Testing PCI-E Slot Functionality

[antaresix]

Asus P8Z68-V
i5 2500k
Asus GTX 560 Ti
Corsair HX750

Is there any way to test the functionality of PCI-E slots that are suspected to be bad, other than trying to plug in different GPUs to see if they work?

Also, is it at all common for both PCI-E slots to go bad on a board when only one has ever been used? Do they have some shared pathway on the board that could explain this?

 

[alexoiu]

No.
Apart from a damaged slot, a CPU socket bent pin, a faulty CPU, a short or broken traces.
If the board is under warranty, try RMA-ing it.

 

[antaresix]

The board is under warranty, and I called to RMA it and the Asus tech said that since I'm getting a BOOT_DEVICE_LED with my discrete card plugged in and no video, yet when I run straight off of the integrated graphics with the card removed I can boot successfully, he believes that I have an IRQ conflict. He said that he is concerned that if I RMA the mobo that they will test it and it will run fine for them.

I have never done any modifications of interrupt requests, so I'm kind of lost here.

 

[alexoiu]

Can you test the system with a different power supply, or test the card in a different PC?

 

[antaresix]

I actually RMAd my power supply because someone convinced me that was the issue, and even with the new PSU I'm getting the same exact issue. Before that I had RMAd my GPU and got it sent back to me and was told it was fine. Just yesterday I went out and purchased a brand new GPU to try to rule it out for certain, and I'm getting the same issue.

I was advised to re-install Windows to see if that would resolve any IRQ conflicts on its own, so I guess that's what I'm going to try before I RMA the board.

After the tech initially told me to me try to resolve the IRQ conflicts (I have a PCI-E controller and a SATA controller both on IRQ 19) I scoured Google and I can't seem to find any solid information on doing this. It seems beyond the scope of my experience and seems like a very uncommon issue.

 

[alexoiu]

Since the board does not POST with only memory and graphics card (no other PCI or USB devices), it can't be an IRQ conflict.
And it used to work, right?
Try clearing the CMOS (unplug the power cord and remove the CMOS battery for few minutes). Leave single RAM stick installed, graphics card, DVD drive, keyboard and mouse.
If still the same, try removing the CPU and check for bent pins.
If no bent pins, I would RMA the board.

 

[antaresix]

Yep, this same build worked since I built it about two years ago. I've tried clearing the CMOS about 3 times now over the last few months but I can try it again. I'll check for bent pins as well and post back.

 

[antaresix]

Since the board does not POST with only memory and graphics card (no other PCI or USB devices), it can't be an IRQ conflict.
And it used to work, right?
Try clearing the CMOS (unplug the power cord and remove the CMOS battery for few minutes). Leave single RAM stick installed, graphics card, DVD drive, keyboard and mouse.
If still the same, try removing the CPU and check for bent pins.
If no bent pins, I would RMA the board.

No bent pins (I didn't expect any, this system has not even been moved since it was built 2 years ago except for the recent handling it's endured from trying to troubleshoot this issue. CMOS reset also produced no different result.

The tech said that it was an IRQ conflict between the SATA and PCI-E controller as they're both on the same IRQ. But I can't even get a splash screen with just RAM and the GPU, but it boots up and runs just fine from the motherboard. At this point I can't see how it's anything other than a PCI-E slot issue. I'm just going to RMA the board and I'm not going to reformat and re-install, it seems like it would be a waste of my time. Thank you for your help alexiou.

 

[alexoiu]

You're welcome. Please post back the result.