Help! Updated to macOS 10.12.4, MBP randomly shuts off

[farw1d]

Here is a little tutorial: https://www.dropbox.com/s/nfvwkqwemcwpu0p/Macbook Shutdown Fix.pdf?dl=0

Would someone be so kind to repost this tutorial please? Thanks so much in advance!

 

[delboy27]

A quick update… the random shut downs still happen, albeit a lot less frequently.

It seems to only do it first thing in the morning, after having been in sleep mode for the whole night. It's happened to me 3-4 times in the last two weeks (lost count). If I remember correctly it only happens when it wakes up and is not plugged it.

Have to turn it off and reset PRAM then it works fine again. Not ideal and it has already cost me some mild embarrassment at a client meeting. Anyone has the same problem?

Just signed up to add that this just happened yesterday to my mid-2012 non-retina MacBook Pro running 10.12.6.

Took me hours going through the SMC and PRAM etc to no avail. Finally stumbled upon this article. Removing thunderboltnhi.ktext did the trick.

I’m travelling and working for the next two weeks so it would have been very frustrating if I couldn’t get it to work.

Just wanted to thank you guys for saving me lots of hassle and potentially money!

Del

P.S. so many people experiencing it in the last week or two.. coincidence??

 

[JohnMcR]

Hi there unfortunately I'm another victim of the Apple MacBook Pro Shutdown Syndrome. I have a MBP Late 2013 Retnia Laptop and have just installed macOS Mojave and out of nowhere the system has stared these random shutdowns.

Was working completely fine until I installed the new macOS Mojave I thought I had escaped the curse of High Sierra and dodged a bullet but no its got a hold of my system on Mojave.

Is anyone else experiencing this on OS Mojave or is there any other new fixes I don't know about. Mines only seems too do it in the morning when trying too login in or at night time when the system hasn't been on for a while.

 

[bivalvegruff]

Hi there unfortunately I'm another victim of the Apple MacBook Pro Shutdown Syndrome. I have a MBP Late 2013 Retnia Laptop and have just installed macOS Mojave and out of nowhere the system has stared these random shutdowns.

Was working completely fine until I installed the new macOS Mojave I thought I had escaped the curse of High Sierra and dodged a bullet but no its got a hold of my system on Mojave.

Is anyone else experiencing this on OS Mojave or is there any other new fixes I don't know about. Mines only seems too do it in the morning when trying too login in or at night time when the system hasn't been on for a while.

Yep my MacBook Air from 2013 still has this issue with Mojave. Mostly when not connected to a charger. Now it seems like the battery won't take any charge, so the shop that replaced it probably put a crappy battery in it. Hopefully that's what's causing my issues.

 

[js4250764]

Hi there unfortunately I'm another victim of the Apple MacBook Pro Shutdown Syndrome. I have a MBP Late 2013 Retnia Laptop and have just installed macOS Mojave and out of nowhere the system has stared these random shutdowns.

Was working completely fine until I installed the new macOS Mojave I thought I had escaped the curse of High Sierra and dodged a bullet but no its got a hold of my system on Mojave.

Is anyone else experiencing this on OS Mojave or is there any other new fixes I don't know about. Mines only seems too do it in the morning when trying too login in or at night time when the system hasn't been on for a while.

Same thing here, I have an MBP Late-2013 A1398 (2.0GHz 8G ram 256gb SSD) This issue started today and getting worse and shut down by itself after few seconds making it unusable... until the thunderbolt fix (lucky, I had time to do the trick before it shut down by itself). Using a generic battery since last summer, no problem with the battery. Two weeks ago, I've installed Mojave on SD CARD but didn't like it and now back on mavericks 10.9.5 on SSD. I have 2 partitions of Mavericks (split in 185 Gb & 65 Gb). The 185G one, I used on a daily basis and another one a fresh install of Mavericks that I rarely boot into. At first I thought it was some kind of new virus, scan with free avast & disconnect from wifi (still shut down). Then, boots in recovery mode shutdown after few minutes was barely able to read the thunderbolt fix on the internet. The thunderbolt fix seems to work for now but had shutdown after few minutes on the partition that has a fresh Mavericks install (UPDATE: I renamed it with Finder, i realise it used ".BAK.kext" as extension so I renamed it correctly... now Thunderbolt fix does seem to work for me). It has been 5 years without a problem (except for battery replacement) but like you said, I also suspect something has changed since the Mojave installation, could this be related to EFI firmware update? I'm willing to downgrade the firmware to test this out (Has anyone tried this?). I've tested in the pass High Sierra, Sierra, El Capitan, Yosemite but always restore it back to Mavericks which I prefer.

