MacBook Pro Heat Issue (A challenging problem...)

[TheRealNick]

Hi, long time reader first time poster. If this post is in the wrong section, apologies in advance and may the mods move it somewhere more appropriate. I will try to cut my issue as short as possible:

One day I realised my 2011 15" MBP wouldn't start. Through a series of coincidence, trial and error and luck I worked out that if I put my MBP on heat mat for a few minutes. (Thus gently warming the bottom of the machine) I could then press the button and it would turn on without an issue.

So, when the power button is pressed the MBP is doing a temperature check and deciding, wrongly, that the temperature is too low and thus won't turn on. Heat the unit up a bit, it does the heat test and all is well. (At least I assume that's what is happening)

Is there any way that I can fix this? Maybe turn off the heat test on startup? And if so, how? I have tried to search for this issue elsewhere and can't even find anyone else with the same issue. The Genius Bar were also stunned by this bizarre development. I wonder if many people have the issue and it simply won't turn on, but I am the only person loony enough to try heating the bottom of the unit...

Things I have tried:
- Resetting PRAM NVRAM etc
- New battery (since it needed one anyway, and this was the Genius Bar suggestion)
- Wiping and reinstalling from scratch
- Searching on google for the issue

Any help would be met with extreme gratitude.

Many thanks!

 

[BrianBaughn]

Did anyone with Apple confirm a "temperature check" before booting? I've never heard of that.

Could be a power connection is wonky and expansion from heat is completing it.

 

[TheRealNick]

Hi. Thanks for your reply.

Do you mean, for example, the pin(s) on the battery connection might be very slightly bent?

If so, this is something I did not consider, I will open up and check it out. Are there any other power connections that you could direct me to that I could check out?

 

[kohlson]

Without knowing more about the environment situation, hard to really know on the "heat check." But most consumer products have a stated operating temperature range within 0-30 C. Spinning disk drives are usually the gating items. After the battery (which I see has been replaced) I would check the disk drive.

 

[TheRealNick]

Thanks for your reply, Kohlson.

I took the bottom off... and that was more difficult than I remember! The constant heat on the bottom must have something to the screws / screw holder because my god, that was hard work!

I have an SSD in it and I seem to recall one of the first things I tried was putting the original HDD back in there which didn't solve the problem. Sorry I didn't mention in the OP, I have tried so many things it is hard to remember them all! I will also mention that I did the same with the original RAM.

 

[BrianBaughn]

For clarification, the misbehavior is that the MBP doesn't give any reaction when you press the power button?

 

[TheRealNick]

If the MBP is 'not warm enough' then yes, pressing the power button gets no reaction, at least none that I can see or hear.

 

[ZapNZs]

My guess is that your GPU has failed. Quite a few Users (myself included) have found that intentionally overheating a 2011 with a failed dGPU that will not start/will not fully complete boot often allows it to start and successfully complete the boot process.

If you took it to an AASP and asked them to perform the VST, this could confirm the issue.

 

[TheRealNick]

They did do some sort of hardware test when I went to the Apple Store though I could not confirm what sort of hardware test this was. Everything came back fine. (Except the SSD, which eh said was normal for a third party SSD)

I have also has the logic board replaced previously under the GPU recall program but that was a long time before the issue began.

What you said about heating the GPU does seem to mirror my problem though, is there any way of testing for the issue without going to an Apple Store? If it is indeed the GPU, I will continue using the unit as I am now until it dies completely, as to get the logic board replaced at this stage wouldn't really be worth the money vs a new 13" MPB which would now be sufficient for my needs. (And the GPU program has now ended)

 

[TheRealNick]

Ok, an update:

I found this article: http://tinyurl.com/m7tn87q

I am right in inferring from the above article that if my Mac is warmer, then it will try and boot from the integrated graphics card rather than from the discrete one?

Could it be that my discrete graphics card does have an issue that is not quite the same as described in the above article, but that whenever I press the power button it's actually telling the MBP not to start for whatever reason and when the MBP it's warm it is booting from the integrated card instead. As described in the above article?

I feel like I've made a few assumptions here, so if someone more knowledgeable on the subject than myself could advise whether I should try and apply the above fix and adapt it to my situation I'd be very grateful. Also, if there is any way to test the discrete graphics to see if the fix is more likely to work, that would also be awesome!

 

[ZapNZs]

They did do some sort of hardware test when I went to the Apple Store though I could not confirm what sort of hardware test this was. Everything came back fine. (Except the SSD, which eh said was normal for a third party SSD)

I have also has the logic board replaced previously under the GPU recall program but that was a long time before the issue began.

What you said about heating the GPU does seem to mirror my problem though, is there any way of testing for the issue without going to an Apple Store? If it is indeed the GPU, I will continue using the unit as I am now until it dies completely, as to get the logic board replaced at this stage wouldn't really be worth the money vs a new 13" MPB which would now be sufficient for my needs. (And the GPU program has now ended)

They likely ran MRI's (Mac Resource Inspector) basic diagnostic, which does not test the GPU switching. The test they need to run to confirm GPU failure is called the VST, or Video Switching Test. And IIRC, the full Apple Service Toolkit has several more in-depth tools for testing the dGPU as it's not unheard of for a 2011 with a bad dGPU to pass the VST despite still being faulty. I cannot think of any at-home test that will easily confirm this issue, but other Members smarter than I may have some ideas.

If it is the GPU, you will want to turn the computer on, use GFXcard status to set the system to always use the integrated GPU (and try to avoid programs that activate the dGPU as GFXcardstatus doesn't work as well as it once did on previous OS X versions), disable any sleep/hibernate settings you have (so the computer only turns off the screen when idle for X amount of time), and try to never shut the computer down or restart it. I have a 2011 that's been on for about the past 8 months without a restart.

The replacement logic boards Apple used during the program were never modified from the original design to any notable extent. They have the same faulty design as the original, and a huge portion of the repaired systems would fail again with the same issue in less than six months after service (some much sooner.)

If I was a betting man, I would bet on this being the dGPU.

 

[TheRealNick]

Thanks again for your reply, ZapNZs.

Based on what you have said and my research following your suggestion I am now fairly convinced that the issue lies with the GPU. Leaving the machine on permanently may be an option - it is odd though that the GPU itself *appears* to be working fine once the MBP is turned on.

I knew the replacement logic boards were reconditioned but I didn't realise that they were also being churned out with the same defect - I guess Apple bet on a good number of people thinking their MBPs had died and never bothering to get them checked out after the warranty has expired. It makes it even more foolish for me to go down the repair route at this stage, as it will once again be reconditioned parts and will inevitably fail again given enough time.