AS5 is JUNK!

R

RogueTrip

2[H]4U
2FA
Joined
Feb 22, 2003
Messages
2,881
Ok, here is my story, same parts, same week, Specs in SIG! I recieved my X2 4200+ 6 days ago. Installed stock HSF with AS5 per directions (not my first time by a long shot). Fast forward to today where i decide to start O/C'ing a touch, well running 220(2420MHz) and running Orthos after 20mins my CPU is at 67c (still running ok but temps too high for me!). So i have a Tower112 Thermaltake HSF that i was going to install until i find a bunch of SMD caps on the back of the mobo (ASUS A8N32-SLI DLX) so that screws that up due to the mounting method. So while the stock HSF was out, i cleaned up the cpu and the stock HSF again with alcohol. This time i decide to use the paste that came with the Tower112 (high performance white paste (probably generic). Well here i am with same setup after 36minutes going strong at 57c!!!!

Been a long time user of Artic Silver 3&5 and after doing my own lil bench, don't think i'll be buying anymore AS5 again!

If anyone wants to know, system temp is 38C load. and when idle it is about 30-32c for the cpu, and the sytem is around 32-34c during idle.

Guess what, 39minutes and still 57c!

Edit, no voltage bumps done at all! Oh, and 57c @ 44mins ;)
 
I don't think it's your thermal paste. I think you didn't have a good mount on your cooler and now you do. That or maybe you had really old AS5?
 
Sideroxylon said:
I don't think it's your thermal paste. I think you didn't have a good mount on your cooler and now you do. That or maybe you had really old AS5?
Click to expand...

There isn't a wrong way to mount the K8 stock cooler, My idle temps with the AS5 was around 28c idle at stock cpu settings and 38c when o/cing at idle, then at 2.42ghz it was running 67c with AS5, so it wasn't the mounting.

I bought my AS5 on 8/9/2006 which i don't think qualifies as old.
 
Applied too much, applied too little, didn't have uniform coverage, HSF wasn't making uniform contact, etc etc etc. Could have been a dozen reasons why you saw what you did. Before you write off AS5 as junk, perhaps you should reapply it and try again. If you get the same results as before, you've validated that it isn't well suited for your setup. Until you do that, you're jumping to conclusions.
 
Did you remove the thermal pad under the stock HSF before the first application of AS5?
 
oh DADGUMN! i DID NOT remove the thermal pad on the stock HSF before applying the thermal paste, is this why my temps skyrocketed to 70C!???
 
what are you using to read your temps? there is no way your cpu temp can be lower than your system, even if you are idling
 
jnick said:
Did you remove the thermal pad under the stock HSF before the first application of AS5?
Click to expand...

Yep, cleaned and removed before fist install.

N H O said:
what are you using to read your temps? there is no way your cpu temp can be lower than your system, even if you are idling
Click to expand...

Yes, many peoples can, even my PIII 1GHz runs cooler than the mobo it's on!

I haven't had time to try anymore remount with AS5, maybe sometime in the next week i can do it. I've been building systems for over 10 years, so i think i know how to setup it up and get it going without the noob mistakes.
 
RogueTrip said:
I haven't had time to try anymore remount with AS5, maybe sometime in the next week i can do it. I've been building systems for over 10 years, so i think i know how to setup it up and get it going without the noob mistakes.
Click to expand...
Your results are pretty odd...but I'll check them in a couple of weeks (doing a thermal paste performance comparison test...I'll post the results when I'm done)
 
RogueTrip said:
Yep, cleaned and removed before fist install.

N H O said:
what are you using to read your temps? there is no way your cpu temp can be lower than your system, even if you are idling
Click to expand...

Yes, many peoples can, even my PIII 1GHz runs cooler than the mobo it's on!
Click to expand...

I've recently built a Celeron D system that runs cooler than its motherboard at Idle. The BIOS and Software Measurements both display about 40C for the proc and 45C for the mobo at idle. This is partly because of bad cooling for the motherboard I'm sure, but nevertheless, it is possible.
 
As5 i noticed tend to be picky to little/much will vary in temps.. It has to be just right i keep re applying till i get a good consistant temp im happy with or until im sick of taking off the hsf lol :D
 
(cf)Eclipse said:
well within margin of error for me :D
Click to expand...

3c is within your error of margin?! Oh Eclipse :p

Mine is 1-2c. But 3c is 3c either way, irregardless of it's source.

I'm also a Master of AS5 application, so don't go blaming it on application differences!!
 
Well of course it must be the AS5. Sounds like your infallible so that only leaves room for the AS5 to be "junk" as you said. We can all now just use the super good thermal material that comes with our heatsinks and disregard all tests and studies done to show that AS5 is superior to most thermal interface materials. Thank goodness you've opened our eyes to the truth.
 
