Originally Posted by Pygo You mean I've been doing it wrong all along? How do I remove paste from the cpu socket and the pins, as per image linked below??
Is this really all that bad? My friends are saying I'm stupid now... :( http://i87.photobucket.com/albums/k151/cburn25/42prosessor1.jpg
OMG i just pissed myself :)
This is just epic!
It also demonstrates how little the difference is between all these methods, and how rediculous these guys with their "nanotube compounds" are. Soon there will be another company selling us nanotube re-alignment kits, and we'll be writing articles and discussing the best methods of using them to make that last 0,05 degree difference, so we can up our CPU by 0.1 Hertz
Seriously though, getting it at least marginally right is important for beginning system builders, and for that, this article gives some good advice. It's a method that's sure to work, and easy to do, with little chance for serious error. Not to mention you solved Pygo's problem! ;)
OKAY guys, I may have been a bit harsch by saying that the method they are using in this article is "faulthy" (which basically means wrong, someone who evidently didn't understand that). But this is the old method of applying TIM. And as I said, and have been backup up by, air might get trapped between the heatsink and cpu. This is ultimately just common sence. Also, you don't need the TIM to be covered over every single mm^2 on the HS because the corners of the HS don't really help cooling the chip. Having the tim perfectly "vacuumed" between the HS and heatsink will give you a little, say very little, temp drop.
Nice article, but I think this is not a proper method, though. I think the manual-spreading method described in this article causes the following problems.
1. Even though the grease spreads evenly by careful scraping ON THE SURFACE OF IHS, It will not spread evenly if one mount a heatsink because of poor surface(microscopically) condition of the heatsink base. According to bit-tech, they recommended that only the amount of a grain of rice is appropriate. When you try this, it means that the thickness of grease will be only tens of micrometer (if you have a doubt, then calculate it with proper dimension). This thickness will not be sufficient to compensate and fill in any possible cracks, which may be deep up to 0.1 milimeter, on the base of the heatsink unless heatsink are produced in the finest quality. As a result, air can be trapped in which the grease can't fill.
2. Evenness of grease will be screwed up after you mount the heatsink. The technique that Bit-tech has introduced is somewhat idealized and abstract. The reality is that you cannot avoid any little movements and slides between the two surfaces when you mount the heatsink. In this sense, manually spreading method loses its advantages. Microscopically, some parts of grease will be slightly amassed due to this movements and this spoils the evenness of grease.
I hope, in the next time, Bit-tech can publish a better article which completes the stage of heatsink installation. Only meaningful is that the method containing both this procedure and applying TIM. Without describing how to mount it, the whole article written in this time is nothing but a mere method how to evenly spread MIT BEFORE INSTALL THE HEATSINK. Nothing is meaningful more or less than that. Hence, we cannot say spreading MIT in this way is effective.
You can't avoid having small deviations in the thickness of the compound throughout, but having a decent spread before you start is better than just leaving it in a big lump in the centre.
It's always an idealized scenario with these kinds of things, because in an ideal world your CPU would come with a heatsink on it already.
As for fitting instructions, anything beyond the obvious manufacturer's recommendations would be immensely hard to write about over different sockets etc. Be realistic. :p
Originally Posted by Veles Or you're just following the advice of Arctic Silver, since they you know, make the stuff, I would imagine they know their stuff.
I'm just following the laws of physics. The greater the area of contact between two surfaces, the greater the thermal transfer between them will be. Therefore ensuring the entire surface has a coating of the stuff at an appropriate thickness BEFORE you apply the heatsink has got to be better than applying a blob and HOPING it covers the entire surface AFTER you apply the heatsink. Think about it. You can either apply a grain of rice sized blob and hope the heatsink pressure spreads it evenly, or you can apply the same size blob and make sure it covers the entire surface manually before you apply the heatsink. The absolute worse that can happen is there will be no difference whatsoever... the best that can happen is you will get better heat transfer.
I challenge anyone to argue with that logic.
I couldn't give a monkey's what arctic silver say.
Quote:
Originally Posted by Veles My CPU hasn't burnt out yet, if you think about it, the majority of the heat would be directly over the chip, the periphery of the heat spreader which wouldn't get covered would be at a much lower temperature than the centre unless the heat spreader is crazy efficient.
Like you say.. the majority will be directly over the chip, yes... the majority.. not all. Ensuring the entire heat spreader is coated will ensure ALL of the heat is transferred, especially as many third party HFS have a surface as large, or larger than the chip.
Originally Posted by Pookeyhead I'm just following the laws of physics. The greater the area of contact between two surfaces, the greater the thermal transfer between them will be.
hm... no one can deny this fact. :)
Quote:
Originally Posted by LeMaltor I've always spread it about on the CPU and heatsink, never had a problem with temps.
