Q: 2011 MacBook Pro and Discrete Graphics Card

Exactly apple isnt going to do anything about it. Those who are affected that use mac only apps are just going to get another mac due to no other choice. The rest apple will consider collateral damage.

Still extremely frustrated.

Does anyone have experience caling into Apple Relations? I'm about to try that today and maybe even try to get a discount on a "temporary" fix aka logic board replacement.

I was curious, so I looked into the discussion groups about the iMac 27" 6970 problem reports.

This may be one going back to Sept 23, 2011.

https://discussions.apple.com/message/16212463#16212463

If so, I am sure that owner was beside himself as the 'fix' didn't arrive for nearly 2 years.

Our problem is complex, in the sense that mine actually began failing some time ago, I now realize... well before it became a critical problem.  In fact my system would lock up and crash going way-back into 2011.  I am a careful but heavy user and there were many-many days I had 16-18 hours on the notebook; I assumed it was just life; I am now convinced it was GPU lockup of some kind.

Even when the computer became unusable, I was getting improper diagnosis or none at all.... from no less than 3 different trained and authorized repair centers.  If the hands-on techs couldn't 'get it'; the notion that the Apple exec who must make such decisions, and then run it by lawyers, is somehow sitting there all-knowing is simply not true (IMHO).

I mentioned that this is, of course, a business and economic decision for Apple; the idea that the cost of it is a problem for them is amusing.  Not only that, they can expense it all.  This is about getting the proper attention from the proper person(s) at the proper moment.

Apple could take 1000 refurb motherboards and sent them to a reball shop(s) for nickels; getting them to see that is the right thing to do (assuming its the permanent fix) is what we can all only hope for.

Color me stupid; I still have faith this will be resolved.

There no official support from Apple, they didn't aknowledge the problem, so they might not offer anything but please try, they need to be hammered with this until something happens. In my experience senior managers can't do much, other departements (customer relations?) seem to have more power but even then...

Hey folks.

Here I am back on day 2 after havin my MBP successfully repaired by a BGA chip replacement

and wanna share my results and thoughts.

May some of You find it interresting or helpful on making a decision how to proceed.

1st of all to all affected owners from Germany or even Europe who search for a trusted company

to get a chip replacement I wanna highly recommend the guys of:

Laptop Reparatur Thüringen:

http://stores.ebay.de/Laptop-Reparatur-Zentrum-Thuringen

They made me a happy Mac user (but of course not customer) again.

Perfect work, perfect communication, (for me) perfect pricing.

I won't hesitate a second to get them If I ever will be in need of such repair.

But now to what happened under the hood of my silverboy.

One root cause only, leading into several different next issues made my MBP die

and from my pov Apple is responsible!

In may case (to assume not only in my case) the thermal compound between chip surface and copper cooler

mutated to a drained heap without any functionality!

This started a silent chainreaction of death.

Without working thermal compound the heatflow gets reversed and goes directly

through the chip making its way through solder balls and finally spreads over the logic board.

As we all now a logic board consists of several layers and from thermal view each layer

has its own coefficient of expansion what produces enormous tension on the board until

solder balls are cracked.

In my case the heat not only damaged my GPU as initially thought but diagnosis of my unit

showed that GPU Vcore and SMC  (both are placed quite close to GPU) were affected too!! 

And here comes the thing. GPU was not the first faulty component. It startet with SMC chip

in my unit..

Obviously my MBP recognized a problem first time over a year ago

via messages in system.log like:

"SMC::smcReadKeyAction ERROR"  and "SMC::Fall back to old SMC mode"

That was all hints regarding a possible problem. I could not recognize any faulty

behavior till finally the chain reaction hit my GPU and its Vcore chip.

The messages above indicate the system seems able to read SMC related info

from somewhere else than SMC itself.

Nice feature if at least some warnings were represented or system protection would work.

One cannot expect users to dig through system log all day long searching for messages they might not

understand or react correctly.

Btw... @Apple

Why in gods name your $2k design wonder is not able to do

what every crappy piece of sh**ty PC hardware is capable of ????

Prevent the system from hardware damage by overheating, give an audio alert and do a emergency shutdown???

This is the second issue I say it is in Apples responsibility if I manage somehow

to accept that they save 1/10 cent production cost per unit

on using the worst thermal paste ever seen..

So to all of You currently running the common gfxCardStatus "Fix" without checkin thermal compound inside

are running a ticking bomb...

If not ensuring a proper working thermal flow system you will burn more than only GPU !!

My advise:

Get a BGA repair with photo proofed chip replacement.

You will end up with 200,- EUR at a maximum and get way better service on such repairs.

Check for bga_repairs posts.. 

cherio

Thanks for the info. Can u apply paste yourself?

Thank you very much Jens!

I don't recall any SMC messages, but the 'ticking bomb...' is what I do know; I had graphics lockups going way back that I ignored as 'normal'.

Did the shop say anything about the solder ball/joints themselves?  That is, was the whole problem the thermal paste or did they not say?

Do I understand the new GPU/SMC chps and the work were ~200 Euro?

Sorry for all of the questions.

Thanks

You will need to follow steps like replacing logic board to get there.

