Won't boot, screeching, magic fixes

[NightrainSrt4]

E6300
Asus P5N-SLI
8800GTS 320
2Gb OCZ DDR2-800

That out of the way, the issue:

PC will not boot. Push power, absolutely nothing. No fans spin up, nothing. The light of the motherboard is still on however.

If you pull the 6-pin out of the 8800, then try and push power, a LOUD screech comes from the motherboard, and persists until flicking the power switch on psu. This is not coming from the PSU, the screech is coming from the motherboard. Not beeps, or tones, but a loud screech. Mind you, everything comes on and spins up when this occurs.

If you take the motherboard, gpu, etc, out of the case, reseat the sli select card then put the motherboard back into the case and hook everything up again, it boots perfectly fine. After a few days of not being used, the issue occurs exactly the same once more.

This is completely repeatable, as I have done this three times.

Another thing, there are no issues when the select card is set to sli, and 2 7600GS are in place.

Any ideas?

Edit: This screeching seems to be an 8800 power warning thing when the plug isn't in. I still don't understand why everything works fine after being taken out and put back in. I'm questioning if something is touching the case, but can't find anything. Also, everything works completely solid when its reseated. Hours of TF2 does nothing to phase it.

[Luke122]

I suspect the power supply.. have you got a replacement you can try?

[nevermind1534]

I was thinking either the PSU, the mobo, or both.

[NightrainSrt4]

My issue with the power supply is that its a 650W BFG, that was just powering the same 8800, an overclocked q6600, watercooling, 3 times the fans, lighting, 3 more hard drives and another optical, perfectly fine and never hiccupped once.

Another thing I just noticed is that the mobo light works, but when its not going to power on, the nic lights aren't on. After I reseat it all, the nic lights aren't on until a full boot is completed, then they stay on as usual when the rig is off until the not wanting to power up starts again.

I'm really starting to question the mobo, as I remember when I first got it, and tried just using one 7600 with the card in single mode it was having booting issues. Swap it over to dual mode with another and it worked fine.

With mobo's with selector cards, can you run a single card when its under dual. I know it will be limited to x8, but I didn't know if there were any other issues. That would give me a signal if it was something with the pci lanes or something.

[nevermind1534]

I have a bad (I think) motherboard that was acting very similar from the time I started using it, but it got progressively worse. It was an open box biostar.

[NightrainSrt4]

Hmmm, I really hope that its not the psu or motherboard, then again I'm just hoping for no issues, when there obviously is one.

It's just I take real good care of my crap and yet it's always mine that has issues. I know people that treat their stuff like poo and yet it lives forever.

I think Asus has a 3 year warranty on mobo's, and its been 2 and a half I think. Perhaps I should send them an email. First I think I am going to try and run the card in x8, as that way if the issue goes away then it would be the mobo, no?

I don't know, and I honestly don't want to deal with it, but when do we ever?

[FuzzyPlushroom]

Every time I've seen that problem and the board or PSU wasn't completely cactus, killing the power to the PSU for a minute or so has worked. I have no idea why, mind.

[SXRguyinMA]

the rig I picked up for project monster was doing that...press power, the cpu fan would spin for a sec, then nothing but the mobo light. tried a different psu, different ram, removed everything, no go. suspect bad mobo. now I've got a spare celeron d to put into a new mobo once I find one :D

[NightrainSrt4]

Every time I've seen that problem and the board or PSU wasn't completely cactus, killing the power to the PSU for a minute or so has worked. I have no idea why, mind.

Hmmm... That could make sense, perhaps its not working after I get everything reseated, but simply because the PSU has been unplugged for a while.

I just was checking with the breakers in the apartment, and it seems that everything on that wall and the other side in the bedroom runs on the same breaker. Its never enough to need to reset the breaker, but occasionally I do because it starts making a buzzing sound if too much is on at one time.

I just realized perhaps its something in the PSU that after it turns off after a gaming session with that rig and an overclocked q6600/GTX280 rig. Undervolt protection or something? They are running on different outlets, but perhaps being on the same breaker with half the other stuff in the apartment is too much and its trying to protect itself?

