You simply get 3 different resistance readings across the probe for approximately known temps, then you draw a graph with a line (curve if necessary) joining these points. After that you simply measure the resistance and read off the corresponding temp from the graph.Originally posted by ROSSCO_2004@12 June 2004 - 04:34
Multimeters work as temp probes?
how? can i do this with mine too or do i need a special one\special attachment???
Don't forget, I'm interested in comparative differences anyway, accurate temps aren't important. What I need to know is when I get the best results, not what that result is.
It's the same with the motherboard question, does the absolute temp matter (within reason) as long as you know that any cooling changes you are making are improving the situation, or overclocking changes are not about to cause chip failure.
BTW, the article suggests that the software could be modified to give a more accurate representation of the true temp. In practice the sensor is probably common across many motherboards, so software such as MBM5 has no choice but to interpret the reading from the sensor in the same way for all boards. A software solution is impractical.
However, the board provides a reference voltage to the sensor. Varying this reference voltage changes the reading provided by the sensor so it is VERY easy for the board maker to alter the temps shown by their board.
I know clocker got his ASUS board some time before I got mine, and his temps on that board have been consistently much lower than mine. It could be that ASUS have already addressed the under-reading problem. It would not make sense for them to knowingly under-read the temp since it would screw up their vaunted A.C.O.P system. Perhaps the sensor chip on that board was new to them, and on early boards they didn't realise it needed a different reference.
Bookmarks