Water cooling running too hot.

[Stolid]

Hi everyone, been a while since I've been by Overclockers. Life's busy and I've been too short on time to put any effort into my hardware. Anyway, I'm back.

For those of you who may remember, I built a water cooled system (Athlon 1400, 256 PC150 DDR, ABit KT7a-RAID). I had Shadow do some case mods for me, and after both Shadow and myself had thousands of problems with the project, I got the case, installed the hardware, and got the box working.

My water loop uses a Maze2 (classic, before the .1 .2 revision sequences), Cube radiator, and a nice little pump Shadow found for me (I can't remember the name off hand, but I think she's enough).

Anyway, the system runs *way* too hot. I've barely O/Ced her (1435 MHz off the normal 1400) - and if I go any higher stability gets wierd and unpredictable.

I think this is due to my temperatures, they seem way too high for having water cooled this Athlon - 40 celcius idle, 45-50 load. I've checked my loop for kinks and can't find any. The water is distilled with water wetter and just a touch of bleach for safety. The cube radiator has a Panaflo L1A 120mm running @ 12 volts (She's on a 12/7/0 v bus, but with temps this high - they're running full bore almost always). I'm considering upgrading the L1A to an H1A - but am looking for any other ideas.

Thanks in advance, and good to be back!
Stolid

[Dissolved]

Originally posted by Stolid
Hi everyone, been a while since I've been by Overclockers. Life's busy and I've been too short on time to put any effort into my hardware. Anyway, I'm back.

For those of you who may remember, I built a water cooled system (Athlon 1400, 256 PC150 DDR, ABit KT7a-RAID). I had Shadow do some case mods for me, and after both Shadow and myself had thousands of problems with the project, I got the case, installed the hardware, and got the box working.

My water loop uses a Maze2 (classic, before the .1 .2 revision sequences), Cube radiator, and a nice little pump Shadow found for me (I can't remember the name off hand, but I think she's enough).

Anyway, the system runs *way* too hot. I've barely O/Ced her (1435 MHz off the normal 1400) - and if I go any higher stability gets wierd and unpredictable.

I think this is due to my temperatures, they seem way too high for having water cooled this Athlon - 40 celcius idle, 45-50 load. I've checked my loop for kinks and can't find any. The water is distilled with water wetter and just a touch of bleach for safety. The cube radiator has a Panaflo L1A 120mm running @ 12 volts (She's on a 12/7/0 v bus, but with temps this high - they're running full bore almost always). I'm considering upgrading the L1A to an H1A - but am looking for any other ideas.

Thanks in advance, and good to be back!
Stolid

welcome back.. i dont think those temps should make u that unstable.. i run 45~60C all the time, and im very stable. so it might be ur hardware, or a software issue...

[Colin]

I ran a Maze 2 and a Cooling Cube for a while. With a KT7 series thermistor pressed up against the ceramic plate, your load temps should not be any higher than the low 40s. Knowing your water temperature would help diagnose the problem.

First, is your block making good contact with the CPU?

A second fan on the Cube will drop your temps 2 C to 3 C at full load. Also if the radiator is pulling warm air from the case your temps will be high. With proper airflow, you should be able to keep the case temps within 1 C of room temp.

Cubes and Maze 2s are flow killers. You may be under pumped. In my system, a Danner Mag 7 with no flow restriction and a plumbing order of pump/waterblock/radiator worked best.

[Stolid]

Thanks for the replies.
As for the temps being normal (reply 1) - keep in mind, this is at default voltage. If I up voltage anymore my temps get so high that I don't even bother seeing how high the chip will run.

Reply two: my plumbing order is 'pump->radiator->cpu->pump' so that the water hitting the CPU is fresh out of the radiator cooler. As for the pump's flow, Shadow had a really nice setup for it that I didn't really want to try and redo with a larger one, further - this is my first WC system - so I have little first hand knowledge.
As for CPU core contact: I did the best I could, my previous CPUs were so cool that I just epoxied a plastic heatsink (not even needing a real fan) on there with some thermal epoxy radioshak stuff; so my plate connection might not be the greatest.
Since I was scared of ruining 2 CPUs (I had already overheated the 1st one because of a bad power switch that didn't let me turn it off when I ran it for testing - by the time I dived for the powersupply it was too late. Grrr. >.<) - used Artic Silver, followed their directions for cleaning the maze2/CPU and applying, then when setting the maze2 plate up I tried for 'firm but not difficult' - then we booted to bios, finger on the powersupply, and watched PC health. It hovered from 27-30 degrees for 2 minutes so we thought that was good. If I boot her 100% cold she gets up to 40 within about 5 minutes.

If I can stomach the risk of ruining the 1400 MHz chip, I might try to reapply if you think that's it. Sorry, I'm kind of queasy about working with this system. I've thus far had to return a radiator because of a drill hole, a hard drive and sound card from a power surge, and the CPU from the heating. That's - (estimates) about 500 of 'junk' on this first project, minus the hard drive's cost because it was warrentied - so 400ish.

Anyway, I've got a friend who wants to replace their P3-600 which runs at 50 celcius load, 45 idle (and it's not OCed). I was considering selling them my mem/mobo/proc and replacing it with a DDR/XP setup. So we'll see what happens. Thanks for the advice, keep it coming ^_^ I'm truely appreciative.

Stolid

[Colin]

If your boot temps are low and then rise considerably after five minutes, your waterblock/CPU interface is fine. It's the water temps that are killing you. More airflow over the radiator, more coolant flow and or a different pumping order. One would think the plumbing order of pump/radiator/block would work best but my temps went up 3C over pump/waterblock/radiator.

I should also note that over 5% Water Wetter will increase your temps.

[IFMU]

I would have to agree that those temps arent really that high. I would suggest tryin to find out what the water temps are in the loop however. The only thing I can think of is bad contact with the cpu. Have you tried lappin the block? Might help. Or as was already suggested, check your internal air flow, you might need to make sure that the airflow inside the case it working pretty good.

Well thats all I can really think of that might cause the temps up there without any volt mod...

[Stolid]

I did lap the block, contact seems pretty good. I can *barely* I mean *BARELY* feel airflow out the back of the radiator, which is why I was thinking about replacing the L1A with an H1A. and moving the L1A outside to help 'pull' while th H is 'pushing'. The internal temps are great, I've got 2 92s taking in fresh air, 1 120 exhaust, and the 120 exhaust over the radiator. All Panaflo L1 models.

Stolid

[Colin]

If the temps start out low, the waterblock has good contact with the CPU. The problem is the water temps rise.

[Vega]

I wouldn't question your system stability with relation to those temps, i'm running anything up to 58 c stable ( I know this is not good, needs to be cooler, but there is a point)

I would really check through that hardware