• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Improving water cooling performance (360 Radiator)

aphasia

[H]ard|Gawd
Joined
Jan 15, 2011
Messages
1,137
howdy

i'm looking to improve to cooling performace of my water loop. reason being i've decided to skip upgrading to SB (adopting a wait a see till BD / IB arrives) and partly due to these threads.

-[Review] Triple Radiators V2
http://hardforum.com/showthread.php?t=1467444

-PUSH or PULL for single 120mm radiator exhaust fan?
http://hardforum.com/showthread.php?t=1537776

after reading the 'Triple Radiators V2' topic i realised my loop kinda sucks (have a 10 to 15 degree delta btw ambient temp and idle on q6600 cores).
Note: My q6600 is clocked @ 3.6GHz (400x9) on vcore 1.4v. all power management feats. turned on. idle clocks @ 2.4Ghz (400x6)
also i'm not sure if i should have my rad fans in a push or pull orientation after reading 'PUSH or PULL for single 120mm radiator exhaust fan' thread.


the parts in my loop (in sig) all on 1/2 ID loop tubing;

-swiftech MCP 655 pump >
-swiftech MCR360 radiator (w- 3x120mm NBlocker 1800rpm fans @ 1400rpm, push config) >
-Apogee GTX CPU block >
-MCW30 block (SB) >
-MCW30 block (NB) >
-Dual 5 1/4 bay res >
-swiftech MCP 655 pump >

See pics below (i know, not exactly pretty at the moment...)

Left side of case;
mozart%20WC%20loop1%20(left).jpg


Right side of case;
mozart%20WC%20loop2%20(right).jpg


Right side (rear top);
mozart%20WC%20loop3%20(rear%20top%20right).jpg


in the 'PUSH or PULL' thread a suggestion was made to install shrouds between the fans and radiator. i have 4 or 5 thermaltake tt-1225 120mm fans (never used, came with the case) which i can gut for this.
i couldn't find much in the way of info for the TT fans apart from here;
http://www.hardwarecanucks.com/foru...rmaltake-big-typhoon-cpu-cooler-review-5.html
they seem to think the fan is ok. perhaps it may be better to use them conjunction with my Noise Blockers' in a push-pull config rather than gut them for shrouds...

in anycase i'm worried about introducing more noise. i recently replaced the rad fans. the old fans were Scythe Kaze 2000rpm 38mm fans and they were a little too loud even when reduced to 1300/1400 rpm.

anyway any suggestions would be great.

cheers.
 
Do push/pull with some 1200 silent fans (gentle typhoons are nice, but there are many options) and you will be good to go. Don't worry about shrouds.
 
thanks caniba

i'm trying to keep costs down (just replaced 3 rad fans @ just under $100AUD). i could try the TT fans i already have on hand.

if i was to use my NoiseBlockers with the Thermaltake fans, does it matter which type does the pushing / pulling. the NB's are a far better quality fan than the TT's.
 
The measurements in the test tested the temperature difference of the WATER, not the heat source. The water temperature will always be lower than the heat source temperature, that's why it can take heat from the cpu.

To test how your loop really is doing, you need to measure the temperature of the water as it is coming out of the cpu, and the water temperature as it is coming out of the radiator. That is how you find the deltas with which you can compare to the review testing.

Additionally, cpu temps as reported in windows or w/e are not always accurate. To be accurate, you need an actual temperature probe on the cpu to measure the temperature.

My system has a 3x140 radiator. Fully loaded, my core can reach approximately 20-25 C above ambient. However, the water in my loop never gets above 10-12 C above ambient.
 
Last edited:
You already have a triple rad and only on a CPU+Mobo loop the only improvements would be fans, switching out the water blocks for better ones if any, and a thicker 360 rad. But upgrading the blocks and rad would be overkill for just that loop.

Are you using coolant or dye? How long have you had it in the loop as well? Could be gunking up the blocks and making temps worse. What are your hardware temps and ambient?

Your temps also can't be that bad either I have a 360 rad for my 930 and 2 470's.
 
