Thermal Stress Test All MacBooks for Peace of Mind

Question

Level 1

44 points

Thermal Stress Test All MacBooks for Peace of Mind

Given the now documented propensity of some MacBooks to melt the insulation of a thermal sensor wire and develop random, intermittent shutdowns after a month or two of normal usage, I think it advisable for all MacBook owners to test their machines and expose any weakness while it is WITHIN warranty. A newly purchased MacBook probably should be thermal stress tested well before the 14 day replacement period expires.

Opening terminal and running yes in two windows maximizes CPU utilization and temperature. Several overnight runs or a 24 hour run without arranging extra ventilation should expose any MacBooks that are physically prone to melting their thermal sensor wire insulation. The goal of the run is to heat the heatsink to as high a temperature as it will ever experience and hold it there for several hours. If the insulation holds, it should continue to do so. If it melts, the machine will start experiencing intermittent shutdowns during normal use.

A few days of normal usage after that session would give good peace of mind that the machine is less likely to fail later. Intentionally thermal testing early gives owners a better chance to have a fault prone machine repaired under warranty or replaced.

This would also be useful for newly repaired MacBooks prior to putting them back into critical usage.

macbook, Mac OS X (10.4.7)

Posted on Oct 7, 2006 8:16 AM

Reply

Answer 1

RADe8

Level 4

1,342 points

Oct 7, 2006 8:48 AM in response to guykuo

Several
overnight runs or a 24 hour run without arranging
extra ventilation should expose any MacBooks that are
physically prone to melting their thermal sensor wire
insulation.

While we're at it, let's "stress test" the LCD. Several swift but forceful cracks with a standard ball-peen hammer should also expose those Macbooks that are physically prone to LCD damage.

Reply

Answer 2

MRCUR

Level 2

426 points

Oct 7, 2006 9:04 AM in response to RADe8

Haha, I have to agree. No one should be "stress testing" their macBooks like you have said to do. It isn't smart. No MacBook under normal use would have this happen to it, If RSd becomes and even larger problem, Apple will probably start an extended repair program as they have done in the past.

-MRCUR

Reply

Answer 3

guykuo Author

Level 1

44 points

Oct 7, 2006 9:19 AM in response to MRCUR

I think computer should be able to "withstand" running its CPU without destroying itself. A laptop is a computer. I expect it to be able to handle any CPU intensive task like rendering overnight. The yes command is a convenient means of simulating a high CPU load application. This is not an abuse of the machine like hitting it with a hammer, but a simulation of high usage.

Why should MacBook owners gingerly treat their CPU so it never gets warm? That is akin to stating that the notebook is only rated for light web browsing and word processing, but a video encode isn't advisable because it could damage the machine. No. That isn't sensible.

On the other hand, delaying the eventual heating of the system means a MacBook owner may delay onset of RSD until after their warranty expires. It's better to expose any potential fault and have it fixed under warranty than delay its appearance and have it repaired at the owners expense. Well... better for the owner anyway.

Reply

Answer 4

Oct 7, 2006 9:50 AM in response to guykuo

One of the biggest problems, that I can see, with your argument is that notebook computers are designed for convenience more than heavy use. Just look at the difference in ventilation and the size of the fans in a PowerMac G5 or MacPro vs any notebook computer, particularly the MacBook with it's emphasis on form factor.

Heat is always going to be a concern in a notebook and there is a compromise made in the design process between heat and size/weight. The goal to provide enough heat dissipation for normal use, not intentional overheating.

additionally,
Because there is a compromise, you can never satisfy everyone's needs. If Apple were to design the MacBook to handle overnight or 24 hour renders with no additional ventilation provided by the user, it would mean they'd either:
a- Slow down the processor (people complain computer is too slow)
b- Put in a bigger/more powerful fan (people complain computer is too loud/too big)
Thus the choice falls to the user.. Those that want to use their computer for continuous CPU intensive tasks should either get a computer designed to handle that type of workload-like the MacPro- or take steps to ensure that their computer remains cool during the process- using it in a cool (air conditioned) room or providing extra airflow around it.

