New build help, Rampage V Edition 10 + 6900K CPU, Stock speed overheat with WC

[equifox]

Alright, so I usually try to fix stuff by myself before asking for help but I gotta admit I don't know what to do at this point.

This is my current build: https://pcpartpicker.com/user/equifox17/saved/VMcxrH

Running a custom watercooling loop which is basically RES -> PUMP -> RAD 1 -> GPU -> RAD 2 -> CPU -> RES

I didn't touch the BIOS at all, everything is stock, even my 3200 RAM is running at 2166.

So when I boot up Win 10 and check temps, GPU idle temp is 24-25 degrees and CPU idle temp at around 33-35.

Now when I put load on the GPU at 100% i will read max 35-38 degrees, but the CPU in realbenchmark will sky high to 80-95 degrees in no time as soon as it get to video encoding. It will begin to throttle. I tried to reseat my waterblock but Evo Supremacy are kind of noob proof on that side. Also replied thermal paste different times, made sure there was no air in my loop as well (GPU temp would be affected anywaya)

I'm out of option, I tried pushing real hard on my waterblock while stressing the CPU to see if it makes any difference, nope. Might it be a bad luck and should RMA the CPU ? Cause I'm getting real tired of this and this computer is used for some casual gaming but mainly for programming and in its current state I can't use it for work.

Thanks for your help !

 

[EarthDog]

What clocks and voltage is it shooting up to on load? Have you updated the BIOS on that massively overkill motherboard you have? ALl the air bubbles out of the block (tilt the case around, etc)?

 

[trents]

I would check the waterblock contact on the CPU and the TIM spread pattern. It sounds like it something may be wrong there.

Also, what are your CPU voltages like? Auto settings will often severely over volt things. Have you actually checked them with say HWMonitor?

 

[Witchdoctor]

What are you using to measure temps ?
If the software is accurate make sure it is set for Celsius
I agree, the EK blocks are next to impossible to screw up , but as trents suggests check anyway and try a different TIM to rule that out

Like Joe said grab latest and greatest BIOS

This may sound ridiculous, make sure the protective plastic has been removed from the block

 

[equifox]

What clocks and voltage is it shooting up to on load? Have you updated the BIOS on that massively overkill motherboard you have? ALl the air bubbles out of the block (tilt the case around, etc)?

I've attached a GIF image showing a few seconds of the load where we see voltage and temp rise up. As for my MB yeah I agree it is overkill, I chose it mostly because it was mainly black (Stupid yeah but hey )
Anyway I'm running the latest BIOS version 1502. Bubbles are out I moved the case around like crazy.

I would check the waterblock contact on the CPU and the TIM spread pattern. It sounds like it something may be wrong there.

Also, what are your CPU voltages like? Auto settings will often severely over volt things. Have you actually checked them with say HWMonitor?

Pretty sure the contact is good, the way these EK blocks are mounted is really noob proof and I screwed it down using a star pattern. I've replaced the thermal paste a few times already, tried blob, line and X.
I've attached the HWMonitor log.

What are you using to measure temps ?
If the software is accurate make sure it is set for Celsius
I agree, the EK blocks are next to impossible to screw up , but as trents suggests check anyway and try a different TIM to rule that out

Like Joe said grab latest and greatest BIOS

This may sound ridiculous, make sure the protective plastic has been removed from the block

I'm using HWMonitor in celcius yeah. Room temp is at 21 degrees too.

BIOS is 1502 which is the latest according to ASUS website and yeah the plastic is off haha I made sure of it.

Image showing HWMonitor while running Realbenchmark:

https://imgur.com/VDIU3Aw

- - - Updated - - -

Correct fullscreen link:

https://imgur.com/VDIU3Aw

 

[EarthDog]

I bet that voltage is wrong...I assume your voltage is on auto???

Please post up a screenshot of CPUz showing what it says the voltage is under load. No need for a fancy GIF, just capture what CPUz says is the load voltage. Please attach it to the forums instead of 3rd party (the images will go away after time).

Honestly, you need to start it all over. Leave the strap on auto and adjust the voltage manually.

 

[equifox]

Sorry about the GIF, I've attached the output of CPU-Z when under load.



Start it all over ? Even for stock speed ? I mean this is out of the box, I would have expected more from both CPU and MB at this point :S

 

[EarthDog]

Interesting... CPUz shows the same voltage... hmm. For giggles, try setting the voltage MANUALLY in the BIOS to 1.19 and see if that puts temps back in order.

 

[equifox]

Just rebooted and switched from auto to manual, set to 1.19. Idle temps are a bit higher, load temps are the same.

 

[EarthDog]

So it isn't false voltage reporting.. hmm..

I would imagine its something in the loop/contact. It shouldn't be doing that with a so much radiator...

 

[trents]

By some chance you don't have the inlet and outlet ports on the CPU water block switched do you?

Hows about attaching a pic of the interior of the case so we can see the layout?

 

[equifox]

By some chance you don't have the inlet and outlet ports on the CPU water block switched do you?

Hows about attaching a pic of the interior of the case so we can see the layout?

I double checked and nope IN and OUT are correctly used.

This is the only picture I have for now, it's not the best one but you get the idea.

 

[trents]

What does monitoring software say is the rpm of the pump under load?

And I'm not sure it would make much difference but it seems to me a more logical order would be pump>CPU>first radiator>GPU>second radiator>reservoir. Like I say, not sure it would make much difference since the whole system kind of equalizes itself because it is in a loop. And the way you did it may have made the most sense from a routing of the tubing standpoint.

 

[trents]

Sorry about the GIF, I've attached the output of CPU-Z when under load.

View attachment 187925

Start it all over ? Even for stock speed ? I mean this is out of the box, I would have expected more from both CPU and MB at this point :S

vcore looks about what it should be.

Which way are the radiator fans blowing?

 

[equifox]

Pump is around 4500-4800 RPM at load.

As for the fans, top is PUSH and front is PULL.

Thanks for your time

 

[Witchdoctor]

EK CPU blocks have a jet plate inside to disperse the flow accordingly

They typically come from the factory with the 1151 jet plate installed, you have to change that one to the 2011 plate

Make sure you put it in the proper orientation,

This is the only thing I can think of as your GPU is where is should be ruling out something that would affect the entire loop

 

[EarthDog]

Pump is around 4500-4800 RPM at load.

As for the fans, top is PUSH and front is PULL.

Thanks for your time

so.... top is exhaust and front is intake? Push/pull doesn't tell us that.

The plates, worth a try, but, doubting that is it.. the difference was a couple of C.

 

[Witchdoctor]

so.... top is exhaust and front is intake? Push/pull doesn't tell us that.

The plates, worth a try, but, doubting that is it.. the difference was a couple of C.

I doubt it to, but with the GPU temps in line it really is odd ???

 

[EarthDog]

I can't wrap my head around it... honestly... its got to be in the CPU or the mount. He's tried a couple of methods and mounts too. I wonder what the spread actually looked like.......

 

[Witchdoctor]

me neither ? I agree, has to be the mount or flat out a defective block ??? or the reporting software is just wrong