|
|
Heatsink/ Heat Pipe / ThermoSiphon Cooling The cat will only make the mistake of putting its paw by your HSF once. :) Also the place to discuss the new high end heat pipe goodness. |
Thread Tools |
03-06-2006, 09:15 AM | #1 |
Cooling Neophyte
Join Date: Mar 2006
Location: South Central PA
Posts: 3
|
CPU Temperature Sensing
I have an NF7-S MB on which the temperature sensor does not work. My case has two LED temperature readouts on the front attached to two temp probes.
I spliced a new probe end on the one which is made to be treaded between the CPU chip and the socket and taped it to the back of the CPU. The second lead is stuffed down between the heatsink and the CPU nearly touching the die. At idle, with a room temp of 20c, the probe on the CPU reads about 72c and the heatsink probe reads about 28c. Under load, Prime 95, they read 78c and 31c respectively. I have read on the AMD site proceedures for calculating an estimated die temp by taking readings from the back of the CPU and readings of ambiant air at the intake of the CPU fan. However, the paper stated that the given equasion only applied to CPUs with a ceramic package. I believe mine is not ceramic. (Athlon XP-M 2600+) My question does anyone, on this board, have any experience measuring temps in a similar fashion? If so, any guidance or input would be appreciated. Thanks |
03-08-2006, 06:46 PM | #2 |
Cooling Neophyte
Join Date: Mar 2006
Location: South Central PA
Posts: 3
|
Re: CPU Temperature Sensing
Am I asking in the wrong area of the forum? Anyone in here use anything other than the standard motherboard sensor to measure the CPU temperature?
|
03-08-2006, 11:45 PM | #3 |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
Re: CPU Temperature Sensing
Did you check the probes before you installed them on the PC? It would be a pretty good idea to get both probes together and see what they read side by side. Wrap some plastic around them (water proof them) and get a cup of warm water and stir them around in the cup for a minute or so and see how they read together.
Also make absolutely sure that probe is not keeping the cooler off the CPU. 70C+ seems insanely hot. My stock AMD cooler keeps my CPU around 40C load. |
03-11-2006, 02:09 AM | #4 |
Cooling Neophyte
Join Date: Mar 2006
Location: South Central PA
Posts: 3
|
Re: CPU Temperature Sensing
Thanks for the reply jaydee. I did not check the probes side by side, good idea. I know 70C+ seems really hot but the probe is strapped to the back of the CPU! I've never measured temp that way before, so is it hot?!?
The second probe is not interfering with the heatsink, I stuffed it down there after mounting the heatsink and it is close to the die. The computer is stable, Prime 95 for 8.5 hrs, memtest no problem. Runs all the benchmarks to completion and has no problems in any of the games I have. I'd like to overclock it a little bit more but just not sure where I'm at with the temperature, for real. That's why I wondered if anyone on here had actually measured temperature from the back of their CPU, something to guide me. I'm using an XP-M processor and AMD says 100C max operating temperature, I guess I somewhere under that. |
03-11-2006, 09:38 AM | #5 |
Cooling Savant
Join Date: Sep 2003
Location: Vallentuna, Sweden
Posts: 410
|
Re: CPU Temperature Sensing
70° C is a little warm.
Have you tested the probes together (in a plastic bag in water for example) and checked that they read the same? I used to measure CPU temp this way, on the Tbirds, and your numbers are a bit too high. You would definitely want to get them lower if you want to overclock. I used to thead thin wires under the actual socket, not between the CPU and socket. You are running the risk of tilting the CPU and thereby causing a bad contact with the heatsink. I am guessing that this is happening. If that is not possible, cut a groove in the top of the socket, a channel for the thermistor wires. Void your warrantee. |
03-11-2006, 11:24 AM | #6 | |
Put up or Shut Up
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
|
Re: CPU Temperature Sensing
Quote:
You will have to to verify the probes are reading somewhat accurately by using the methods me and Inco suggested. You will want to try and record temps over cool down period so you know the probes a liner. Meaning they change temp the same as each other. I had a set once both read the same at 25C but as it got warmer one would go to 50C and the other 35C. I would bet you are not running 70C if everything is stable and working fine. |
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|