PCA Forums

It was almost 2 mm up higher. Meaning, when I put the processor on, the temp probe was pushed down around 1-2mm. That would explain the high readings. But, where is the temp diode SUPPOSED to go. I now have it even, so it BARELY touches it. (I am using an edge of a credit card to see, by holding it sideways on top of the socket).

Is it suppose to NOT touch the core at all?

Thanks.

I assume you got one of those flat temp probes and you want to measure the exact temp on the die, not relying on the mobo sensor, right?
In that case your temp probe has to touch the cpu-die on the side, lemme show a nice picture I made:



Just make sure the probe doesn't lift the HSF itself, you might have to modify the temp-probe a lil.

Nope, don't have one of those, otherwise I would use it.

i meant the mobo probe

In that case just bend it up as high as you can to make sure it touches the back of your CPU.
It is natural that it'll read more when it touched the backside of your CPU as when it is 1-2 mm away ...
The method shown in my pick should give you even more accurate temp readings.

ALrighty, willdo.

I hope you understand whatever temp you get from ANY mobo probe isn't reliable ... the mobo manufacturer developed some weird algorithm that goes with some testing and that's supposed to show SOME temperature.
It also happened that the manufacturer changed the temp-reading algorithm from one BIOS to the next bringing up to a 10 degree celsius difference.
The only accurate tempreading for a CPU will be a temp probe placed the way I showed in my masterful picture (lol) with an independent temperatur reader (Compunurse or comparable).

Ok, that sounds resonable to me.

Just to make my point look at the readings of some HSF reviews on overclockers.com like this one:
http://www.overclockers.com/articles539/
Just compare the numbers between "CPU die temp" and "MBM".