MBP 1,1-10,2 Permanently Disable 2011 15"/17" MacBook Pro Dedicated GPU - gMux IC Bypass

[mirceac]

@chaosdesigns : there is a megathread regarding this:
https://forums.macrumors.com/thread...ntel-integrated-gpu-efi-variable-fix.2037591/

 

[FlorisVN]

Retina machines don't have a defective GPU, but rather only bad solder joints on the GPU VCORE regulator IC (U8900). Removing the chip, applying new solder, and soldering it back on will completely fix the issue.

if im correct, the A1286 mid-2012 15" mbp's, are not affected at all with bad vcore regulator's nor bad Nvidia GT650m kepler chips.. ?

So this could be the most reliable A1286 2012 15" model there was ever made.. ?

 

[mirceac]

if im correct, the A1286 mid-2012 15" mbp's, are not affected at all with bad vcore regulator's nor bad Nvidia GT650m kepler chips.. ?

So this could be the most reliable A1286 2012 15" model there was ever made.. ?

True, the worst think happening there is some voltage regulator getting un-soldered, but this is pretty rare, so you may be right, that these are the most reliable A1286.

 

[FlorisVN]

True, the worst think happening there is some voltage regulator getting un-soldered, but this is pretty rare, so you may be right, that these are the most reliable A1286.

Ok thx for your info

 

[theredcloud]

True, the worst think happening there is some voltage regulator getting un-soldered, but this is pretty rare, so you may be right, that these are the most reliable A1286.

That’s right got me a 2012 15” recently and it’s fantastic running Mojave....Don’t think the 2011 is worth the stress of repairing in 2019

 

[mirceac]

That’s right got me a 2012 15” recently and it’s fantastic running Mojave....Don’t think the 2011 is worth the stress of repairing in 2019

The "stress" consist in a minimal investment and a bit of "elbow grease", that's really nothing. Of course you're not going to become a next Louis R. repairing these guys, but they are far away from being garbage, especially the I7 with matte screens.

@dosdude1 How is the full demux progressing, any updates ?

 

[theredcloud]

The "stress" consist in a minimal investment and a bit of "elbow grease", that's really nothing. Of course you're not going to become a next Louis R. repairing these guys, but they are far away from being garbage, especially the I7 with matte screens.

@dosdude1 How is the full demux progressing, any updates ?

Actually I do board level repairs just like Louis R. and start a business from his teaching....

 

[mirceac]

@theredcloud IMHO, the investment for repairing a 2011 Macbook Pro with dGPU issues is really minimal for a repair shop, and if I, almost blind guy close to 60, I was able to do the soldering and everything in less than 20 min, you should be able to do it as well, considering tat you'll have a microscope and profi tools.

The idea is to explain clearly to the customer what are the consequences ( no external display anymore, no brightness control until the next version of the deMux ) and let him choose.

If he wants to throw away his 16GB, Samsung EVO SSD, matte screen Macbook Pro laptop in the garbage, he's free to do it (I will kindly ask to shipp it here, we can still make use of them ) , but even if he wants to sell it, it is worth selling a repaired device instead of defective POS.

It also depends of the price you're asking for it, in NY where people seem to wipe their behinds with 100$ bills, it will probably be a 250USD repair, in all other sane places with 100$ I think most people will take it.

To be honest, on my old I7, after I fully removed the old dGPU (a bit of hot air and wick after) and disabled the power IC for it, the battery life is so much better. IMMV.

 

[TheElectrocat]

Another update: PCBs for the custom backlight controller implementation have arrived, and they look great! Within the next week, we will get the boards all soldered and ready to go. You can read more about the project on my friend's webpage, who I am working with on this project. Once the PCBs have been soldered, we will be selling the remaining ones of the 17 that we have.

View attachment 790178 View attachment 790179 View attachment 790180

Regard friend, where is posible buy this module today please, need this modules for macbook pro solutions, thanks for your time. (sorry for mi english, is not mi native language).