 

[AJVAJV]

I have a theory as to why this is happening to at least some people with this issue:

- There are sudden power draws in the Macbook Pro that draw large amounts of current. The MacBook Pro requires the battery to supply these large current spikes. When a battery is old, or of poorer generic quality, it can have a high internal resistance, even though it may still have lots of charge, which means that its voltage will dip during these spikes.

- I expect, that perhaps coupled with this aging/life problem above, there could be parts in the DC\DC power supply (capacitors specifically), that are also aging and make the unit more susceptible to a shutdown from these power spikes. (Replacing the logic board replaces some DC\DC converters in the unit.)

How I have arrived at this theory:

1) If you pull the battery from your Macbook Pro, it will still run with just the charger attached, albeit quite a bit slower. It will also stop crashing.

2) Reports online of people with generic batteries having more frequent random crashes. Generic batteries can have good MAh ratings, but have poor internal resistance.

3) MY Short-Term Fix: Connect an external monitor (in my case HDMI), and set the laptop brightness level to the lowest level (and/or off). This reduce the crash rate from 5-30 minutes to running hours+ without any issues (though still the odd random crash).

FYI - I redid the thermal paste on the processor/GPU, ran different versions of the OS, and "thought" at times it was better, but would always get to a state where it was crashing regularly, then realized it was screen brightness.

 

[AJVAJV]

Curious, when I went to look at the Trade-In value of my Macbook Pro, it pretty clearly states that if your Macbook has signs of battery swelling, that it is not eligible for trade-in. Given the cost to Apple of a battery is perhaps $50-60, is this not essentially a tacit statement by Apple that battery swelling will damage the unit beyond a simple low cost repair? I have to imagine when they refurbish units, they install new batteries?

 

[gmac34]

I have a theory as to why this is happening to at least some people with this issue:

- There are sudden power draws in the Macbook Pro that draw large amounts of current. The MacBook Pro requires the battery to supply these large current spikes. When a battery is old, or of poorer generic quality, it can have a high internal resistance, even though it may still have lots of charge, which means that its voltage will dip during these spikes.

- I expect, that perhaps coupled with this aging/life problem above, there could be parts in the DC\DC power supply (capacitors specifically), that are also aging and make the unit more susceptible to a shutdown from these power spikes.

How I have arrived at this theory:

1) If you pull the battery from your Macbook Pro, it will still run with just the charger attached, albeit quite a bit slower. It will also stop crashing.

2) Reports online of people with generic batteries having more frequent random crashes. Generic batteries can have good MAh ratings, but have poor internal resistance.

3) MY Short-Term Fix: Connect an external monitor (in my case HDMI), and set the laptop brightness level to the lowest level (and/or off). This reduce the crash rate from 5-30 minutes to running hours+ without any issues (though still the odd random crash).

Non sure if it is going to be

I have a theory as to why this is happening to at least some people with this issue:

- There are sudden power draws in the Macbook Pro that draw large amounts of current. The MacBook Pro requires the battery to supply these large current spikes. When a battery is old, or of poorer generic quality, it can have a high internal resistance, even though it may still have lots of charge, which means that its voltage will dip during these spikes.

- I expect, that perhaps coupled with this aging/life problem above, there could be parts in the DC\DC power supply (capacitors specifically), that are also aging and make the unit more susceptible to a shutdown from these power spikes. (Replacing the logic board replaces some DC\DC converters in the unit.)

How I have arrived at this theory:

1) If you pull the battery from your Macbook Pro, it will still run with just the charger attached, albeit quite a bit slower. It will also stop crashing.

2) Reports online of people with generic batteries having more frequent random crashes. Generic batteries can have good MAh ratings, but have poor internal resistance.

