Welcome to the MacNN Forums.

noskaj · May 1, 2006, 01:39 PM

http://forums.somethingawful.com/sho...readid=1864582

APPLE DIDN'T APPLY THE HEATPASTE TOO WELL

REDONING IT REDUCED THE TEMP. BY APR. 10 DEGREES!!!
SEE LINK ABOVE FOR HOW TO DO

CHEERS

ksloan2 · May 1, 2006, 02:06 PM

That's really impressive if it is true, but I don't think I dare to break open my Macbook Pro case (even though I've built my own PCs since the early 90s and am actually manical about applying thermal paste carefully myself).

Also, there aren't any step-by-step picture instructions, so I'm not sure this is a complete "solution" yet.

(And again, why the heck does the Powerbook have dozens of temp indicatiors, while the Macbook Pro has only one, and in the wrong place?)

EDIT: Oh, crap... there was a link at the top with instructions! I still don't dare to do it though. Perhaps if I get Apple Care and ask them if I void my warranty first...

Tuoder · May 1, 2006, 02:12 PM

Awesome fix, if true. I'd use some Arctic Silver 5, though.

lowdaksport · May 1, 2006, 02:57 PM

I dont want to shoot this down or anythign cuz for one that is an ungodly amount of paste they used which is definatley a factor in the heat problem. But if they used some crap on them from apple then just cleaning it and switchign to somehtign like arctic silver, or any higher quality paste, usually drops processor temps about 10 degrees anyways. On every computer i have had i have noticed around ten degrees difference on the processor but I dont know if this is different for laptops or anything or if that has any effect on the total overall temperature. Not sure if that makes sence, basically im trying to get at the fact that just switching from crappy thermal paste to higher qulaity might be the reson for the temp drops.

SpaceMonkey · May 1, 2006, 03:12 PM

http://forums.macnn.com/showthread.php?t=293777

ksloan2 · May 2, 2006, 03:45 AM

Originally Posted by lowdaksport

But if they used some crap on them from apple then just cleaning it and switchign to somehtign like arctic silver, or any higher quality paste, usually drops processor temps about 10 degrees anyways.

That's also an excellent point. I wonder what would happen if I manage to get my Macbook Pro sent to be serviced because of the heat, and include some printouts from Artic Silver's website on how to apply paste sparingly and then also include a tube of Artic Silver 5?

Do you think they would use it?

jhonizzle · May 2, 2006, 08:43 AM

i learned my lesson about the MBP "fixes"...i will wait for a recall or a software fix. i will not go through the hell of another fix.

lowdaksport · May 2, 2006, 11:14 AM

Originally Posted by eobet

That's also an excellent point. I wonder what would happen if I manage to get my Macbook Pro sent to be serviced because of the heat, and include some printouts from Artic Silver's website on how to apply paste sparingly and then also include a tube of Artic Silver 5?

Do you think they would use it?

If you request it and supply it i dont see why not. But who knows

aehaas · May 2, 2006, 06:50 PM

Differences in thermal paste will have no effect on the temperature of your computer in your lap. The only thing to vary is the Td between the chip and the heat sink on the other side of the paste. For example, it the chip is running at 20 watts heat output, steady state, then that is what you will feel from the computer.

With poor conductance of heat the temperature of the chip may be 150 C and the heat sink on the other side of the paste may be 120 C. If the heat sink gives off 20 watts of heat at 120 C. With a good conducting paste the chip may sit at 125 C with the heat sink at 120 C, still giving off 20 watts.

The chip in each case gives off 20 watts but has to do it from a higher baseline temperature with the poor conducting paste. In either case the temperature of the computer in your lap is the same as the energy given off from the heat sink, 20 watts.

aehaas

ksloan2 · May 3, 2006, 06:01 AM

Except that the report is that the fans go on more often with proper thermal paste applied, so what seems to be happening is that the heat is actually being lead through the heat pipes, onto the fans which blows it out of the case. Hence the heat goes out into the air instead of into the case and onto your lap.

Elixir · May 3, 2006, 10:33 AM

ok this is way out of my comprehenstion.

is what the guy posted a true fix? i dont see how apple could screw up what seems to be such a fundemental and usually simple task of applying the right amount of thermal paste.

than again didn't the xbox 360 screw up the same thing?

phantomo · May 3, 2006, 10:47 AM

As per my previous post, no wonder I couldn't hear the MBP fan. There aren't any hot air coming from the back hinge area unlike my PB 12" when it is hot, the fan turns on and hot air comes out at the back.

Hopefully Apple will provide a fix.

