Odd temperature issues.

[scoizzle]

About to give up on liquid cooling, patience wearing thin... I put together my first custom loop, very simple... currently just trying to cool the cpu.

 

While testing the loop I am getting unreasonably high temperatures and can't seem to figure out why, I've re-seated the cpu block multiple times, tried changing pump speed, made sure there were no air bubbles as best as I can.

 

Specs that apply:

i7 4790k @ Stock

ASUS z97-WS

Corsair Dominator Platinum DDR3 2400 CL11 (2x4GB)

EVGA Supernova 1200 P2

 

Any suggestions, comments, and tips will be greatly appreciated.

 

This is under load from Prime95 26.6 Blended test.

 

Current picture of the build:

 

[AlwaysFSX]

Did you remove the sticker? Hows your thermal paste spread?

[Cameron_Messed]

what are the specs for the pump?
how long has the system been running for? could there still be bubbles in the loop?

[scoizzle]

Did you remove the sticker? Hows your thermal paste spread?

Yes I removed the sticker, the thermal paste (Arctic Silver 5) spread well and I don't think I've been applying too much, used about the size of an uncooked grain of rice in the middle of the IHS.

[scoizzle]

what are the specs for the pump?

how long has the system been running for? could there still be bubbles in the loop?

 

It's a standard D5 pump, system has been running for a week being turned off during the night.

 

http://www.ekwb.com/shop/reservoirs-and-acc/res-pump-combo/ek-d5/ek-dbay-d5-vario-incl-pump.html

[AlwaysFSX]

Yes I removed the sticker, the thermal paste (Arctic Silver 5) spread well and I don't think I've been applying too much, used about the size of an uncooked grain of rice in the middle of the IHS.

Look at it..

[Cameron_Messed]

It's a standard D5 pump, system has been running for a week being turned off during the night.

 

http://www.ekwb.com/shop/reservoirs-and-acc/res-pump-combo/ek-d5/ek-dbay-d5-vario-incl-pump.html

are you getting an RPM reading from the pump? could put it on full speed and see if its actually reading 4800rpm

[scoizzle]

are you getting an RPM reading from the pump? could put it on full speed and see if its actually reading 4800rpm

 

Confirmed, it is reading 4838 RPM max.

 

Look at it..

 

Was trying to avoid having to remove and re-apply the thermal compound again... but I will do this soon.

[Cameron_Messed]

what is the internal temp of the case when you were using prime 95? could it be that your graphics cards were heating up the air that it meant to be cooling down your cpu?

[airdeano]

About to give up on liquid cooling, patience wearing thin... I put together my first custom loop, very simple... currently just trying to cool the cpu.

 

 

what is your loop order?

 

res/pump>rad>CPU>res/pump?

res/pump>CPU>rad>res/pump?

 

as per the instructions, your top fitting of the CPU block should be the outlet

and the middle is the inlet.

 

 

if correct, you prolly have a large air pocket in the CPU block or radiator, assuming your fans are 100% speed.

[scoizzle]

what is the internal temp of the case when you were using prime 95? could it be that your graphics cards were heating up the air that it meant to be cooling down your cpu?

 

Internal case temp was ~ 38C. Could be the graphics cards... will re-test with them out of the system soon.

[scoizzle]

what is your loop order?

 

res/pump>rad>CPU>res/pump?

res/pump>CPU>rad>res/pump?

 

as per the instructions, your top fitting of the CPU block should be the outlet

and the middle is the inlet.

 

ek evo route.jpg

 

if correct, you prolly have a large air pocket in the CPU block or radiator, assuming your fans are 100% speed.

 

Top fitting is the outlet, Res/Pump > cpu > rad > Res/Pump

[airdeano]

then you have a air pocket lodged in the radiator (rad outlet will not pass large

air downwards). tilt the case backwards to allow the air to escape from radiator

and purge to the res. make sure the res does not run dry while tipped.

 

when the rad doesn't have enough coolant, high temps are expected.

air is a lousy thermal conductor.

[runit3]

-snip-

Good ol' P95 on Haswell...Is your voltage on manual? What is your voltage?

