I got my new toy today (CGK)

[yo2tup]

stopped by taisol's main office as i was coming home from the dorm, and picked it up. compared to my alpha, it looks cheap, but it also looks like it could get the job done, especially with it's size, so i like it http://www.geocities.com/civic27/other/Dcp00589.jpg


i was disappointed with the base of the heatsink. it had the texture like that of the grip part of the cap of a 2 liter coke bottle. well a little bit finer, but you can really feel it. so i got out the 1000 grit and 2000 grit paper out and lapped that sumb!itch till i could see my face in it
it doesn't bling bling like the alpha's copper base tho :frown:
http://www.geocities.com/civic27/other/Dcp00590.jpg


my tbird system is at my dorm, and it's thanksgiving break so i don't get to install the cgk until i go back on sunday :frown:

insane-hardware's review showed that it performed better than the fop38 (i dont trust anand's review, too inaccarute). but i don't think either of them lapped it...so i wonder what the results are going to be..

right now i have a tbird 750@1050mhz, and runs 36C idle and 53C underload i hope this cgk makes a difference.

 

[Mikewarrior2]

Man, mine didn't show up today .

The reason the bottom of the heatsink isn't as smooth as an alpha is because socket A heatsinks use a PCTC(phase-change) pad/tim, and this stuff usually performs much better than any silicone/generic grease out there. Part of the reason being that the stuff is heat-fused onto the heatsink.

A lapped one should perform very, very nicely. I'll post results on a p3/internal diode reading hopefully friday(if it shows up ).


Mike

 

[ErikS]

Neither did mine - & I was hoping to having something to play with on turkey day!

Mike -

I read the HSF thread & had a question. Isn't the point of these type of reviews to give you an idea relative performance? I know when I read them I don't pay much attention to the actual number, I just look at the relative performance of each (to each other). I was also interested in the figures for the chrome orb, where did they come from (I'm not questioning, just suprised they were that bad)? I get okay performance from the ch-orb, but then again I have a big ass (see sig) case that stays very cool. Even with the fans @ 7v I can keep the chip at a stable temp. I get a reading of 45c, including a generous error margin I'm figuring 53-55c which is still way below the 90c AMD says is max. Of course I think it's too high which is why I've ordered the Taisol

I do appreciate the information in the discussion, for kindergarteners like myself the information is very helpful.

 

[Mikewarrior2]

Well, for one thing, if you've made it to the bbs, you're no longer a kindergardner. My biggest problem with the Anandtech review is the number of new computer users that show up to the only the front page and see this review.

While the ordering in the review may well be correct ordering, you see a ton of heatsinks grouped in the 45-47C range. This is due to the messed up temp reading, and the difference between those guys is not shown whatsoever. So while it reads only a 2C spread between those 4 heatsinks, the actual spread may well be 8-12C. The problem is, there is no real way to know.

I got the c-orb scores off a p3/internal diode reading. A taisol cek733 in My Case runs the cpu roughly 14C over ambient. The c-orb runs the cpu roughly 19C over ambient. I go into detail about how I did this in the thread in my sig. I ran the tests 4 times, both heatsinks were lapped, and I used Artic Silver for both.

The best results equaled roughly .48C/w for the taisol, and .63C/W for the c-orb. Now, I don't believe tillman when he says the cpu wasn't putting out maximum heat, but to give him the benefit of the doubt, I calculated theoretical temps for his t-bird putting out 45W of heat. Hence, the alpha w/ys-tech is about right, whereas the c-orb is percieved to be 2C warmer than the alpha, whereas the core temp is actually 12C warmer.


Mike

 

[ErikS]

I see you like to have the inside of the case neat & tidy. I also like the paint job.

I guess the reviews would make it hard on a newbie. There are so many factors that influence the performance of a HSF. I mean things like the case size, case shape, airflow, and neatness all factor into the cooling picture. I think HSF ratings are one of the hardest to accomplish. Compare them to, oh, say a vid card where you have x resolution, running y program, with z drivers.

I repeat, thanks for the information. Threads like the one you started allow for a good discourse & we can all learn from them (if nothing else they point out all the factors involved).

 

[Mikewarrior2]

that's the whole point of being a member to a bbs... glad you found the information helpful.


Mike

 

[Maverick2002]

oh man oh man........i cant wait to see those results...btw mike, what would you recommend: a digitaldoc or a compunurse (ill use a 5.25 bay either way so size isnt a factor)