poor water cooling temps

needmorstuff · 19 Sep 2007 at 08:46

Hi all - I was using a Thermalright ultra 120 and idle temp was 47.. with an e6300 @ 3ghz on default volts, I have installed watercooling (list below) and the temps are practically the same.. I know none of my watercooling kit is the best at the moment but would have thought temps would be better.. I have wc'd many times before but this is my first venture back in a year or so.. after the flood!!

swifty micro res
dtek db-1
dd txd
7/16 masterkleer tubing
black ice pro (single) with 2 antec tri-cool fans in push pull (lowest setting)

I only had the same 2 fans on my ultra 120.. 1 on the ultra and one on the exhaust of the case..

what gives guys! and if you changed one component what would it be?

get a laing ddc
get a black ice 120.2 gt steath and mount externally using a rad box or bolts.
get a black ice PRO 120.2 and mount externally using a rad box or bolts.
get a fuzion cpu block
I would get a pa120.2 but they are sooo big and sooo expensive.

Tooks · 19 Sep 2007 at 09:18

First off, a watercooling kit won't necessarily lower your idle temps... What is the change in your load temps?

Second, if I was to change anything in your setup it would be the block.

I've also got an OC'd E2140 (3.2Ghz) that's being cooled by a black ice 120.1 and a single Amber fan at 5v, with a D-Tek fusion block and that is 29c/50c idle/load.

It's actually a lot better than my main setup at 3.4Ghz which has both a 120.2 and 120.1 rad with a DDC pump and an Apogee GT cpu block/MCW 30 chipset block, which achieves 37c/60c.

Nothing I've ever changed on the second setup has really lowered the idle temps, the results have always been seen at the load end of things.

needmorstuff · 19 Sep 2007 at 10:00

my load temps before where about +10c... I'll do some testing and see what they are now.

needmorstuff · 19 Sep 2007 at 10:24

quick dirty load test indicates that load temps are also +10c...

w3bbo · 19 Sep 2007 at 10:33

Change the block and then the rad - preferably both.

Its like trying to get a good sound system. You will still get crap sound if you have a good amp with crap speakers or a crap amp with good speakers.

You really need to change both at the same time to reap the benefits.

Marci · 19 Sep 2007 at 11:59

That's because ultimately the water rig you've built has the same c/w as your previous aircooler.

Quick and easy way for better temps short term is more airflow.

Long term... Black Ice ProII will beat GTSStealthII... DDC will outperform DB-1... D-Tek FuZion will outperform TDX... upgrade the lot.

Low-end water generally equals high-end air... and that's what you've got there...

rome87 · 19 Sep 2007 at 12:22

Is the Aprogee GT that bad at wc ?

Marci · 19 Sep 2007 at 12:27

Eh?? The OP doesn't mention the Apogee GT anywhere... anyways...

needmorstuff · 19 Sep 2007 at 12:32

Marci said:
That's because ultimately the water rig you've built has the same c/w as your previous aircooler.

Quick and easy way for better temps short term is more airflow.

Long term... Black Ice ProII will beat GTSStealthII... DDC will outperform DB-1... D-Tek FuZion will outperform TDX... upgrade the lot.

Low-end water generally equals high-end air... and that's what you've got there...

why aint you pimpin the pa120.2 marci ;-)

also you have previously stated that cathar says as long as you have over 1lpm that you will only see a minimal diff from changing pump..

rome87 · 19 Sep 2007 at 12:51

Marci said:
Eh?? The OP doesn't mention the Apogee GT anywhere... anyways...

@marci what cpu are those comparisons for ?

Marci · 19 Sep 2007 at 16:59

@marci what cpu are those comparisons for ?

Core2Duo

why aint you pimpin the pa120.2 marci also you have previously stated that cathar says as long as you have over 1lpm that you will only see a minimal diff from changing pump..

Cos you already said it was too big and too expensive in the OP so why bother... ?

Indeed, as long as pumping power is above a certain threshold difference is minimal, but there is still a difference. Cathar's statement was in relation to current blocks... TDX doesn't fall within that bracket - it's a block that partially relies upon impingement that was designed for old naked-die processors, so for THAT block headpressure and flowrate will generally improve performance by more than the minimal amount that you'd see with FuZion or ApoGTX.

Either way, didn't say that that threshold was 1lpm...

"With everyone using IHS capped CPU's nowadays, pumping power has less of an influence than it used to. From what I'm seeing, both in theory and actual testing (which are aligning nicely), you can pick almost any single pump with 2-5mH2O of pressure head, and 8-15lpm of flow-rate, and the temperature differences at the CPU at the end of the day will not vary by more than 0.3C between them. We are now in an era of relative flow-agnosticity with respect to pump choice. On bare-die CPU's, low-flow, small tubing, weak pumped w/c systems lost out. On modern IHS capped CPU's, the differences are now small enough to be counted as being insignificant by many."

Raikiri · 19 Sep 2007 at 17:08

Hmm, should be lower...

I get better temps on my e6300 with a Black Ice GT, Apogee GT and a crappy Hydor pump. It was 39c earlier playing CSS and I've never seen it go above 44c.

I'm only using two 1000rpm fans (Nexus) in a push / pull set up like yours. Based on your temps (with air and water) I would say it isn't making very good contact, you could try lapping the CPU. Other than that the CPU block would probably be the weakest part of that system compared to mine but even then it isn't much worse.

needmorstuff · 19 Sep 2007 at 18:09

Marci said:
Core2Duo

Cos you already said it was too big and too expensive in the OP so why bother... ?

Indeed, as long as pumping power is above a certain threshold difference is minimal, but there is still a difference. Cathar's statement was in relation to current blocks... TDX doesn't fall within that bracket - it's a block that partially relies upon impingement that was designed for old naked-die processors, so for THAT block headpressure and flowrate will generally improve performance by more than the minimal amount that you'd see with FuZion or ApoGTX.

Either way, didn't say that that threshold was 1lpm...

I remebered about my OP after I posted... indeed you are correct.. but that was when i was keeping my current case (read small) now considering DD waterbox plus (any experience with these?) - and reckon a pa120.2 is right up my street.

1lpm? don't know where I pulled that from lol.

good to know about the naked core and current block issue.. maybe thats why my temps suck.

Marci · 20 Sep 2007 at 14:59

If getting a DD Waterbox Plus, ensure you order the ThermoChill-specific edition... regular Waterbox Plus is spaced for HWLabs radiators...

needmorstuff · 20 Sep 2007 at 17:41

Marci said:
If getting a DD Waterbox Plus, ensure you order the ThermoChill-specific edition... regular Waterbox Plus is spaced for HWLabs radiators...

ah excellent - thanks m8. If overclockers don't sell it (and i am sure they don't) then I presume I can order one from you guys and pop in for it?