Reply

Answer 5

guykuo Author

Level 1

44 points

Oct 7, 2006 9:48 AM in response to Mike-N-nahyunil

It should not be an issue if the MacBook's has the new heatsink and thermal sensor system. After all, the machine does step down the CPU when there is a high load. If we had to intentionally keep all tasks on notebooks "light," that would be tantamount to prohibiting all CPU intensive tasks like video compression, 3d rendering, hires video playback. Eventually, everyone is going to do some CPU intensive tasks and if the thermal sensor wire happens to be one which is prone to melting the machine will fail. My point is that it should NOT fail.

Saying this type of test is dangerous also implies that the MacBook should never be used for any CPU intensive tasks for prolonged periods. Saying that CPU intensive tasks are okay also means that yes commands in terminal are also safe. Can't have it both ways.

Reply

Answer 6

Oct 7, 2006 9:50 AM in response to guykuo

This post has got to be removed now. This is not a good idea and the damage that it causes will probably not be covered by Applecare.

When was the last time that your laptops 2 processors were at 100 percent for several days straight? The last time they ever worked again.

Reply

Answer 7

Oct 7, 2006 9:56 AM in response to guykuo

I have done some of those CPU intensive tasks (video/3D rendering) on the MacBook.

What is dangerous is the step of doing it without providing additional cooling/ventilation.

"Several overnight runs or a 24 hour run without arranging extra ventilation"

Reply

Answer 8

guykuo Author

Level 1

44 points

Oct 7, 2006 9:54 AM in response to thecompkid

Yes, doesn't keep the CPU's pegged at 100%. The MacBook steps down the CPU when the temp goes up. This really isn't that different from running a large ray trace on the MacBook.

Is anyone really meaning that the MacBook should categorically be kept from all long, CPU intensive tasks? I hope not.

Reply

Answer 9

guykuo Author

Level 1

44 points

Oct 7, 2006 10:03 AM in response to Mike-N-nahyunil

You're right. I should clarify that portion.

What I mean by "without arranging extra ventilation" is that no extra external means of cooling added like a fan or AC blowing across the MacBook. This definitely does NOT mean nor imply intentionally smothering the machine.

The machine's own heat dissipation system should be allowed to work. That way the machine gets to its actual maximum working temperature.

Reply

Answer 10

Willett58

Level 2

220 points

Oct 7, 2006 11:48 AM in response to guykuo

I agree this is not a good idea!. I was considering dunking my macbook in the bathtub to check it's water resistance. Maybe try the 25' drop test onto concrete, any other ideas? I appreciate suggestions to help with the RSD, but trying to make the computer fail is not a good idea.

Reply

Answer 11

tcdm1972

Level 2

240 points

Oct 7, 2006 11:50 AM in response to guykuo

I have to agree with guykuo. I should be able to run my MacBook at full load overninght (8hr or so) without any malfunction. I'm able to do so with any desktop - why should my MacBook be any different. Notebooks are already underpowered compared to desktops - thus treating like a normal PC shouldn't be an issue.
I will not try this test - as I am personally satisfied with 30 min or an hour of running CPU at 100%. But bottom line is - I didn't buy a criple - I spend the extra cash for a product that I should be able to trust!!

Reply

Answer 12

Oct 7, 2006 12:56 PM in response to tcdm1972

why should my MacBook be any different

Maybe because your desktop has a big fan, large heatsink, and plenty of space to allow air to circulate.

Again, you can't really expect a notebook to work like a desktop... There's a reason that Desktop PCs are as big as they are and cooling has a lot to do with it. If you are going to stress your MacBook for extended periods trying to make it fail, you'll probably succeed.. But for those that want to keep their computers working without problems, the goal should be to keep it as cool as possible.

This means if you are going to work it like a Desktop, you should provide cooling comparable to that of a Desktop (particularly greater airflow around the case with the assistance of a fan, and air conditioning to keep the ambient temperature low to further assist heat dissipation).