3) MY Short-Term Fix: Connect an external monitor (in my case HDMI), and set the laptop brightness level to the lowest level (and/or off). This reduce the crash rate from 5-30 minutes to running hours+ without any issues (though still the odd random crash).

FYI - I redid the thermal paste on the processor/GPU, ran different versions of the OS, and "thought" at times it was better, but would always get to a state where it was crashing regularly, then realized it was screen brightness.

I don't think sudden power surges are creating the problem, it seems correlated to the "power saving" profiles of the CPU, I was using intel power gadget to monitor the CPU and noticed that the crashes were occurring always when the CPU clock was decreasing under a certain threshold, (e.g. you close an application and the load on the cpu goes down). So it seems that different power states of the machine have an impact on which cpu profile it is used.
That is why it doesn't happen as often when the power is connected (the power saving is less aggressive), or when the battery is removed (the mac goes in limp mode and the cpu clock is stable at 0.8ghz).
Not sure how this is related to the Thunderbolt/PCI but it goes all down to unstable cpu voltage. and it is probably an issue with the cpu power supply or the cpu itself
https://www.rossmanngroup.com/board...hooting/28288-820-3662-sudden-power-off/page3

 

[AJVAJV]

Non sure if it is going to be


I don't think sudden power surges are creating the problem, it seems correlated to the "power saving" profiles of the CPU, I was using intel power gadget to monitor the CPU and noticed that the crashes were occurring always when the CPU clock was decreasing under a certain threshold, (e.g. you close an application and the load on the cpu goes down). So it seems that different power states of the machine have an impact on which cpu profile it is used.
That is why it doesn't happen as often when the power is connected (the power saving is less aggressive), or when the battery is removed (the mac goes in limp mode and the cpu clock is stable at 0.8ghz).
Not sure how this is related to the Thunderbolt/PCI but it goes all down to unstable cpu voltage. and it is probably an issue with the cpu power supply or the cpu itself
https://www.rossmanngroup.com/board...hooting/28288-820-3662-sudden-power-off/page3

I read the thread you presented, but unfortunately I could not open the scope trace links to see if the scope being used was set up for catching glitches on the power good signals. It appeared they also used a multimeter to check the power supply voltages. If there are supply glitches, that would require an oscilloscope to detect.

That it occurs at light CPU load does not mean there are no power surges, however short, which could destabilize a poorly performing power supply. Rapid loading and unloading of a DC-DC can both cause issues.

- I do not notice less crashes on AC power, though there may be fewer crashes with an external monitor
- Lowering screen brightness on the laptop definitely reduces crashes (on AC or battery)
- Thunderbolt hardware would be yet another "random" power draw, possibly powering on/off to detect external equipment.
- These problems seem to be cropping up for many people all about the same time (aging issue), and seem worse with 3rd party batteries (poor response to transients).

Is it an issue with unstable CPU voltage? ... that is more likely than the CPU itself, and would be more likely in an aging scenario. That brings us back to the most likely culprit which would not be the ICs in the power supply failing, but the capacitors in the supply and what is supplying them (battery). This laptop needs to keep working a few more days, so I can't do hardware probing on it, but when I can, I may take off a few capacitors and test them.

 

[gmac34]

I read the thread you presented, but unfortunately I could not open the scope trace links to see if the scope being used was set up for catching glitches on the power good signals. It appeared they also used a multimeter to check the power supply voltages. If there are supply glitches, that would require an oscilloscope to detect.

That it occurs at light CPU load does not mean there are no power surges, however short, which could destabilize a poorly performing power supply. Rapid loading and unloading of a DC-DC can both cause issues.

- I do not notice less crashes on AC power, though there may be fewer crashes with an external monitor
- Lowering screen brightness on the laptop definitely reduces crashes (on AC or battery)
- Thunderbolt hardware would be yet another "random" power draw, possibly powering on/off to detect external equipment.
- These problems seem to be cropping up for many people all about the same time (aging issue), and seem worse with 3rd party batteries (poor response to transients).

