Q: 2011 MacBook Pro and Discrete Graphics Card

Hello,

Do you use an external display ?

This simple fact of plugging a display :

- Forces the use of the discrete AMD chip

- Increases the GPU temperature significantly (even with simple web browsing)

That's the reason why since I know about this problem I limit the use of an external screen (maximum 1 hour per week when it's strictly necessary). And I use only integrated graphics.

I know that I run my macbook pro well under its capability.

SkyFox. I take it your machine is not exhibiting and problems?

*any* problems

Degger, your reaction on that post of me. In fact, I did mean the same.

When I wrote "Apple repair", those exclamation marks, dunno if that applies for English language too, but they were ment to make it sarcastic.

As in "Apple doesn't really repair themselves"

So, in fact, maybe we finally agree to something ;-).

Degger, your reaction on that post of me. In fact, I did mean the same.

I know what you meant, I just wasn't sure that you understood what I meant.

So, in fact, maybe we finally agree to something ;-).

We agree on lots of things. The major difference is that I would suggest a new GPU for a replacement instead of a reball job because the latter is much trickier to get right and an outsider cannot know which reball service is capable and which one to stay away from.

ps3specialist wrote:

 

Part of the manufacturing and assembly process of the logic board is inspection and of course for a comapny like Apple their logic boards go through X-ray inspection befor they go into a computer so beleive me there is a reason that caused the problem to develop into your computer, it could be dust, Humidity, dry weather , usage time is one factor but not all the factors. In cars there is a milage counter so they were able to input that into cars warranty conditions, I lately started to see that kind of counters in smart TVs so I expect so see this usage time counters in computers soon if it is not implemented already somewhere. 

I can't help but think that the prime reason is the 5 year exemption on using lead free solder on high temperature products expired at the beginning of 2011.  I know through friends who manufacture products cooling very high temperature semis that they ran into similar problems in 2011 with the replacement lead free products breaking their joints prematurely.  Add to that the combination of the Sandy Bridge CPU/AMD GPU generating the most heat out of any combination of the whole unibody range from 2008-2012 and the fact that stress testing the CPU heat can be transferred back towards the GPU as I've tested that happening on many of these models.

I am of the firm opinion that the only true long term fix for these models is a replacement GPU reballed onto the logic board, preferably lead soldered, the heatsink contact plates for CPU/GPU polished for a smooth surface and using much less and improved thermal paste such as AS-5/Ceramique.

So far out of all the replacement GPU's in these 2011 models I have advised my clients to have done only one out of now 15 units have failed after replacing and that was fairly quickly after with the second attempt now working fine.  A pretty good success rate of a fix so far, and when certified Apple engineers I know personally who are aware of the spike of failed units coming into their centres happen to agree wholeheartedly with me 'off the record' that this is by far the best solution for these Macbook Pro's I'm quite sure I'm not wrong!

I have been monitoring this thread for many months. I do not want to get into a forum battle of words but maybe 'ps3specialist' should stick to repairing PS3's as you are clearly an expert in this field.. however maybe your understanding of apple machines is a little less. I would have to disagree with alot of your opinions. (this does not of course mean they are wrong)

The problems apple are experiencing with these failures are '100%' a design flaw.

Whether it is the GPU itself, the BGA balls it sits on or the VRAM.

We repair Apple units everyday and a BGA reball is not always suffucient to permantly repair.

Sure for PS3's this generally resolves it but as we know in the past that GPU's have had internal issues with the materials inside the GPU die.

You offer reballs for 2007/08 models but you are not aware of the revised chipset Nvidia specifically made to cure the problem???

It was a materials change as the solder materials inside the die were unsuitable for the application (a design flaw) so over 95% of the 2007/08 models you repair will fail again as nearlyu all are due to the incorrect materials used in the die. Sure a reflow or reball (or even hairdryer) will make it work again as the heat can re-join the balls in the die substrate - but only temporary.

It needs to be noted that ATI may have the same problems with their internal die. We have had alot of 6750M in for repair from the iMac also. This has a huge heatsink and sufficent cooling but still they fail.

We need to ask ourselves why.

This is why I recommend to customers to fit a brand new GPU in every case. In cost comparison it is worth paying that little extra.

There is no way to test if the old GPU is fully working as after a reball - if the internal die had broken it, the heat used in the reball process will melt the balls in the die and make it work again. Whether it is weeks, months or years.

I don't want to list services directly here as I want this to be a balanced opinion but I am sure some are aware of users who have used our services. We have chassis for nearly every Apple model for users to send in logic boards only as we repair alot of Apples.

Food for thought...

Using leaded solder is much easier than using leadfree due to the fact that leaded solder has a much lower melting point but from my experience with at least a hundred of them leadfree solder gives a much longer term repair , about replacing the GPU I don't think it will make any difference , what makes a huge difference is the thermal profile that was used in the soldering process and the hands that performed the reballing process from start to finish that of corse in addition to all the materials used in the process, after all , manufacturers can not choose which kind of solder to use, Environmental regulations now give them only one choice which is leadfree solder.

Again I think you are missing the point. Why would you still use lead free if it is known that lead solder is more suitable and reliable.

This is why the Automotive, Aviation and Military are still permitted to use lead solder due to the increased reliability needed.

If it is due to your personal health concerns fair enough but otherwise the advantages of lead is overwhelming.

Yes lead is easier to work with as you can solder at lower temps. This is a huge advantage as the new GPU gets put under less thermal stress when fitting.

Yes thermal profiles have their places but are not the only reason for success or failure.