[scoizzle]

Good ol' P95 on Haswell...Is your voltage on manual? What is your voltage?

 

1.218V

[runit3]

1.218V

Assuming that it's on manual and the voltage isn't spiking up to something obscene (1.5v+) I would do this:

With nothing running unmount the radiator but leave everything else in place.

Rotate it so the fittings are above the bottom of the rad (lengthwise) and hold the rad below everything else, having the case on a desk edge would be good for this.

See if you can get any air to come up into the res or purge out into the rest of the loop. You may have to hold the long lengths of tube up above the rad with the other hand.

Run it for a couple minutes in this position (just boot to desktop) while monitoring temps then shut it down. do not run P95 w/fans disconnected

 

Fasten everything back up and see if anything has improved. Hopefully you'll see a drop in the water level of the res indicating some air got out of the rad.

 

Nice lookin build btw. Are those 280x windforces I see?

[scoizzle]

Assuming that it's on manual and the voltage isn't spiking up to something obscene (1.5v+) I would do this:

With nothing running unmount the radiator but leave everything else in place.

Rotate it so the fittings are above the bottom of the rad (lengthwise) and hold the rad below everything else, having the case on a desk edge would be good for this.

See if you can get any air to come up into the res or purge out into the rest of the loop. You may have to hold the long lengths of tube up above the rad with the other hand.

Run it for a couple minutes in this position (just boot to desktop) while monitoring temps then shut it down. do not run P95 w/fans disconnected

 

Fasten everything back up and see if anything has improved. Hopefully you'll see a drop in the water level of the res indicating some air got out of the rad.

 

Nice lookin build btw. Are those 280x windforces I see?

 

I've been working with the air a lot, thought about that whole taking the rad out idea, will probably do that tomorrow as I have some replacement fans coming in (Noctua iPPC 2000 x6 ;D). And I just set manual voltage to 1.150V, with stock clocks and I'm getting ~ 63C. I'll post some glam pics once I get the cooling figured out, and the cards are Windforce Gtx 770 4GB cards... going to replace them both with a single Titan X one of these days.

[runit3]

I've been working with the air a lot, thought about that whole taking the rad out idea, will probably do that tomorrow as I have some replacement fans coming in (Noctua iPPC 2000 x6 ;D). And I just set manual voltage to 1.150V, with stock clocks and I'm getting ~ 63C. I'll post some glam pics once I get the cooling figured out, and the cards are Windforce Gtx 770 4GB cards... going to replace them both with a single Titan X one of these days.

<3 Windforce

If memory serves me correctly I don't believe the Supremacy MX has jet plate adjustment, which is my only other idea aside from a giant air bubble in the rad. I had to do the same thing with my 360x60mm, I'm sure you can imagine how fun that adventure was <_<

[airdeano]

can the bay res be removed from the ODD bay? meaning towards the inside of

the case? if so, remove bay res, lay case on its back, allow the bay res to be

highest point (air travels up), keep res full of coolant. cycle pump 30sec on,

10sec off and vary the pump speed (if vario no faster than step 3).

[scoizzle]

I was reading on the arctic silver website that Arctic Silver 5 has a 200 hour break in period, I'm going to go ahead and give this compound some time. Have managed to get quite a bit of air out of the loop, but I still don't think it's quite optimal yet.

[airdeano]

break-in if cured usually only yields 2-4° and quicker transitions. not like it'll be

a 10° drop under loaded situations.

[scoizzle]

break-in if cured usually only yields 2-4° and quicker transitions. not like it'll be

a 10° drop under loaded situations.

 

I guess that was just wishful thinking then. Might have solved the issue now, maybe not. Currently have it running @ 4.6, 1.281V (1.312 according to AI Suite 3) under Prime95 26.6 Small FFT max temp is 80C, seems to be holding ~ 78C.

[airdeano]

have you tried using Intel XTU? you might find as loaded, but not as heated over P95.

[Wirerat]

70c p95 blend? Run a loop of x264 and report temps. There may not be an issue here at all.

P95 is hot on thesechips.

[scoizzle]

Issue is resolved it seems, stock load temps went down to max 51C, thank god.