I was skeptical until I tried it.....

[Tophinator]

Iv read in a lot of posts in this forum about different methods of applying thermal paste and what works best and not. And many people have tried "put a big glob on there and squish your heatsink down" method and faced some critizism instead of doing the thin layer stuff like so many suggest. Well I use artic silver 5 and have always followed the directions of applying. And doing that for my current comp my temps were always haning around 50 degrees at idle with no OCing and would hit around 55 at full load. I always thought this was pretty high for the hardware im using so today I tried the other technique. I took my HS off, put a glob about the size of a peas of AS5 on my processor then just smashed it down when I put my HS back on and fired it back up.(NOTE: i did not clean, or adjust anything, I didnt even clean off the old AS5 that was on there in the first place which I applied just as the website tells you, so the only variable is how I applied the compound) Once I let my computer run for an hour or so I noticed an imidiate drop in temperature. My CPU now idles around 30-35 degrees and never goes above 40 under full load.


Just wanted to post my results cause they seemed very interesting to me. Anyone else do it the glob and smash way. :)

[sataraid0]

Gosh, that goes against everything they tell ya. Are you sure?

I'm considering trying it. Have a LGA 775 socket and runs at 50c at idle and cant get it below. Tried everything except that. Even went to the thread that shows how to mount the HS with the MB not mounted and it bending and so forth.

Just glob more on and mash it down, huh? :confused:

Regards.................Sterling

[Tophinator]

I would give it a shot. Im going try it next on my graphics card to see any temp drops.

[Panama Red]

Maybe I missed something. I thought the "glob" method was recommended for P4's and PIII's cuz they are large and flat as opposed to the AMD cpu's that have the small rectagular die in the center. BTW, is your processor a 333 or 400 fsb?

[sataraid0]

Panama Red, my processor is a 800 FSB.

well, the only instructions I ever saw was to put a small amount of artic silver, the size of a grain of rice, then mount the HS.

Never did get how to spread it out with a credit card to 1mm thick. Gosh, how do they expect you to measure that?

Anyway, gonna try the glob method and see how it works.

Never had a temp problem untill got this lga 775 socket. the p4 is 3.6 so expected it to run hotter, but not that much hotter.

Regards..................................Sterling :)

[Panama Red]

BTW, is your processor a 333 or 400 fsb?

Sorry, Sataraid. That question was directed at Tophinator. ;)

[Tophinator]

its the 333, but I used to have it OCed to about 385.

[Panama Red]

its the 333, but I used to have it OCed to about 385.


Just curious cuz I have a 3000+ 400fsb oem chip on an Asus A7N8X-E Deluxe and it runs mid 40's C idle. It has an unlocked multiplier but when I played with oc'ing it temps went up REAL fast. :eek:

[glc]

You need more compound when the CPU has a spreader cap - more surface area. The grain of rice is for exposed core processors.

[Stryker]

I'm pretty sure that if you tried the "Glob & Smash" method on something like my AXP 2800+, it would short out the CPU because it would squish out onto the circuts surrounding the thermal contact point (which is only a tiny square in the center of the chip). I have never tried it mind you, but it seems logical that this would happen when the are circuts only a few mm's away from the thermal contact point.

Stryker :)

[Tophinator]

stryker theres nothing there to short out. The only exposed on the top side of the chip is the core silicon. Also many companies like thermaltake make copper spacers that you put on the CPU that covers almost the entire back side except for the core. And if there was anything there to short out, those spacers would definatly do it.

[HAL9000]

Umm.. there have been MANY articles about shorting on the small laser cut traces that are on the top side of the CPU. THose copper spacers have a cutout to clear those and there have been many instances where one was not sitting right and shorted it out.

I would suspect that the reason you are getting better temps is that for one reason or another, your heatsink wasn't sitting quite flat and is sitting better now. I've always found that proper application produces lower temps.

[RPA]

I'm pretty sure that if you tried the "Glob & Smash" method on something like my AXP 2800+, it would short out the CPU because it would squish out onto the circuts surrounding the thermal contact point (which is only a tiny square in the center of the chip). I have never tried it mind you, but it seems logical that this would happen when the are circuts only a few mm's away from the thermal contact point.

Stryker :)

I had a friend do that to his AXP. Brand new, installing it the first time. He put way too much on the die and it went all over the CPU's PCB. Suffice to say the thing never started.

[fcoral]

i have an xp2500 on an a7n8x-e deluxe with stock hsf and my temps are always around 50-55c just from browsing the web. It was up to 56c a while ago with screen saver running. So i'm really thinking of trying some of that stuff also.

