Are your water temps for REAL??

[Flash]

Seriously, I'm beginning to wonder at some of the temps I see posted here. Why, you ask? Well, I've screwed around with my setup extensively, and I can't come close to some of the results I've seen posted here, despite the fact tha I'm using mostly top-of-the-line equipment, with very good system temperatures (23-24C).

My setup: 1.33 T-bird at 1.54, 2.0v core (Epox 8K7A). Maze 2 rev.2; DD Cube w/ 120mm Sunon pulling; Eheim 1250 inline; 1 gallon reservoir (h20 + Red Line); 3/8" throughout; fully blead; properly seated with ASII, tightened down hard.

My results based on die-side thermistor: idle 45C, Load 49C.
Based on MBM: idle 40C, Load 44C.

OK, so not bad and I'm not complaining, but I keep seeing folks with similar setups reporting temps up to 10C lower! What gives?! It's hard to believe anyone could do *that* much better with the same/similar equipment. In fact, with my current setup I'm only about 5C lower than with my cheapo Leufkin kit.

Do different boards/cpus run at different temps running at the same speed? Not from what I've read, but....

[Tacoman667]

Here's my setup. I have the DD MAZE2 version 1 on my CPU. My pump is outside the case with the DD Cube on top of the case. I have 2 120mm 131cfm YS-Tech fans on my cube, one to push air through and the other pulls it out. That way I should get as close to the 131cfm of air flow at all times. That is what is getting me my low low temps. When I had only the one fan on it I had temps of about 39C idle and 43C load. I keep my room temps at a constant 71F. That is why I have such low temps. I also use the 1250 Eheim pump and love it with no more then a foot high from the pump to the cube (the highest point in the loop). Experiment with your setup and you just might get those lower temps like some of us have.

[Flash]

I suppose that increasing airflow would shave a few C, but right now I see very little difference (.1-.3C) between running my Sunon at 7v and at full speed. Anyways, even if I could lower my temps 2-3C by running two powerful 120s, I'd rather have a little peace and quiet. So I guess I shouldn't complain, eh? :)

BTW, I LOVE my Eheim pump! Super-quiet, strong, dry ... what a difference from my ill-fated Mag7.

[Colin]

You may be interested in this. (http://mikewarrior.freeservers.com/) . Different motherboards report temps differently. On top of this, the algorithm used to interpret the socket thermistors signal changes to suit higher speed processors as new BIOS are introduced. It looks like the temps are increasing compared to an older BIOS. Also some of the BIOS on some motherboards do not always fully implement the CPU idle functions. This would explain the small range of temps between idle and load, even with your side of die thermistor. The fact that your temps have dropped 5C over the Leufken system leads me to believe your current system is working fine. Don’t believe the temp, believe your overclock. The Palomino with an in-die diode will help this situation but the signal from the diode will still be subject to interpretation by the motherboard.

[cjtune]

If your WC setup is the same as other ppls but you are getting hotter temps, chances are it's the mobo probe -as Colin pointed out. The relative height of the pump has very little to affect flowrates. Yes, your pump has to 'fight' the higher elevation to get water to the waterblock, but the water 'falling down' after exiting the waterblock will help suck water into the waterblock and both effects cancel out. The only thing here is that you have longer tubing and this saps pump pressure (and thus flowrate). Maybe the others used larger diameter tubes?

[Flash]

I suppose alot of the difference in reported temps could be attributed to in-socket thermistors, but like I said, I'm measuring by both in-socket and external (Enermax dual LCD) thermistors. So maybe those reporting super-low temps just have boards that read 10+C too low. I know that my 8K7A reads 4-5C too low.

Also, I do agree that the 8K7A--or at least my sample--does not employ much in the way of cpu idle. With all of my various setups I've never had a differential > 5-6C between idle and load. But it's amazing how fast the temp reockets down once I turn the system off.

Oh, I found this interesting. The quick response of the slug-mounted thermistor allows you to see minute, almost real-time changes in cpu temps. Well, you see alot of folks touting extreme stability because they can run P95 at the same time as other demanding apps. But, I find that running P95 along with other apps actually *lowers* cpu temps by as much as .3-.5C. Seems like W2K is splitting the cpu load, and since (with one or two exceptions) other apps will run at lower temps, overall temperature goes down when the processor is occupied with both demanding and somewhat less demanding applications.

[Quaky]

I never believed in any of these CPU heat-up apps, if I want to see if my system is stable, I encode a DivX. The CPU runs very hot then and if you are able to complete the encoding (4-5 hours) you can be sure the system is pretty stable :D

[Tacoman667]

My 2 YS-Tech 120's aren't even close to being as loud as a Delta 38. I know, I have one in my second system. I have a Delta 80mm 68cfm for my case intake and that's louder then BOTH of my 120's and then some more. YS-Tech is not completely quiet but they are EXCELLENT fans.

[SteenkyBastage]

just thought i'd ask...is your radiator in your case?

i'd imagine that HAS to make a large diff in temps. for myself, i have the raidator sitting out of the case next to the AC vent, so it gets a little turbocharge of 56 degree air every 15 or 20 minutes, heh.

but even without that, if you have the radiator inside the case you are doing one of two things. either heating up the air in your case by pushing the radiator air into the case, or you are using warm, in-case air to cool your radiator. either way to me seems like it's not a good solution.

i have all my fans on rheostats and turned almost all the way down (even the radiator fan), so my case stays at around 4 to 5 deg. C above room temp (not counting when the ac's on) so my processor will actually idle BELOW system temperature. there is no way that would happen with the radiator inside the case.

anyhow, just something to take into consideration...

btw, my setup:
DD maze 2
danner mag 7
DD cooling cube
half gallon reservoir (approx)

my room stays about 78 F the past week or so, while my case temp is usually 86 F. the processor idles at 78 to 80 F, and full load i can SOMETIMES hit 100 F, altho usually at the end of the day i look at the MBM readout and my high temp is 98 F.

[Flash]

Basically all of my waterworks are outside of my case, except of course the water block.