See http://www.ifixit.com/Guide/MacBook+Pro+17-Inch+Unibody+Logic+Board+Replacement/ 3447

I cannot recommend to do it by yourself If you never gone through hardware,

build up few computers from scratch, or repaired some.

It is kinda tricky and there are some steps You definitely do not want to do errors...

The shop has a flat pricing of EUR 148,-

Therfor u get:

- new GPU

- new thermal paste by "Prolimatech"

- clean up

- component stress test

- 48 hours longterm test

I advised the shop to do the repair only after calling me back.

They sent analysis report and microscope photos of faulty solder joints and heat marks

round the GPU die back and offered me repairing as followed:

- new GPU                                                       148,00   <- GPU burned + cracked solder joints

- new GPU Vcore                                               48,60

- SMC resoldering                                                0,00   <- no replace required - "only" few solder balls cracked

- new thermal paste by "Prolimatech"                  0,00

- clean up                                                             0,00

- component stress test

- 48 hours longterm test                                       

- shipping                                                              0,00

My question if they use leaded balls unfortunately kept unanswered but I guess

they used unlead again.

P.S.

They did not explicitely blame one component to be the bad guy but they confirmed in

written words error was caused by overheating.

As I am into computer hardware and its thermal flows for over 20 years now

there only my statement is left.. additionally I received my faulty chips back and I definitely know when

I see a flamed chip...

but at last you're right .. thats my thaughts and expierience only..

@jensgoehler,Can you give us bga_repairs contact info?

Thank you for the response.  I recall reading about excess thermal paste on the 2011's back in 2011 and not really paying attention.

I will get my system back from Applecare with a refurb (possibly new I suppose) mobo and 4 more months of Applecare; I want to be understand all of this now, not later.

148 Euro is modest in the context of cost of ownership of an expensive tool.  I think that when my Applecare expires I would try to have it serviced that way; even if I ended up selling it, at least have someone competent reapply the thermal paste. 

I am in Colombia which means more work finding a shop I feel comfortable with is all.

I would probably need to source the chips myself if it came to that.

Sincerely, thanks for the info. 

Page 79 is the only I have.

<Email Edited by Host>

jensgoehler wrote:

 

In my case the heat not only damaged my GPU as initially thought but diagnosis of my unit showed that GPU Vcore and SMC  (both are placed quite close to GPU) were affected too!! 

 

And here comes the thing. GPU was not the first faulty component. It started with SMC chip in my unit..

 

Obviously my MBP recognized a problem first time over a year ago via messages in system.log like:

"SMC::smcReadKeyAction ERROR"  and "SMC::Fall back to old SMC mode"

 

That was all hints regarding a possible problem. I could not recognize any faulty behavior till finally the chain reaction hit my GPU and its Vcore chip.

 

The messages above indicate the system seems able to read SMC related info from somewhere else than SMC itself.

 

So to all of You currently running the common gfxCardStatus "Fix" without checkin thermal compound inside are running a ticking bomb...If not ensuring a proper working thermal flow system you will burn more than only GPU!!

 

My advise:

• Get a BGA repair with photo proofed chip replacement.
• You will end up with 200,- EUR at a maximum and get way better service on such repairs.
• Check for bga_repairs posts.. 

 

cherio

Thanks for the super informative/helpful post Jens. I just did a quick search of my logs and also show the SMC error message you flagged:

12/2/13 9:28:36.000 PM kernel[0]: SMC::smcInitHelper ERROR: MMIO regMap == NULL - fall back to old SMC mode

11/29/13 11:41:51.000 AM kernel[0]: SMC::smcInitHelper ERROR: MMIO regMap == NULL - fall back to old SMC mode

11/29/13 11:42:26.000 AM kernel[0]: SMC::smcReadKeyAction ERROR: smcReadData8 failed for key B0OS (kSMCKeyNotFound)

3 questions:

1. Are there any other error messages I should look for? How do I search logs beyond the most recent week?
   
   
2. Are you/BGA confident that the solution they provided is the best (only?) way to fully fix the GPU failures long-term
   
   
3. What *exactly/specifically* did BGA do that Apple isn't doing to fix your MBP (**update, this you answered  above while I was typing out this message)

I'm one of the many poor/frustrated souls currently stuck in LIMBO (re: integrated only mode... which has made design work very difficult since the setting does not allow for use of an external monitor) because I can't afford to turn over my MBP to apple for 3 to 7+ days PLUS pay over $300+ for them to replace the entire logic board when doing so doesn't actually "fix" the GPU issue (as has been reported by  numerous individuals on this forum and others). But if the repairs BGA made = a real, long-term solution, that would certainly be the first bit of good *actionable* news I've read since first experiencing GPU failures (shortly after upgrading to Mavericks).

Thanks!

Apple quoted $500 for a new logic board. Apple Certified repair quoted $654.

Sent e-mails out to some mac repair shops in Toronto, Canada. Going to look into getting mine properly diagnosed and possibly reballed with new thermal on the CPU as well.

wow thanks for the report, interesting infos

from Ifixit in 2011 when they tore down one macbook they explicitely said:

«Holy thermal paste! Time will tell if the gobs of thermal paste applied to the CPU and GPU will cause overheating issues down the road.»

see here; step 10

http://www.ifixit.com/Teardown/MacBook+Pro+15-Inch+Unibody+Early+2011+Teardown/4 990