 

[lenselijer]

Tried flashing an early 2011 but even with the L2406 removed it still wont flash
Already flashed multiple late 2011 so i know my setup is ok.
Anyone in Europe can help me out, i can ship you the logic board?

 

[mirceac]

Tried flashing an early 2011 but even with the L2406 removed it still wont flash
Already flashed multiple late 2011 so i know my setup is ok.
Anyone in Europe can help me out, i can ship you the logic board?

I can help you, if it's possible to ship it to Germany, PM if you want. BTW is the board working normally, besides dGPU issues, it's actually pretty strange that it doesn't work with a proven setup ?

 

[WilYawn]

Just brought a 2011 MacBook Pro 17" back from the dead. Flash worked on the first try. Soldered the wires directly to the pcb without any connections in between.

Also doubled the ram to 8GB and replaced the tired old hdd with an ssd. Installing Mojave rigth now.

Thank you Collin for all your hard work

 

[mirceac]

Just brought a 2011 MacBook Pro 17" back from the dead. Flash worked on the first try. Soldered the wires directly to the pcb without any connections in between.

Also doubled the ram to 8GB and replaced the tired old hdd with an ssd. Installing Mojave rigth now.

Thank you Collin for all your hard work

Is a bit risky for the pads but with a bit of care it works . Now if Collin will bring us some good news, a full implementation of the mux, with brightness control by Christmas time, now that will be even better

 

[bobnugget]

if im correct, the A1286 mid-2012 15" mbp's, are not affected at all with bad vcore regulator's nor bad Nvidia GT650m kepler chips.. ?

So this could be the most reliable A1286 2012 15" model there was ever made.. ?

Definitely, had all of them and both the 2012's are still running (2010/11's all had board replacements, these have failed in 2011's). Worryingly, my wife's one has the occasional GT650 kernel panic that the retina model was recalled for. It has only panicked twice in the last couple of years though.

 

[KvnTM]

Can I just inject 3.3v to PP3V3_SO to reduce noise on the jtag lines? I just can't get it to work. It drives me crazy at this point. I guess only the FPGA has to powered to perform the flashing.

 

[mirceac]

Can I just inject 3.3v to PP3V3_SO to reduce noise on the jtag lines? I just can't get it to work. It drives me crazy at this point. I guess only the FPGA has to powered to perform the flashing.

If there is noise on the JTAG lines it doesn't come form the other power supplies on the board, but eventually from your charger(s), I've seen it with some no-name chargers and sometimes even with Apple branded ones. There's also some mains current transfer from the crappy chargers most laptops have, like the one you've attempted to program with.

My bet is if the both device ( Mac and programming laptop ) have batteries holding for 10-15 Minutes you should try with both devices battery powered only, no charges connected, and it will work. This should take care of any electrical noise if that's the cause.

 

[KvnTM]

If there is noise on the JTAG lines it doesn't come form the other power supplies on the board, but eventually from your charger(s), I've seen it with some no-name chargers and sometimes even with Apple branded ones. There's also some mains current transfer from the crappy chargers most laptops have, like the one you've attempted to program with.

My bet is if the both device ( Mac and programming laptop ) have batteries holding for 10-15 Minutes you should try with both devices battery powered only, no charges connected, and it will work. This should take care of any electrical noise if that's the cause.

I already tried to power both from battery only. It would always fail. JTAG Connections are 100% correct. I desoldered them 4 or 5 times already. Also going vertically from the board. Some people said they had luck by removing the inductor under the JTAG connector.

 

[mirceac]

@KvnTM Then it's something else, you don't need re-solder the wires without knowing what caused the issue, and I'm reasonable sure that the JTAG bus exposed there is to be used without removing any parts form the board.

My best bet is to download the Lattice programming software and see if the FPGA is getting detected when scanning JTAG. I did and I was able to see it on the bus, of course it was not possible to read it back because of the code protection bit was on but at least the model and serial number of the FPGA should be visible.

http://www.latticesemi.com/programmer#_20C94305815A4B3AAAFEA8B83943B751

Also this will test your drivers and programmer, if it's not seen correctly by the Lattice programming application, then it won't work anyway with Collin's app.