Elixir · May 3, 2006, 10:48 AM

how can i test my MBP for heat?

sometimes its hot, sometimes its not.

what is a good test?

aristotles · May 3, 2006, 01:37 PM

Originally Posted by eobet

Except that the report is that the fans go on more often with proper thermal paste applied, so what seems to be happening is that the heat is actually being lead through the heat pipes, onto the fans which blows it out of the case. Hence the heat goes out into the air instead of into the case and onto your lap.

Guys, I might not be an expert with hardware but these guys talking about their MBP running cooler are messuring the heat output from the screen hinge right? Well isn't that exactly where the heat pipes transfer the heat to for the fans to blow it out?

If they applied the paste as a thinner layer and the heat output decreased at the the fan exhaust, wouldn't that imply that more heat is now being trapped within the case now?

It seems to me that the hotter the hinge area is, the better your default thermal paste is working.

eevyl · May 3, 2006, 02:10 PM

Yeah the fans turning on more often to me is an indicative that the chips are getting actually hotter, not cooler.

The guy could just be doing it wrong, so it doesn't dissipate as well. That's nice because you get a cooler case, but the chips could be getting hot hot hot, requiring more fan work.

Until an utility to get the thermal readings of the chips is released, I won't trust this kind of experiment.

ksloan2 · May 3, 2006, 05:02 PM

I don't buy that heat gets trapped INSIDE the case. If that would happen, then the heat wouldn't reach the fans and they wouldn't turn on.

Since they now turn on more often, according to the reports, it must mean that more heat is REACHING the fans instead of the case, and that it gets properly transported out of it.

Didn't the guy say that the temp sensors where located near the fans, and not the chips? And the only heat that reaches the fans goes through the heat pipes. Since the case is cooler, and fans are working more, it must mean that better thermal paste application means that the heat pipes actually get to do their job, instead of bleeding it all off into the case.

eevyl · May 3, 2006, 07:20 PM

What I don't buy is that the heat reaches more easy the exterior of the computer than the internal sensors, no matter how many thermal paste you put wrong there is more stuff between the chips and the external case than between the chips and the thermal pipes.

I say the heat gets trapped inside THE CHIPS, a dangerous thing in my book.

Anyway, until we get to measure chip temperatures, this is highly dangerous to do on a laptop on your own,

mduell · May 3, 2006, 08:19 PM

All of Intel's chips have thermal sensors built in. Another "system" temperature may be reported (who knows where that sensor is), but the CPU temp is from the internal diode.

You can read all about it in chapter 5 of this PDF.

ksloan2 · May 4, 2006, 11:13 AM

Ok, I can't find the link now, but one of these home-fixers specifically said that he saw temp sensors (or some wiring from it) by the fans only, nowhere else. So that means that the heat pipes ARE NOT WORKING with the current application of the thermal paste, hence the fans never turn on because heat never reaches them, rather it bleeds off everywhere and the case gets hot instead (remember, the fans are on the edge of the Macbook, and the chips are in the center).

There is no ****ing chance in hell that proper thermal paste application will trap the heat inside the chip. That's just silly thinking! And even if that happens, it would be easy to see, because the Intel CPU would throttle itself down so OS X would crawl along at turtle speed, or in case of an emergency, the chip would shut itself down.

Ok, so granted, this is heresay, but imo, these are the logical conclusions.

eevyl · May 4, 2006, 11:33 AM

Been following the Apple discussions thread, now that there IS an utility to measure the chips IN DIE sensors. So far there are reports of people "fixing" the paste with high temperatures similar to those who had "untouched" MacBooks Pro.

Overall, I think it does not make a difference, the differences in temperatures seems to root in different clockspeeds (1.83, 2.00 and 2.16) and enviroment temperature, and of course crappy testing like "running iTunes Visualizer". I've been taking care of more kind of easy to follow test like launching two "yes > /dev/null" processes.

I still stand in the "no help in reapplying the thermal" camp.

breakbeat46 · May 4, 2006, 05:15 PM

So if I buy a macbookpro today, would it still have the heat and whine issues?

eevyl · May 5, 2006, 02:34 AM

I wouldn't say the heat is an "issue" more than a fundamental "flaw". The whine is an issue, and probably you will still get it with a new one, there is no official word on getting it fixed.

Nice to remember there are two kinds of whines, the so called CPU whine and the Inverter whine, the latter is indeed fixable and way worse than the first one.

Elixir · May 5, 2006, 07:04 AM

i think the inverter whine has been fixed for new models.

cpu whine however has not.

aristobrat · May 5, 2006, 10:30 AM

I exchanged a Week 11 for a Week 14 hoping that it'd be cooler.

The week 11 MBP didn't have the invertor (LCD) buzz, but the week 14 does.