Sideroxylon said:
Well of course it must be the AS5. Sounds like your infallible so that only leaves room for the AS5 to be "junk" as you said. We can all now just use the super good thermal material that comes with our heatsinks and disregard all tests and studies done to show that AS5 is superior to most thermal interface materials. Thank goodness you've opened our eyes to the truth.
Click to expand...

Umm yeah he already knows he made a mistake, how about trying not to rub it in?
 
Sideroxylon said:
Well of course it must be the AS5. Sounds like your infallible so that only leaves room for the AS5 to be "junk" as you said. We can all now just use the super good thermal material that comes with our heatsinks and disregard all tests and studies done to show that AS5 is superior to most thermal interface materials. Thank goodness you've opened our eyes to the truth.
Click to expand...

My gosh, didn't know i was ruffling the feathers of so many people about AS5, The compound i used came with my 45$ aftermarket HS (thermaltake tower112) and as i said its probably not much better than the generic compound you can get. You can always believe what you read, or you could actually test for yourself to see whats better, sure i could to re-apply ten more times looking for that great amount when i already used a credit card to spread a thin layer of AS5 on my X2, but after one application use with the Tt compound which i did the same way i still have better temps!

Here is a pic of 55c load after an hour running 2.42GHz!


TekSomniaK said:
Umm yeah he already knows he made a mistake, how about trying not to rub it in?
Click to expand...

Hope your not referring to me, i have made no mistake, maybe you mean OsageCowboy still running a pad with thermal material, as i've already posted i cleaned both HS and CPU before my first install.

The last time i made a noob mistake like that was back when i got my Celeron 366 and didn't put ANY paste on it, i was posting around 75c, got some paste on it and went on my marry way running @ 550! Even my P1 166MMX @ 250MHz ran fine with generic paste!
 
Well, I sorta agree with Rogue trip.

I mean, I've PERSONALLY seen a few temp decreases by changing the paste to AS5... Alot of times when customers come to me with "overclocking issues" (I help profile a customer's system and help them overclock it for a very reasonable price) I always add AS5 rather than ceramique or the white junk they use.

I generally see a 1-2c temperature decrease with AS5. Now, considering alot of these decreases in temperature are completely due to the misapplication of previous thermal paste, I can only (confidently) say that it is a BIT better than the generic white stuff.

Where AS5 really shines is it's "curing" ability, which helps decrease temperatures another 1-2c over a month after application. While this is technically in the margin of error (for Eclipse anyways) it's still a significant drop. An initial 2-3c drop, followed by another 1-2c drop after a bit of use. Ceramique does not do this, synthetic pastes do not do this, pure silver pastes don't do this either (even though their inital temperatures are about the same as AS5).

The only compound that's come close to beating out AS5 IMO, is the coolaboratory Liquid pro, which netted me a 6c drop on my current system. Get the chip a bit warm (hairdryer) and the stuff makes an impossibly thin layer.

Since all thermal paste (except for the pure metal liquids, like mercury or gallium) have such high thermal resistances, that the best and worst pastes are only about a w/mk or two apart... Sure, that's a difference of maybe 50% total heat conduction, but when you think of it... 50% of about 4w is only an extra 2w.

How do i know this? Well...

AS5 has a thermal conductivity of 35 000w/m^2 at a layer thickness of 0.01 inches.

Considering a thickness of 1mm is about the typical thickness of this stuff...

35000/4 = 8750w/m^2

Now, since 1m^2 = 1 550 square inches (and the hotspot of a processor is about 1 inch ^2 any more than that and you're only covering the IHS, which does nothing) we divide 8750 by 1550 to get something like 5w of thermal conductivity in a "real world" application.

Not a huge difference, when "generic white pastes" have a thermal conductivity of ~18000w/m^2.

So, yeah, technique is most important for initial application, but as I said, AS5 outedges the competition after the "curing" period, which helps by a degree celcius or two.

That's a very simplistic conversion between W/mk and w/inch^2 which leaves alot out, so don't quote me on it... But it certainly gives you the idea without using complex formulae that I can't type in Safari ;)
 
I'll share one of my nooblet TIM bloopers with ya. My first time applying TIM was the ol' RatShack sillly-cone + zinc oxide tubes. It was on a DX4 that I had overclocked to 120MHz (which fooled Diablo into thinking it was a Pentium 66 so I was able to play it okay). I used my bare finger to apply it, and yeah it looked fine at the time, but boy was I wrong. Luckily it was just one of those little orange HSF and CPU's didn't put out much heat so there was no damage or even any stability issues. These days guides are everywhere, one doesn't need to be shown in person to learn, but the stakes are higher. Glad you got it sorted.
 
You must log in or register to reply here.
Back
Top