For information only...............

Hagar · 30 May 2011 at 11:00

..............ever wondered how warm your heatsink gets.

Test using an LED display and thermocouple attached to a heatpipe just above the CPU. Thermalright Ultra extreme with 120mm Apache fan PWM.

At idle the CPU temperature using HWmonitor was 34C in a warm room and the display read 33.6C at the heatpipe, a reasonable correlation I thought. Suggested accuracy of testing +/- 1C for the range.

Now running prime for max heat.

HWmonitor is at 51C the heatpipe temp is 38C, only a slight rise.

This goes to prove I suppose that the heatsink is performing properly removing heat at source.

The heatsink temperature is far cooler than the CPU temperature under load.

You cannot gauge CPU temperature from touching your heatsink which will just feel warm under any loading

olivier renault · 30 May 2011 at 12:07

nkata said:
..............ever wondered how warm your heatsink gets.

Now running prime for max heat.

HWmonitor is at 51C the heatpipe temp is 38C, only a slight rise.

This goes to prove I suppose that the heatsink is performing properly removing heat at source.

Or is it the opposite

I don't honestly know, thermodynamics and all, but I would expect the heatpipes at least to rise much higher as the CPU temps rise, so the heat is transferred from CPU to the pipes, then to the radiator. At least on a passive system. Sort of an equilibrium between ambient and surface temps.

Hagar · 30 May 2011 at 12:37

olivier renault said:
Or is it the opposite I don't honestly know, thermodynamics and all, but I would expect the heatpipes at least to rise much higher as the CPU temps rise, so the heat is transferred from CPU to the pipes, then to the radiator. At least on a passive system. Sort of an equilibrium between ambient and surface temps.

I suggest that the radiator effect of the fan + finning is stabilising the heatpipe temperatures so that they never achieve the temperature of the CPU.
The thermocouple was attached to the heatpipe as low to the CPU as possible below the finning.

joey2cool · 30 May 2011 at 13:48

From what I remember, heat pipes don't transfer heat as temperature, but as latent energy from the phase changes of the fluid within the pipes. There is a small amount of heat conducted through the copper, but hardly anything in comparison to the fluid. Hence measuring the pipe won't reveal the heat passed through it per se