XP-120 Orientation Tests

Hi

I did some crude tests on the orientation of my Thermalright XP-120 heatsink (fitted with Sunon 100cfm fan) on my Asus P5WD2 Premium motherboard with Intel 820D cpu.

I fitted the heatsink, and rotated the case to various orientations to see what the effect was on cpu/system temps. After each rotation I turned down the cpu fan speed (using a Zalman fanmate) to make the cpu very hot, then turned the fan speed back to full to see what the temp would go down to.

My case is a full tower so the motherboard is mounted vertically.

I used the temps displayed in the bios. All temps are in Celsius.

The results were:

XP-120 (mounted vertical with curved section of heatpipes facing upwards) cpu:63 system:43
XP-120 (mounted vertical with curved section of heatpipes facing downwards) cpu:62 system:43
XP-120 (mounted vertical with curved section of heatpipes facing sideways) cpu:62 system:43
XP-120 (mounted horizontal as it would be in a non-tower case) cpu:62 system:43

In other words, all orientations appear to be equal apart from having the curved section of the heatpipes facing upwards which is only 1 degree worse (and may just be an error on my part)

Then I took the side of the case, and temps dropped to: cpu:55 system 40

Time to do some work on getting more airflow thru my case.

Hope that helps someone.

John

Thanks for that, I actually had wondered which setup was better. I think I'll go with my SI-120 mounted so the curved section of the heatpipes is facing downwards so the RAM will be cooled also.

i hear turning the fan around also makes a difference
you wanna test that too? :D

62C lol, i suppose as you said that was with fan turned down, whats your normal temps on load, ie do you stick to the rule of thumb of 55C max ?.

Hi flytek. Yep. I'll test that. I'll get back to you with the results. The PC is in the "workshop" at the moment as I'm cutting a hole in the side of the case adjacent the cpu fan. And also fitting a second 120mm fan into the side of the case.

Hi humeyboy. The 62C was with the fan at FULL speed. CPU at 3.72Ghz. CPU volts at 1.4v. That's why I'm cutting the holes in the case and fitting the extra fans. I spent a lot of time carefully fitting the heatsink, so I'm fairly sure it's well fitted. Ambient temp here would have been about 30C when I ran those tests. I don't have any "normal" readings yet because I haven't really got that far yet.

I can't expect much better than with the case side off. So it's a worry.

I dont suppose you want to try with the CPU inverted?

I doubt your onboard sensors would be accurate enough to measure any difference in temps from the heatpipe orientation..
But as its going to me a minimal change (if any) ... its not very important in the scheme of things.

i noticed a full 4ºC drop from changing my xp-120 from heatpipes facing up to heatipies facing down with my 3.0E ona DFI Pro875B board

i remounted it several times, and it came up with roughly the same temp drop when i switched it back to the heatpipes pointing down



IMO the test would be more accurate by doing dual prime 95's and using a fast-setting generic thermal paste and remounting in each direction atleast 2 times to verify your results, rather than mounting once in each position and letting the CPU heat up and then turn the fan on to cool it down (never seen any HSF review done that way before)