The FrostyTech Synthetic Temperature Test Platform consists of an aluminum block heated by a 120V 250 Ohm precision resistor producing 50W of heat. That heat is transmitted to the heatsink via the aluminum block and up through special copper die templates with thermistors embedded directly in them.

A temperature reading for a particular heatsink is taken once the temperature stabilizes in the copper die template. Since each of the die templates have the surface area of either a silicon topped or heatspreader-topped processor the flow of heat to the base of the heatsink accurately mimics a real life configuration.

Shown with the copper die template for a silicon-topped processor, the synthetic heatsink test apparatus awaits a coating of AOS29KY non-silicon thermal compound and the Thermoengine V60-4210 heatsink. A slightly larger copper die template was used to test the effective cooling ability of the V60-4210 atop a heatspreader-based processor like an overclocked PGA Celeron processor.

We received some emails regarding the effects on the placement of thermistor used in testing the Thermoengine. Unlike traditional heatsinks which have their exhaust air going out to the sides, the thermoengine has the same cross section along its entire height. This leaves the thermistor used for recording the temperature from within the copper die template in the exhaust air of the heatsinks' fan. While the actual thermistor is embedded in the copper die template, the lead wires are not, and as a result they get cool air moving over them.

The air moving over those lead wires will always be cooler than the actual thermistor so there is the potential for the temperature readings to be indirectly influenced. This might help explain the large improvements in the Thermoengine's performance when larger, higher RPM fans are mounted on it. With the larger fans, more and more cool air is moving over the lead wires, which would normally be protected from direct exhaust by the base of the heatsink.

To gauge the impact of this unique configuration we set the heatsink up on our synthetic test platform for another test. In the first test we used a 1 inch long piece of closed cell foam around the thermistors lead wires to insulate them from the exhaust air and surrounding environment. We took our measurements then removed the insulating foam from the lead wires. Once the reading stabilized, we took another temperature measurement.

According to our test results there was around 3 degrees difference between the two temperature measurements. This would seem to indicate that because of the Thermoengines unique design, the downward exhaust air can indirectly influence the thermistors' reading.

The results on the next page were taken before the effect of insulated lead wires was tested for however. Since all other temperatures were also recorded with uninsulated lead wires, the results have not been augmented.