System won't boot after GPU Waterblock install

[infamous_last]

OK guys this is frustrating me so I thought I'd throw it out to the forum for some other things to check.

After many many years with PCs I decided to build my first custom water loop PC.
Main parts list consists of

• 2x Gigabyte G1 gaming 980ti - EKWB WF3 GPU block + back plate.
• Rampage Gene III + x5670 (Xeon / LGA1366)
• EVGA 1000W PSU

Currently I've got everything up and running but the PC won't post with PCIE power attached, but will post with no external power (with the normal no power connected warning), or I've also found only either 1 only plugged in .
Stripped down one of the cards and nothing changes, tried either with no backplate no difference, each card individually also.

No noticeable damage to either card, everything worked on air prior.

Hoping I've missed something simple...

[Bwithnewcast]

if you used a air cooler then the fan was connected to the cpu_fan pin on your motherboard and if you dont have a fan connected to the port or changed the bios/uefi setting then you will have problems

[SmashinMachine]

Your MB does not support this XEON CPU. 

Your GPU Cards cannot connect with the lanes from CPU

It can cause this problem.

[infamous_last]

1 minute ago, SmashinMachine said:

Your MB does not support this XEON CPU. 

Your GPU Cards cannot connect with the lanes from CPU

It can cause this problem.

It is supported with firmware update and has already run with all this hardware on air.
I have a second PC with single 780 running the same RAM / CPU / MOBO

[infamous_last]

8 minutes ago, Bwithnewcast said:

if you used a air cooler then the fan was connected to the cpu_fan pin on your motherboard and if you dont have a fan connected to the port or changed the bios/uefi setting then you will have problems

had already accounted for that unfortunately. 

[Bwithnewcast]

Have your pc ever booted?

[infamous_last]

Yes, used it for 6+ months with a cpu air cooler and its factory gpu cooler on a test bench (e.g. cardboard box) before starting the custom loop build.
 

[ApolloX75]

7 minutes ago, infamous_last said:

It is supported with firmware update and has already run with all this hardware on air.
I have a second PC with single 780 running the same RAM / CPU / MOBO

Have you tried swapping that 780 into the system to test? And when you tried the cards on their own they did or didn't work at all? That wasn't very clear in your original post, just want to clarify it. 

 

Seems odd that the simple addition of water blocks would have this effect on a healthy system. What else (if anything) did you change?

[infamous_last]

Just now, ApolloX75 said:

Have you tried swapping that 780 into the system to test? And when you tried the cards on their own they did or didn't work at all? That wasn't very clear in your original post, just want to clarify it. 

 

Seems odd that the simple addition of water blocks would have this effect on a healthy system. What else (if anything) did you change?

Its a PITA to swap the 780 in as it has AIO 280mm rad squeezed into node 304... do have a radeon 5450 in my HTPC which will probably be easier to move for testing... will give that a go.

The case was new, so everything has been pulled apart a few times since it was working on the test bench to get working, so not just a simple pull and replace.

I've tried both cards by themselves on either x16 slot with no change in behaviour so far.

[ApolloX75]

2 minutes ago, infamous_last said:

Its a PITA to swap the 780 in as it has AIO 280mm rad squeezed into node 304... do have a radeon 5450 in my HTPC which will probably be easier to move for testing... will give that a go.

The case was new, so everything has been pulled apart a few times since it was working on the test bench to get working, so not just a simple pull and replace.

I've tried both cards by themselves on either x16 slot with no change in behaviour so far.

Ahh okay. Definitely try that 5450 just to be sure nothing happened to the motherboard. If it's in a case now, maybe something is touching on the underside and causing the issue. It's rare but it definitely does happen, so check all the usual problem areas like stand offs and edges of the board for anything at all that could cause a short.

 

IF all else fails, try removing the system from the case and running it on the bench again. If that solves it you know something is wrong with the case. I've had cases with slightly twisted frames before that caused this kind of problem.

 

I'm dealing with a similar issue myself right now with an old Athlon 64 X2 system. It's definitely a head scratcher.