So I thought i'd share my Speedfan experience, after setting up and identifying the temperatures on my motherboard (Shuttle SX38P2) and CPU (Q6600) I ran through a few tests running Prime95 Small FFTS, General use... recording the maximum Tcase (Kind of an average CPU reading, BIOS etc) plus the Tjunction values (Each individual core temperature.).
Firstly I primarily read through this thread (http://www.overclock.net/intel-cpus/263821-c2d-quad-temp-guide-computronix.html)
to get a feel for the actual chip I was using, determining the "Tcase-Tjunction Delta value"
. (I did not know about this prior to reading this article to props to the author, it also covers setting offsets for the core and tcase values if they are required. I found after vigerous testing of my chip with the help of an analog thermometer!, I needed no calibration.)
I would sincerely advise
to read through this for a guide.
Noting that my chip is a Q6600
it runs on the following scale:
Scale 2: Quad
Q9x50: Tcase Max 71c, Stepping C1
Q9300: Tcase Max 71c, Stepping M1
Q6x00: Tcase Max 71c, Stepping G0
These are the values intel have set to ensure longevity in the CPU life, if the values are surpassed the chip may start to throttle itself. Seeing my Tcase max
100% load values do surpass the 71 degrees celcius
I also noted that my chip did not start throttling so don't worry too much. My CPU is seldom used at 100% load anyway, my Video playback and simultaneous Firefox browsing test showed a max Tcase value of just 68 degrees celsius meaning that my CPU would indeed my fine and dandy.
The Tjunction max on the q6600 go0 stepping is 100 degrees c, the older version is 75 degrees c. At this temperature your chip will shut itself down to protect itself. Do not exceed this value. For other intel or AMD chips this may differ.Test Prerequisites:1.
Install Speedfan - Identify Temperatures, Manually Control Fan Speed dependant on your setup.2.
Read this guide
if your new to this: http://www.overclock.net/intel-cpus/263821-c2d-quad-temp-guide-computronix.html3.
1. GTX260, 8GB DDR2 PC6400 RAM, 2.4GHZ Stock speed, 30% PSU Speed.
2. Ambient 22 Degrees celcius.Test 1: 10 minutes at 100% Fans 100% Load - 4 Thread Prime 95 Small FFT's.
Cpu Tcase: 78/79
Core4: 74/75Test 2: 10 minutes at 100% Fans 50% Load - 2 Thread Prime 95 Small FFT's.
CPU Tcase: 64/66
Core4: 65/69Test 3: 10 minutes at 40% Fans 0% Load - Idle.
CPU Tcase: 52/53
Core4: 57/60Test 4: 10 Minutes at 40% fans 20-30% average load playing 720P MKV through VLC + Firefox Browsing.
CPU Tcase: 63/65
By recording the max values for the Tcase and Tjunction values whilst doing moderature computing. I feel that setting the minimum cpu fan speed manually to 40% (at which point the noise is no problem, you shuttle SX38P2 users know what i'm talking about.) should be fair as the max temperatures of all cores+tcase did not reach to the maximum stable operating temperatures noted in the 100% load, 100% fan test. Once temperatures now reach this limit value, I have set Speedfan to increase the fan flow speed by 10% to bring the temperature back down or as close to this temperature as possible.
Knowing that I can now play video without the fan kicking in all the time is awesome. A correct speedfan setup at last! Only when under intense load will my fans need to kick up in order to keep the temperatures down to a minimum.
These shuttles do sure run hot, i'm wondering whether reversing the ICE fans will be worthwhile as now the GTX260 I installed today completely blocks off an air intake from that side of the case. If I reversed the ice fans wouldn't the air flow surely be increased as there will be cool air flow for the cpu, and also extraction through the gtx260 exhaust.
But i'm happy atm, although running hotter than i'd usually like to run hardware. The temperatures are still operatable under Intel Thermal Specifications, just about lol.