Macbook Pro A2141 020-01700-A GPU Artifacts then no S5 rails

[skwarszczow]

Hello everyone,

This is my first post on the forum

I recently purchased a MacBook A2141 from a colleague at work, which had issues with the dedicated GPU. The system occasionally displayed artifacts on the screen and sometimes experienced kernel panics related to the GPU. Interestingly, applying pressure to the VRAM module would temporarily resolve the artifact issues, which made me optimistic about the ease of repair.

Board: 020-01700, Config: Intel i7, 512GB SSD, 16GB Ram.



Before completing the purchase, we decided to erase all data using the Network Recovery then using Disk utilities wiped data followed by a fresh system installation.
The installation process completed successfully, but the system continued to encounter kernel panics upon boot.
I decided to close the laptop and take it home for further diagnosis.

However, a new problem emerged after returning home: the laptop no longer powers on, despite using both the battery and/or original power adapter.

I have checked the power rails and confirmed the presence of all voltages up to the S5 power states, including SSD and RTC sections. The following voltages read:

• PP3V3_S5 = 0.00V
• PP1V8_S5 = 0.00V
• PP1V05_PRIM = 0.00V
• And next related to these rails are also 0.00V

No short circuits were detected on these rails.

Also there is indication that T2 chip is alive:

• PPBUS_G3H = 12.3V
• Current draw on USB-C is approximately 0.08A at 20V

Based on forum suggestions, the issue may be related to a corrupted T2 firmware.

I am able to enter both Recovery and DFU modes without any issues. The flashing process proceeds up to approximately 50%, at which point the target device typically reboots to perform system boot - tested on other machines such as A1990.

But in this scenario DFU Recovery results in an error code 0x09 (9).

Notably, the fans spin/ticks (2-3 times) during the flashing process when it stalls at 50% just before error prompt appears.
An oscilloscope measurement shows that voltage appears on each rail for about 1-5ms before dropping back to 0V.

Has anyone else experienced a similar issue?
Everything suggests that something may be blocking the power rails and blocking CPU/GPU to start normally, but I have not observed any short circuits.

[mon2]

T2 has not yet booted. When T2 boots, the PPBUS_G3H will be boosted to 12v6.

Power delivery from external power adapter is with a low current draw.

https://repair.wiki/w/MacBook_Pro_A2..._at_20V_repair

and review this video:

https://www.youtube.com/watch?v=7RRnqb_62u4

[reformatt]

Check SSD power rails for shorts. Common on this model. You might be in luck for it to be a shorted cap on the ssd power rails, as PPBUS_G3H isn't shorted as well.

[skwarszczow]

Thank you for your answer.

Apologies about PPBUS_G3H rail typo, My actual readings are:

• PPBUS_G3H: 12.62 V

And of course it is about board 820-01700

According to repair wiki steps:

My measurements:

[Diode Mode]:

• PP2v5_NAND_SSD1: 0.439 V
• PP2v5_NAND_SSD0: 0.395 V
• PP1v8_SSD1: 0.425 V
• PP1v8_SSD0: 0.372 V
• PP0v9_SSD1: 0.490 V
• PP0v9_SSD0: 0.423 V
• PP3v3_S5: 0.333 V

[Resistance to GND]:

• PP2v5_NAND_SSD1: 29.8 kΩ
• PP2v5_NAND_SSD0: 29.5 kΩ
• PP1v8_SSD1: 28.6 kΩ
• PP1v8_SSD0: 18.2 kΩ
• PP0v9_SSD1: 591 Ω
• PP0v9_SSD0: 168 Ω
• PP3v3_S5: 17.5 kΩ

[Voltages when Original PSU Attached]:

• PP2v5_NAND_SSD1: 2.525 V
• PP2v5_NAND_SSD0: 2.530 V
• PP1v8_SSD1: 1.807 V
• PP1v8_SSD0: 1.806 V
• PP0v9_SSD0: 0.903 V
• PP0v9_SSD1: 0.905 V
• PP3v3_S5: 0.0025 V

Resistance on PP0V9_SSD0 seems to be lower than on SSD1 but voltage is generated properly when PSU is attached.

No signs of corrosion of U9000 or any other signs of damage in this area.
Howewer... I see that something was done previously with this motherboard.
There is Kynar wire soldered between R7107 (PP5V_G3S pin) and L8504 (PP5V_G3S) - It seems that it was liquid damaged previously and one track was missing for J8500 connector.

Anyway, Voltage PP5V_G3S is also present = 5.17V

[reformatt]

This model (and the A1990) often have corrosion on each wing on the board where the fan intakes are. I would inspect these areas closely for corrosion. Note DFU should be a last resort, not the first one. If you have an existing hardware issue, this makes troubleshooting so much harder as the DFU restore fails and you have incomplete or missing BridgeOS. So two faults and it makes it a lot harder. This used to be in the old RepairWiki notes, not sure about the new one.

Refer to the wiki for DFU errors:

https://logi.wiki/index.php/DFU_Mode...roubleshooting


BTW, I hate the new RepairWiki and don't contribute to it at all. I find the interface unintuitive and impossible to follow. The old Wiki I could just add an additional box at the bottom. I'd say this is why it doesn't get updated very often. I hate it.

[skwarszczow]

Yea... unfortunately, I am aware of what I did. I attempted a DFU restore, and before that, the laptop was at least "attempting" to start when pressing the power button. Now, I need to do that by a DFU update.

That being said, there are no signs of corrosion or damage, including the resistance measurements in wing sectors.

Anyway, based on my experience with the DFU update process, when Apple Configurator reaches step 4/4 (update), it first writes the image/firmware to the SSD partition and then attempts to start the processor and the rest of the system (around 45-50% of the progress bar) and
at that point, it completes the DFU process but it requires CPU, PCH etc. to be running.

I approach this as a personal non profit hobby project, fully aware that it may end in failure, but I am also searching for a path to success.
I know that similar designs based on I7-I9750H can run without a external GPU, so I decided to experiment by desoldering the RAM and GPU and configuring the U9850 mux chip by grounding the GPU_SEL output.
This should permanently set the Intel graphics as the primary GPU - as this is the problem which laptop comes to me.

However, I am experiencing the same issue—the laptop does not power on, DFU responds, the process reaches 50%, there is a brief "tick" of power on the lines and fan spins, and then nothing further happens. I am increasingly concerned that the CPU itself might be the cause of the problem.

Do you have any experience with specific points that should be checked regarding the CPU and its power supply? Or maybe other ideas...

[reformatt]

If you feel T2/BridgeOS are fine, then you would check all your voltages up to CPU Vcore are present during the Intel power sequence. Presence of Vcore indicates the preceding ones are present and that your PCH is likely ok. The EFI BIOS is part of the DFU restore and located on the NAND's so you can rule that out.

If the CPU is attempting to boot, you should see that in your USB-C current consumption (without battery of course). Would expect over 1 amp for this board on the bench when plugged in.