Is it an issue with unstable CPU voltage? ... that is more likely than the CPU itself, and would be more likely in an aging scenario. That brings us back to the most likely culprit which would not be the ICs in the power supply failing, but the capacitors in the supply and what is supplying them (battery). This laptop needs to keep working a few more days, so I can't do hardware probing on it, but when I can, I may take off a few capacitors and test them.

My mac and a few others around here do not crash when connected to the power supply, but that's not very consistent. And in my case removing the thunderbolt made the symptoms disappear completely. I also noticed that if I leave the thunderbolt eternet adaper plugged in at boot it will be recognized as a PCI device and in that case the mac will start crashing again. In the same forum I also read (need to search the topic) that replacing the CPU mosfets have apparently (client did not come back to protest) solved the problem.

 

[AJVAJV]

I have read a bunch of threads on this problem. Some seem to be better when connected to the charger, some it does not seem to matter. Most seem to improve by removing Thunderbolt (mine did), but the odd one does not. I expect the Thunderbolt fix is temporary, as in extends the life, but not a permanent fix.

Interesting about the MOSFETs, but unexpected. Normally those are a "works" or doesn't part unless they are operating near their voltage limit ... always possible.

 

[js4250764]

Very interesting reading, thanks for the input.

I've upgraded to a used logic board i7 2.8GHz mid-2014 16 Gb DG 2Gb that I purchased on eBay, problem still remains unless I remove AppleThunderboltNHI.kext. I've been using Ubuntu 18.04LTS on a 3rd partition (with and without TurboBoost) no shutdown occurs within the few days (but once it boots back to Mavericks, shutdown reappears with kext loaded after a couple minutes sometimes few hours later). I've managed to dump EFI-ROM in Linux a pre-Mojave and post-Mojave firmware. The new EFI-ROM firmware has 913 removals, 4024 insertions, 116 changes (with Araxis Merge) of the 8Mb. I've messed around with UEFITool but its way beyond my level of expertise, lol. Also, I’ve noticed that undervolting by -50 mV with Volta under Mojave with TurboBoost on/off have no effect on the shutdown issue. Running without battery does seem to stop the shutdown (haven’t tested it long enough though but like mentioned the CPU is underclock at 800 MHz (deleting the .plist for my model inside IOPlatformPluginFamily.kext didn’t work for me).

Could the new firmware update add some kind of power management profile that supports newer models but brings problems to older models? Why Linux not affected at all, ethernet thunderbolt adapter works. On the other hand, could new code in firmware detect non-genuine battery and some kind time bomb (plugged or not)? I'm very tempted and curious to flash it to original stock unlocked firmware (clean ME) with my programmer. Having known of this issue beforehand, I would have kept my genuine battery to test it out (I believe it had 87% of original capacity).

 

[alex191]

I want to share my observation with you that may help to pinpoint the issue:

I have a MacBook retina 15 from late 2013 and have the same shutdown issue: The screen freezes or turns off, the fans run high for about 20sec and then it shuts down.

A while back, I connected to my desktop pc directly via an Ethernet thunderbolt Adapter to transfer some files. This caused the MacBook to shut down immediately. This was the first day when the shutdown issue started.

Deleting the Ethernet driver solved it for me. So I assume, I have broken something by an overcurrent via the Ethernet?

 

[AJVAJV]

I had the same behaviour in El Capitan, Sierra, and Mojave which I transitioned to in a fairly short period of time. When the issue started with El Capitan, I thought maybe an OS upgrade would help (fix broken files, etc.) I was having random crashes probably in the last year, but all of a sudden they started happening all the time. It could have been as simple as a security update, etc. that was rolled out across all the OS versions that changed something in how multiple parts of the H/W were operating at the same time that caused "stress" on a system that was already on the edge.

I would question why your Logic Board was on Ebay .... perhaps it was having random issues.

Undervolting would reduce load on the DC-DC so it could just as well harm or improve things. It is not necessarily a low voltage, could just be a glitch on a supply. I am going to see if anyone has a block diagram of the power supply sections/capacitors to save me some time. I have the equipment capacitors at the component level ... FETs as well to some degree, though expect this is not the issue. From what I have read, putting in a new logic board fixes the issue.

