Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
kainjow said:
Uh, I thought one of the huge selling points of the Core Duo was that it runs much cooler? :confused:
Maybe you were thinking of "performance per watt"? That's one of the Core Duo's selling points...
 
kingtj said:
This was discussed earlier and on other forums - but service techs. who work on other brands of notebooks reported seeing similar, incorrect instructions for application of thermal paste in their manuals too. This isn't just an Apple goof-up.

There are also quite a few reports on the Apple message forums from people with overheating Macbook Pros who removed the excessive thermal paste, re-applied a proper amount, and *still* saw very little temperature change.

I'm not trying to defend Apple here, or claim that "too much thermal paste" isn't an issue.

I'd like to make that claim, though.

In the much quoted arctic silver manual the image that has been shown in this thread is not for the Intel type processors with small heat sinks. It's for the large AMD type heat sinks.

For processors with small heat sinks, they recommend a fairly thick layer of goo spread out on the pad.
as2closeup.jpg


And the only reason they give for not applying too much goo in those situations is: "Even though Arctic Silver thermal compound is specifically engineered for high electrical resistance, you should keep the compound away from processor, memory, and motherboard traces and pins. There is a possibility that dust or metal particles and/or shavings carried by the airflow inside the computer case could contaminate the compound and increase its electrical conductivity."

Which is good advice for home service, but may not apply as much for more sterile places, and in either case the quoted concern is not reduced heat conduction.

Even the relatively small amount that is applied in the picture above will squish out over the sides, and since excess goo will also be squished out if you apply much more goo, the thermal conductivity will be the same.

I'd like to call myth busted. Let the flaming begin.
 
gekko513 said:
I'd like to make that claim, though.

In the much quoted arctic silver manual the image that has been shown in this thread is not for the Intel type processors with small heat sinks. It's for the large AMD type heat sinks.

For processors with small heat sinks, they recommend a fairly thick layer of goo spread out on the pad.
as2closeup.jpg


And the only reason they give for not applying too much goo in those situations is: "Even though Arctic Silver thermal compound is specifically engineered for high electrical resistance, you should keep the compound away from processor, memory, and motherboard traces and pins. There is a possibility that dust or metal particles and/or shavings carried by the airflow inside the computer case could contaminate the compound and increase its electrical conductivity."

Which is good advice for home service, but may not apply as much for more sterile places, and in either case the quoted concern is not reduced heat conduction.

Even the relatively small amount that is applied in the picture above will squish out over the sides, and since excess goo will also be squished out if you apply much more goo, the thermal conductivity will be the same.

I'd like to call myth busted. Let the flaming begin.



Hmm. And I just bought arctic silver. Waiting for more proof now.
 
gekko513 said:
I'd like to make that claim, though.
Even the relatively small amount that is applied in the picture above will squish out over the sides, and since excess goo will also be squished out if you apply much more goo, the thermal conductivity will be the same.

I don't think the heatsink (just looking at the way it is installed--as in not coming down perfectly straight) does not apply enough pressure to the diode to spread it evenly enough. Looks like it sorta just sits on top. Just given many pictures that have been taken already of an incredibly think amount of grease layered on there unevenly after disassembly... There's a couple in this thread as well as this one (https://forums.macrumors.com/threads/199840/)

@macenforcer... ONE way or another, that arctic silver is extremely high quality.. you will notice an improvement even if it is "myth busted".
 
gekko513 said:
I'd like to make that claim, though.
[...]
Even the relatively small amount that is applied in the picture above will squish out over the sides, and since excess goo will also be squished out if you apply much more goo, the thermal conductivity will be the same.[/COLOR]

No, the thermal conductivity will not be the same. The problem is that the some of the heat generated by the CPU will be conducted to the excess paste instead of the heat sink. The excess paste will thus heat up, but it has no way to disperse this heat but radiation, warming up the air around it. The heat transferred to the excess can not be dissipated anywhere near as efficiently as the heat that is conducted to the heat sink and then air-cooled.

As for your other comment regarding the Arctic Silver manual: I think you are reading it selectively, or not interpreting it correctly. Indeed they recommend keeping it away from certain parts to prevent short-circuits. However, that does not mean, as you suggest, that AS says there are no thermal disadvantages to excess application of thermal paste. There are, which is why the manual explicitly states "On the CPU, Arctic Silver thermal compound should only be applied to the top of the core". Furthermore, a bit further in the manual they recommend that this layer should only be 0.003" to 0.005" thick. AMD seems to agree with this, and I would expect they know a thing or two about cooling CPUs.
 
mkaake said:
you know, rants like this don't bother me too much.



as long as the person making the rant actually has the computer.

well, I'll get flamed for saying that, but that's my take on it. There are people who go digging around the web to find problems with other people's computers, and then complain about it.

This is called prepurchase research.

