WHAT? Am I missing something????

[itnetpro]

Ok Guys... I just built a system
Athlon Thunderbird 1 gig
Abit K7T Raid
256 Meg Cas 2 133 Sdram
Creative Labs GeForce 2 Ultra 64 Meg 2 40 gig Western Digital 7200 ata100 hard drives
Exc Exc
When I started the system with the Default Heatsink my Temp
under full load for 24 hours max was 71c

Then I went out and put a Coolmaster on with Default Fan
24 hours full load 55c

Then I ordered a HedgeHog and placed the fan from the cool master on it. 24hours full load 44c

I used the same Thermial Goo on both bought Heatsinks.
Same Fan. Now why does the Coolmaster out perform the Hedgehog? Something is not right. I understand what Tom is saying about the weight, yea he has a point but hey I do not plan on transporting my PC anytime soon.
Now why on Tom's Chart does the Coolmaster out perform the Hedgehog??? NOT HERE! and not on 3 of the boards I have seen test on.

Now look Tom. I have trusted you as a reliable resource for over two years now. After the Pent 4 test and now this I am starting to wonder. Please be more carefull in the future.

John Mc Laughlin

 

[gregorarch]

Maybe they made a poor job of applying the thermal grease on that test? It can make a big difference.

 

[Guest]

For every cooler we used thermal grease between CPU and cooler. The setup remained the same during all tests.

Uwe von der Weyden

 

[Guest]

You definately did something wrong!!!!!
Compare the results you achieved with all the others on the web, yours stands out as being total rubbish.

 

[Guest]

Dear SoldierBlue,
i am open minded for suggestion regarding the setup for the follow up review.
Everyone can send me tips, hint and suggestion, but "rubbish" feedback doesn't help.
Uwe

 

[gregorarch]

It's more or less impossible to be absolutely consistent with the application of thermal grease - all you need is one uneven glob to louse up your contact and you're getting way under optimum cooling (I know, I've done it). I don't believe for an instant that I could set up 17 heatsinks and get them all just right first time. That could explain the poor hedgehog result. Another possibility is just that it was a badly machined piece - I guess most people buying one of these would lap it before installing.

Heatsink testing is something that I've read and thought quite a bit about recently. I don't think that the methodology used for this test was ideal for a number of reasons. Fistly, I don't think you can control the test well enough in a real-world situation (see above). Secondly, measuring the temperature at the back of the CPU package isn't ideal, because you're measuring a differnt heat path to the one we're really interested in. Perhaps a better approach would be to use a peltier or two wired to a quality, controlable PSU as a simulated heat load - you can then measure cold-side temeratures, plug them into the peltier power formulae and get a pretty good measurement of thermal resistance, which is what we really need to know.

However, I can understand that from an author's point of view you probably wanted to produce an accessible article, that anyone could read and understand without getting into degC/W and discussions of thermal interfaces.

 

[sesquipedalian]

My temp was too high, so I removed my hedgehog and put it back on. The difference was cooler by 15 degrees C. I think it really makes a huge difference with the clip, that it is applied in such a way as to be of equal pressure along the top of the CPU. This is with a Celeron 600@900 that I have, the die size being such a small surface area. These heatsinks can tip ever so slightly and not sit perfectly flat on the CPU surface.

 

[Guest]

OK you want more specifics then just rubbish?
Every review I have seen put the hedgehog above the fop38 in terms of cooling efficency, in fact they all rated it above all others tested alongside it, one reviewer even ran his tbird without a fan, just the fins.
Seeing your review rate the hedgehog so differently must of raised your eyebrows?
If I was a reviewer I would certainly keep an eye on the competition, least of all as an alarm should differences arise. There were many reviews published before yours so even if you were the first to test, last to publish, an addition to the review at least acknowledging the differences would of gone a long way to preventing whinging from the likes of me.
What if the differences in the review of the p3 1.133 had gone unreported?
Further investigation on your part will be my only solace.