H100i Tubes are Room Temperature Under Load?

Guinibee

Distinguished
Nov 28, 2010
237
0
18,690
Hello all,

For a while I've been wondering if my h100i is working properly. I'm getting temps in the mid 70s during x264 testing with my 4790K OC'd to 4.5ghz at 1.22v. That seems alright, but everything I read said both tube lines should be approximately the same temperature in order to know the h100i is working and well, they both are the same temp, but they're not really hot at all. I heard they're supposed to be somewhat hot to touch and these are basically room temp. The tubes are vibrating so I don't know if that's fluid moving or just the system vibration.

I'm wondering if maybe something isn't working right? I'm not very proficient with liquid cooling, so I'm not too sure what to look for to know if my cooler isn't working right, but I was hoping someone might have some information on this. If I put my hand on the exhaust fans during a stress test they appear to give out some slightly warmer heat, but the tubes themselves don't feel hot. Corsair Link also says that my pump is moving at 2160rpm and it always appears to move at 2160rpm.

The header is plugged into the mobo and the sata is also plugged in. As far as I can tell it's working, but I've seen people get much better numbers with this cooler and honestly all the tests outside of x264 seem to spike the heat too high. Prime95 (the newest version) and linx are scary high...so I don't run those...although I know they're super synthetic tests. AIDA64 usually gets me high 70s/low 80s.

Any help appreciated! smile.gif
 
Solution
Do not use "newer" or "newest" versions of Prime95. You should only use version 26.6, here: http://windows-downloads-center.blogspot.com/2011/04/prime95-266.html


Or else you're running AVX on your cpu which is not desirable. Version 26.6 on Small FFT does not run AVX on your chip and presents a steady state 100% load, which is the correct method for establishing thermal compliance and a 12-24 hour run is necessary to establish stability on overclocked chips. 10-20 minutes to establish thermal compliance.

If it stays below TJmax or within thermal boundaries for that period of time, it's fine.

Do not run Blend or Large FFT, as those are entirely different test methods and do not represent the best or desired method of testing for...
Do not use "newer" or "newest" versions of Prime95. You should only use version 26.6, here: http://windows-downloads-center.blogspot.com/2011/04/prime95-266.html


Or else you're running AVX on your cpu which is not desirable. Version 26.6 on Small FFT does not run AVX on your chip and presents a steady state 100% load, which is the correct method for establishing thermal compliance and a 12-24 hour run is necessary to establish stability on overclocked chips. 10-20 minutes to establish thermal compliance.

If it stays below TJmax or within thermal boundaries for that period of time, it's fine.

Do not run Blend or Large FFT, as those are entirely different test methods and do not represent the best or desired method of testing for thermal compliance.
 
Solution


Ok will do...I did just run through 5 loops of x264 and was in the mid 70s...

I will do the prime 26.6 in the near future.

 
I can't say for sure as to whether your system is properly working but water can absorb a lot of heat without changing temperature by much (water has very high specific heat). I haven't gotten to water cooling myself yet but from my research, the temperature difference in the water should only be a few degrees. Maybe too little to notice by touch. There is also the possibility that the pipes are insulated because otherwise the pipes will radiate heat inside the case.
 
The pipes and water block are generally still cool enough to touch. In fact, it's a specific test for mounting pressure to apply medium pressure to the heatsink/waterblock to see if temps drop when cooling performance is in question. If it drops, there is either not enough mounting pressure or an issue with the thermal paste application. Therefore, it should be warm, but not too warm to touch.