There is NO software that will monitor the PSU directly. There are hardware monitors and VOMs to test power supplies. The DC supply voltages (3.3, 5.0 and 12.0) acceptable to motherboards and processors usually have a +/-5% tolerance. Outside of this range could cause problems so the meter you use must be able to read at least down to 1/10 (.1) volts.
I measured the voltages while idle and they all were good + or - 10%. When I get time I'm going to measure while under load. Thanks for the info Dwarf the pin outs were spot on. I will let you know what I find under load.
I take it you used a digital multimeter for this. It will be interesting to see what the actual voltage is when TOverclocker is reporting a low value as you reported above.
Yes and yes it will. Sorry I did not tell you I have a digital volt meter. I hope the low voltage shows up. It might be too fast for my meter to see it, but I'm going to give it a shot. My PSU has a 5 year warranty and I really would like to verify it.
As has been mentioned above, if your 12V line was indeed falling as low as 3.6V then your system would in all likelihood just abruptly turn off. It doesn't though, which is why I think that there isn't anything wrong with the PSU, rather the on-board sensor(s) is/are faulty.
Don't wait until you get the low voltage reading to connect your meter - you can run the tests with it connected.
I did a two hour load test this morning with IntelBurnTest and FurMark running at the same time. The multimeter stayed between 12.17-12.19 volts the whole time. TOverclocker had a Min of 12.044 and a Max of 12.148. This test was inconclusive and I plan on doing a longer test when I have more time.
Those results look good and stable.
I agree that longer tests are sometimes needed to get problems to manifest themselves (hence the advice you'll often see about allowing Memtest to run overnight).
However, 2hrs is a reasonable period of time for this, and I would be quite happy with the result you got. Should you wish to test further, then you will need to put more load on the 12V line to check if it still remains stable. The two programs you ran are excellent for this, just ensure that you are using the extreme settings in both cases. In addition, if you have a DVD/CD/BD drive, play an audio CD at the same time. Don't worry if the audio is choppy and broken up, you are merely using the drive motor to put a bit more load on the 12V line.
Make sure that you keep an eye on the temperatures, and be prepared to stop the programs if the temperature on either climbs too high.
I ran IntelBurnTest using max memory and FurMark in a window 1024 x 768 (so I could see the other programs). I could make the window a bit larger and add AA, but the thing is with the case door off my video card (HD 6570) is getting pretty warm (hit 88 c). With the door on I have a fan blowing right on the video card. I will run a video on my TV at the sametime next time. Man that should be quite a load! :)
I ran a much longer test the other day and saw no drop in voltage on the voltmeter, but TOverclocker did show a drop at some point and I did not see the drop on the voltmeter. If the 12 volts is dropping it is dropping faster then I can see it or faster then the digital voltmeter can pick it up or I took my eyes off the voltmeter just long enough not to see it on the voltmeter. Also I played Far Cry 2 for several hours without a restart or shutdown. Then yesterday I was on-line trying to get a item ready for sale on ebay when all the sudden the computer shutoff. I have an UPS on the system, so it was not commercial power. At the time I was not doing anything, I was almost ready to post it on ebay when the computer just shut off without warning. In event viewer all is said was: 'The previous system shutdown was unexpected.' Seems it could be getting worst, what ever the problem is. Any ideas on how to isolate what the exact issue is? Can memory timings also cause this issue?
Quote