PciE slot saying "empty"

[xxNAPxx]

*Everybody, I'm having recently a big issue with my GPU/Mobo.*

During some maintenance on my system, I had the GPU connected to the mobo through a vertical GPU holder (my case was a mini itx Raven Rvz01). Accidentally while System was On i've broken 2 rear pins of the holder. The pc after that froze. Initially I thought, well, it's just the holder, so i bought a new case for my rig to finally use my gpu connected directly into the PCI-E slot (a bought a NZXT Manta). But connecting the GPU to it ended up in a blank screen, initially I was like, ok the gpu is dead, so what I did was to connect the the PC through the iGPU, but after I starded thinking it wasn't gpu fault but mobo, in fact after managing to boot the pc up through it I realised, also the network card and the Bluetooth stopped working, plus in the board explorer of the bios, at the PCI-E slot it was saying empty, like it wasn't detecting the gpu, so I thought the accident with the gpu holder caused a short that killed the PCI-E slot and network card. So I RMAed the board and got a new "repaired" one, but same problem again, GPU is not detected on the PCI-E slot saying it is empty. With the new board the network card is working fine now, the Bluetooth instead is working odd (difficulties in connecting and findings devices), so I was like, ok also the GPU is dead so let me check the post debug leds, and I noticed that the leds are not detecting any problem, they are supposed to stay lit when a fault is detected but debug just works fine as expected, either I boot up through the iGPU and the GPU not detected. So I thought even if is statistically difficult to happen, that I received another faulty motherboard. In your opinion what is the problem? GPU or mobo? Why it doesn't detect at all the GPU, with the board explorer saying it is empty?*

My rig is
I7 7700k*
Avexir DDR4 RAM 3600 xmp cl17
Samsung 960pro nvme
GTX 1080 g1 gaming by gigabyte*
MSI Z270 pro Carbon AC mini itx motherboard*
Corsair sfx600 pSU

 

[EarthDog]

Try the GPU in another PC and see if it is recognized there?

 

[xxNAPxx]

This would narrow things down certainly, Unfortunately is a test I cannot do

 

[Evilsizer]

if it froze then there is a chance for shorting out or a voltage/current surge that either blew out the PWM section or a surge to the GPU/RAM that is now causing the issue. if you have a friend with a pc, see if it works there, if not sounds like a call for warrenty replacement if they can/will.

 

[xxNAPxx]

So you think the short has caused the gpu to fail?
RAM is absolutely fine, the issue is just about the GPU, what I don't understand, if there is a fault, why the mobo does not detect the fault on post

 

[EarthDog]

It may be because there is a shorten the GPU itself. The only way to test is to try that GPU in another PC. Any pcrepair shops around? They can do it easily. You've got to narrow it down.

 

[trents]

By "vertical GPU holder" I think you mean what we usually call a riser card. Riser cards adapt motherboard slots so as to change the angle.

It's either a partially fried video card or a bad replacement motherboard or some of both. Are you certain the replacement motherboard is just not the same one you sent back to the manufacturer? There are occasions where motherboard manufacturers will not test the one you sent them thoroughly enough to encounter the problem the customer was having and just send it back.

Take the video card to a PC shop and have them test it in another PC. That shouldn't cost you very much.

I'm not clear on if you get video when using just the IGP, when the video card is removed. Some of your sentences kind of run together and I found your post hard to follow. Is English your first language?

 

[xxNAPxx]

Unfortunately the components are connected in a watercooling loop, to send the card to someone, I should take the loop apart and reconnect the airblock. Quite annoying, also don't have time

 

[xxNAPxx]

Initially I thought The replacement mobo was mine, but after a little test it looked was different.
Sorry for the confusion created. Basically even if I unplug or plug the gpu the pc boots up only through the IGPU, and in both cases the EZ debug leds do not recognize any fault

 

[trents]

Okay, looks like you're using your cell phone and Tapatalk to communicate so that explains some things. Sorry, no insult intended about the question of English being your first language.

