820-00840 no ppbus_g3h stuck on 5v liquid damage

Roan

Member
A A1706 came in with liquid damage, alot of it. Tried cleaning the CD3215 and ISL but no luck. Removed both CD3215s after seeing brown/black goo coming from underneath it, but after replacing still no luck.

Have not worked with USB-C macbooks before so one thing is not completely clear. Is it true that PP3v3_G3H need PPBUS_G3H to function and, the CD3215s need the PP3v3 voltage line to make 5v into 20 volts? Does this mean that I start with the PPBUS short and after that start solving the CD3215 problem?

If my reasoning is bad I apologise, I wanted to show my thought process to make it easier to see when I am going wrong.
 

2informaticos

Administrator
Staff member
3V3_G3H MUST appear first, with 5V input.
Once SMC boots, will talk with CD3215 chips, to switch input to 20V.
It will also tell to U7000 when to generate PPBUS_G3H.

Post diode mode to ground for the 4 LDO outputs of CD3215 chips.
 

Roan

Member
4 outputs of L3200
1 - 645
2 - 645
3 - 1507
4 - 1507

couldnt find another LDO connected to the other CD3215, not experienced with these boards Im sorry
 

2informaticos

Administrator
Staff member
Each CD3215 has 4 (four) LDO rails:
- 1x 3V3.
- 2x 1V8,
- 1x 1V1.
Consult schematic, look inside CD3215 block notations.
 

Roan

Member
Jep found em!

PP1V1_UPC_XA_LDO_BMC = 0V (diode mode 540)
PP1V8_UPC_XA_LDOD = 1.77V
PP1V8_UPC_XA_LDOA = 1.80V
PP3V3_UPC_XA_LDO = 3.35V

PP1V1_UPC_XB_LDO_BMC = 0V (diode mode OL)
PP1V8_UPC_XB_LDOD = 1.78V
PP1V8_UPC_XB_LDOA = 1.80V
PP3V3_UPC_XB_LDO = 3.35V
 

2informaticos

Administrator
Staff member
"PP1V1_UPC_XB_LDO_BMC = 0V (diode mode OL)"
There should be a bad solder; you must get similar diode mode reading as the other chip.
I hope you've measured both chips with red probe to ground.

BTW, 1V1 LDO voltage should appear only on the powered port.
 

Roan

Member
diode mode fixed
measured PP3v3_G3H at R3001 and it's there
you mentioned the 3V3 rail should start with input 5V is that correct?
I see alot of bidirectional datalines on the CD3215's, what output line(s) can I check to know it's working to a large extent? which are the last ones to come on?

PP1V1_UPC_XA_LDO_BMC = 540
PP1V8_UPC_XA_LDOD = 500
PP1V8_UPC_XA_LDOA = 600
PP3V3_UPC_XA_LDO = OL

PP1V1_UPC_XB_LDO_BMC = 540
PP1V8_UPC_XB_LDOD = 1850
PP1V8_UPC_XB_LDOA = 1900
PP3V3_UPC_XB_LDO = OL

Is this an indication that the B chip is badly soldered? Since some the diodemode readings are relatively high?
 

2informaticos

Administrator
Staff member
You can't get OL on any LDO rail.
Be sure to check (ALWAYS) diode mode with red probe connected to ground.
 

Roan

Member
PP1V1_UPC_XA_LDO_BMC = 530
PP1V8_UPC_XA_LDOD = 490
PP1V8_UPC_XA_LDOA = 550
PP3V3_UPC_XA_LDO = 530

PP1V1_UPC_XB_LDO_BMC = 570
PP1V8_UPC_XB_LDOD = 520
PP1V8_UPC_XB_LDOA = 570
PP3V3_UPC_XB_LDO = 550

now I payed attention to my probes :)
 

2informaticos

Administrator
Staff member
The values are bit higher.
But this could be caused by bad chip, or bad solders.
Can be even normal difference.
Did you solder same CD3215C00 chip?

I bet for more damage caused by liquid spill.
U2890 and surrounding components, look clean?
 
Last edited:

Roan

Member
U2890 does look quite dirty. Cleaned with alcohol and did a reflow, but the diode number dont seem to change much from earlier. These are CD3215s from the same IC strip. bad solder could be the culprit, would you advice to put on 2 new CD3215s or first rule out other problems? Maybe when removing CD3215s checking underneath the IC could uncover some damaged traces? The solder points when I put em on seemed fine for both at the first replacement
 

2informaticos

Administrator
Staff member
U2890 is not related with diode mode readings on CD3215 chips.
But these chips need to read information from TBT ROM.
If any pull-up resistor is bad, or other surrounding components, board cannot switch to 20V.

Liquid could also affected Thunderbolt controller (U2800), which is directly connected with both CD3215 chips.
 

Roan

Member
couldnt find any out of the ordinary resistor reading as in no bad pull-up resistors

at U2800 the lower right resistors are affected by liquid damage: R3097 R3094 R2829, replacing those right now
 
Top