C/Ws not a linear measurement......

I've noticed in my situation this is not the case. With watercooling a Duron 650 @ 800 1.625V my C/W was .15. With the same testing condition and the Duron at 952Mhz 1.85V my C/W went to .20? Should the C/W remain constant??

No I don't think so. The real governing heat transfer equations are WAY more complicated than just the simple 1-dimensional Q=kA(dT/dX). Don't expect the ratio of deltaT to Q to be constant under different situations. But for most pratical O/Cing purposes I think the C/W rating is good enough a guide to help someone size a heatsink for use.

AMDGuy (Jun 20, 2001 08:10 a.m.):
I've noticed in my situation this is not the case. With watercooling a Duron 650 @ 800 1.625V my C/W was .15. With the same testing condition and the Duron at 952Mhz 1.85V my C/W went to .20? Should the C/W remain constant??

Click to expand...

Hmmm... I would hazard a guess that it has something to do with efficiency of heat transfer at different loads and under different conditions. (I am not a mechanical engineer, I am only guessing.)

Did your measurements take the temperature of the chip less the room temperature?

Actually, I'm not even sure that the Celcius scale is linear... are there any mechanical engineers out there?

ttyl
tpb

TPB (Jun 20, 2001 10:32 a.m.):

AMDGuy (Jun 20, 2001 08:10 a.m.):
I've noticed in my situation this is not the case. With watercooling a Duron 650 @ 800 1.625V my C/W was .15. With the same testing condition and the Duron at 952Mhz 1.85V my C/W went to .20? Should the C/W remain constant??

Click to expand...

Hmmm... I would hazard a guess that it has something to do with efficiency of heat transfer at different loads and under different conditions. (I am not a mechanical engineer, I am only guessing.)

Did your measurements take the temperature of the chip less the room temperature?

Actually, I'm not even sure that the Celcius scale is linear... are there any mechanical engineers out there?

ttyl
tpb

Click to expand...

The Celsius scale is linear.

The only thing I can think of is that at different Watt loads there will be different C/Ws. It just suprised me a little. All calculations I've done are correct. I checked them 3 times.

I used the following formulae
Radiate to determine watts of processor.
Delta = Processor - Ambient
C/W = Delta\Watts

Temperature, is a measure of how much thermal energy a material can hold per unit mass. The heat capacity of materials will change according to temperature, so for a given constant heat transfer, the temperature will not rise or fall linearly. Aluminium, for example, has specific heat capacities of 482, 798, 949 J/(kg.K) for 100, 200, and 400 K respectively. But for the sake of this argument, let it be know that the non-linear change of temperature due to this phenomena is not the only factor that is responsible for the different C/Ws AMDguy has observed. I'm not very sure of the other contributing factors... but, you can be sure all the heat drawn from your CPU can be constant if the load is constant and will all passed on to your WC system. Otherwise, all the excess heat the WC (or HSF) cannot remove will go on to heat and heat your CPU till it reaches the melting point of something under that protective ceramic core of the CPU.