CPU Die temperature testing
 

I always wanted to get ahold or even make one of these but I don't have a great knowledge of going about doing this. Haven't found much help from buying one so i dont know.

http://www.overclockers.com/tips263/index02.asp
http://www.frostytech.com/articlevie...id=1355&page=3 (look at pic)


these are modules used to simulate a CPU Die temperature as opposed to using an acutal CPU to test heatsinks/blocks. I found on P4's it's somewhat less dangerous to test heatsinks/blocks because of the heatspreader but in the past month alone I burned one tbred, cracked one athlon core and spent over $150 well earned dollars on CPU's alone. Alot of the heatsinks don't cause much problems but once I hit the high end heavy copper heatsinks/blocks then the chances of kiling the CPU increases. I need a solution where I won't need to worry about overheating or cracking the core in any way and this seems to be the best solution at testing heatsinks/blocks. Could anyone help me out either guiding me through or even making one for me? Either is preferred, of course i'd take on all costs. I'd prefer a setup to monitor temperatures and not calculate c/w (like overclockers.com does).

So can anyone help me and my ever so light wallet out?

ouikikazz,

I would guess you realize much of this but I'll start without assuming that. I use OC Joe top picture as a reference for now.

The top of the block represents the cpu and its mounting board (I might argue the board shouldn't be copper but lets not for now). The wings are the mounting lugs, not sufficient for some water blocks.

After that, the shape of the block is somewhat arbitrary but OC Joe has just made the basic block in the shape of the cpu board. Basically, the bigger the block the more stable (slow) your system. Down side of that is it takes a long time to change the blocks temperature.

Insulation is to minimize heat losses, the more insulation the better. Don't know how OC Joe calibrates his block for heat losses but there are straight forward procedures.

There should be a reasonable amount of flexibility in selecting heaters. Bigger heaters will of course allow fast heating of the block to a desired temperature. Heat supplied to the block must be more than the heat loss. Once at temp then less power to the heaters is required. You do need to consider electrical isolation of the heaters from the block, this is a consideration as you start trying to measure cpu temperatures using TC/thermistors.

That is a quick summary of the basics, I believe.

By the way, I don't build these things.

http://www.dansdata.com/coolercomp.htm

Dan has a good (54 of them) cooler comparison, and a description of his testing methods, and " cpu die test " building .

Extra: BigBen2k found a link, that also had an ideia :

http://www.overclex.com/content.php3...D50%26page%3D2

It's french, but babelfish is your friend :) .