Aller au contenu principal

Processeur quad-core Intel Core i7 à 2,2 GHz (Turbo Boost jusqu'à 3,4 GHz), 2,5 GHz (Turbo Boost jusqu'à 3,7 GHz), ou 2,8 GHz (Turbo Boost jusqu'à 4,0 GHz) avec 6Mo de cache L3 partagé.

781 questions Voir tout

2015 MacBook Pro Thermal Throttling (Kernel_Task) After New Battery

Seeking the assistance of all the iFixiters out there to solve a mystery I’m having with my MacBook Pro.

So to preface, I have run a iPhone, iPad, & Mac repair business for 10 years now and have extensive experience with Apple devices, however this one is stumping me.

So, a few months back a friend of mine hired me to fix his MacBook Pro. He stated that it was running slow so he decided to do an OS update and half way it failed and now won’t turn on.

Initial diagnosis was that I could not get the machine to boot. I then noticed that the trackpad would not click down. In hindsight I recognize this is because this model has a Force Touch trackpad that only clicks when the computer is on.

However I assumed that this could be an indicator that the battery was swollen. I removed the back plate and sure enough the battery was swollen.

I told my friend that the computer needed a new battery desperately and this could very likely be what caused the computer to die during the update. This being said, I ordered him a new battery from iFixit to be installed.

Jump 2 weeks later when I have the battery in hand, he drops of the MacBook Pro to have me complete the replacement. Unfortunately, in the time waiting for the battery to arrive the battery has bloated even MORE.

Given that this model has the battery glued in, & that the battery is bloated to this extent, I told the customer that unfortunately I could not move forward with the repair due to the risky nature of LiPo batteries when bloated.

He decided that he wanted to give it a go on his own so he bought the battery off of me and borrowed my tools. He was able to successfully replace the battery with no fires 😮‍💨 .

This is where it gets weird. Post battery install the computer turns on, however the only key on the keyboard that works in the power key and the trackpad is non functional. The fans also ran full blast and the computer ran really slow (I was able to test the computer with external mouse and keyboard but as mentioned VERY SLOW).

At this point, I took the laptop and fully disassembled it again to check all cables and insure all was connect correctly. It was.

It should be mentioned that many years in the past this computer came into contact with a small soda spill on the keyboard. Apparently the spill never caused any issues. I also didn’t see any corrosion or signs of liquid damage at all inside the machine, which I took as a good sign.

I did some extensive research and found a great many iFixit articles about this issue and the consensus was that the fix was a new trackpad cable.

I assumed that perhaps the trackpad cable was damaged during repair and now that the trackpad doesn’t work it can’t read the thermal sensor in the trackpad which is causing it to run at max fan and thermal throttle.

At this point however my friend was eager to get a laptop up and running, and frustrated with the time and money spent on the 2015 MBP, so he decided to pull the trigger and upgrade to a new MacBook Air.

For all my time guiding him through the process and answering questions he paid me with the semi functional 2015 MBP to tinker with at my expense.

I decided to order a new trackpad cable, trackpad, and uppercase with keyboard.

I installed all of these and Voila, it booted up, keyboard works, trackpad works, no fan noise! So, I called it a success. Only when I started using it did I realize truly how slow it was, beach balling, could barely run 2 apps at once, activity monitor says 12ish percent idle. This is with fresh install of Big Sur.

When I checked my activity monitor I noticed that system was using 70+ percent of the overall cpu and that kernal_task was using 500+ percent of the cpu at idle. However still no fans.

I did extensive research about kernel_task and did all the recommendations as follows:

SMC reset, PRAM reset, fresh install of OS, Apple diagnostics (no codes except “could test power adapter” since my power adapter was connected at the time of diagnostics), I followed some instructions to attempt to delete .kext files from my library/extensions to fix kernel task but the files mentioned in the tutorial don’t exist on my machine??

Then I started having this issue where the computer would not boot at all unless I held down the power key for at least 5 seconds. Apparently, this preforms an SMC bypass. So I can’t boot if I don’t do an SMC BYPASS or I can do an SMC reset each time I boot then I don’t have to hold the power key.

Either way to boot I have to preform either the SMC BYPASS OR RESET THE SMC. If I remember correctly this booting up issue started after the first SMC reset I performed.

I even went so far as to do PRAM & SMC resets with the battery disconnected too. No dice, however I found out something interesting, when booting on AC power with no battery connected I have no issue booting, no need for SMC bypass or reset.

So the battery is implicated in this somehow. Then I read online that Mac’s sometimes get weirded out by non OEM batteries and cause issues like this. Even cause kernel_task.

I read also that you could update your SMC firmware and that helped out a guy online with this issue post battery install, but I can only find an update file for mid 2012 retina MBPs not 2015 models. So I don’t think it will help me, maybe I’m wrong.

See update link for 2012:

https://support.apple.com/kb/dl1559?loca...

So I am totally stumped.

I really want to get this thing up and running & any help would be greatly appreciated.

Saludos from Mexico 🇲🇽,

-Milo-

Update (08/27/23)

@danj Removed heat sink and now understand Fb what you mean by the black shield. Removed the shield and there is a large amount of paste covering the chips, both old and new paste.

