higher flow rates = warmer temps

Hi,
I wanted to share this little tidbit.
I have a pelt water chiller
http://westech.home.mindspring.com/chiller/mychiller.htm
that I use to keep my cpu and nb cool.
I have been using a no name 145 gph (rated) pump for some time. Recently the cpu temp has been reported as 29.5~30.5 (idle). I had a Via-Aqua pump that was rated at 290gph so I thought I would swap them around and see what happened.
First….I confirmed that the VA pump did pump more fluid. I pumped into a measuring container. The outcome was the VA pumped around 30% more fluid than the no-name.
So +30%
I swapped out the pumps….same hoses no changes other than the pump. The outcome?
My cpu temps report .5 degrees warmer (at idle)!
I also noticed that my chiller cools the water just a tiny bit quicker. This is with the water temp set at 21.5.
Now then, under load the cpu temps are .5 degrees COOLER!
I surmise that this indicates that for every situation there are different outcomes.
wj

yeh that doesnt surprise me. i dont see why someone would use a pump over 150gph. the key to cooler temps is in the water temp, not how much water is pushed over it. people that buy those high gph pumps are just pumping heat back in the water and doing the reverse of what they are trying to do.
with that extra money you save from buying a 150gph pump than a 350 you could buy a pelt or something and cool alot lower.


laters.
t

I thought it was interesting that with the SAME water temp I saw these changes.
wj

Where are you taking the temps from? You have to remember that bigger/faster pumps means more heat. I guess this is proof. I'm trying to find the smallest pump I can find around 150gph.

I am taking the temp measurement from the output of the pump.
This is a peltier chiller system, with a set output temperature, not a water cooled system whose temperature floats around a bit.
wj
the chiller

I just brushed past your pics briefly so I missed a couple of things.

But: Hows you waterloop constructed?
Which parts go where, i.e. pump before res, and TECs + cooler goes where??

A consideration is: The change was 0.5C. The temp measurement on both water/cpu, how stabile is that? What's the instruments resolution? Precision? Deviation?
For home measurement a 0.5C difference is very small, considering the cheap instruments.

Also, remember that hoses picks up heat on the way to cpu, so on a warmer day, it will be warmer water reachin' your cpu.

A higher velocity throughput might force the warm water to spray towards the exit of res before it has thoroughly mixed with cold water (yes, I know that you measure outgoing temp), and that might tip the scale a 0.5 C.

I have to agree that it ain't funny to upgrade and notice no (or worse) difference.

Hi,
Well after running it for a week I see that the deviation is just the lowest increment of temp measure that is displayed. One of those things that you accept and don't worry about.
The overall pumping performance is better, if under load the cpu is either at the same temp as before or a tad (.5) cooler, this is with the 'same' water temp as before the pump swap. The obvious improvement is that the recovery time is lowered. Now then, FWIW, it doesn't really matter. I keep the water around 5~7 (C) below ambient, the cpu always stayed cool, it still does, it doen't overclock any better (at the same water temp) so........
in my situation the original pump did the job as well as this more powerful one.
wj