Best way to measure power consumption on each rail?

**parksgm** · 11-29-2006, 02:28 PM

Hi all,

After yet another thread debating the failure rate of the M2 vs. the likelihood of user error, I've grown weary of debating theoretical power consumption specs. So, I'd like to be able to directly measure power consumption on each individual rail of my computer while it is operating.

What is the best way to do this?

It is possible to cut and combine the multiple 12+ leads from a standard ATX power supply together, place an ammeter in series in the circuit, and then split the leads to reconnect with in original ATX harness locations? Would I fry the power supply doing this? If so, why?

**MatrixPC** · 11-29-2006, 04:44 PM

In theory, it should work by joining all 12V rail together and put an meter in series then split like you said. It may also not work because of the PSU design.
To make sure, double check for continuity between the same rail. For example, if 2 12V wires coming out of the PSU and they have continuity between them, the 2 wires are just split inside the PSU. In that case, you can just use 1 wire or join the 2 together.
If the 2 wires don't have continuity, they are probably coming out of 2 separate regulator inside the PSU. If you join that 2 wires together, you change of frying that PSU is very high.
The same case/senario also apply to other voltage rail as well.

**parksgm** · 11-29-2006, 08:16 PM

Originally Posted by MatrixPC

If the 2 wires don't have continuity, they are probably coming out of 2 separate regulator inside the PSU. If you join that 2 wires together, you change of frying that PSU is very high.
The same case/senario also apply to other voltage rail as well.

Well, that's probably why it didn't work on the first PSU I attempted to do it with. Good thing it was a cheapo spare. I'll remember to check for continuity before attempting it again.

If the various 12V+ feeds are coming from seperate regulators and I can't use a single meter to measure them all at once, I doubt that measuring them individually and summing them would be accurate either.

**sama** · 11-29-2006, 08:21 PM

I made a supply that used an M2, where all the 5v outs connected to one thick 5v wire, and all the 12v outs connected to one thick 12v wire.

I spoke to the guy that made the m2's and he said that would be ok, which in my mind, says the different rails have seperate regulators, but all 12v should have the same regulator, and all the 5v have the same regulator (and the 3.3 for that matter, which I used to power a fan for the PSU which is in a box

)

**Wiredwrx** · 11-29-2006, 09:22 PM

Originally Posted by parksgm

Well, that's probably why it didn't work on the first PSU I attempted to do it with. Good thing it was a cheapo spare. I'll remember to check for continuity before attempting it again.

If the various 12V+ feeds are coming from seperate regulators and I can't use a single meter to measure them all at once, I doubt that measuring them individually and summing them would be accurate either.

You should be able to sum them, though if you can combine and test each rail at once, that would be easier.

Michael

**billmee** · 11-30-2006, 08:30 PM

If you want to spend some bucks get a DC amp probe aka clamp on meter.

You can clamp it around one or all of the 12 v wires and read the current flowing thru.

Older ones only measured AC amps.
Newer ones have a hall effects sensor that is snesitive down to milliamps.

Sears sells a few different models.

**parksgm** · 12-01-2006, 12:39 AM

Originally Posted by billmee

You can clamp it around one or all of the 12 v wires and read the current flowing thru.Older ones only measured AC amps. Newer ones have a hall effects sensor that is snesitive down to milliamps.

Huh. I wasn't aware they now had DC capability. I'll investigate...thanks for the tip. :-)