A1989 2018 13” Touchbar MacBook Pro

Confirm Touch Bar is causing crash by checking Panic Log, look for DFRD. Usually you will see "panic(cpu 1 caller 0xfffffff02856d6ec): userspace watchdog timeout: no successful checkins from dfrd in 180 seconds".

Usually will occur soon after booting up. The trackpad will freeze, then macOS will restart. Restarting results in the same issue. You may also experience:

1. TouchBar not working,
2. No known damage,
3. No damage detected in Apple Diagnostics or Disk Utility,
4. Same issue in safe mode.

Solutions:

1. Create a cronjob to restart the Touch Bar every 3 minutes,
2. Disconnect Touch Bar from logic board (this may not solve the issue in some cases),
3. Replace the Touch Bar or MacBook Pro enclosure.

Set Up Cronjob

1. Open Terminal,
2. Type crontab -e, then press i to 'insert' using vim,
3. Type */3 * * * * /usr/bin/pkill "Touch Bar agent" &>/dev/null; /usr/bin/killall "ControlStrip" &>/dev/null;
4. Then to save press ESC (you will need an external keyboard as the Escape key is on the Touch Bar), then type :wq
5. All done, you can check the cronjob saved in Terminal by typing crontab -l

Inject 1 V (increase if no reaction), watch for hot spots with thermal camera. Usually it's a bad tantalum capacitor.

Measure PPBUS_G3H to Vcore coils (coils around CPU/GPU) IF those measure under 1 Ω (0.3–0.8 Ω) then you most likely have a direct 12 V short to CPU/GPU. On this board this situation is always a no fix.

• 5V 250mA on all ports and T2 warming slightly. If you've replaced a CD3215 and everything else seems fixed up it may be a fake CD3215.

• 5V 0.01A Check Short on PPBUS_G3H.
• PP3V3_G3H missing due to a short circuit on the input to PP3V3_G3H creation circuit.
• PP3V3_G3H missing due to a short circuit on the rail itself(refer to symptoms for repairing short to ground on PP3V3_G3H below).

video: https://www.youtube.com/watch?v=eL_5YDRWqGE

• Corroded LDO cap for PP3V3_G3H around a CD3215. On right side of board, check under shielding by SSD under the right CD3215 for hidden corroded cap.
• PP3V3_G3H missing due to CHGR_EN_MVR missing due to bad ISL9240. ISL9240 is not available for purchase. Donor would be required. Watch this troubleshooting video on similar 15" model (A1990) https://www.youtube.com/watch?v=HJ2jyo7pAmE
• Bad CD3215 video: https://www.youtube.com/watch?v=7n-R-TUrohg
• Bad CD3215 ROM chip (U2890)
• Corrupted T2 firmware.

If the issue is a bad CD3215, one way to get an idea, is to see which does not boot loop. When you plug the charger in with the USB-C current meter, it will turn on for 2–3 seconds, then turn off and turn on again. Whichever port doesn't boot loop usually has an associated CD3215 that is bad. Also, the bad CD3215 will occasionally get hotter than the rest.

T2 firmware can also cause the machine to be stuck at 5 V. It is however, important to rule out other possible causes for the issue before re-flashing T2 firmware. Typically machines that have corrupted T2 firmware will be stuck at 5 V 0.0 A and typically will have PPBUS present and PP3v3_S5 missing. PP5V_S5 will often be present. T2 firmware can be re-flashed via Apple Configurator 2, however this will erase data in most cases. Try a "soft" re-flash first by clicking Advanced options → Revive Device in Apple Configurator 2.

I had an 820-00850 board with 5v / 0.1A steady on all four ports. Tracked it down to a damaged end cap on R7075 which is CHGR_VDDP line to U7000. Replaced the resistor and now have 20v on all ports and fully booting. I later confirmed this on another board by removing R7075 and noting the same exact result.

If PP3V3_G3H had a short circuit, but the board stays at 5V and PP3V3_G3H is still missing, it can be due to R7900 (0 Ohm Resistor between PP3V3_G3H_RTC and U7800 PMIC, that takes PP3V3_G3H_RTC input and outputs it as PP3V3_G3H). Short cuircuit on PP3V3_G3H can burn out R7900 (may be visually intact). Check R7900 resistance.

• PPBUS_G3H at 12.60 v, 400-500ma

Burnt U3000, there was a hole in U3000. The chip directly connects to the USB-C connector J3300, a bad charger or cable will fry it.

• Trackpad not clicking, force DFU mode not working.Faulty U3000 (but not short).

• A CD3215 is short circuited to ground, will show up instantly on thermal camera which one it is.
• CB300 corroded, but you can't see it because it is hiding under the metal around the NAND. LOOK CLOSELY!
• Some other random component, but check CB300 first.

Check PP20V_USBC for each port's resistance to ground.

USB-C charge port is shorted to ground, inside the charge port. Replace the charge port, it is modular. See if it changes.

First, try a known good battery.

If it doesn't work, check the battery cable going to the board. Sometimes there is corrosion here, sometimes the customer tore the bar on the connector off the connector.

In some rare cases the SMC still is a factor in battery recognition on this board.

Make sure the issue is no backlight, and not no image!

• Blown backlight fuse rarely if ever happens. This board has a current sensing circuit in place that keeps the backlight fuse from blowing if there is a short to ground on the backlight line.
• Short to ground on backlight output due to bad capacitor on output.
• Bad screen. If backlight voltage is 49–55 V, bad screen.
• Corrosion on current sense vias/traces for backlight to U8400 LP8548

Check the LCD cable and the connector on both the screen side and the logic board side. The backlight power line is right next to the eDP data line for the image. Any minor corrosion here results in 30 volts or more going to the DisplayPort mux, U9850. U9850 is what the dedicated GPU and the CPU's integrated GPU go to before it goes to the screen.

This is a good thing, unlike the 13" machine where the backlight voltage destroys the CPU, here it destroys a cheap chip. Replace U9850, and the screen cable, and the connector(on the screen side and the logic board side) and it will usually be fine.