Le Sigh

clbell · 08:08 PM

"I still stand in the "no help in reapplying the thermal" camp."

What would it take to convince you otherwise? BTW, here is the result after redoing the paste on mine...

http://www.pbase.com/eclecticphoto/image/59709053

The case used to get VERY hot. Now it's just slightly warm, even above the F keys. Before I did this, I never heard the fans. Now I hear then spinning at various speeds depending on the load. I can wear shorts and put this thing on my lap! If I rest my fingers on the grey rubber part of the hinge I can feel the hot air rising but there are no hot metal surfaces.

The major problem with too much paste is that the sensor on the heat pipe is far to the right, away from the chips. If a proper connection is not made then the sensor will not kick the fans on.

Notice how low the conductivity of even the best thermal paste is when compared to copper or silver. It's essentially an insulator. The compund Apple is using is most definately not as effective as Artcic Silver 5. They are piling it on. That would be a major expense! It's probably in the cheap range I listed below. This should make it clear that a VERY thin layer is critical.

Thermal Conductivity:

Silver - 430 W/mK
Copper - 400 W/mK
Arctic Silver 5 - 7.5 W/mK
Cheap Thermal Compound ~ 1-3 W/mK
Plywood - .13 W/mK

maxhedrum · May 6, 2006, 03:16 PM

Just had an odd experience with my two week old MBP.

For the first week or so it would get very hot just web surfing and installing a few apps and downloading new versions of apps and drivers. Considering all the skuttlebutt on these machines I thought I'd hold off installing actual files 'till I saw how things would shake out.

After a call to apple tech support and a trip to the apple store where the machine was purchased, the determination was that it was "within spec". After reading this and other threads on the thermal paste situation, I started doing some stress tests to see just how hot it would get under load and to see if the rumored fans would ever kick in. After three days there was no change in behavior. Then, after running it about three hours, I checked on it and found to my amazement that it was actually just fairly warm and the fans were on full. So far the fans have been coming on and the temps been reasonable since then. As I had made no changes, I'm not sure why it started functioning better.

Not sure if it's related, but there is one thing that I did run into. Early on I was having problems maintaining connection with my base station, so I did a HD verification with the disk utility and it came up with an error. After doing the repair, it didn't seem to make much difference at the time. Unfortunately I didn't document just what the error was and now can't recall it. Sorry.

MORT A POTTY · May 6, 2006, 05:21 PM

http://macbricol.free.fr/coreduotemp/

you can get CPU temps real easy. no terminal or .kext hackage... just launch the app.

mine gets 56-59 when it's burning my lap. anyone else care to report numbers?

Apple has a solution to this... "don't use it on your lap" and they insist it's NOT a laptop...

maxhedrum · May 6, 2006, 09:19 PM

With the CPU bouncing between 20% and 40%, I'm getting 51C, with SMART showing the HD at 38C. It's been working continuously with just airport and web surfing for about two and a half hours. The fans are working and while warm, lap use is not uncomfortable while wearing pants. This thing is running a lot cooler than two days ago.

Thanks for the link. It'd be nice to also have access to the readings coming from the two sensors for the fans.

MORT A POTTY · May 7, 2006, 09:44 PM

my HD is running near 50c IIRC (hardware monitor) . damn, I'll trade you

ksloan2 · May 9, 2006, 08:59 AM

Originally Posted by MORT A POTTY

http://macbricol.free.fr/coreduotemp/

you can get CPU temps real easy. no terminal or .kext hackage... just launch the app.

mine gets 56-59 when it's burning my lap. anyone else care to report numbers?

Apple has a solution to this... "don't use it on your lap" and they insist it's NOT a laptop...

It says on the application that it uses the framework from those who did the .kext, though.

However, I launched the app and didn't get a reading... then, after awhile, the app asked me for my admin password. Nope, no way. I don't trust it. Sorry.

aristobrat · May 9, 2006, 10:52 AM

FWIW, it needs your admin password to temporarily install the .kext

ksloan2 · May 9, 2006, 12:08 PM

Oh, I see. And it uninstalls and remove it when I quit the application?

EDIT: Just tried that application. It said 25C. AHAHAHA! My Macbook is twice that on the outside even! Which means, since the .kext from those Increw people seems to get a correct reading, does this mean that the CoreDuoTemp application installed a trojan when it asked for the sytem password?

****!

Voch · May 9, 2006, 12:23 PM

Seriously...is the MBP completely unusable as a *lap*-top?

Even if I always wear jeans like I do with my TiBook? I think the difference is the MBP gets heated on the sides whereas my TiBook gets warm on the bottom middle (and I can keep space between my knees for air flow).

Voch