Macbook Air A2337 820-02016 No Power

[Rezid]

i've got an A2337 820-02016 that has no power. No sign of liquid damage and no obviously damaged components. The board looks brand new. USBc amp meter does not turn on when plugged into either port. I've tried multiple known good chargers and usbc boards. No shorts on any of the large coils. Not super familiar with diagnosing Macbook boards so any help is appreciated.


USBC1_CC1 = 1.6v but fluctuates to .9v
PPVBUS_USBC1 = 0v
PP3v3_UPC1_LDO = 0v
PPDCIN_AON_CHGR_R = 0v
PPVBAT_AON_CHGR_REG = 0.07v
PPBUS_AON = 0.07v


I'm thinking bad CD3217 but just hoping someone more knowledgeable can help me confirm that.

[mon2]

Hi. It will help to pick up a USB Type C meter (bidirectional type) to gauge the current draw and the voltage being selected by the power delivery interface (PD). You can get these on Amazon.

Remove all power. Disconnect the battery. Measure the resistance to ground of PPBUS_AON @ fuse F5200. Measure both sides of this fuse.

If there is a low resistance to ground then there is a shorted part on this major power rail. Post the measurement in ohms.

[Rezid]

Hey thanks I've got one already. Screen on there remains black and doesn't show current draw.

Both sides of F5200 are about 2.7M ohms. Diode mode is .48

[mon2]

No short there.

For each CD3217 that is onboard, measure the resistance to ground of each LDO rail. Respectively, when you make use of a type C port, that local CD3217 PD controller must create the local LDO rails. Each CD3217 must be functional to allow for the 20V power delivery contract. We are now checking to see if there is a LDO rail that is shorted.

Do inspect the contacts of the USB Type C paddle boards which are wear out items. Often we see macbooks in the shop with cat urine or similar substance on the contacts. If in doubt, clean the contacts with IPA (alcohol from Costco pharmacy is great for this) with Q-tips. Better to just replace with a fresh paddle board if available to you. You can check Mobile Sentrix in VA for such parts with short lead time. Otherwise Aliexpress stores.

If your display on the Type C meter is dark then this implies that even the basic 5 volt rail that is offered upon power up by the power adapter is being shunted to ground. Test that the display is not dark by connecting the meter to the power adapter but do not mate with the logic board. Then the meter should show 5v with 0mA current draw (since there is no real load).

Review this great webpage if you have not done so already:

CD3215 Bootup Sequence - LogiWiki

[Rezid]

I don't know how I missed it as I measured it in diode mode before, but found that PPVBUS_USB1 was shorted. Injected 1v and DF602 lit up under the thermal cam. Removed it and measured PPVBUS_USB1 again and the short was gone. If I replace it with one from a donor board the short comes back. If left empty, the other port attached to PPVBUS_USB0 draws 20v 0a.

[mon2]

DF602 (TVS2200) is an unidirectional TVS diode so there is a polarity. Be 100% sure that you are placing the replacement in the same orientation as the original. Review the notch on the bellyside of the parts for the key.

Respectively, it should be possible to test the part before placing onto the board using diode mode. Meter in diode mode. Red meter lead onto the anode (+) = backside of the arrowhead = ground pins on the PCB; black meter probe onto the cathode (-) = signal line to be protected on the PCB.

[Rezid]

Looks like the DF602 I took from my donor board was also faulty. Swapped it again and everything is working. Thank you for your help.

[mon2]

Well done!