I have a genuine battery and have this issue. My battery is still showing about 75% of original capacity, but it looks like a children's balloon at this point! I am not going to invest money in a battery unless I can be sure it will solve the issue. At best, the 2013MBP will become a backup. I have to have something reliable.

Very interesting reading, thanks for the input.

I've upgraded to a used logic board i7 2.8GHz mid-2014 16 Gb DG 2Gb that I purchased on eBay, problem still remains unless I remove AppleThunderboltNHI.kext. I've been using Ubuntu 18.04LTS on a 3rd partition (with and without TurboBoost) no shutdown occurs within the few days (but once it boots back to Mavericks, shutdown reappears with kext loaded after a couple minutes sometimes few hours later). I've managed to dump EFI-ROM in Linux a pre-Mojave and post-Mojave firmware. The new EFI-ROM firmware has 913 removals, 4024 insertions, 116 changes (with Araxis Merge) of the 8Mb. I've messed around with UEFITool but its way beyond my level of expertise, lol. Also, I’ve noticed that undervolting by -50 mV with Volta under Mojave with TurboBoost on/off have no effect on the shutdown issue. Running without battery does seem to stop the shutdown (haven’t tested it long enough though but like mentioned the CPU is underclock at 800 MHz (deleting the .plist for my model inside IOPlatformPluginFamily.kext didn’t work for me).

Could the new firmware update add some kind of power management profile that supports newer models but brings problems to older models? Why Linux not affected at all, ethernet thunderbolt adapter works. On the other hand, could new code in firmware detect non-genuine battery and some kind time bomb (plugged or not)? I'm very tempted and curious to flash it to original stock unlocked firmware (clean ME) with my programmer. Having known of this issue beforehand, I would have kept my genuine battery to test it out (I believe it had 87% of original capacity).

 

[Marlow17]

Hi i recently did the security update 2018-002, i had being putting all updates off as disabling thunderbolt was working well for me. i did the update and my macbook (late 2013) seemed to work ok without taking the file out of extensions. Today it has gone crazy with crashes. I tried disabling the the applethunderboltnhi.kext but it crashed whilst doing that and now after rebooting the file is no longer there but the crashes are persisting. Any ideas, or should I try going to Mojave.

Thanks Mark

 

[crowhill]

I have first posted to this thread now just over 1 year ago. I have a Mid 2014 MBP Retina with 2.5 GHz Intel Core i7, 16 GB 1600 MHz DDR3, NVIDIA, Intel Iris Pro graphics processors, original battery and currently running Mojave 10.14.1 with an external Apple monitor (at the time of my first post it was High Sierra 10.13.2 Beta).

Since I use my MBP most of the time connected to the external monitor and the power adapter, I face the sudden shutdowns only when disconnected from both, but then within a few minutes of disconnecting. The only way I can operate the MBP while mobile is by removing AppleThunderboltNHI.kext from the Extensions folder (dragging the file to the trash bin) prior to shutting down my device and then to startup the device again. When I get back to my desk, I then need to drag AppleThunderboltNHI.kext back into the Extensions folder, shutdown my MBP, connect it to the monitor and power adapter and start up again.

I was hoping that going from High Sierra to Mojave would fix the problem, but it did nothing for me. I don't think Apple is ever going to address this and will fore us to buy new laptops or live with frustration.

 

[patbsh]

Hello everyone, I've a MBP 15" mid 2014 and I was experiencing the same exact problem, I have given my MacBook to apple so they fix it an they replaced the battery (as they thought that it was the issue) and it didn't fix the problem even though I've spent a lot of money on the service.

Fortunately I've found this thread as disabling it completely solved the issue (over two weeks of constant everyday use without a SINGLE shutdown).

However, does anyone know if there's a fix to that problem other than disabling the driver? I need its functionality as I need to use virtual machines for my job and it looks like this driver is required for them in order to run (tried VMware, VirtualBox and Parallels) and I always get similar network problem that causes an instant crash of the machine.

If not the driver, does anyone know if theres a way to run VMs while having this driver disabled? My MBP still works perfectly so buying a new one bc of some stupid driver is the last thing I want to do...

 

[timelessbeing]

Another victim here. Wow. This problem is almost two years old? How come it only started happening to me a few days ago?