[Stryker]

i have an xp2500 on an a7n8x-e deluxe with stock hsf and my temps are always around 50-55c just from browsing the web. It was up to 56c a while ago with screen saver running. So i'm really thinking of trying some of that stuff also.
Hey fcoral,
The first thing I would try if I were you is check for dust on your Heatsink. That can make a HUGE difference in temps, as THIS THREAD (http://forum.pcmech.com/showthread.php?t=124146) will explain. I'd make sure you're all cleaned up before you begin fiddling around with anything else. Hope this helps,
Stryker :)

[timbit]

today was one of my "spring cleaning" days, and i took my computer apart. got all the dust blown out of there, finally did some cable routing, and reapplied thermal compound. i used a lot more compound this time (about the size of 2 or 3 grains of rice), because i noticed that it was really really thin and some spots appeared not to be covered. my cpu temp is down to 40-42 celsius from 50-54 (running fans in silent mode, folding @ home 24/7). my case temp is also down about 6 degrees. i figure a combo of the cleaning, rewiring, and thermal paste did it.

[spyder003]

I don't think I could get my temps down much farther if I tried. Running an A64 Winchester core w/stock HSF, and it's idling at 26C right now. I thought about scraping off the thermal pad and using some AS, but I thought I would give the thermal pad a try. These Winchesters really do run cool.

[fcoral]

hey stryker it's a pretty new build was put together just before xmas and i have a clear side window that i can see and it's all pretty clean.I have only one side exhaust fan. I think i'll get a fan for the back also and try to tidy up some ide cables on the inside. I may replace them with some round ones.

[Tophinator]

Umm.. there have been MANY articles about shorting on the small laser cut traces that are on the top side of the CPU. THose copper spacers have a cutout to clear those and there have been many instances where one was not sitting right and shorted it out.

I would suspect that the reason you are getting better temps is that for one reason or another, your heatsink wasn't sitting quite flat and is sitting better now. I've always found that proper application produces lower temps.

Hal9000 I have heard of the shims shorting stuff out as well as Arctic Silver 2 on a lot of older chips like the durons. But not Arctic Silver 5 because as to my recolection it does not conduct electricity(unlike AS 2) which is why I didnt really worry about shorting out my chip.

And also I did some research on it and AS5 does infact not conduct electricity because it is an Alumina based compound. Heres my source:
http://www.arcticsilver.com/as5.htm

[HAL9000]

But note this;

Not Electrically Conductive:
Arctic Silver 5 was formulated to conduct heat, not electricity.
(While much safer than electrically conductive silver and copper greases, Arctic Silver 5 should be kept away from electrical traces, pins, and leads. While it is not electrically conductive, the compound is very slightly capacitive and could potentially cause problems if it bridges two close-proximity electrical paths.)

... and as such, easily explains the systems that I have had run thru my shop with shorted out processors and AS5.

[Stryker]

Hal9000 I have heard of the shims shorting stuff out as well as Arctic Silver 2 on a lot of older chips like the durons. But not Arctic Silver 5 because as to my recolection it does not conduct electricity(unlike AS 2) which is why I didnt really worry about shorting out my chip.

And also I did some research on it and AS5 does infact not conduct electricity because it is an Alumina based compound. Heres my source:
http://www.arcticsilver.com/as5.htm
I still wouldn't take any chances...
(While much safer than electrically conductive silver and copper greases, Arctic Silver 5 should be kept away from electrical traces, pins, and leads. While it is not electrically conductive, the compound is very slightly capacitive and could potentially cause problems if it bridges two close-proximity electrical paths.)

[HAL9000]

Beat ya ;)

[Stryker]

Great minds think alike... and apparently so do Canadians!! :D

[Tophinator]

as long as it isnt directly between two conducters that are carrying a charger of 1+V there shouldnt be much a of a problem. If you spread a desent amount of the stuff over the better part of a chip I could see it building a charge but probably nothing more than few couloms. Which really isnt enough to do anything. Probably not even enough to effect a single input/output of a gate.


Edit; Typo

[Stryker]

...there should be much a of a problem...
Note quite definitive enough for me. I don't know about you but I don't have spare CPUs kickin' around that I can experiment with and risk cooking. Why not just apply it properly so there is no excess on the CPU and not risk any zappy-zap?

And besides... if the MANUFACTURER has a warning, why press your luck. They put that there so when someone does cook a CPU, they aren't liable.

[Tophinator]

Im not saying its the best way to go. If you dont want to take the risk then by all means dont do it. I dont have a bunch of spare CPUs lying around either but I weighed the risks and gave it a shot and im happy with the results. Im able to OC even farther then before so now I dont have to wait the extra minute or so for rending or simulations in Catia.

[shadowbreaker513]

I don't think I could get my temps down much farther if I tried. Running an A64 Winchester core w/stock HSF, and it's idling at 26C right now. I thought about scraping off the thermal pad and using some AS, but I thought I would give the thermal pad a try. These Winchesters really do run cool.
My newcastle gets down to that with nothing running with stock heatsink. [/offtopic comment]