 

[KvnTM]

@KvnTM Then it's something else, you don't need re-solder the wires without knowing what caused the issue, and I'm reasonable sure that the JTAG bus exposed there is to be used without removing any parts form the board.

My best bet is to download the Lattice programming software and see if the FPGA is getting detected when scanning JTAG. I did and I was able to see it on the bus, of course it was not possible to read it back because of the code protection bit was on but at least the model and serial number of the FPGA should be visible.

http://www.latticesemi.com/programmer#_20C94305815A4B3AAAFEA8B83943B751

Also this will test your drivers and programmer, if it's not seen correctly by the Lattice programming application, then it won't work anyway with Collin's app.

Already downloaded their Diamond Flashing Tool thing. Does not detect the FPGA. Always said something like failed to scan the board iirc. Drivers are fine and it detects the Programmer. It always fails to read the ID of the FPGA.

 

[mirceac]

Well then, there not much else to do, the only things remaining are: desolder the wires and measure each pad in diode mode, see if the ground pad goes to ground and 3v3 has really 3v3 and the signal pads shows the diode on the inputs and also if possible check the programmer itself, either on a different Mac or if you can get access to a Lattice development board.

 

[KvnTM]

Well then, there not much else to do, the only things remaining are: desolder the wires and measure each pad in diode mode, see if the ground pad goes to ground and 3v3 has really 3v3 and the signal pads shows the diode on the inputs and also if possible check the programmer itself, either on a different Mac or if you can get access to a Lattice development board.

All pads and connections are good. Tried with a 2nd mb and even installed Windows 7. It still fails. Maybe I got a dodgy programmer? I used the debug tool to toggle the pins and each pin does what it's supposed to do. Sadly I don't have access to a dev board and I don't want to order one. Maybe someone near Würzburg in Germany can borrow me his just to confirm a bad programmer? I'm pissed off and wasted 20$ for the programmer and 20$ for one license.

 

[dosdude1]

All pads and connections are good. Tried with a 2nd mb and even installed Windows 7. It still fails. Maybe I got a dodgy programmer? I used the debug tool to toggle the pins and each pin does what it's supposed to do. Sadly I don't have access to a dev board and I don't want to order one. Maybe someone near Würzburg in Germany can borrow me his just to confirm a bad programmer? I'm pissed off and wasted 20$ for the programmer and 20$ for one license.

Just to verify here, you're ensuring that the machine is NOT booted into an OS when flashing it, correct? Also, you're using a physical Windows machine and NOT a VM?

If you still cannot manage to get it to work, I can refund you for the license key.

 

[KvnTM]

Just to verify here, you're ensuring that the machine is NOT booted into an OS when flashing it, correct? Also, you're using a physical Windows machine and NOT a VM?

If you still cannot manage to get it to work, I can refund you for the license key.

HDD disconnected. Even tried with two bare logic boards only. Physical Machine. No VM. I'll send you an email.
Spent around 10 hours now to fix this and I can't get it to work. Tried a ton of different cables. Soldered directly to the programmer. Nothing.

I've sent you an email to your doslab gmail address.

 

[WilYawn]

HDD disconnected. Even tried with two bare logic boards only. Physical Machine. No VM. I'll send you an email.
Spent around 10 hours now to fix this and I can't get it to work. Tried a ton of different cables. Soldered directly to the programmer. Nothing.

Sounds like you have tried everything. My best bet is a faulty programmer.
Have you tried different versions of windows? I use Windows 8.1 X64 and it works great. No antivirus or firewall blocking the flash program from talking to Collins server?

 

[KvnTM]

Sounds like you have tried everything. My best bet is a faulty programmer.
Have you tried different versions of windows? I use Windows 8.1 X64 and it works great. No antivirus or firewall blocking the flash program from talking to Collins server?

Yup. Windows 7 & 10 without AntiVirus / Firewall
It fails while flashing. Server connections are fine.

It fails to read the device id. So it probably is the programmers fault.
Well what can you expect from cheap Chinese knockoff programmers. Am I right?