And yes, actually everybody does this, especially if it is for a laptop, unless you are Bill Gate's son, then perhaps you will buy a specially custom ordered diamond plated Thinkpad from Leveno preinstalled with the latest Windows Vista on it.

People who are keen on getting the laptop, but come across information like these floating around the net, coupled with negative experiencs encountered by their peers who purchased the earlier model Macbook Pros, are not going to be buying this new MB if they are remotely sane.

Sure, Apple may market to idiots, the bottom 4% of the population in fact, but it will just be a matter of time before these same idiots start whining on these very same forums unless Apple's engineers decide to grow a brain in their collective asses and realise that not only is a glob of thermal paste not innovative at all but is actually detrimental to the user experience.
 
Just to compare. My Sony laptop runs at 35-45C, the fan rarely comes on and the case is cool to touch. The same is with MBP's when new thermal paste is applied.

Too much paste causes the die to get very hot as the HSF does not draw away the heat. The heat builds up inside and dissipates through the case rather than the HSF.

A lot of paste is very well documented indeed.
 
I just reapplied Arctic Silver grease in my PowerBook G4 and it's made a noticable difference in temperature. Right now (with just Firefox, Dashboard, Finder, Temperature Monitor, iTunes, and Adium running), it's at about 35-37ºC. With a few more processor intensive apps running, it runs at about 40-45ºC. If I open 20-30 apps (including F@H so that the CPU stays at 100%), I can reach as high as 50-52ºC.

Before making the change, just surfing the net and chatting online would run the CPU to around 40-45ºC. Right now, the "Processor Bottomside" reading is 33.2ºC.
 
Heb1228 said:
I don't get it... why and how does this thread make it "official"? :rolleyes:

Apple listed in their official service manual to apply too much thermal grease. That's what makes it official.
 
odedia said:
Apple listed in their official service manual to apply too much thermal grease. That's what makes it official.
We've known that for a few weeks at least now. Official would be Apple admitting there was a problem with it. And that's probably not going to happen. So there will never be an official thermal paste thread.
 
all these thermal paste saga make me think that portable Mac don't really have an elegant solution to the heat problem after all.

with almost all other components updated / refreshed... e.g. Glossy screen, new keyboard, magnetic power cord, easily accessible HD / RAM replacement, magnetic latch etc...

the very last thing that is still left unsolved is the heat problem....

i hope to see this problem going away by early next year when 'cooler' chips are used.
 
kingtj said:
This was discussed earlier and on other forums - but service techs. who work on other brands of notebooks reported seeing similar, incorrect instructions for application of thermal paste in their manuals too. This isn't just an Apple goof-up.

There are also quite a few reports on the Apple message forums from people with overheating Macbook Pros who removed the excessive thermal paste, re-applied a proper amount, and *still* saw very little temperature change.

I'm not trying to defend Apple here, or claim that "too much thermal paste" isn't an issue. But what I am saying is, I doubt it's as big a deal as some people are making out of it. Apple's recent SMC BIOS update changes around the behavior of the cooling fans on Macbook Pros it has been used on. Sounds to me like Apple just did a software "work-around" for the problem, telling the temp. sensors to turn up the fans when they read lower temperatures than the threshhold they were programmed to trigger at before.

(Most of the people complaining about this thermal paste problem were arguing not so much that the CPU isn't receiving enough cooling and might "burn out" or freeze up - but rather, that the CPU, the video chip and bus controller chip are *all* touching a metal plate that has sensors on it that try to figure out the overall temperature of things. If the thermal paste is causing the sensors to see things 5 or 6 degrees cooler than they really are, due to heat transfer not getting to the metal plate quite so well - then a recalibration of the sensors by that amount in software would seem to get things back to the intended mode of operation.)

Stop it, you're making too much sense!!
 
I'm curious to see some numbers (i.e: before and after core temps with the new paste) on this for the MacBook (not pro). I expect the heat dissipation in the Pro is better because even the conductive metallic case will help the process (also why it was so damn hot to the touch, I suppose). So to see someone hitting 40 on the Pro doesn't surprise me, but I wonder what "normal" should be on the MacBook.

My MacBook has been running lower temps (45-52 at idle, light workload), and has never gone above 75, although I'm thinking about throwing more at it to see what will happen.

The case gets warm, probably warmer than my old G3 iBook (haven't tried them together yet), but that probably shouldn't come as a surprise...

Still very interested, though. I don't really want to buy Applecare, but if this saga continues and starts to look like a fatal problem for some users, I might take the plunge.
 
stefan15 said:
GREAT RESOURCE FROM ARCTIC SILVER
http://www.arcticsilver.com/arctic_silver_instructions.htm

Note the size needed:
p4_as_dab1.jpg

That's really not an even comparison, as that CPU has a heat spreader. I believe you want this photo instaed (from that Arctic Silver page).

as2glob.jpg



NOTE: You don't actually have to spread the grease, you just have to put the bead of it on the die, then firmly place your heatsink on top, being careful not to lift it again once you've put it on (doing so would create air bubbles in the grease). Placing the heatsink firmly on the bead of grease will naturally spread it perfectly to size.