Guessing this is what I should remove…?

Block Image

Update (10/14/23)

@danj

UPDATE: I had the opportunity yesterday to have a friend who works at an Apple Store to run Apple Diagnostics on my MacBook Pro. It showed many errors regarding sensors and power/voltage readings on the logic board. He recommended an SMC reset, but I explained I had already tried that.

The concluded that there is a chance that the issues are caused by the non OEM iFixit brand battery. This is something I also read online regarding it being caused by a non OEM battery.

These issues started after I replaced the battery, which would back up that hypothesis.

I attached the images of the Apple Diagnosis thinking they might be is interest to you.

Interesting side note. I also own a mid-2013 13inch MacBook Air. Several years ago I replaced the original battery with an iFixit one and it has been working perfectly ever since.

THEN, about a week ago, to disprove the idea that it was a failed ssd causing issues in my 15 inch MacBook Pro, I swapped the SSDs (MacBook Air SSD into MacBook Pro and vice versa) since I knew the MacBook Air ssd was good and they are cross compatible.

It made no difference for the MacBook Pro, it still had the kernel_task issue, so I put the SSDs back to their proper positions, I forgot about it.

3 days later, my MacBook Air won’t turn on without the AC adapter connected, doesn’t recognize the iFixit battery (which had been working for years) and NOW has the same kernel_task issue as the MacBook Pro.

The ONLY two things these Macs have in common is that for an hour I swapped the SSD from the MacBook Pro into the MacBook Air and I also use the same AC adapter for both Macs.

Could the SSD from the MacBook Pro which had the kernel_task issue originally have somehow corrupted the SMC on the MacBook Air after I swapped them temporarily.

The MacBook Air has the proper SSD reinstalled and I have attempted SMC resets on the MacBook Air, however it still does not recognize the battery and has the kernel_task issue.

P.S. you will see in the diagnostic pics that there were issues flagged regarding the “power port”. Could this be related to a bad AC adapter. I have several MagSafe 2 adapters and have tried different ones on both computers to now avail.

Just wanted to give you an update since it is a very strange issue.

Saludos,

-Milo

Block Image

Block Image

Block Image

Répondre à cette question J'ai le même problème

Cette question est-elle utile ?

Indice 1
4 commentaires

Good morning,

I am no expert here, however that looks like a hell of a lot of thermal paste. As @danj said, i would start trying to remove as much of it off the capacitors as possible. This could be resulting in the capacitors overheating because they are design to run on airflow, not thermal compound.

Good luck :)

par

@mfirestar and @np_f1_2021 - Yep TIM is not something you want around the capacitors The TIM effects the capacitors inductance.

Here's a fun experiment to tryout take an aluminum tube which a steel ball can travel down. Carefully drop the ball down the tube it should travel quite quickly as it shouldn't be magnetized. Now a coil to induce magnetism into the ball (it needs to be a strong force) not sent the ball down the same tube what happened? Here the magnetic flux interfered with the aluminum tube.

While different here the same kind of effect happens as the TIM covering the caps alters their effectiveness! This is what I use loaded with 90% Isopropyl alcohol Refillable Aerosol Spray Can Sprayer and a soft tooth brush to fully clean the green carrier and the SMD devices on it.

par

@np_f1_2021 @danj Thank you both for your responses. I had the opportunity to clean off much of that excess Thermal Paste. I didn’t have the opportunity to acquire isopropyl alcohol, however I am using my iFixit anti-static brush with the Articlean Thermal Compound Removal product and it is looking much better.

Interestingly, a lot of the paste was from my most recent application. I recognize I used an excessive amount of paste the most recent time I reapplied it, however much of it also looks very old and dry as if it leaked under the black shield long ago.

I am going to give it a final clean tomorrow and will reapply a much smaller amount of paste and will let you both know how it goes.

Thank you for all the info!

-M

par

@np_f1_2021 @danj Had the opportunity this morning to clean off the CPU/GPU chips and remove the extra TIM from the chips around them. I then reapplied Arctic Silver but a more reasonable amount (I used your tissue technique to estimate to amount) and reinstalled the heat sink.

Unfortunately, kernel_task is still chewing upward of 600% of my cpu at idle. I downloaded iStat Menus and verified the temperatures of all my sensors. Nothing above 40 Celsius.

The fans are running at an appropriate speed however I even tried forcing them to spin at max. The temps dropped but kernel_task took no mind.

In activity monitor it still says that “system” is using upwards of 70% of the CPU at idle.

Thank you for all of you help and if you have any ideas, lemme know!

-M

par

Ajouter un commentaire

1 Réponse

Réponse la plus utile

The next time you get a bloated battery put it into a plastic bag sucking the air out and then putting it into the fridge (no colder than 34F) that will slow the chemical reaction so you can safely remove the battery before it warms up. And of course run the battery down first.