Note:
I don't know what type of desktop computer you have, but my PowerMac G4 has a 4" fan blowing across the RAM and CPU which has a 3x3x2" heat sink. There is also a 2" fan pulling hot air from the power suppy and pushing it across the PCI cards and out the back of the case. Compare that to the minimal internal space of the MacBook and I think it's pretty clear that there should be a difference in use expected.

Note2:
To restate the point, a MacBook can be used under load for extended periods of time, but it is up to the user to take steps to keep the computer as cool as possible to avoid damaging the internal components. This means not pushing it in hot climates and providing additional, artificial cooling when necessary.

Reply

Answer 13

cazabam

Level 1

75 points

Oct 7, 2006 1:16 PM in response to Mike-N-nahyunil

One thing nobody else has mentioned is that if you run the 'double yes' test overnight, or do other things to deliberately drive the machine to the hottest it can go, then it WILL shut down. This is the thermal cut out and it's there to prevent damage - I bet people will try this, it WILL shut down and they WILL think that have RSD. IT ISN'T. All laptops have a thermal cut off where it will protect itself by shutting down cold. It is NOT RSD.

This topic is dangerous because it will definitely give false positives, and it may actually cause physical damage to the machine.

Reply

Answer 14

Oct 7, 2006 1:23 PM in response to cazabam

One thing nobody else has mentioned is that if you
run the 'double yes' test overnight, or do other
things to deliberately drive the machine to the
hottest it can go, then it WILL shut down.

This is not my experience. Simply running the double yes test for 10 minutes will show that the fan ramps up, the temperature hits a maximum and then decreases to a lower steady-state value for as long as you leave the processes running. For this reason, I think both that the stress test is useful to verify proper operation and that it shouldn't be necessary to run the test for hours to do so. Unless you want to test if your fan fails or some such.

Edit: Actually the point of the original post IS to test the "some such", e.g. sensor and/or wires that may be in contact with parts that are hotter than the temp readings indicate. That seems a valid point.

Reply

Answer 15

guykuo Author

Level 1

44 points

Oct 7, 2006 3:50 PM in response to guykuo

You aren't forced to do the test, but here is my logic.

1. The early MacBooks have a heatsink/thermal sensor wire assembly that is prone to shorting if the insulation on the wire melts. Once the machine melts through the insulation during a high temperature episode, the machine can begin unpredictable, sudden shutdowns due to an intermittent short of the wire.

2. Once the insulation is compromised, the insulation doesnt heal itself.

3. Not all early MacBooks will develop this problem, but over time the risk of it appearing increases as the machine eventually will do some tasks which cause it to heat up. Each heating episode may further degrade the insulation.

4. The failure can occur after months of unremarkable usage. One may well use the machine without it happening until after the warranty expires.

5. Keeping a machine cooler by altering the firmware to make the fan come on sooner will slow the insulation degradation but does not actually make the wire insulation more resistant to heat. It only slows the onset if a machine is destined to have that failure. This is advantageous to Apple because the total number of failures within the warranty period should decrease and the failures spread out over a longer period of time.

6. Running Yes heats up the machine, but no more so than any other heavy CPU task like video compression or 3D rendering. The machine is designed to slow down the CPU and increase fan speed so running a heavy CPU usage application doesn't cause thermal runaway.

7. No all users will do enough heavy CPU tasks to thermally stress the wire insulation, but are likely to eventually do something that heats up the machine. It isn't reasonable to make users responsible for limiting their MacBook only to low CPU usage tasks.

8. An out of warranty repair is expensive.

9. A random shutdown event can cause the loss of hours of work even if one saves frequently. Imagine what happens if the machine is in the process of writing out a file and the sudden shutdown occurs. If it is overwriting the same file, both your original and you new work are destroyed. If it is during a file directory write, the directory structure could be munged. You usually get away without too much damage, but RSD events become more and more frequent once they begin and soon it becomes impossible to use a machine reliably.

10. I think it better to know a machine is prone to a failure before warranty coverage is over. Once the flaw is exposed a permanent fix can be done to prevent the problem and one can trust the machine. If a machine fails after a thermal load test, then it was probably destined to fail at a future date anyway. If it passes without incident, then you know your MacBook is less likely to fail in this manner.

Reply