Watercooling The Conroe E6600

[wevsspot]

Was looking for some input from any one of my fellow forum members who are running an E6600 clocked in the 3.4ghz - 3.6ghz range AND using a good water cooling setup.

 

I just finished my Swiftech H20-220 Apex Ultra installation this weekend and I'm less than impressed by the end results. Ambient temp. outside the case is 23C. Ambient temp. inside the case ranges from 43 - 46C.

 

I'm using the 120X2 radiator included with the Swiftech kit as well as the Apogee GT water block.

 

At 3.4ghz and 1.45vcore my cpu readings are (using coretemp .95 and everest ultimate);

 

cpu = 39

core 1 = 46

core 2 = 41

 

After an eight hour run of orthos I browsed through the temperature logs and found several temperature readings of 63 - 65C on the individual cores deep into the eight hour run.

 

I've played with various configurations of the fans, including pushing, pulling etc. and settled on a push configuration with the fans blowing through the radiator. Contact between my cpu and water block looks good. Appears that I have a good flow rate and there aren't any kinks in any of my hoses.

 

Mainly I was hoping someone else here was using a similar setup and could share with me what kind of temperatures you are seeing.

 

If this is normal I guess I can live with it, but after spending $250 bills on this setup, this will bug me to no end if my water cooling setup should be performing better.

 

I am open to any and all input or suggestions.

 

Right now my circuit goes like this (per the Swiftech installation recommendations)

 

Pump out > northbridge in > northbridge out > radiator in > radiator out > cpu in > cpu out to resvr. in > resvr. out > pump in

 

I'm gonna take my Fluke Digital Infrared Thermometer home with me tonight and try and get some readings from the various hoses and the water blocks themselves.

 

Thanks.

[Praz]

Wevs

 

Do you have any baseline temps from aircooling the same system? Loaded C2d temps vary quite a bit from one to the next.

[wevsspot]

Add about 2C to the results from the water cooling setup and thats about where I was at using the Zalman 9700 - fan at full speed and the Thermalright HR-05.

 

I think I will know more tonight or tomorrow night when I do some analysis with my Fluke.

 

Maybe I left a couple of my hose lengths way too long, but I didn't want to cut any short.

 

I'll get some pics of the finished setup and post them so you can give me your valuable input.

 

I realize that my idle temps are very close to the ambient air temp. in the closed case, but I really thought that hanging the radiator outside would improve the performance of the water cooling circuit.

[PeTeZZ]

hi Wevsspot

When I had my E6600 I was clocked at 3.420 ghz vcore 1.4250v. During Orthos, which went 9H+ my temps ranged between 46ºC(low) - 53ºC (high) both cores generally the same. room temps = 22c. My Stacker is side panel-less. lots of fresh air!

is lowering your volts an option?

NB gets really hot , maybe air cool it? or invest in a triple rad?

[bcbooter]

basically all the sensors are good for is to let you know you didnt screw up the contact on the cpu with the block they really aren't very reliable, sorta like judging the rails by the bios readings....

i mean, at least on my board, with liek 10+ different cpu mounts all the temps were within 1-2c but when i got a different board it was like 5-7 lower.....

 

basically if your readings went down from the previous air cooling your most likely getting the most out of ur h20 system. Did ur OC's Improve?

[Elano]

I've read in the Apogee GT review that this block comes with a thicker o-ring so you can bow the block to fit Intel concav IHS:

 

"The block is also provided with a thicker o-ring that allows you to bow the block. By bowing the block, you alter the shape of the base, so that it is no longer flat, but convex. This should make it "fit" Intel's CPUs better, which usually have a concave IHS.

 

That should help lowering the temps a bit.

[Clay]

Hummm.... Looks like a lot of the stuff I was gonna' say has allready been said.

(Apogee O-ring,board sensors.)

 

What I've gathered about the O-ring:

Fir IHS apps the apogee gt is reputed to be about as good as any other water- block. As allready mentioned: Using the thick O-ring will warp the waterblocks base.(Convex) To better mate with the intels now becoming infamous IHS's.(concave) Once done though it can't be undone.Flatter IHS's won't mate up nice again. People are reporting better temps by doing that though.

