Part 2. Test Results (temperature, thermal resistance, noise level)

Part 1. Budget AVC solutions, new models from Thermaltake, Titan and Zalman

Our updated testing technique for Socket A coolers is given in the GlacialTech Igloo 2310 and Igloo 2400 Coolers Review (the general issues and the original version of the testing technique can be found in the Coolers Comparative Testing Technique). The procedure hasn't changed since that time.  

The testbed is traditionally based on ABIT KD7-S mainboard.  
 

This board features the temperature monitoring system based on National Semiconductor LM90 chip. The chip is located in socket's heart which minimizes various distortions and makes the measurements very precise.  

Test system:    

• ABIT KD7-S V1.0 mainboard
• AMD Athlon XP 3000+ CPU (Barton)
• OS Microsoft Windows XP

CPUBurn burnk7 utility is used for modeling a nearly maximum load, and Motherboard Monitor utility is used for temperature control.  
   

Note 
Each cooler was tested with the Stars 420 thermal grease 
The diagram demonstrates the complex result 
 

Diagram 2. Thermal resistance 

Note 
Thermal resistance θ_ja  can be found from the following correlation: θ_ja = (T_j - T_a)/P_h, where T_j  is CPU core temperature, T_a is ambient temperature (in this case it's 33°C), P_h  is CPU thermal power (in this case it's 70W). 

Finally, here are the noise measurement results (the technique is described in the Coolers noise characteristics and noise level measurement technique).  

Note: The background noise is 19 dBA

Part 3. Ratings (technical-economical coolers map, performance and economical ratings)

Appendix. Coolers consumer qualities assessment (New Edition)