As to your problem, I would see what the onboard diagnostics tells you. Restart the system and see if you can get it to fire up. Apple messed it up so it might fail but worth the try. Press the D key to enter you may want to try the internet version as well. Review this Use Apple Diagnostics to test your Mac and review this guide for the other startup key options Mac startup key combinations. Tell us what you get for error messages.

The reason the system is slow is one or more sensors are telling SMC the system needs to run in CPU Safe Mode which cranks up the fans and lowers the CPUs clocking to prevent it from over heating.

I would inspect all of the cable connections to the logic board as well here as one could be not seated correctly or the ZIF latch is not pressing on the ribbon cable or even missing as they to tend to fly off nearer to be found! Reference this guide Remplacement de la carte mère du MacBook Pro 15" Retina mi-2015

Hopefully it’s just a simple cable issue. Let us know how it goes! Good luck 🤞

Update (10/14/23)

@mfirestar - Sadly, I think you still have a TIM issue 😩

I use a loadable spray can which I load with ISO and using a soft tooth brush fully clean the green chip carrier of the paste. It doesn’t take much to mess things up.

Now jumping to your MacBook Air… swapping the SSD wouldn’t be the cause and if you had kernel virus that wouldn’t infect SMC and that would require firmware level access which only Apple has the certificate key to allow via their OS updater process. Basically very remote and you wouldn’t be the only one.

So why is my Air acting like my MacBook? Some symptoms can be caused by different problems so it’s important to fully diagnose the root issue. At this point see if your friend can run the diagnostics on this system. I would look at Activity Monitor to see what process is running hard as a badly running App can be the problem. Compare your systems besides the Apple processes what else is running that is common?

Lastly, let’s check the battery, install if you haven’t yet CoconutBattery let’s see what it tells us.

Cette réponse est-elle utile ?

Indice 1

7 commentaires:

Thank you for your swift response @danj.

However as I mentioned, I already ran Apple diagnostics and it didn’t throw any codes except for the fact that it couldn’t test my power adapter which is to be expected since the adapter was not connected at the time of diagnosis.

As for the Logic Board connectors, this is the 3rd time disassembling and reassembling it & making sure to check each ZIF connector, but still no luck.

By CoU safe mode I’m assuming you mean CPU safe mode. This is interesting, but interestingly the fans are no longer running full blast, that stopped once I replaced the trackpad cable, trackpad, and uppercase with keyboard. However I do think you are right, the computer is convinced it is hot even though it isn’t, and thus it is throttling.

1 extra thing to mention is that I replaced the thermal paste as well with new Arctic Silver but the thermal issues were happening both before and after the thermal paste so that rules that out as an issue.

Gracias

-M

par

@mfirestar - Sorry I missed you had tried Diagnostics. And yes that CoU was CPU.

There are two different SMC circuits that will effect things here Temp and Voltage. So a bad or missing temp sensor will force the Fans to ramp, a Voltage sensor will reduce the CPUs clocking. And if both are affected you’ll get both reactions.

Mmm… You may have done your self in. While on the surface refreshing the thermal paste is not a big issue as long as you don’t use the wrong kind as in the case of liquid metal which is very tricky stuff!

Some TIM is conductive and some inductive! Which might be your problem here. The amount of paste used is the issue as it takes so little we tend to over do it. So it gets all over the place! Here we need on make sure it doesn’t pollute the SMD devices on the chip carrier. I would pop off the heat sink and carefully clean the green carrier fully. To be careful as you can pull the tiny SMD chips off.

par

@danj WOW! Thank you for all this info! This sounds promising. I definitely think I may have out too much Thermal Paste on the CPU dye because after the computer was on for a few minutes after replacing the paste and the paste presumably got hot for the first time, a small amount excess was visible leaking around the edge of the heat sink. I’m a bit of a noob when it comes to logic board chip terminology. Can you more simply explain what you would recommend carefully cleaning, please?

What is the chip carrier in respect to the CPU/GPU area of the board and what are the SMD devices that need to be un polluted?

Thx!

par

@mfirestar - after you lift the heat sink you’ll find a black shield which covers the green carrier area. The shiny black element in the middle is the chip it’s what’s under this black shield we are worried about as the paste tends to get under it then effecting the small SMD devices underneath.

I use a fillable spray can with a snorkel tube to focus a stream of reagent grade isopropyl alcohol (90%) with a small brush to wash away the paste fully and then apply very tiny amount. To get an idea of size take a piece of Tissue (one layer) cut it to the same size as the chip area, now wet your fingers and roll this piece into a ball tightly. Even that is too big! But you get the idea of how little that’s needed.

par

@danj sorry for my delayed response! I am going to try this later today after work. I was wondering, however, I don’t know where to acquire 90% isopropyl alcohol but I do have the arcticlean products, both the thermal material remover & thermal surface purifier. Can the thermal material remover be used in place of isopropyl alcohol to remove the paste?

I will keep you posted!

Thanks all the help!

-M

par

Voir 2 commentaires en plus

Ajouter un commentaire

Ajouter une réponse

Milo Firestar sera éternellement reconnaissant.
Nombre de vues :

Dernières 24 heures : 2

7 derniers jours : 11

30 derniers jours : 31

Total : 438