Some info here...

http://www.xtremesystems.org/forums/showthread.php?t=142998

 

I must say I was abit disappointed when I first saw my MCR220. It's only about 1/3rd the size of the PA.. And my idle temps are quite abit higher than with the volcano12 Extreme with a 92mm tornado.(Actually 2# tornados) But my load temps are somewhat nicer. (And lots quiter.. ) Plus my gpu is now cooled.The 47cfm fans mine shipped with were abit too weak to cut the mustard.(At least to my concern) Though the swiftech kit ships with 81cfm fans which are comparible to the fans I'm now using. It seems okay for the moment. But I've taken it to mind to order an MCR320. $55.00 +shpng. Mounted with 88cfm fans Just to maybe knock down the temps a couple-few more degrees. Currently Ambient 29.2c,idle 34c cpu load 37-8c.

 

Another thing might make some diffrence is if the computer is stuffed under a desk or in a confined space temps can go up if air doesn't circulate fairly freely.

[danzgpgt97]

My temps were almost exactly the same as PeTeZZs were. As I got to the 3.4ghz and higher v-core area my temps were about where yours are during orthos. It didnt take much for me to see alot more heat so if your not to happy with it you could try backing it down a hair. My intels seem to run hotter than my amd cpus but that may be cause i push them harder. Dan

[Elano]

Hummm.... Looks like a lot of the stuff I was gonna' say has allready been said.

(Apogee O-ring,board sensors.)

 

What I've gathered about the O-ring:

Fir IHS apps the apogee gt is reputed to be about as good as any other water- block. As allready mentioned: Using the thick O-ring will warp the waterblocks base.(Convex) To better mate with the intels now becoming infamous IHS's.(concave) Once done though it can't be undone.Flatter IHS's won't mate up nice again. People are reporting better temps by doing that though.

Some info here...

http://www.xtremesystems.org/forums/showthread.php?t=142998

 

Another thing might make some diffrence is if the computer is stuffed under a desk or in a confined space temps can go up if air doesn't circulate fairly freely.

 

Clay,

 

Have you ever done that to your block? I am going to do that on a new system I got and the Apogee's manual doesn't say nothing about "how to perform the bow move".

 

Sorry for hijacking the thread:-)

[wevsspot]

Elano,

 

No offense taken, you aren't hijacking this thread. Your input and questions are welcome. The larger O-ring trick is detailed in a technical white paper available at Swiftech.com, and the O-ring should be included with your installation kit.

 

Interesting read. I'll update everyone on the progress I made this weekend. Last week I conveniently knocked my NZXT case off my workbench and onto the floor (it was an accident REALLY!) Thankfully everything except optical drives and power supply were already removed from the case.

 

So I ordered an Antec Nine Hundred which arrived last Friday. All Friday night and half a day on Saturday I spent reinstalling all of my gear into the new Antec case (extremely tight fit by the way).

 

Since I had my water block off I went ahead and swapped out the original O-ring for the thicker one that came in the kit. I also did a good lap job (no I wasn't in a titty bar!) on the Conroe IHS. With that completed I went about stuffing all of my gear into this new Antec case. While I was re-attaching the fan wires to the Swiftech dual 120 radiator I noticed that the molex plugs going to the fans on the radiator were wired to the 5V line instead of the 12V line. So I re-did the molex connectors making sure that the positive lead to the fans was connected to the 12v line of my power supply molex. Wa-La. Dual 120mm fans on the radiator sounded like jet engines. I pulled out the 12V - 7V resistors from the Swiftech installation pack and installed them. Fans are much more silent, but at 7V still moving a bunch more air than when the fans were hard wired to the 5V lines.

 

I hope all of this is making sense?

 

To put it in the readers digest version . . .

 

Got a new Antec Nine Hundred that has awesome case ventilation

 

Installed the fat O-ring

 

Lapped my E6600 IHS

 

Upped fan voltage from 5V to 7V (anyone with the same setup you might want to check that your fan wires aren't screwed up like mine were)

 

Ambient temperature outside the case right now is 25.6C

 

Everest reports the following;

 

Motherboard - 38C

CPU - 32C

Core1 - 41C

Core2 - 36C

North Bridge 38C

 

I'll run Orthos overnight and review my temp logs in the morning.

[wevsspot]

Then I'll follow up with some photos of the completed rig.

[Elano]

Installed the fat O-ring

 

Lapped my E6600 IHS

 

Upped fan voltage from 5V to 7V (anyone with the same setup you might want to check that your fan wires aren't screwed up like mine were)

 

Ambient temperature outside the case right now is 25.6C

 

Everest reports the following;

 

Motherboard - 38C

CPU - 32C

Core1 - 41C

Core2 - 36C

North Bridge 38C

 

I'll run Orthos overnight and review my temp logs in the morning.

 

Thanks mate!

Just a thought:

If you lap the IHS it won't be concave anymore so I think you won't be needing to aply the bow in the Apogee.