Hey guys. Its been awhile since ive posted but I need some help.

I have senior design this semester and for our project we are investigating liquid cooling for PC's. I have volunteered my machine as the guinea pig for the project but I am having a hard time getting some heat information about my pc components. Heres what I have.

P4 550
925X chipset (ASUS P5AD2 premium)
ATI x800 XT (from MSI)

I found the TDP of my processor which is 115W. I assume this is about how much the processor will put out in heat at full load? I cant for the life of me find any information about either my chipset or the graphics card. Anyone know where I could find such information?

Once our project takes off I think I will be making it a project log. Its gonna be sweet.

Hmm... a tough one. Well, you can try looking up the voltage and amperage requirements of the chipset and the videocard and use this equation to figure the wattage:

W = I x V ... where W is watts, I is amps, and V is volts.

You could also just get a sensor or even a thermometer, perhaps and electronic or infrared thermometer. Using a sensor or thermometer won't give you measurements in watts, which watts is possibly more ideal for you from a physics standpoint... but then again you can always convert.

Hope that helps a little.

The heat flux across the surface of a CPU die varies with time even under constant load and so does current draw. This is one of the largest sources of error in any technical heatsink / water clock test, which is why they're modeled using uniform resistor elements instead. Online wattage calculators are completely useless. The best you can do is get a figure for average power output in a given time period....and you'll have to measure it. Basically isolate the heat output of the CPU into a source of known specific heat so that you can quantify the amount of energy dissipated. Off the top of my head:

-Construct a calorimeter-type experiment on the IHS where the CPU is the heat source and monitor water temp.
-Same as above but instead use a mass of ice and measure time until melted.
-Construct an insulated box and isolate the heat-output device (rad, heatsink) and monitor air temp inside....basically another version of the above....you get the idea.
-Insert thermistors in the baseplate of a heatsink and compare heat flux values of CPU vs. that of known power-output sources and interpolate.
-Since energy in = energy out, measure the dT across a heat exchanger (probably radiator in this case and water temp) of known C/W value....this will be difficult without an environmental chamber.

Anway, you get the idea. Regardless of what you choose you're looking at pretty big uncertainties and alot of tedious testing since you'll need a decent sample size. Do you actually need to know accurate power outputs for these components or just estimate or only need to know them relative to one another? Sounds fun either way. Good luck.