Testing: performance and noise

Test Setup
CPU	Intel Pentium 3 866MHz
Cooling	Thermaltake Volcano 9 heatsink (no fan)
Video Card	GeForce 3 Ti200
Mainbord	DFI CA64-TN
Memory	256Mb PC133 SD-RAM
Other	Octek 235W ATX PSU Seagate 30Gb PATA HDD Aopen DVD reader

First I disabled all case fans. It turned out that I could cool all processors/chipsets passive which is a good thing if we want to measure noise. The Octek 235W PSU's fan has been modded to 7V so that it would only have a minor impact on our test results. The noise level was recorded with SmartSensor SL4001A. The sensor was placed ~60cm away from the front of the case, and a second time only ~20cm away from the front of the case. The lowest dBA reading in the test room was ~37,8 dBA.

I booted the system until Windows was loaded. At that moment the HDD will be doing almost nothing. Then I tested every fan separately and finally all the fans together. When you install the M.Play - Home Center software you will see that the fan speed are only set to 50% each. For our dBA test we cranked it up to 100%.



While our test setup was rather silent, the two case fans were not, especially the top fan. Luckily Zalman added their software so that the user can adjust the noise level to his needs. Both fans at 50% were already noisy enough for this HTPC case so I don't see the point in testing anything higher then that. This is the performance with the case fans set @ 50% while the system was running idle:



Both fans have some impact, but it's clear that the top fan has the best position to effective cool down our system. For load testing I ran Toast and 3D Mark 03 together:



Again the top fan is the most effective, the difference is even bigger.

Let?s wrap things up ->