26 replies to this topic

[billt79]

So, built my nice new shiny watercooling loop today and im not impressed by the temps. i have an i5 3570k @ 4.4ghz and under IBT at max stress it hits 72 degrees C, vcore is at 1.236v which is the same as my old cooler which was a noctua nh-d14. its a new loop, specs are alphacool nexxos ut60 in 280mm with some akasa fans with a high static pressure, so plenty of flow, EK dcp 4.0 pump EK nickel supremecy plexi, phobya single bay res, photos here, http://postimage.org...in2xy/ce5e1fff/ the nzxt m59 wasn't an easy case to build into i had a lot of problems with routing the hoses. Anyways any ideas about the temperatures?

 

EDIT: pump-block-radiator-res-pump

 

Cheers, Bill.

Edited by billt79, 08 March 2013 - 12:54 PM.

[red1776]

Are you sure that you have bleed it 100%? air pockets like to hide in the block(s)

[Puck]

In all honesty watercooling isn't a huge benefit to IB. The high temps are from very concentrated heat on the die - it actually outputs LESS heat total then a SB but the cores still run hotter. Even with the cold plate of my block holding 6-7c my core temps are ~65c in IBT @ 4.65ghz/1.38v. No way water gets you those temps. Ivy Bridge has a very bad IHS Temp/Core Temp ratio. (Normal gaming won't break 45c though, IBT is a terribly inaccurate tool for temperature information.)

 

They are just hot, leaky chips. Water cooling may not get you much better temps or any more of an OC over high end air.

[mattyamdfanboi]

i beg to differ.. my 3570k @ 4.5ghz @ 1.25v runs at 58 degs under water.. im running the XSPC water loop system with the twin fans but extra thickness rad..  idles at around 30.. and thats at super hot temps.. i mean.. its been mid 30s deg c in melbourne for like a week now..

[Waco]

The thermal paste instead of solder really kills any benefits for super-high clocks and volts but for more reasonable clocks there's a lot to be gained.

 

I'd bet there's an air pocket there somewhere or your block isn't mounted perfectly.  What speed are you running your fans/pump?

[wevsspot]

Also check the mounting contact between the water block and IHS.  Even if contact looks good I'll pull out my old metal engineer's ruler and run it across both checking for a concave or convex surface.  There has to be something going on there.  Also, depending on what TIM you used it may require some cure time.

[Drdeath]

 

It is hard to see if you have the correct line in on the block. The directions specifically say the port closest to the middle. Something looks different compared to the instructions. The EK logo looks diffrent position...

Edited by Drdeath, 08 March 2013 - 10:30 PM.

[billt79]

The whole block looks different, the TIM is phobya hegrease, and the instructions did'nt say anything about which port to use as inlet, maybe thats for a different block? There is plenty of pressure and it looks like its mounted fine

 

il post a better photo of the block

 

http://postimage.org/gallery/pbzm0p4/

 

edit: my block is not a supremacy, its a supreme ltx nickel csq, so it won't have the same inlet/outlet configuration as the supremacy, although good idea

 

Fans are at max and the pump is running directly off a 12v molex, so should be at max, also 2 cores are running at 58 degrees and 2 at 68 degrees, i expected some variation between cores but not that much.

Edited by billt79, 09 March 2013 - 07:52 AM.

[Puck]

The whole block looks different, the TIM is phobya hegrease, and the instructions did'nt say anything about which port to use as inlet, maybe thats for a different block? There is plenty of pressure and it looks like its mounted fine

il post a better photo of the block

http://postimage.org/gallery/pbzm0p4/

edit: my block is not a supremacy, its a supreme ltx nickel csq, so it won't have the same inlet/outlet configuration as the supremacy, although good idea

Fans are at max and the pump is running directly off a 12v molex, so should be at max, also 2 cores are running at 58 degrees and 2 at 68 degrees, i expected some variation between cores but not that much.

The two middle cores always run hotter since they have cores on both sides of them. Delidding to replace the IHS Tim helps, but its widely known that IB runs hot because of its high core density. Some chips just run hotter then others, and IBT is unrealistic for core temps.

[Drdeath]

The whole block looks different, the TIM is phobya hegrease, and the instructions did'nt say anything about which port to use as inlet, maybe thats for a different block? There is plenty of pressure and it looks like its mounted fine

 

il post a better photo of the block

 

http://postimage.org/gallery/pbzm0p4/

 

edit: my block is not a supremacy, its a supreme ltx nickel csq, so it won't have the same inlet/outlet configuration as the supremacy, although good idea

 

Fans are at max and the pump is running directly off a 12v molex, so should be at max, also 2 cores are running at 58 degrees and 2 at 68 degrees, i expected some variation between cores but not that much.

The photo is right from EK's website and all EK nickel supremecy are the same design. There is a difference what port is the inlet on almost all waterblocks and that could be the culprit.

[Hyper]

looks like the block is actually an EK-Supreme LTX, you might try switching the inlet/outlet hoses on the block to see if you get better temps.. I cant realy tell from the picture but that block actually works better with one of the restrictor plates in it, cant remember which one does the best off hand though...

[Drdeath]

looks like the block is actually an EK-Supreme LTX, you might try switching the inlet/outlet hoses on the block to see if you get better temps.. I cant realy tell from the picture but that block actually works better with one of the restrictor plates in it, cant remember which one does the best off hand though...

Thats what I was trying to get at.....