Water cooled system overheating?!

[ManillaForce]

CPU is overheating on boot! Can't get past the BIOS, temps are 65+.
Leak tested for hours and ran good. Using distilled water. Any suggestions?

 

[[Ion]]

I'm almost certain that there isn't proper contact between the CPU block and the CPU. I'd make sure that it's mounted properly and that the thermal paste is spreading properly.

 

[cadaveca]

What is posted above. BUt...

CPU is overheating on boot! Can't get past the BIOS, temps are 65+.
Leak tested for hours and ran good. Using distilled water. Any suggestions?

Looks like loop goes from GPU, then CPU?

Honestly, I think that your loop is too complex for that single pump. I'd re-route it, res-pump-CPU-RAD-GPU-RAD, or CPU-GPU-RAD-RAD.

 

[ManillaForce]

I'm almost certain that there isn't proper contact between the CPU block and the CPU. I'd make sure that it's mounted properly and that the thermal paste is spreading properly.

That was my first thought, shall give it a try.

 

[ManillaForce]

What is posted above. BUt...



Looks like loop goes from GPU, then CPU?

Honestly, I think that your loop is too complex for that single pump. I'd re-route it, res-pump-CPU-RAD-GPU-RAD, or CPU-GPU-RAD-RAD.

Loop goes CPU-GPU-PUMP-RAD-RES-RAD

 

[cadaveca]

Loop goes CPU-GPU-PUMP-RAD-RES-RAD

Normally res goes before pump?

 

[drdeathx]

Normally res goes before pump?

+1 loop is not configured correctly.

 

[ManillaForce]

Normally res goes before pump?

Mmm ok. I'll remount and reapply my CPU first to see if temps drops and if not I'll look into redoing the loop

 

[MT Alex]

That's a DDC style pump with an after market top, so it should have enough head for the two rads and blocks. I'm assuming since you did all the leak testing that you can see movement in your res, so water is flowing. The res before the pump is a great rule of thumb, but with the location of your pump it's not going to be starved of water. I'm betting things will be just fine after a remount, but honestly, things have to be really jacked up mounting wise for the cpu to get that hot so quickly, the problem should be visible to the naked eye. I don't think too much TIM would screw things up that badly, your block would have to be cockeyed, or the sticker would still have to be on the bottom of the block. You have fans on the top of your upper rad?

 

[ManillaForce]

That's a DDC style pump with an after market top, so it should have enough head for the two rads and blocks. I'm assuming since you did all the leak testing that you can see movement in your res, so water is flowing. The res before the pump is a great rule of thumb, but with the location of your pump it's not going to be starved of water. I'm betting things will be just fine after a remount, but honestly, things have to be really jacked up mounting wise for the cpu to get that hot so quickly, the problem should be visible to the naked eye. I don't think too much TIM would screw things up that badly, your block would have to be cockeyed, or the sticker would still have to be on the bottom of the block. You have fans on the top of your upper rad?

I took the sticker off before hand, and yes I have 2 fan for the top 240 rad.

 

[the54thvoid]

I think you've got air in that loop.

It's a very poor logical flow system you have there. You fill the reservoir which drops water into the pump via radiator. But... the water wont simply fill the loop up to reservoir height - it'll get caught by air pockets in the radiator. And when you switch the pump on it'll just suck air which will end up being circulated around that reservoir (especially if it's a bay res).

Unless you spent hours trying to fill it to the brim, I think air in loop. Ideally your pump should be fed cleanly by the fill point (reservoir or fill port). Otherwise air is going to be a problem.

So if you reseat your cpu block and still have the problem, change your loop orientation.

 

[MT Alex]

It's a closed loop, and once you put the cap on the reservoir it is pressurized. I imagine both the intake and exits of the res are at the bottom, so air won't magically get sucked into the loop. Regardless, even if the pump was below the res, most loops have the res in or around that position in the case .Sure, it's not ideal, but there isn't any hydraulic flaw with the loop.

 

[drdeathx]

Take the side of the case off and post a photo.

 

[sneekypeet]

Take the side of the case off and post a photo.

There is an attachment in the OP

@ OP...Do you hear and gurgling when the loop is running?

 

[Finners]

What rad is that in the front? seems to fit well.

Is the pump defiantly coming on? and the quickest way I can see that would put the loop into a better order would be to swap the hose coming off the bottom of the GPU with the hose going from the RES to the big Radiator at the front

 

[cadaveca]

There is an attachment in the OP

@ OP...Do you hear and gurgling when the loop is running?

Read loop order.


Pump-rad-res. He's pumping water into the res only, the rest of the loop, the 240 rad up top, the CPU, and the GPU, are running off of pump vacuum and gravity. I don't see how it is even working at all.

IF he flipped the pump around...OK...

 

[sneekypeet]

Res will still build pressure against the trapped air wouldn't it?
In no way would this be the way I ran a loop, but if the res is showing movement I'd go with that. IMHO the draw of the pump would collapse the tube if the flow was insufficient.

 

[cadaveca]

Res will still build pressure against the trapped air wouldn't it?
In no way would this be the way I ran a loop, but if the res is showing movement, the draw of the pump would collapse the tube if the flow was insufficient.

Yeah, but how much pressure? enough to suck minimal water into the res. I bet the pump isn't even delivering half it's flow.

Like, it's so odd...I dunno wtf is going on there, but high temps, to me, are unsurprising.

 

[MT Alex]

Res will still build pressure against the trapped air wouldn't it?
In no way would this be the way I ran a loop, but if the res is showing movement I'd go with that. IMHO the draw of the pump would collapse the tube if the flow was insufficient.

Yup. I agree.

Remember: It's a pressurized loop. Pump placement doesn't mean doodley squat.

 

[cadaveca]

Yup. I agree.

Not if the water is sitting there idle keeping the tubing rigid. There's no way with that restriction it's pulling much water.

pump placement matter for flow rate, no?

 

[MT Alex]

Not if the water is sitting there idle keeping the tubing rigid.

Check my edit.

 

[cadaveca]

Sneaking this one in before your new edit

 

[sneekypeet]

I still say if the tube from GPU to pump isn't collapsing he has mounting issues or trapped air. Even with minimal flow he should be able to boot without overheating it. I could see under stress, but I have booted air coolers with the plastic still on the base and they boot to windows at 50some degrees.

 

[cadaveca]

I wonder how much coolant is in the loop...

 

[Finners]

excuse the crude edit but its late over here, this would be the easiest fix for your loop order imo