Sign in with
Sign up | Sign in
Your question
Solved

System no longer booting. Thinking PSU issues

Last response: in Systems
Share
May 24, 2011 11:31:16 PM

Whats going on now?
When I press the power button, the debug code on my motherboard nearly immediately goes to FF (boot), the fans spin up, and then the power drops out (usually within a second or two of turning it on). No beep code.

Current System:
EVGA 750i FTW motherboard, 2x Crucial Ballistix 2GB 1066 MHz RAM, Intel Quad Core 2.67 GHz, ASUS 480 GTX, Rosewill 950W power supply. 3 hard drives, and a DVD burner are also installed.

Problem History:
I previously had a Gigabyte 3D Mercury case (huge monstrosity with integrated liquid cooling) and the pump has been dieing for a year now. Last week I tried turning it on and it would near immediately power off. Figuring it finally died, I got a new case (This case has a built in circuit preventing it from being run without the liquid cooling.) and a new video card (I modified my EVGA 8800 GTX for liquid cooling.).

I got all the components in, installed everything in the new case. It booted but the video didn't display. With a lot of fiddling (very finicky seating), I got it to display and updated my drivers. Reset the computer and it was running fine for the rest of the night. I left it on when I went to bed at midnight, and it was still on when I checked at 3 am. I woke up at 7 am and it was no longer on. I tried the power and that's where I'm at now.

What Have I Done?
I have replaced the video card and have demonstrated that it worked (My old card worked as well, but I couldn't keep it running without coolant.) prior to last night. I haven't tried reseating any components since this latest issue (focusing on the power supply now). I have tried multiple outlets, with and without surge protectors (The electricians are rewiring my building - just wanted to make sure it wasn't something stupid.).

I have pulled out the power supply and checked each pin on the cables. My 12 volt line is reading ~12.43 volts; my 5 volt line is reading ~5.10 volts; and my 3.3 volt line is reading ~3.6 volts. The 12 and 5 volt lines are within the tolerance of the ATX spec, but the 3.6 volt falls outside the +/- 5%. I'm not sure if that is still acceptable or not or whether it would cause this problem. The power supply is a Rosewill 950W (not a great brand) and is over 3 years old. I do not have an PSU I can drop into the system to test.

----

If I need to buy a new power supply, so be it. If that will fix the problem, I'll do it and get it over with. But before I get one, I would like some assurance that this is indeed causing the problem. I have already spent much more than I would like on this and don't want to drop more cash on something that won't fix it.

Any help is appreciated. I've been working on this for about a week now and it's driving me insane.
a b B Homebuilt system
May 24, 2011 11:40:35 PM

Well to be honest there are no assurances... you can buy a tester or bring it to a repair shop... I would just buy a new PSU... how old is the old one?
May 24, 2011 11:46:21 PM

I realize I can't be assured that it is indeed the problem - I'm more looking for something more than it being my hunch.

There is a local computer store nearby - I'll try to check with them to see if their prices/turnarounds are reasonable.

I got my old PSU back in September 2007.
Related resources

Best solution

May 25, 2011 12:25:02 AM
Share

This is a different picture of your problem than the one you asked an hour earlier in another section http://www.tomshardware.com/forum/305606-28-rail.

I have done a little searching and found this http://www.evga.com/forumsarchive/printable.asp?m=70666... which seems similar to your problem.
May 25, 2011 12:36:03 AM

Yes it is. I initially the 3.6v was my concern. But after I thought about it more, I began to think that in the case it wasn't a big problem, a more thorough history would be necessary. Thus I created this thread in a more appropriate category in the forums.

Going through your link now and I'll post what I come up with.
a b B Homebuilt system
May 25, 2011 12:44:11 AM

Measuring power supply voltages or using a power supply tester is a waste of time in determining if the power supply is good or not, substitution is the only reliable test. If the fault is not the power supply then it would most likely be a faulty motherboard I know which one I would like to change first.
May 25, 2011 12:52:04 AM

pjmelect said:
Measuring power supply voltages or using a power supply tester is a waste of time in determining if the power supply is good or not, substitution is the only reliable test. If the fault is not the power supply then it would most likely be a faulty motherboard I know which one I would like to change first.


