power supply problem

[matanzas]

Is it possible for the wrong type of ram in a machine to blow a power supply. I think I am reaching but I cannot figure out why it would have blown.
The machine belongs to someone I know. It is an AOpen with a 80 gig HD but that's about all I know about it at this point.
TIA

 

[dglienna]

No, but blowing a PSU *might* take out a RAM stick. Also, many other components.

-David
2006 Microsoft Most Valueable Professional (MVP)
2006 Dell Certified System Professional (CSP)

 

[matanzas]

Hi David:
I have not had much experience with power supplies and, while I wouldn't pretend to be able to fully diagnose and fix the problem, it would be nice to learn how to track down the situation that took it out. I find out apparently that the owner took it upon himself to install another ps and it worked until he shut the machine down and tried to restart it.
Can you diect me to a stepped troubleshooting process that one could follow?
Thanks

 

[edfair]

It is unlikely that a proper functioning memory stick working one time with a power supply would then take the power supply out on the next power-up cycle. But there are always qualifiers about blanket statements. If it happened it is more likely that the memory suffered "infant mortality" on the second power-up cycle or failed when the power supply glitched.
This is supposing that the memory in question is of the wrong specs and not some other type that was "a square peg forced into a round hole".
Troubleshooting at the point is pretty much the "sniff" test. Or take the M/B to minimum functionality and go for POST.

Ed Fair
Give the wrong symptoms, get the wrong solutions.

 

[matanzas]

The sniff test on the first ps revealed a definate eau du singe...

The second unit powered up the pc, ran just fine for a time, shut down normally and then wouldn't start up again. No sent of burning though. I removed this ps and put it in another pc and it started normally.

I am hesitant to try to try it again for fear of blowing another ps but must admit, I really don't know where to start looking. I am game to test anyting that I can though.

 

[satrow]

Time for some workbench bare-bones troubleshooting then

Setup the PSU and mainboard on a bench/table, add one stick of ram, gfx card, connect keyboard and display and power it up. If it starts, enter the BIOS and reset all to default.

Read faq602-2731 for more details.

 

[matanzas]

I'm back... it's been a tough day in the mines.

I thought I would take a closer look at the interior of the box and remove the dust jungle first. When I touched the fan on the cooling fins on top of the CPU, it moved... actually the fan and fins lifted right out (at least to the end of the fan's power cable). It looks like the body of the cooling fin was glued to the top of the CPU with the operative word being "was". The CPU is still fully attached to the mobo but is suspect that any cooling qualities that the fins might have had ceased when it became "unglued".

This is obviously another problem but an over heated CPU wouldn't cause the ps to fail... would it?

 

[edfair]

If the CPU is still functional, no.
How about putting the power supply back in that failed on the second try. Then try it again. If it powers up, then shutdown and try powering up again.
Attempting to see if this is possibly a power supply/motherboard combo that doesn't want to power up a second time without pulling the power cord.

Also might help to know what type PS you are dealing with. ATX type power up with a signal on the green wire on the molex connector. And the power up signal comes from a circuit on the M/B. Those can fail and I suspect that failing to take the signal to full voltage might give your problem.

This is all speculation. You troubleshoot one step at a time.

 

[matanzas]

Thanks Ed.

I don't build machines so I haven't ventured beyond altering the RAM, HDD, FDD and optical drives. I take it that the heat sink (proper term?) is to be securly attached to the CPU in order to transfer heat? If so, I suspect it's toast or in need of a good heat glue.

Both power supplies, old and new are ATX type.

I tried the second ps in another machine and it fired up without problem. I then placed it back into the suspect machine and once again and there was no activity in the computer at all.

If the CPU is fried would that prevent the computer from displaying any activity at all?

Notwithstanding the CPU, if there should be some activity, what might I look at next?

 

[edfair]

Yeah, heatsinks are usually firmly mounted, some more firmly than others. Later models mounted with pressure and some with enough pressure that they do major bowing of the board.
If the power supply works in the 2nd machine you can discount a gross failure.

Next test would normally require a VOM and a test for voltage on pin 14 and one of the black wires on the MOLEX power block. Can't recall how it goes, suggest you measure the voltage on the good machine for comparison. The power block (this is from memory so check it out) with latch down pins 1-10 left to right on top row 11 to 20 L/R on bottom and pin 14 should be green.

It is beginning to sound like there is a problem with the standby circuitry on the M/B. The power switch on the front of the case gets to the standby circuits to energise the power up circuits to whack 14 which kicks the PS in the butt.

Ed Fair
Give the wrong symptoms, get the wrong solutions.

 

[matanzas]

Did the test Ed and found power on the black wires on the MOLEX. So, if there is a problem with the standby circuitry is there a test that can be performed with the VOM?
Thanks

 

[matanzas]

I plugged the ps into the suspect machine once again and toggled the on/off swirch on the back of the ps. The fan on the CPU actually turned a few times and then quit. Can't get it to repeat the feat.
I get just under 3 volts, AC across the black and red in the Moldex (4 prong) plug and just under 9 from the 22 pin plug to the black on the 4 prong.
Am I reading that correctly?

 

[matanzas]

duh, should have been looking at the DC scale.
the above should read 9 volts, DC across the black and red in the Moldex (4 prong) plug and 27 from the 22 pin plug to the black on the 4 prong.

 

[edfair]

Do a google search for "atx power specificatons" and link to the bluemax.net site. Pictures and all.
Spec for power on is green on pin 14, 0v on black wires and standby power purple on 9. Until you have convinced me otherwise I suspect that you will have the same.
Your voltages are out of line so , since the power supplies work sometimes, it is likely you are confused about pin orientation or VOM setup.
When you initially add power, particularly after everything is drained down, there isn't any specification on how the fan should operate, so sometimes they come on and rotate a while until the standby stuff gets sorted out. This is longer than the rotation when you have a short and the power supply crowbars at somewhere around 1/4 of a rotation.

Are you talking about a 4 pin Molex plug for the M/B, a 4 pin AMP plug for the hard drive, or a 4 pin Molex plug for the floppy? For the drives red/black should be 5v yellow/black 12v.

Ed Fair
Give the wrong symptoms, get the wrong solutions.