Help: Bad Motherboard?

[WiKiD]

Just doing a sanity check before i rip out this mobo and rebuild tomorrow.

Last week my PC locked up, red and purple vertical bars on the screen. When i rebooted the GPU fans were at full speed and i had no video.

Powercyling a couple of times got it back to running like normal. However, last night the issue became worse, i could not run the PC for 10 min without crashing. When it does boot as normal, vibrations or movement of the card will cause artifacting and an eventual crash.

Today i noticed after a crash, the card is scaled back almost 60% on core and memory clock speeds, ramping them back to normal and stress testing would cause a crash.

Tried two different PCI-e slots and had the same behavior, tried two different GPU's in both slots and also had the same issue. Also re ran new power cables for all components and tried different ports/rails in the PSU without luck

While in my last case the GPU was in the middle PCI-e slot and in the new case it is in the top slot. I have had to make several 600+ mile drives with the PC in my car in both configurations.

At this point i can only conclude that shaking and vibrations from the road trips have put a lot of wear and tear on the PCI-e contact pins and the port itself.

On boot i am getting:
4F
62
A2
99

4F i cannot find an explanation for
62 is Memory Issue: reseated ram, cleared CMOS, error persists. Also tried different DDR3 memory, cleared CMOS, error code stayed
A2 is SATA/IDE. Only clears when disconnecting both my OS SSD and my HDD
99 is PCI-e issue: Removed all PCI cards and cleared CMOS. Error still shows on boot.



PC:
i5-4690k
Asrock Z97 Extreme 4 mobo
Gigabyte Radeon 7970 (used MSI Radeon 6870 for testing as well)
8GB Corsair DDR3
Corsair RM1000x PSU.

 

[VirtualLarry]

Before you junk the board, I would try a known-good 500W+ PSU. (You don't need 1000W unless you are going to CF, even then you could probably get by with 800W.)

Btw, are you overclocked at all?

 

[WiKiD]

Thank you for the reply. I will have to see if i can scrounge up a PSU from a friend to test.

I am currently not overclocked so the RM1000x is overkill. Picked it up on BF for a great price so figured why not.


Update: Decided i should pay a bit more attention to the Dr Debug post codes.

On boot i am getting:
4F
62
A2
99

4F i cannot find an explanation for
62 is Memory Issue: reseated ram, cleared CMOS, error persists. Also tried different DDR3 memory, cleared CMOS, error code stayed
A2 is SATA/IDE. Only clears when disconnecting both my OS SSD and my HDD
99 is PCI-e issue: Removed all PCI cards and cleared CMOS. Error still shows on boot.

 

[Arcanedeath]

Based on the symptoms I'd say that it's probably the power supply, but if you can rule that out, then it's most likely the Mother board

 

[tortillasoup]

OP, open up the computer and look at the electrolytic capacitors. If any are slightly bulging/swollen, leaking, either from the top or bottom, that is your issue. If you find the computer has difficulty booting up unless you 'warm it up' before fulling booting, then it's likely the board.


One way to test is to turn on the computer from a cold boot, press the pause button and wait 10 minutes, reduce cpu fan speed so that it warms up the computer quite a bit (disable case fans!). Hit the enter key to resume boot, you may need to do a hard reboot once, but if the computer boots normally, it's likely the board, possibly the capacitors.



This logic also could be applied to the power supply, so one thing you could do is go into the bios and see if the voltage rails are low or out of spec.

 

[WiKiD]

Hey everyone. Thank you so much for all the tips.

After trying a different power supply and brand new motherboard i have limited the problem down to both of the graphics cards i tested with being damaged. Co worker will be bringing me a different card to test on wednesday.

It appears that due to the long road trips, the PCI connections on both (!) cards were worn through the metal down to the PCB. The card that had the worse of the bite marks was the 7970 which was failing most often.

The 6870 had similar marks but was only worn to the PCB on a few rows. Similar experience though. I was able to get the card to boot reliably every time by applying a slight amount of inward and upwards pressure to the right side of the GPU.

Visual inspection of the MOBO looks good, the bios error codes on boot are a bit worrisome still though. I was able to run the PC for several hours without issue on onboard graphics.