Right. However, virtually nobody has problems with the temperature by doing whatever methods they'd like to do. What we should pay attention is that the differences in tiny scales. Of course, not many user would concern this but the problem is that this article deals as like this methods is a recommended one, though nobody can sure it worths such recommendation.
Originally Posted by ksiryx Right. However, virtually nobody has problems with the temperature by doing whatever methods they'd like to do. What we should pay attention is that the differences in tiny scales. Of course, not many user would concern this but the problem is that this article deals as like this methods is a recommended one, though nobody can sure it worths such recommendation.
Taking the 5 seconds required to smear some paste over your CPU isn't worth the chance that your temps will be lower, even if only by 1 degree?
How lazy are you? It's not as if the article suggests undertaking 5 hours work or anything... jeez!
So far as I'm concerned it IS a recommended method... just by applying logic and physics.
I stress again: The worst that can happen is no improvement at all... and at best, an improvement.
Why would you NOT want to do it? You're all just being stubborn on some principle, or because the god like figures at Arctic Silver haven't told you so. If this info suddenly appeared on the AS website you'd all be citing it as gospel.. strangely, it appears on a well respected hardware website that is run, edited, and produced by some of the leading experts in the field, and it's suddenly a pile of crap advice and you all complain.
i'm gona stick to the rice/pea method, get me iree, and hold it daaaahn blud
jacob know what best for he heatsink ya kno?
(apologies, wanted to add some light-hearted comedy to this otherwise stress-filled thread. in my eyes, some of you need to get out more, just a thought)
Originally Posted by Pookeyhead Taking the 5 seconds required to smear some paste over your CPU isn't worth the chance that your temps will be lower, even if only by 1 degree?
How lazy are you? It's not as if the article suggests undertaking 5 hours work or anything... jeez!
Don't be too provocative. I said your logic is right as in the above and moreover I didn't mention that conventional squeezing method works better than the method described in this article. Read carefully what I wrote again and plz don't do overact.
Your argument maybe along with the logic behind this article, but as I said we cannot say it will hold after installing heatsink. Why don't you try to apply your physics laws and see what happen after mounting it? I'm majored physics and hence understand your point well enough but I still can't sure it will be effective in the end. I strictly mention again. You and bit-tech's approach work great before mounting it. But after do that, it becomes meaningless due to the surface conditions of the heatsink base and unavoidable movements during the mounting procedure.
If you still keep wanting to argue your firm, god-tone logic, plz give me counter-evidences of the problem; applying your idea to the whole procedure including not only pasting TIM but also installing heatsink. I don't think my logic has some pitfalls either, like you. :( Without that, I don't think Bit-tech finishes the job, and also, your saying.
In theory i think the "blob" method will give you a better spread. Here's why:
If you spread out the paste on the cpu before applying the heatsink, there will be little pockets of air between the thermal paste and the heatsink at the first moment the two touch. These pockets are caused by imperfections on the side of the HSF, but much more so on the fact that the "finger method" does not leave an even finish on the top.
The "blob" method also has a lot of air between the Heatsink and the IHS.
At this stage, both the "finger and the "blob"method have a lot of air to push out. However, the blob method offers a free escape path for this air, whereas the finger method needs to push this air out through parts where paste is already in place. It would need to release this air much in the same fasion that you and i release unwanted air from our biological system from day to day :D
[mode=Dr.Evil]I shall call this release of air the "fingerfart". [/mode]
Thus, on a thoretical level the blob method is far superior. However, this assumes that neither your heatsink nor the IHS is convex or concave, and that your heatsink mounting movement is perfectly vertical, and that the viscosity of the thermal paste is perfectly evenly distributed, and that the blob has no air pockets in it. If you fail to meet one or more of these requirements, the only valid conclusion can be:
Quote:
Originally Posted by Brulath Why I care is beyond my comprehension.
WoW I would take this article down, you dont want to mis-inform people on how to apply TIM correctly. The rest was decent but the application of TIM is way off.
+1 for unnecessarily complicated recommended method - not necessarily worse but just not anymore likely to get better results.
My theory:
Surface of CPU and HSF are uneven, TIM is there to fill in gaps. Therefore logic dictates that using another uneven surface (i.e. your finger) to spread the TIM will not fully compensate for previous unevenness. Only logical solution is mild oversaturation of TIM between surfaces (to fill in gaps) and an application method that forces air outwards. I.e. the blob in middle method.