 

[bivalvegruff]

Another victim here. Wow. This problem is almost two years old? How come it only started happening to me a few days ago?

Something probably broke in your computer.

I've removed macOS and installed Linux on my affected MacBook Air, and while the shutdowns perhaps are more rare now, they still occur. So it must be a hardware fault, at least in my case. A connector somewhere getting loose perhaps. In my case, the problem started after I had my battery replaced by a third party workshop – the battery seems to be original, but maybe they didn't quite install it right.

 

[the_adamk]

Hi guys! Just registered to thank you for helping me not to trash my Late 2013 15" MBP (2,3GHz i7 + Nvidia GT750M). I have tried everything and started to consider hardware repairs. It was quite odd for me that after fresh install shut-downs were less frequent. Then I started googling even more and found this thread.

To be honest, I was not hoping for such ridiculous solution as deleting driver to work - I'm using Macs for 15 years now and actually I didn't know where the drivers are located - I thought that it was a Windows thing to hustle with drivers...

Imagine my amazement when the shutdowns stopped completely. It is 2nd day now.

Have any of you changed the battery after the whole script/driver thing? Does it change the computer in any way? Does the solution work same as before the change?

Btw, I do not use thunderbolt at all (USB3.0 -> gigabit ethernet adapter and DP external monitors), so I do not need thunderbolt to work, but it is a shame and Mac is for sure not for sale for market price...

 

[dqrf29]

The fix was working fine since this problem occurred until I upgraded to Mojave like an idiot and now it's back to restarting randomly..

 

[bittered]

The fix was working fine since this problem occurred until I upgraded to Mojave like an idiot and now it's back to restarting randomly..

I'm on 10.12 (Sierra) and I'm considering upgrading to 10.14 (Mojave). I'm interested in hearing the experiences of anybody who has a shut down problem that was fixed by deleting AppleThunderboltNHI.kext. After you upgraded and ran the usual fix, did everything work as normal or did you still have issues?

 

[Vu Man Chu]

Okay ladies and gentlemen after 4 months of testing I have isolated the issue.

-updated to mojave problem still persisted

-removed thunderbolt port in Network Preferences problem still persisted

-reset PRAM problem still persisted

-reset SMC problem still persisted

-took my MBP to the genius bar, all hardware passed diagnostics problem still persisted

-took my MBP to mac specialist, said it was a software issue or logic board issue. roll back OS to previous state problem still persisted

-rolled back to previous state problem still persisted

-rolled back to Mavericks original factory OS problem still persisted

3 things that could be wrong after this trouble shooting


1) CPU Core voltage is causing the random shutdown

2) iGPU is bad (requires logic board replacement)

3) Logic board needs to be replaced


before throwing in the towel to replace the logic board i attempted to try to code the python script posted by anders wiggers myself with no success as i am not a coder and spent 3 hours wrecking my brain. so i decided to use the mac mod software for 10$ https://realmacmods.com/product/macbook-... and ran the software which creates a gear spinning on the bar and voila the laptop no longer randomly shuts down!!! updated to Mojave and just an fyi the gear is no longer in the bar but the laptop does not randomly shut down. The problem is CPU core voltage related which is corrected by the python script or the macmods software for 10$. if this does not work it is most likely a logic board issue.


Good Luck

 

[vrose04]

15 inch 2014 Macbook Pro here... Has been like this since day one, we bought it on offerup.

Without running the Python script it would turn off multiple times per day, sometimes multiple times in a 10 minute period. Now that the Python script is running, it still does happen but less frequently, even while plugged in. I took it to the Apple store and they ran the diagnostics, even keeping it for a few days and found absolutely nothing.

The rep at the store said that they guarantee all work for 90 days so if we did try to get it fixed and the problem persisted, we could send it back again, but I don't think it's very helpful since the problem doesn't show up in their diagnostics. If I could pay $600 to make this go away I would.

 

[bivalvegruff]

The rep at the store said that they guarantee all work for 90 days so if we did try to get it fixed and the problem persisted, we could send it back again, but I don't think it's very helpful since the problem doesn't show up in their diagnostics. If I could pay $600 to make this go away I would.

Did you try the possible $10 solution in the post above yours?