Will -11V on the -12V pin prevent startup?

[tgrundman]

I'm a noob when it comes to electricity. No class had ever taught it to me throughout high school and college. I know the basics (self taught), and this isn't the basics (to me).

All readings except for the +3.3V (which are spot on) are coming up slightly short. The +5V are coming up as ~+4.9V, which I presume are ok since it's not too far off. However, the -12V is coming up around -11V, too low (high?) I think. Also, pin 10 (20 pin connector) is coming up as +12.5V instead of +12V.

Summary (20 pin connector):

• 
  [*]+5V reading as +4.9V
  [*]+3.3V reading as +3.31V
  [*]Pin 10 (+12V): reading as +12.5V
  [*]Pin 12 (-12V): reading as -11.1V

300W PSU.

Thanks!

Edit 2: There is no load. PSU is by itself, PWR_ON is grounded so that it starts up. Used multimeter for readings.
Edit 3: The mobo does not POST, just silence

 

[Doctor Micro]

There may be several reasons why the PC won't boot or post, however, to answer your question...

Normally, a reading of .1 or .2VDC low or high on any rail shouldn't make much difference, other than to cause a bit more current to be drawn. 11.1VDC on the 12V rail does concern me, though... especially since you're reading it under zero load. If it's already 11.1VDC with no load, it's likely even lower with a load attached.

If this machine were on my bench and I read those voltages, I'd plug in a known-good PSU and see if it booted. Naturally, you'll want to do a visual check for blown caps and make sure all connections and components are tight. I'd also temporarily remove or disconnect anything that's not necessary for the system to boot (extra hard drives, optical drives, modems, NIC, floppy, front and rear-panel USB/Firewire headers, etc.) If it boots, plug everything else back in and check again. If it still boots okay, the PSU was probably your problem. If not, unplug everything again and plug them in one at a time. What you're doing there is testing the possibility that you may have a short on one or more of the components.

 

[angry_geek]

I've seen this with power surges and lightning strikes. As Doc Micro said, try another psu that you know is good.

 

[studiot]

You should never test a switching power supply off load as this can destroy them. Some have protection, most will refuse to turn on so I'm suprised you got any readings at all.

Certainly none of your off load readings are meaningful.

The handy psu testers are designed to provide the minimum required loads to indicate correct operation and are very cheap.

Substitution is, of course, a good test - unless there is an overload fault on the mainboard that has caused psu failure. A known good psu may be damaged by this, although this is a rare occurrence - I have only seen it happen twice. Usually a bad mainboard will simply drag down the supplies without damage.

If you are going to practice hardware servicing you need to acquire a reasonable working knowledge of electrics, if only for safety reasons.

Standard tolerances are plus or minus 10%, high spec tolerances +/- 5%. But remember your voltmeter may also be inaccurate so go a bit tighter than these on measurements.

 

[Doctor Micro]

You should never test a switching power supply off load as this can destroy them. Some have protection, most will refuse to turn on so I'm suprised you got any readings at all.

Certainly none of your off load readings are meaningful.

The handy psu testers are designed to provide the minimum required loads to indicate correct operation and are very cheap.

Substitution is, of course, a good test - unless there is an overload fault on the mainboard that has caused psu failure. A known good psu may be damaged by this, although this is a rare occurrence - I have only seen it happen twice. Usually a bad mainboard will simply drag down the supplies without damage.

If you are going to practice hardware servicing you need to acquire a reasonable working knowledge of electrics, if only for safety reasons.

+1 Good advice and good catch. I forgot to include a recommendation to buy a PSU tester.

 

[tgrundman]

You should get a PSU tester

I have a PSU tester. My PSU tester said it was bad.
Then I tested the PSU with my multimeter, it said it was good.
Confused by the different readings, I went to the computer store to get another power supply anyway. Better safe than sorry. The guy said "Let me test it to be sure" and used the PSU tester there. He led me to the back room and showed me the PSU tester. It registered good. He wasn't trying to swindle me but rather on the contrary, he was dissuading me from purchasing anything.

I believe my PSU tester is faulty.

You should never test a switching power supply off load as this can destroy them. Some have protection, most will refuse to turn on so I'm suprised you got any readings at all.

I tested the PSU with my multimeter by grounding the PWR_ON rail because it wouldn't turn on otherwise. I read plenty of places that said that it was a perfectly acceptable method. Perhaps this is false information.

So it's two against one that says the PSU is good, excluding the fact that I maybe ruined the PSU after grounding the PWR_ON rail (even though it still worked after I did my testing).


I believe I found the ultimate cause of failure (maybe). The power cable itself. I used the multimeter to test the AC wall outlet, it registered 120V. Great. I plugged the cable in and measured the female end, it registered anywhere from 2V to 18V. If I jiggled the cable, the volts would change. I think that the cord was sitting just perfectly to allow X number of volts through to the machine (barely enough to boot, probably under 120V) and the underpowering of the machine eventually caused motherboard failure. To clarify, the machine used to power up but not POST, I messed with the machine (hence the power cable) trying to fix it, and now it doesn't power up at all. I can't seem to get the the cable to sit perfectly anymore. The cable was not hot anywhere, no obvious shorting. Once I swapped in a known good cable, the machine powered up again, but no POSTing occured.

Has anyone every experienced this before? The cable was just one of these: http://www.johnraul.com/images/computer-power-cord.jpg I tested the computer with another known good cable and it measured 120V at the female end, contrary to the bad one. The bad cord would jump around, averaging 12V-15V (I only tested the cable with the multimeter after the machine was failing to power up). The wall outlet was good, the power strip was also good.

Thanks for the input!

 

[studiot]

You are learning the art of 'signal tracing'

Well done for checking the power cable, many forget to do this, and in the UK you also have to check the fuse.

I once had an American client over here and the 'repair' comprised changing a blown fuse to a laptop power supply adapter.

 

[Doctor Micro]

You are learning the art of 'signal tracing'

Well done for checking the power cable, many forget to do this, and in the UK you also have to check the fuse.

I once had an American client over here and the 'repair' comprised changing a blown fuse to a laptop power supply adapter.

Ah yes... the old "220v not same as 110v" trick...