According to the Arctic Silver article you ARE supposed to spread it. Like this:

as2spread.jpg


Quote:
Spread the Arctic Silver thermal compound over the CPU core as shown in the photo to the right. The small amount from the photo in step 5 above has been carefully spread over the top of the core using a single edge razor blade. A razor blade or the clean edge of a credit card can be used as the application tool. You may use whatever tool you choose as long as it is CLEAN and allows you to control the application area and thickness.
 
Thanks for the heads up Seafox.
Although I did mention you can spread it with a plastic bag. Your picture is obviously more accurate. Even so, that's a very very thin layer.
 
stefan15 said:
Thanks for the heads up Seafox.
Although I did mention you can spread it with a plastic bag. Your picture is obviously more accurate. Even so, that's a very very thin layer.

Indeed, the instrcutions specify "the thickness of a sheet of paper". :eek: If the CPU core is manufacturer polished smooth (like a mirror) it's supposed to be translucent.

The spreading with a plastic bag instructions are actually a little old (as they only apply to Arctic Silver 1-3), it seems you had to apply it to both the CPU and the heat sink and then put them together (rather like contact cement if you're ever used it). But the Artic Silver 5 instructions say to only apply it to the CPU, so maybe they improved the formula.

I like this part of the AS 1-3 instructions:

Arctic Silver Instructions said:
[Step 5.] For Arctic Silver 1, 2 and 3:
Determine what area on the base of the heatsink will contact the CPU core once the heatsink is mounted. Twist the plunger in the syringe barrel at least 1/2 turn to ensure that it is free. Squeeze enough Arctic Silver thermal compound onto the center of this area to create a small mound as shown in the photo below on the left.
Do not apply the Arctic Silver thermal compound directly to the CPU.


tint1.jpg


Wow, that glob size looks familiar...

On the small core CPUs, use a clean razor, the clean edge of a credit card, a clean knife, or some other appropriate clean tool to pick up a small dab of the Arctic Silver compound from the mound you put on the heatsink as shown in the photo above on the right. Put the dab of Arctic Silver thermal compound you removed from the heatsink base onto one corner of the CPU core as shown in the photo below.

tint2.jpg


So in other words, almost none of what we just squirted out will actually go on the CPU!

The instructions tell us to spread the rest of that huge glob all over the surface of the heatsink, then wipe it off (this leaves a thin film that fills in surface imperfections supposedly). It just looks like a good way to get you to use over 10x the thermal paste you need per job.
 
I'm so sick of hearing about this. If you don't like it, go get a PC that runs cooler. We all know first gen's of anything have problems. What first gen of a car has never had any problems? None. If you really are getting so sick of it, stop complaining and buy a Dell.
 
anti-pcs said:
I'm so sick of hearing about this. If you don't like it, go get a PC that runs cooler. We all know first gen's of anything have problems. What first gen of a car has never had any problems? None. If you really are getting so sick of it, stop complaining and buy a Dell.

Enough with this blind worship of a financial motivated company. If Apple are having issues they are ignoring, people WILL complain and rightfull so. The macbook is the third apple laptop with core duo chips having the SAME problem. It appears from the previous poster that this problem has gone all the way back to the days of the G4 laptops. We WILL talk about end until the it is resolved. Deal with it or read other posts.
 
JamSandwich said:
I'm curious to see some numbers (i.e: before and after core temps with the new paste) on this for the MacBook (not pro). I expect the heat dissipation in the Pro is better because even the conductive metallic case will help the process (also why it was so damn hot to the touch, I suppose). So to see someone hitting 40 on the Pro doesn't surprise me, but I wonder what "normal" should be on the MacBook.

My MacBook has been running lower temps (45-52 at idle, light workload), and has never gone above 75, although I'm thinking about throwing more at it to see what will happen.

The case gets warm, probably warmer than my old G3 iBook (haven't tried them together yet), but that probably shouldn't come as a surprise...

Still very interested, though. I don't really want to buy Applecare, but if this saga continues and starts to look like a fatal problem for some users, I might take the plunge.

You can check out some numbers here: http://mbpro.info

I also saw a very interesting post about someone picturing a thermal gun at his laptop, the before was 56C, the after was 36C. Very impressive.
 
that little rant is misdirected

The thermal paste issue cant be a design problem. I would call it a manufacturing problem. It happens in an production facility. ...but it really should be dealt with.
 
odedia said:
The macbook is the third apple laptop with core duo chips having the SAME problem. It appears from the previous poster that this problem has gone all the way back to the days of the G4 laptops.

Only the Powerbook G4. The iBook G4's used a thermal pad instead of paste, and it's a lot harder to screw something like that up (unless you use more than one! :D).
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.