Indeed. It is a lot easier to swap a power supply but the OP said he did not have a spare. The next logical course of action would be to test the power supply.
May 25, 2011 1:03:43 AM

Yeah, unfortunately I don't have another. I could get access to some old, cheapo power supplies, but I highly doubt they have the connections I'd need to be able to test with.
May 25, 2011 1:13:10 AM

mbecker813 said:
Yeah, unfortunately I don't have another. I could get access to some old, cheapo power supplies, but I highly doubt they have the connections I'd need to be able to test with.


It would be more beneficial to test your power supply in another computer rather than to try to put a lesser power supply in the one your trying to diagnose.
a b B Homebuilt system
May 25, 2011 1:16:06 AM

As you don’t have access to a spare power supply, have you tried the paper clip test? Disconnect the power supply and connect a paper clip between pin 4 (green) and ground of the 20 or 24 pin connector. Does the power supply fan stay on.
May 25, 2011 1:20:21 AM

I didn't use a paper clip but yes. I shorted the PSU ON signal in the 20+4 connector and the PSU did stay on. I then used the DMM to check the voltages on the other pins (8-pin, 2x 8-pin PCIe, 2x 6-pin, 20+4 pin, 4-pin, and a smattering of 4-pin auxiliary power connectors)
May 25, 2011 3:26:44 AM

Alright - the computer is now up and running, so I'll say the problem is tentatively solved. That is, it's working the same way it did the night before it stopped working. I'll have to see if it makes it to morning this time.

I checked out timinator's link that suggested the CPU/CPU fan wasn't seated correctly. I took another look at it, and the CPU fan was able to shift. I took it out, reseated it all and tightened the screws. It then got to the point where it would boot again, but the graphics wouldn't load. This was the same situation I had earlier and I had to futz with it for about 10 minutes before I got it into it's happy space (I have no idea why this card is so finicky while my 8800 GTX wasn't.). It's currently running, but both side panels and a few PCI slot covers are off. If it's still running when I wake up, I'll cautiously try putting those back on and hope I don't mess something up >.<

Oh, and for reference's sake, the now that the computer is running, the 3.3 volt line is now at 3.34 volts, well within the tolerance.
May 25, 2011 3:27:05 AM

Best answer selected by mbecker813.
Anonymous
a b B Homebuilt system
May 25, 2011 3:56:07 AM

I had a similar issue, and it was one of my sticks of RAM. maybe try pulling one and booting with just one, then try the other individually.
May 25, 2011 4:55:10 AM

mbecker813 said:
Alright - the computer is now up and running, so I'll say the problem is tentatively solved. That is, it's working the same way it did the night before it stopped working. I'll have to see if it makes it to morning this time.


Well, at least it's working. These situations can sometimes be a pita to resolve. I'll keep my eye on this thread if you need any more help/advice.
May 25, 2011 10:33:57 AM

...and it's back down. Same problem as before.
a b B Homebuilt system
May 25, 2011 12:46:49 PM

As old as it is, check the mobo for blown or leaking capacitors.
May 25, 2011 1:05:00 PM

Onus said:
As old as it is, check the mobo for blown or leaking capacitors.


Is that something that would let it work for several hours before it shuts down? That seems like something that would prevent it from turning on at all.
May 26, 2011 12:39:27 AM

Throwing in the towel. I brought it to a local computer store to let it be there problem. They'll run tests and (hopefully) find out what's wrong for $35.

I'll post when I get a response with they come up with.
May 26, 2011 7:00:57 PM

Problem was with the motherboard. 2 CPU pins were slightly and were causing the problems. Looking into a new motherboard now.
!