820-01700 persistent power cycle: 20 V, 0 A drop, re-ramp

Intake Condition

• Liquid damage (customer spill).
• On bench: most ports negotiated 5 V ~0.30 A, one port showed amp “bouncing” (unstable current).

Timeline / What I Did

1. Ultrasonic clean (proper dry-out).
   → Behavior changed to boot-then-die loop: negotiates 20 V, fans spin for a few seconds till 2.5 A, hard-off to 0 A (20 V holds), then current ramps again (>3 A peak) but won’t relatch without pressing power or replugging.
2. Set aside a few days. On revisit: stuck at 5 V ~0.30 A, amp-bouncing port symptom gone.
3. Heated the previously suspect CD3217 on that bouncing port.
   → Boot-then-die loop returns (same as #1). That made me suspect that CD3217/BGA.
4. Replaced the suspect CD3217 (side-correct part).
   → No change: still boot-then-die loop as in #1. Can’t force DFU/Revive even using board DFU pads.
5. Found a clearly burnt/open 0 Ω link: RA602 (between PPBUS_HS_GPU ↔ PPVIN_GPU_COREREG_VIN).
   → Replaced RA602 with proper 0 Ω jumper. No change to symptoms.
6. Thermal camera during attempts: only the typical region on the back-side of CPU creeps to ~50 °C; no localized hot spots on SSD, PMU/T2 LDO area, USB-C controllers, or GPU VRM caps/FETs.

Current Symptoms (repeatable)

• Plug USB-C (known-good charger/cable): negotiates 20 V.
• Press power: fans spin, no chime/no image, a few seconds later 0 A (20 V persists), then amps climb again as if it’ll start, but it won’t restart without another power press or replug.
• SMC reset inconclusive; PRAM reset inconclusive.
• DFU won’t assert (even shorting DFU pads); Configurator on host Mac won’t see it.

Parts/Areas Touched

• CD3217: replaced the controller on the formerly unstable port (amp-bouncing one). Side-matched. No change.
• RA602 (0 Ω): replaced (PPBUS_HS_GPU → PPVIN_GPU_COREREG_VIN). No change.

Please help

You should try replacing the other CD3217 too.

Try this for DFU:

Will do, I just ordered more of the chips because I’m out. With Labor Day it may take a little extra time, and I’ll try the DFU method you provided and update you after. Thanks for the advice.

It may be time to give up on this one but I’m seeing if you have any ideas. I swapped the other CD3217, now all ports read 5V 0.3A steady, no fan spin, and DFU still doesn’t work. I swapped it twice in case I messed up or something the first time. Please lmk if you have ideas for something to try or check before giving up. I know you’re not supposed to spend so much time on one job but I recently started doing it for other local businesses and I guess I was really hoping to fix this one and look better.

Looks like something in the board was affected by the hot air.
If you soldered correct CD3217 chips, it should to switch to 20V, as before.