temperature problem

Associate
Joined
15 Nov 2011
Posts
1,237
Hi all

I have recently built a new PC using the i7 3930k processor (see sig for full spec).

I have a custom watercooling loop which runs as follows: Res -> pump -> Radiator 1( Alphacool Monsta 240mm (80mm thick) ) -> GPU1 -> GPU2 -> CPU -> Radiator 2 ( XSPC EX240 (35mm thick) ) -> Res. Both the radiators are using a pair of gentle typhoon AP-15's in a pull configuration. The Monsta is set up as intake and the EX240 is set as exhaust. I also have a 120mm fan in the front of the case as an intake and a 140mm fan on the rear as an exhaust.

The problem I have is that my CPU temperatures seem very high. (numbers that follow are based on stock settings). When I first mounted the block (XSPC Raystorm) I was getting idle temperatures in the low 40's and load temps in the high 70's to low 80's. Yesterday I remounted the block and re-applyed the thermal paste, now I have Idle temperatures ranging between 31 - 36 across the cores and load temperatures between 68 and 74.

Whilst this is a decent drop I still think this is too high, should I remount again? Can anyone who has a similar setup give me an idea of the sort of temps I should be seeing? When I overclock anything past 4ghz the load temperatures are skyrocketing to between 84 and 90 which is just ridiculous.

The ambient in my room is about 19c. Water temperature at idle is about 23c, the highest I have seen it after stressing the PC for a couple of hours is just over 30c. I have tried taking the side off my case to see if it made any difference to the temps but it didn't affect them at all, leading me to believe this is a problem with my block mounting and not case temperature. I used a reasonable pea sized blob in the middle of the CPU and tightened the screws of the block in an X pattern.

Is there a general consensus on best paste application method for LGA2011 processors? I would just like to know what I might be doing wrong as this is really irritating me.

Thermal paste is Gelid GC-Extreme btw. Sorry for the long post but I wanted to give as much info as possible.

Thanks in advance for any help given. Also could someone advise if I am allowed to post this in the watercooling section as well?

edit: I should have mentioned that any time i mentioned 'load temps' I was referring to temps produced by prime 95 using small ffts. I also tried OCCT but the temperatures were much the same between both
 
Last edited:
I see U using aquaero 5 to control watercooling how about upping pump and fan speeds.

Try lower voltages On CPU 8 pack might want to chip in he loves this processor

can't be the 580s dumping extra heat I find they run same temp as the water in my loop when idle

Oh taking the side off won't make any differance you pulling air into the case via pull config try putting your fans in a push config
 
Last edited:
I see U using aquaero 5 to control watercooling how about upping pump and fan speeds.

Try lower voltages On CPU 8 pack might want to chip in he loves this processor

can't be the 580s dumping extra heat I find they run same temp as the water in my loop when idle

In all the examples above fan speeds were on full... Which is still quieter than my last computer lol. Not that that is difficult to achieve. The pump is an xspc d5 vario with a custom top. I have it set to number 3. I did try adjusting it but it didn't seem to make much of a difference

CPU is using stock settings atm so I hadn't thought about lowering voltages.

You are right about the graphics cards, no problems with temps on those at all
 
I don't know much about W/C but would it not make more sense for the CPU to be cooled before the GPUs? Surely with the water hitting the GPUs first you're just pushing hot water to the CPU?
 
I don't know much about W/C but would it not make more sense for the CPU to be cooled before the GPUs? Surely with the water hitting the GPUs first you're just pushing hot water to the CPU?

Loop order doesn;t really matter in watercooling as the temperature of the water will eventually reach a constant level. Also the gpus at idle are running at the same temperature as the water (23c) so they are not really adding any heat to the loop during this testing. Thanks for your input though
 
I am going to try re-applying the waterblock in a minute, using the rice grain method instead of the pea method this time. Also currently I am tightening the screws until they are finger tight, should I be tightening them more than that?
 
Maybe slighly tighter

U could just clean off all Tim and use some contact paper to see how well your block is mounting
 
Well problem sorted... I can't believe I am about to say this, but I forgot to peel the plastic off of the bottom of the block! :eek: I noticed it as I was cleaning the paste off and couldn't believe it, what a plonker lol.

Anyway temps are much more reasonable now, stock load temps are in the high 40's to mid 50's, and a quick overclock using the gamers oc profile on the rive (4.7 @ 1.4v) gives me load temps mainly in the 64 -68 degree range. The hottest core (core 6) goes just over 70, which I think is pretty acceptable overall no?

When I get home today I am going to see if I can tweak the voltages and stuff to get better results.

Thanks for the help guys
 
quite possibly, I have a feeling I have seen this in other threads as well, never thought I would actually do it myself though lol. I think it will take a little while to live this one down
 
I think that is a mistake only made once =) glad to say i did it a while back and hope to not do it again any time soon ( no chips were damaged in the process) =)
 
lol nice to hear I am definitely not the only one. I hope I havent damaged the chip but it seems to be fine so fingers crossed.
 
Back
Top Bottom