3 volts back in before displaying it on the LCD. Same goes for the 12V line, excpet use a 9V Zener. Just remember though: The Zener connects to the line first, then run the current-limiting resistor to ground. Thsi keeps a constant voltage drop across the Zener and leaves the negative side 'floating' for you to measure your voltage with the ADC. Its alot cheaper too, since you can use $3 PIC with a few Zeners that total less than $1.
@Hazer I think he wants to measure current, the voltage will remain fairly constant as it is a voltage rail Rod
Okay, thanks for the comments guys... This seems to be more of a mechanical problem than an electrical one. Using shunts is easy enough, and hall effect sensors are only a little bit more work. The main problem is jamming the things inside the PSU, at a point where all the current for a particular power line goes though. 30A shunts are quite big... Once the shunt or hall effect sensor is in place, it's a simple matter or attaching an op-amp to boot the signal (if necessary), and then feeding it to a PIC with A/D converters. I'll start experiments with a spare PSU, see what I can do. My main concern is that with a shunt the voltage drop might be a problem. However, if it could be put between the output and flyback of the switch mode psu circuit, it might automatically be compensated for anyway. MoJo
All the PSUs I have seen have inductors in series with the 12v, 5v, and 3.3v lines, so it might be as easy as putting a sensor on each inductor.
Can anyone find a good site with info on hall effect sensors being used in this way? Google isn't turning up much at the moment MoJo
