CD3217 and T2 Power on Sequence

From LogiWiki
Jump to navigation Jump to search

T2 power sequence checkpoints

Calpe PMU is a very complex chip with multiple functions and buck converters inside. There are numerous things to check before you replace it if you do not have T2 power.

  1. primary power supply is PP3V3_G3H_RTC. This line is different from PP3V3_G3H supply and it is very important to understand that Calpe passes 3V3_G3H_RTC through to PP3V3_G3H before it issues PMU_VDDMAIN_EN which is a main enable signal for 3/5V regulator. It seems that it is issued once we got 20v on charger and ISL9240 switches into buck mode thus it is T2 firmware dependant.
  2. one of the main checkpoints to look at is Calpe LDO section(Page called PMIC LDOs in schematics). everything except S5 outputs (which are switched proactively by T2 via i2c interface later) should be present.

CD3217 checkpoints

If a CD3217 does not have the LDO outputs, test the lines for shorts, otherwise the chip is missing power or bad. Data sheet of closest Ti chip.

A bad CD3217 can often cause a loop at step 10. All of the CD3217s need to be fully functional and communicate with the SMC before the CD3217 connected to the charger will request 20V

These documents are for specific boards, but will likely apply to most T2 Mac. Any net names and designators will be specific to the board it is from.

All 4 port macs require right side to be sourced from donor. Reason for that is preconfigured custom i2c address which is unlike CD3215 is hardcoded in internal bootrom. Left side uses straps to set i2c address but if you put OTP-Configured cd3217 from right side it wont work.



T2 Power rails resistances (820-01700)


PP0V82_SLPDDR 30-40 Ohm


PP1V1_SLPS2R 1-2 kOhm


PP1V8_SLPS2R 4 kOhm

820-001700 CD3217 Compatibility issues

4TBT port devices have mixed OTP/Strap configured i2C address for CD3217. On CD3215 each controller got its address with debug straps around it, however logicboards with more than 2 CD3217 use OTP (firmware configured) address. There is no known method of changing this configuration however it might be possible with devboard for similar TPS Chip, so most likely, it will not be ever possible to configure chips from other devices to be used on 820-01949 or 820-001700

820-001949 Addresses

Left side of the laptop uses hardware (Strap) configured addresses, right side gets address from OTP config burned into CD3217. Thus, 1949/01700 cd3217 is interchangeable but you need to match right side.

If you have a logic analyzer, you can connect to the i2c line on the debug connector to see which exact port is malfunctioning. This might be also helpful if you have a bunch of randomly sourced cd3217s and you wish to find one with the right address.

Information below might be obsolete/incorrect since it was more or less speculation based on repair experience, not actual board / datasheet analysis

CD3217s from 2020/M1 boards will not work.

Symptoms of compatibility issue: PPBUS ~12.30, all ports are switching to 19V with T2 boot/or one side will turn 19v, another stays on 5v (19v 0.05-0.1A) there will be a visible heat spot on T2 1st pin corner (inner side since T2 itself is approx 2 smaller than IC)

Device will start from charged battery, there will be no issue with usb devices but charger wont be recognized.

Generally 001700 behaviour is very inconsistent. some boards might start with only one cd3217 present on board (should be the one connected to tbt rom of course), some will stuck in 5V 0.2A, some will stuck in 19v 0.1A(basically same power consumption).

Probably there are multiple revisions of the board, images pending. If you can add some more info, please report your revision / cd3217 behaviour below.

T2 PMIC 001700 revision 1(black dot inductors)

Revision 2 (white dots)

820-001700 CD3217 revisions list

Case 1: MacBook Pro 16" 2019 With CD3217 rev. 9AP0NVQ No power with one side of newer 12P1FXQ(took from 2020 board), Symptoms: quick T2 boot, 19V mode + hot T2 Corner on original side, 5v 0.2A on another. PPBUS 12.3V


CD3217 B12 09P09CQ

Possibly alternative revision with TI logo on chip, CD3217 B12 02P04ES


CD3217 B12 97P146Q

Small T2 top side diode measurements

There are more power lines than 1v1slps2r, they just somewhat got lost in the process of uploading. Maybe I'll fix that later. Those are in case you wonder if you have a broken line between T2 and it's RAM, or have some pins torn.


T2 types

А2159, А2289 - Big T2 Package (PMU is also different!)

А1932, А2179, А1989, А1990, А2141, A2251 - Small T2 Package

Important note: T2 have differences in the supported amount of SSD. T2 supporting up to 512GB has 1GB RAM, T2 supporting up to 1TB and above has 2GB RAM. When replacing RAM, it is important to take into account the amount of RAM, otherwise the replaced memory of the wrong size will not work. Also, if you want to increase the size of the SSD (for example, from 512GB to 1TB on the A2141), it will not be enough to replace only the T2 RAM from 1GB to 2GB, you will need to install the T2 that can work with 2GB RAM, in other words, you need a T2 that originally came with 2GB of RAM.

Known types of RAM + APL number:

  • APL1027 with h9hknnn8kumuvr 1GB Hynix Chip.
  • 339S00467 - 2GB
  • 339S00533 - 1GB

There is also BOM table in schematics, but it listed apple part number(339Sxxxxx) does not seem to correlate (based on 5 boards). However, it lists 2 different Ram manufacturers, Hynix and Micron.