I'ma try and get a thick gauge extension cord and run it over from one of the unused outlets on another breaker and see if it does it then.

Edit: The pc has been working fine the last couple days just internet browsing. It only seems to happen the day after both rigs have been cranking TF2 for a while...?...

[SXRguyinMA]

maybe an overheating issue then? :think:

[FuzzyPlushroom]

maybe an overheating issue then? :think:

What bothers me is that (from the impression I'm getting) it's completely stable once it gets running. So no matter what, once it turns on and boots into Windows, it's all set?

[NightrainSrt4]

Yup, perfectly stable. Can sit and run TF2 for hours on end and no stability issues.

Tonight I am going to try and test my theory. I'll have the girlfriend play TF2 for a while, being the only main thing on that breaker, then turn off and see if it can turn back on. If it works, then have both rigs on and playing and the tv and see if it will turn back on after a session of that. If it doesn't after that, then I am going to grab a thick gauge extension and run it over for her rig and see if the issue persists while its on a different breaker.

It is still is working fine right now when its the only rig on. Turns on and off fine.

[Luke122]

You might be able to just add a UPS to stabilize it if it is a power issue.. then there are no more cords to trip over. :)

[SXRguyinMA]

:stupid: after my 500GB drive died not too long ago, I heard from a guyt hat the heads were probably bad (causing the screeching noises I was hearing) it spins up fine now, but you can hear a small chirp every so often and it wont read :think: I'm saving up now so I can get teh data recovered and get a new drive

[NightrainSrt4]

Well it doesn't seem to be a too much on one breaker issue. Was fine overnight then she went on firefox and instant reboots. Put it on the other breaker and seemed okay, until TF2 loaded up, then blue screened.

Running memtest now, and it was running fine, but just pulled up an error on Test 4, address 000192f4488 - 402.2MB. Ughhhh. It's not pulling up errors everywhere like my old dead Corsair, which could explain why its a bit random.

But, no power to absolutely anything except the led caused by, so far, one bad memory address? That doesn't seem very likely to me...

[NightrainSrt4]

Okay, about to eat dinner, but now mmore issues.

It's not popping up errors but it keeps hanging. escape to restart works but it just stops going through the test. Over and over again. I'ma try and reseat the heaksink and and redo the thermal paste.

Overheating maybe? It's not the greatest case, and the added heat of the 8800 over the 7600's?

[NightrainSrt4]

Not to have too many posts, but I know everyone hates it with my walls of text with several edits... Anyway.

Not to count my chickens too soon, but I've gone in and done some wire management, reseated the heatsink and put on some better silver compound. Also added a 120mm fan in the 3 unused 5.25" drive bays, and an 80mm in the side panel right where the 8800 is. I've got another 80mm for the cpu side panel, but can't find my 3pin to molex adapters, and the 3pins on the board are full.

BIOS stated a cpu temp ~15C lower, and motherboard temp ~20C lower. Its running memtest again with no errors so far, and no lockups.

I also realized that the damn rig was put right in front of the baseboard heater. It was never an issue when it was set up as the heater was never on. DOHHH!!! Moved it over past where the heater ends so that shouldn't be affecting it anymore.

Hopefully this fixes things...

Speedfan was registering a temp ~60C idle before when I checked in between crashes. Seems e6300 should never be above ~61C according to intel. Will check once memtest is done with a pass.

[NightrainSrt4]

Ok. After I was done with memtest I tried booting windows. Nope, nada, nothing. It kept hanging at the splash screen after the little bar screen. Every time. Couldn't get into safe mode either. Summed it up to issues with the constant rebooting and crashing.

Instead of doing a repair install I just went with a clean install. Drivers installed and everything.

Temperatures after an hour of TF2 was 42-43C both mobo and cpu. This is in a 27/28C room mind you, so not too bad now. No rebooting, no issues I can see now. Will post if this changes.

Note to everyone:
Remember to move computer away from heater once winter hits! The damn thing was literally pressed up against it.

Now if I could only get my damn P4 524 from running 60C in the Lanbox. Ughh. Issues galore.