Originally Posted by Meanmotion Surface of CPU and HSF are uneven, TIM is there to fill in gaps. Therefore logic dictates that using another uneven surface (i.e. your finger) to spread the TIM will not fully compensate for previous unevenness. Only logical solution is mild oversaturation of TIM between surfaces (to fill in gaps) and an application method that forces air outwards. I.e. the blob in middle method.
yup. Your theory weighs better than that of the article. Maybe the amount of a grain of rice is not sufficient to fill in any possible curvatures between the two surfaces. Certainly, this article looks like misguiding naive people... just my thoughts.
I can see reason behind trying to force the air out with the blob and smoosh method, but it's inconsistent. The method in the article is much more repeatable, and doesn't depend on getting the right amount not to flow out the sides or cover inadequately.
This was the worst part about building for me. I was so nervous when this part came in building my PC, because it was my first time, that I got it virtually everywhere around the socket.. I just really didn't want to put too much or too little on and completely screw up the parts I'd just bought. I used my finger in a plastic bag, which gave me plenty of control to deal with it, though, and It was fine in the end. Shame my RAM had to then bugger up down the line.
Originally Posted by Tim S Looks like they've applied far too much paste. Even the amount of paste in #4 is more than we've used for the method described in our article which I guess is represented the closest by method #1. However, the difference is that we're not using a razor blade or credit card. :)
I have to agree with Tim, the article cpemma links to does not cover the spreading of the TIM in the same method as the article on bit so you can't effectivly compare them.
I've always used the bit article method and never had an issue, small tiny blob spread very thin and evenly across the heatspreader/core then clamp down the heatsink.
However, I don't imagine there will be a great deal of difference in thermal values based on any of the methods cpemma links to or the bit article one.
I've always used the method in the article too. Apply a small blob and work it in. I've tried the method of just putting a blob on there then applying the heatsink, and while I've never had any problems, there's always part of the IHS which isn't covered. To me, even the smallest part without TIM on it just isn't efficient.
At the end of the day, most methods work alright, but if it really bothers you then do it your own way. As long as that way isn't like the Macbook (or whatever model it was) method where almost an entire tube was used per chip :)
In all honesty I've used many different methods for applying TIM including the 'blob' method and the one featured in the article and the difference in temps is non existant.
The only real rules to follow are to ensure both the heatsink and cpu are clean before application, and don't apply too much paste - everything else just comes down to preference.
Comments 51 to 75 of 109
ReplyThis is just epic!
It also demonstrates how little the difference is between all these methods, and how rediculous these guys with their "nanotube compounds" are. Soon there will be another company selling us nanotube re-alignment kits, and we'll be writing articles and discussing the best methods of using them to make that last 0,05 degree difference, so we can up our CPU by 0.1 Hertz
Seriously though, getting it at least marginally right is important for beginning system builders, and for that, this article gives some good advice. It's a method that's sure to work, and easy to do, with little chance for serious error. Not to mention you solved Pygo's problem! ;)
1. Even though the grease spreads evenly by careful scraping ON THE SURFACE OF IHS, It will not spread evenly if one mount a heatsink because of poor surface(microscopically) condition of the heatsink base. According to bit-tech, they recommended that only the amount of a grain of rice is appropriate. When you try this, it means that the thickness of grease will be only tens of micrometer (if you have a doubt, then calculate it with proper dimension). This thickness will not be sufficient to compensate and fill in any possible cracks, which may be deep up to 0.1 milimeter, on the base of the heatsink unless heatsink are produced in the finest quality. As a result, air can be trapped in which the grease can't fill.
2. Evenness of grease will be screwed up after you mount the heatsink. The technique that Bit-tech has introduced is somewhat idealized and abstract. The reality is that you cannot avoid any little movements and slides between the two surfaces when you mount the heatsink. In this sense, manually spreading method loses its advantages. Microscopically, some parts of grease will be slightly amassed due to this movements and this spoils the evenness of grease.
It's always an idealized scenario with these kinds of things, because in an ideal world your CPU would come with a heatsink on it already.
As for fitting instructions, anything beyond the obvious manufacturer's recommendations would be immensely hard to write about over different sockets etc. Be realistic. :p
Just personal experience, mind.
I'm just following the laws of physics. The greater the area of contact between two surfaces, the greater the thermal transfer between them will be. Therefore ensuring the entire surface has a coating of the stuff at an appropriate thickness BEFORE you apply the heatsink has got to be better than applying a blob and HOPING it covers the entire surface AFTER you apply the heatsink. Think about it. You can either apply a grain of rice sized blob and hope the heatsink pressure spreads it evenly, or you can apply the same size blob and make sure it covers the entire surface manually before you apply the heatsink. The absolute worse that can happen is there will be no difference whatsoever... the best that can happen is you will get better heat transfer.
I challenge anyone to argue with that logic.
I couldn't give a monkey's what arctic silver say.
