Skip to main content

Repair Cases Index

Repair Cases Index — Apple Silicon (M1–M4)

Consolidated index of component-level repair cases on Apple Silicon MacBook logic boards, drawn from published independent-repair videos. Cases are limited to non-trivial failure modes — Power Delivery controller handshakes, battery management buses, liquid-induced track and flex damage, solder-joint fatigue, sensor lockouts, and sub-rail regulation faults. Simple "shorted MLCC on PPBUS_AON" cases are deliberately out of scope.

How to read these pages

Every technical statement carries an evidence label:

Label Meaning
[measured] An instrument reading shown on camera (meter, oscilloscope, thermal).
[stated] A claim made by the technician in narration, not independently shown.
[inferred] Reasoning drawn from the observations, not directly demonstrated.
[unverified] Asserted but not substantiated, or contradicted by known engineering practice.

Provenance. These pages were condensed from the audio transcripts of the source videos (extracted with Google AI Studio) and from on-screen meter readings described in those transcripts. Timestamps, reference designators, board numbers and measured values are reproduced as stated in the video and have not been independently verified against schematics or a physical board. Treat every refdes and rail name as a lead to confirm on the actual board revision, not as a fact.

Cases by fault class

USB-C Power Delivery controllers (CD3217 / CD3218)

# Case Model / Board Key finding
11 M1 Air 5V Stall — CD3217 1.1V LDO Short Air M1 A2337 / 820-02016 Shorted 1.1 V internal LDO on one controller locks all ports at 5 V.
12 M1 Air 5V Cycling — CD3217B12 Firmware Mismatch Air M1 A2337 / 820-02016 Mismatched CD3217 silicon revision → I2C abort and 5 V cycling.

Battery management bus (SMBus / I2C)

# Case Model / Board Key finding
5 M1 Max No Charge — Shorted SMBus TVS Diode Pro 16" M1 Max A2485 / 820-02100 Cracked TVS diode clamps SDA to ground; charging stops.
6 M1 Pro No Charge — Open SMBus SCL Trace Pro 16" M1 Pro A2485 / 820-02100 Tool-slip severs SCL trace; open line halts the charger state machine.
8 M2 Air Battery Not Detected — Sheared SMBus Pull-Up Air 13" M2 A2681 / 820-02536 Sheared 0201 pull-up resistor → no battery + severe CPU throttling.

Liquid damage and high-voltage flex

# Case Model / Board Key finding
1 M4 Pro Liquid Damage — USB-C PD Controller Corrosion Pro 14" M4 Pro / 820-03487 Corrosion across PD controllers freezes negotiation on every port.
2 M3 Pro Liquid Spill — 5V Stall, Missing SPMU Power-Good Pro 14" M3 Pro A2992 / 820-02918 Open SPMU feedback resistor stalls the board at 5 V / 50 mA.
3 M1 Pro Liquid Damage — PMU Control Lines & Corrupted Boot Pro 14" M1 Pro A2442 / 820-02098 Liquid on PMU enables → boot-loop + corrupted boot image, needs DFU revive.
10 M1 Pro Backlight — Mini-LED Flex Arc-Through Pro 14" M1 Pro A2442 / 820-02098 44 V mini-LED boost arcs through moisture, vaporising the hinge flex trace.

Display, sensors and power rails

# Case Model / Board Key finding
4 M1 Pro No Display — Lid Angle Sensor Kills Backlight Pro 14" M1 Pro A2442 Damaged Lid Angle Sensor asserts "lid closed" and kills the backlight.
7 M1 Pro Soft Short — PP3V3_S2 Resistive Leakage Pro 16" M1 Pro A2485 / 820-02100 ~120 Ω leakage on PP3V3_S2 holds power-good low, no short-killer trip.
9 Apple Silicon DFU Revive Loop — Inductor Solder Fatigue M1/M2 A2337 / A2338 / 820-02016 Cracked inductor solder joint drops a rail mid-DFU → revive fails.

Recurring thresholds seen across cases

Input signature Typical meaning
5.1 V / 0.00 A Dead short on PPBUS_AON, or a CD3217 internal LDO short.
5.1 V / 0.05 A static PPBUS_AON healthy (~12 V) but a secondary power-good is unasserted.
5.1 V / 0.05–0.25 A cycling PD bus/firmware mismatch, or a rail dropping out during early boot.
20 V / 0.12 A steady Board alive in DFU, awaiting a host.

Standing cautions

  • Chip-off data recovery is not viable. NAND contents on M1–M4 are encrypted with keys fused into the Secure Enclave inside the SoC; reading the NAND elsewhere yields undecryptable ciphertext. Repairing the original board to a bootable state is the only data-recovery path — a point every surveyed channel agrees on.
  • "Revive" preserves user data; "Restore" wipes keys. Revive failures are frequently hardware power dropouts, not software corruption.