Last edited:
AdvanSuper;
Are you using coolant or dye?
-no dye, just 'Swiftech HydrX Coolant'

How long have you had it in the loop as well? Could be gunking up the blocks and making temps worse
-it's been awhile. probably a year. purchased this stuff b/c its supposed to prohibit oxidation, formation of algea etc so i didn't have to spend time draining and flushing.

What are your hardware temps and ambient?
-atm the ambient is 28.5C

Temps...
==============
CPU
--Core0: 42
--Core1: 40
--Core2: 37
--Core3: 36

NB: 40
SB: 39

Voltage...
==============
Vcore: 1.40
NB core: 1.62
SB core: 1.07
 
i wish, those temps are idle.
i'd have to have a friggen awesome setup or my radiator in a bucket of ice next to my beer to have cpu core temps @ 40C loaded while ambient was 30C.
i wish i was that good.
 
Last edited:
You should be posting your load temps so we can see if they're normal or not. Idle doesn't mean anything. And those idle temps look normal.
 
-no dye, just 'Swiftech HydrX Coolant'


-it's been awhile. probably a year. purchased this stuff b/c its supposed to prohibit oxidation, formation of algea etc so i didn't have to spend time draining and flushing.


-atm the ambient is 28.5C

Temps...
==============
CPU
--Core0: 42
--Core1: 40
--Core2: 37
--Core3: 36

NB: 40
SB: 39

Voltage...
==============
Vcore: 1.40
NB core: 1.62
SB core: 1.07

Interesting post, my ambient is 26 and I am averaging 33 on all four cores with i7 965 @ 4.4. My load temps are around 66-70. However, I have a Feser quad and it appears to handle just about anything you can throw at it.
I am looking forward to my new build with the 980, 2 580's and the KL NB-SB kooler.
Just my 2.
 
AdvanSuper
You should be posting your load temps so we can see if they're normal or not. Idle doesn't mean anything. And those idle temps look normal.


-ambient during load 23.5C

-full load w- Prime, In-Place Large FFTs, 4 threads
-360 Radiator. 3x NB 120mm fans throttled down to 1400rpm (push) 2xTT 120mm fans @ 1300rpm (pull)

Temps...
==============
CPU
--Core0: 56
--Core1: 55
--Core2: 51
--Core3: 51
==============

found something odd when glancing at CoreTemp. Core #0 had a exclamation mark (!) next to temp. had a look at the historical High temps;

--Core0: 103 C
--Core1: 97 C
--Core2: 95 C
--Core3: 94 C

WTF...

i've never seen my core temps go anywhere near this.
unless there was bug in the readout, this is very worrying. the only thing i can think off that could of caused this is pump failure.

everything is fine now. any suggestions?

theseeker

my ambient is 26 and I am averaging 33 on all four cores with i7 965 @ 4.4. My load temps are around 66-70

a little jealous of your temps. are you just cooling your cpu on your loop. also how much voltage are you putting through your chip. i've got 3 blocks on my loop (cpu, nb & sb).
 
Your load temps are great adding another rad won't get you huge gains. Play around with fan setups on the current rad if you want.

That also could be a bug. Try using real temp instead.
 
i was a little surprised by the temps as well. haven't tested load temps since i replaced faulty fans. believe i've knocked off a few degrees C off load at roughly the same ambient. as you suggested i might mess with push/pull config. currently the rad fans are mounted to the case in a real ugly fashion (zip ties)

AdvanSuper
That also could be a bug. Try using real temp instead.

is there really much of a difference between CoreTemp & RealTeap. they're still polling the same sensors i'm assuming.
 
Not sure, but I can say I've never seen anything like that with real temp in the time I've used it. Your pump also has a lower failure rate when compared to the 3XX series pumps.
 
--Core0: 103 C
--Core1: 97 C
--Core2: 95 C
--Core3: 94 C

at these temps, core0 must be approaching the kill-switch to shutdown.
anyway i've enabled csv logging in coretemp.
 
Back
Top