Like you say.. the majority will be directly over the chip, yes... the majority.. not all. Ensuring the entire heat spreader is coated will ensure ALL of the heat is transferred, especially as many third party HFS have a surface as large, or larger than the chip.
Again... challenge the logic if you must.
hm... no one can deny this fact. :)
Right. However, virtually nobody has problems with the temperature by doing whatever methods they'd like to do. What we should pay attention is that the differences in tiny scales. Of course, not many user would concern this but the problem is that this article deals as like this methods is a recommended one, though nobody can sure it worths such recommendation.
Taking the 5 seconds required to smear some paste over your CPU isn't worth the chance that your temps will be lower, even if only by 1 degree?
How lazy are you? It's not as if the article suggests undertaking 5 hours work or anything... jeez!
So far as I'm concerned it IS a recommended method... just by applying logic and physics.
I stress again: The worst that can happen is no improvement at all... and at best, an improvement.
Why would you NOT want to do it? You're all just being stubborn on some principle, or because the god like figures at Arctic Silver haven't told you so. If this info suddenly appeared on the AS website you'd all be citing it as gospel.. strangely, it appears on a well respected hardware website that is run, edited, and produced by some of the leading experts in the field, and it's suddenly a pile of crap advice and you all complain.
LOL
SHeesh.. people huh?
jacob know what best for he heatsink ya kno?
(apologies, wanted to add some light-hearted comedy to this otherwise stress-filled thread. in my eyes, some of you need to get out more, just a thought)
Don't be too provocative. I said your logic is right as in the above and moreover I didn't mention that conventional squeezing method works better than the method described in this article. Read carefully what I wrote again and plz don't do overact.
Your argument maybe along with the logic behind this article, but as I said we cannot say it will hold after installing heatsink. Why don't you try to apply your physics laws and see what happen after mounting it? I'm majored physics and hence understand your point well enough but I still can't sure it will be effective in the end. I strictly mention again. You and bit-tech's approach work great before mounting it. But after do that, it becomes meaningless due to the surface conditions of the heatsink base and unavoidable movements during the mounting procedure.
In theory i think the "blob" method will give you a better spread. Here's why:
If you spread out the paste on the cpu before applying the heatsink, there will be little pockets of air between the thermal paste and the heatsink at the first moment the two touch. These pockets are caused by imperfections on the side of the HSF, but much more so on the fact that the "finger method" does not leave an even finish on the top.
The "blob" method also has a lot of air between the Heatsink and the IHS.
At this stage, both the "finger and the "blob"method have a lot of air to push out. However, the blob method offers a free escape path for this air, whereas the finger method needs to push this air out through parts where paste is already in place. It would need to release this air much in the same fasion that you and i release unwanted air from our biological system from day to day :D
[mode=Dr.Evil]I shall call this release of air the "fingerfart". [/mode]
Thus, on a thoretical level the blob method is far superior. However, this assumes that neither your heatsink nor the IHS is convex or concave, and that your heatsink mounting movement is perfectly vertical, and that the viscosity of the thermal paste is perfectly evenly distributed, and that the blob has no air pockets in it. If you fail to meet one or more of these requirements, the only valid conclusion can be:
My theory:
Surface of CPU and HSF are uneven, TIM is there to fill in gaps. Therefore logic dictates that using another uneven surface (i.e. your finger) to spread the TIM will not fully compensate for previous unevenness. Only logical solution is mild oversaturation of TIM between surfaces (to fill in gaps) and an application method that forces air outwards. I.e. the blob in middle method.
My tuppence.
yup. Your theory weighs better than that of the article. Maybe the amount of a grain of rice is not sufficient to fill in any possible curvatures between the two surfaces. Certainly, this article looks like misguiding naive people... just my thoughts.
Not.
I can see reason behind trying to force the air out with the blob and smoosh method, but it's inconsistent. The method in the article is much more repeatable, and doesn't depend on getting the right amount not to flow out the sides or cover inadequately.
I've always used the bit article method and never had an issue, small tiny blob spread very thin and evenly across the heatspreader/core then clamp down the heatsink.
However, I don't imagine there will be a great deal of difference in thermal values based on any of the methods cpemma links to or the bit article one.
At the end of the day, most methods work alright, but if it really bothers you then do it your own way. As long as that way isn't like the Macbook (or whatever model it was) method where almost an entire tube was used per chip :)
The only real rules to follow are to ensure both the heatsink and cpu are clean before application, and don't apply too much paste - everything else just comes down to preference.
Okay great but what movie is the ticket stub for? It won't work if it's a chick flick because the tears contaminate the stub.
-
« Previous
-
1
-
2
-
3
-
4
-
5
-
Next »
Discuss in the forums