MacBook M1 Pro 14" 2021 A2442 (820-02098) — U2020 getting hot, board not powering on

[slimmy182]

Hi everyone,

I have a MacBook Pro M1 14” (board 820-02098) with a broken LCD panel (matrix damaged).
The board does not power on — when connected to power it draws about 5.1 V / 0.474 A (~2.45 W) and stays at that level, with no fan spin or display activity.
Using a thermal camera, I noticed that U2020 heats up to over 70 °C shortly after connecting power.
According to the schematic, U2020 is a SN74AVC2T45RSWR (Texas Instruments) 2-bit bidirectional level shifter.
There are no visible signs of liquid damage anywhere on the board.
Given the broken LCD, the board might have suffered an impact, but I have no confirmation of this.
I haven’t performed other measurements yet — I’d like to ask if anyone has seen this behavior before on this model, and what steps you would recommend to diagnose the cause.
Thanks in advance!

[mon2]

U2020 should not be so toasty.

Measure the voltage to ground for the power rails:

PP1V2_AWAKE
PP1V8_AWAKE

Suggest to flux, remove and replace this part with a fresh one if the power rails are at normal values. Not suspecting the series resistors to be an issue but you can inspect them to be safe.

[slimmy182]

Thanks for your advice mon2.

I checked the rails:

• PP1V2_AWAKE = ~1.2 V (looks normal)
• PP1V8_AWAKE = ~1.55 V (instead of 1.8 V)

With the board off, resistance to ground on PP1V8_AWAKE (measured at C2021 / C2030 / pin 8 of U2020) is about 004 Ω on my meter in continuity mode — basically the same as shorting the probes together.

Thank you!

[mon2]

Flux and carefully remove U2020. Keep note of the pin #1 orientation. To be safe, do not power up without the part but this is only a level shifter between 2 different voltage rails.

Then test the resistance to ground @ C2021 again. What is the resistance in ohms? Be sure to use the resistance scale to measure in ohms.

[slimmy182]

• Removed U2020 (SN74AVC2T45).
• After removal, resistance PP1V8_AWAKE → GND @ C2021 is still ~0.40 Ω.
• Injected ~1 V @ 1 A on PP1V8_AWAKE (C2021).
• With thermal camera, U2010 warms up to ~40 °C, slightly hotter than surroundings, but nothing gets burning hot.

[mon2]

Are you in resistance mode or diode mode? During the injection, did you sink the 1A current? With such a hard short, any current you dial up will be consumed. Try 3A or more. Someone will heat up if truly at this low resistance. Confirm if the meter face is in ohms.

[slimmy182]

Sorry mon2, apologies — earlier I mistakenly reported PP1V8_AWAKE as ~0.40 Ω. That was incorrect (wrong scale/continuity mode).

Corrected measurements:

• Probes short = 0.05 Ω.
• PP1V8_AWAKE → GND = ~4.0 Ω (200 Ω range).
• Power injection:
  • 1.0 V / 3 A limit → ~0.2 A draw (CV).
  • 1.5 V / 3 A limit → ~0.4 A draw (CV).
• No clear hotspot with thermal cam or IPA during injection.

[mon2]

At ~4.0 Ω, this is the wrong resistance for PP1V8_AWAKE so something is wrong. The hunt for the defective part will not be simple.

As a wild guess - flux and remove LR501 and LR502 off the board. Measure the resistance again. In the past month or so, had a macbook where the unit would not power up and the student reported sudden death. In the end, it was the audio amplifier section with a shorted cap. No schematics for the audio board ($110 USD on Aliexpress for a replacement) so plucked off each bulk cap, one at a time. There were only 5 or so on this small board. Found the bad boy and the unit powered up fine after this fix.

The current draw is correct for the injection you are performing. Do not inject a higher voltage than the voltage of this rail.