HSF tests with XP Thermal Diode, Good Read!

[Colin]

Void Your Warranty (http://www.voidyourwarranty.net/review/review.php3?r1=hsroundup1/index.php3) :D

[Hoot]

Nice article Colin. I've been recommending the Dynatron BH-638 to people asking about what HSF to get, but my words have for the most part fallen on deaf ears. What's impressive is how well the Glaciator 2 cools for the sound level it produces. With the advent of the Maxim SMBus chip for building a rather simple monitor of the Thermal Diaode, I believe we will finally see some uniformity in HS testing from the different sites. It is clear (at least to me) that using a Palomino, along with its thermal diode, is the way to go for testing HS performance. No result influencing factors, like air blowing on the in-socket thermistor, drilled HS wicking off some of the heat from the thermocouple, compunurse probe not making good connection to the side of the core, etc. That article has gone into my favorites as a reference page. Thanks Colin for the tip.

73, Hoot

[Newbie_Doo]

Just finished reading it. Thanks Colin. BTW, nice to see you posting. I haven't seen you for a bit, but maybe we just haven't crossed paths. :)

Cheers!

[Colin]

I'm still here. :) It seems the Glaciator should be the HSF of choice for dual CPU system. I can't imagine what dual Deltas would sound like.

[mEKbOY]

Originally posted by Colin
I'm still here. :) It seems the Glaciator should be the HSF of choice for dual CPU system. I can't imagine what dual Deltas would sound like.
You dont want to. Check my sig... its the worst sound because not only do they produce lots of noise but when you have 2 on the same board the begin to vibrate so badly thats its unbearable. I recomend 2 sunon 26cfm 38dba fans on dual's. They push enough air to keep things cool while remaining quiet. The SK6's work very well even with low cfm fans but those new glaciator 2's are very impressive considering the noise they put out.

[pHaestus]

Hey thanks for the kind comments on the review. I am going to try and collect a lot of other popular heatsinks to test and see how they compare. I'll just put up a page like Overclockers' die simulation results page except with results from the diode.

I plan on going back and looking at C/W vs CFM curves for 3-4 high end heatsinks soon: the BM-P and a Swiftech MCXC-370 for 60mms and an Alpha PAL and a MCX462 for the 80mm units. I am interested to see how much of the temp drops that have been noticed before are due to artificial cooling of the thermistor and how much are real core drops. I figured that comparing a thin fin to a more open design would be useful as well.