I think the true reason many repairers do not use new GPU's as it makes them less competiteve on price and also genuine new chipsets are very difficult to source. over 90% of GPU's on ebay are not what they are marked as. The chinese are clever and will change any part number to achieve a few more $$$.

If lead free solder was proven to be more reliable I would switch but we all know the reason it was switched is due to the safety directive NOT reliablility.

Don't forget many of our water supplies in the world are still carries in lead pipes so it is a shame computer parts cannot still use it.

bga_repairs wrote:

 

I have been monitoring this thread for many months. I do not want to get into a forum battle of words but maybe 'ps3specialist' should stick to repairing PS3's as you are clearly an expert in this field.. however maybe your understanding of apple machines is a little less. I would have to disagree with alot of your opinions. (this does not of course mean they are wrong)

 

The problems apple are experiencing with these failures are '100%' a design flaw.

 

Whether it is the GPU itself, the BGA balls it sits on or the VRAM.

 

We repair Apple units everyday and a BGA reball is not always suffucient to permantly repair.

 

Sure for PS3's this generally resolves it but as we know in the past that GPU's have had internal issues with the materials inside the GPU die.

 

You offer reballs for 2007/08 models but you are not aware of the revised chipset Nvidia specifically made to cure the problem???

 

It was a materials change as the solder materials inside the die were unsuitable for the application (a design flaw) so over 95% of the 2007/08 models you repair will fail again as nearlyu all are due to the incorrect materials used in the die. Sure a reflow or reball (or even hairdryer) will make it work again as the heat can re-join the balls in the die substrate - but only temporary.

 

It needs to be noted that ATI may have the same problems with their internal die. We have had alot of 6750M in for repair from the iMac also. This has a huge heatsink and sufficent cooling but still they fail.

 

We need to ask ourselves why.

 

This is why I recommend to customers to fit a brand new GPU in every case. In cost comparison it is worth paying that little extra.

 

There is no way to test if the old GPU is fully working as after a reball - if the internal die had broken it, the heat used in the reball process will melt the balls in the die and make it work again. Whether it is weeks, months or years.

 

I don't want to list services directly here as I want this to be a balanced opinion but I am sure some are aware of users who have used our services. We have chassis for nearly every Apple model for users to send in logic boards only as we repair alot of Apples.

 

Food for thought...

I will only disagree on one point - I would not quite fully call it a design flaw, more like flawed 2011 lead free solder with the usual poor guidelines on over application of thermal paste and rough heatsink surfaces not helping the problem one bit.

I can offer this balanced, independent opinion of your services - Since the beginning of the year I have referred my clients to yourselves directly and after checking the units idle temperatures for CPU and GPU post fix on idle, during renders and after I am very content with your work.  Albeit early doors after only 4 months not one has failed since because of the GPU.   It's a pity that Apple's engineers don't read these 4,000 plus posts fully as the solution for all our 2011 models is all in this thread.

And when my late 2011 17 AMD does surely fail you'll be getting my MBP shipped to you for a replacement 6770, for far less cost than a depot logic board swap with undoubtebly the same duff solder on the GPU.  Therefore I would be happy not to use an official Apple repair program with the current in my opinion flawed solutions but simply Apple just covering my costs of your services which I believe is the best fix for these Sandy Bridge MBP's.

An expert in PS3 makes me more expert in reballing a much much easier chips like the ones in Apple or any other computer, Reballing a 40gm GPU chip with four built in memory chips installed on a 4mm thick board can not be compared to reball a 1gm chip in any other product but any way that is not the point, the point is if what you are saying about the deffective GPU chips that were used is true we should have heared about all on them but we did not, Its always easier to replace than to reball the same chip, it also makes more money !! but I don't force people for it , if someone wants to I don't mind , I did see people on Ebay offering replacing that GPU on these 2007/2008 models for $400 when the chip itself - the newer model - costs only $30 , sadly a lot of people paid them that much for it. Guess what, my own laptop is a MBP 2008 which I reballed its GPU chip about two years ago and still working perfect so why would I replace it? I am not going to make money from myself if I do.

<Link Edited By Host>

Appreciated. Also to mention I have seen the work of 'Apple refurbished boards' and I wouldn't let them out of our door looking like that. Too much heat is sometimes used resulting of slight delaminiation under the GPU and general 'browning' of the board - a sign of too much heat. In the ones we have seen in this condition they have had new GPu's fitted DC13. The other components are DC11.

Also flux not cleaned and in a dirty condition.

I would totally agree with poor heatsink surfaces.

Rant over I will return to my hole .....

ps3specialist wrote:

 

Using leaded solder is much easier than using leadfree due to the fact that leaded solder has a much lower melting point but from my experience with at least a hundred of them leadfree solder gives a much longer term repair , about replacing the GPU I don't think it will make any difference , what makes a huge difference is the thermal profile that was used in the soldering process and the hands that performed the reballing process from start to finish that of corse in addition to all the materials used in the process, after all , manufacturers can not choose which kind of solder to use, Environmental regulations now give them only one choice which is leadfree solder.

It's not a matter of technique, leaded solder is far better than lead free with stability with big temperature variations which in these Sandy Bridge 2011 models have the most heat changes out of all the unibody range from 2008-2012.  My friends company reparing his high temperature boxes uses a third party to lead solder as a repair for his lead free problems as they do not want their equipment to keep failing, they are in remote locations and subject to huge temperature variations, far more than a MBP. 

There's also the 2011 solder inside the GPU die to be taken into account hence why a new GPU lead soldered is in my opinion the way to go for a true, long term fix for our MBP models.

Leaded solder is no longer allowed to be used by any manufacturer of any electronic product.

To bga_repairs:

Hi, how can I contact you?

Thanks.