Unfortunately, I see no other way to troubleshoot your dilemma other than swapping out parts and testing them in a known properly working system, starting with the video card. Do you still have the original cooler and fan for the video card?

 

[xxNAPxx]

Sorry not being clear, English is not my first language, moreover it was difficult to explain all the steps I followed to get a conclusion.
Yes for gpu vertical holder I meant a Riser card (I didn't know the exact name in English for that XD).
Unfortunately I can't do many test, by googling the problem I stumbled across many topics with no conclusion. I'm awaiting a new mobo I purchased already because I am positive the issue is the mobo and not the GPU, at worst the GPU would be dead, I hope not because is the most expensive to replace. What let me think is not, is that the mobo does not detect a fault during post. I also found on other topics that the PSU could be the culprit too, but I would give it a 5% of chances

 

[wingman99]

Did you try the graphics card in a different slot?

 

[xxNAPxx]

The mobo is an itx, so only one slot I'm afraid

 

[xxNAPxx]

Another thing that let me thought about tge PCI-E slot to be faulty and not the gpu is that, even windows does not detect any PCI-E slot (i used hw info)

 

[Johan45]

I feel it's wishful thinking on your part. If the card shorted out then why would the board detect a card? I don't think it's that simple. Maybe the board will fix your issues but I wouldn't hold my breath

 

[xxNAPxx]

Because it runs, if I unplug the 6pin, a little light on the graphic card deyects there is no power, plus the waterblock gets warm, so that means or should mean, it is on and functional powerwise

 

[EarthDog]

Does that mean it works... or that it's just getting power? There is a difference.

My money is on the card if another mobo didnt help.

 

[Zerileous]

I just want to summarize what I think has happened so far, to make sure we're not missing something. You broke your riser card and the PC stopped working. You were able to boot the PC using integrated graphics, but the network card and bluetooth were not working. You RMA'd the first motherboard to replace it with a second motherboard, which you are certain is not the same motherboard you sent them. The replacement works (using integrated graphics), except for the network card? Neither motherboard will boot with the GPU installed.

You have now ordered a third motherboard, to see if that will work with the GPU installed.

I wouldn't worry about the PSU in this case, if the system works without a GPU installed. The SF 600W only as one 12V rail, so it would be unlikely that the CPU would function if the PSU failed in such a way the be incapable of powering a GPU.

 

[Evilsizer]

on a different note, one thing GPU manufactures dont do is build in some kind of surge suppression or over/under voltage or over/under current protection. i think they dont do this because its a function that is built in to the PSU. this instance would not be something they could conceive of happening for them to warrant such voltage or current protection or surge suppression, ect. my only suggestion is to look at the fingers on the card that plug into the slot, see if any of them look slightly burned, maybe take a look at the pwm section if you can see it. i do not see it being anything then the card died.

your story reminds me of what i did long ago with my Abit ST6 Tually celeron 1.2ghz overclocking it to 1.6ghz. i was using a laptop harddrive with a 4 molex to some kind of small pin array, i barely tapped the molex to move it. then it went black, turns out the drive fried, lucky i had more drives on hand.

it is a hard lesson to learn, but if you think of moving things in your pc, power it off first.

*edit*
when i mentioned ram in my last post, i was referring to the GPU's ram. both operate off the pwm section on the gpu, so it could have been either of them got fried or both did at the same time or just the pwm section got fried.

 

[xxNAPxx]

I think you're right guys, it's just me too wishful not believing is the gpu to be dead, cuz it's the most expensive piece to buy brabd new, plus the waterblock and backplate, I've got a brabd new mobo coming up and than I will try the last test. I wished so far was mobo the culprit, since it doesn't detect any fault on post, although is very unlikely it can't be the gpu fried, by the way there aren't any signs of burnt on the GPU itself. What I do not understand is why the mobo doesn't detect a fault during post for dead gpu, sounds pretty odd to me