MacBook Pro A1990 doest turn on

[noz_1974]

Hello ,

I just bought a broken MacBook Pro A1990 , year 2018 board number : 820-01041A.

The previous owner of the MacBook told me that a little bit of water touched the left hand USB C port and the computer shuts off. He took it to the lab and they said that the CD3215C is bad and need to be replaced but they are not good at this .

Anyway , I bought this computer and start examine it , here is what I found ( battery unplugged :

1. All the USB C ports are stuck in 5V / 0.4A
2. PPBUS_G3H is present 12.3V
3. PP3v3_G3H is present 3.3V
4. PP3v3_S5 is 1.3V
5. PP1v8_G3S is missing


I am not familiar with the power sequence of a MacBook Pro.

So, anyone can guide me what do to next ? can it be a problem of T2 chip ? I thought that if there is a problem with the CD3215C so only the damaged chip will show abnormal behaviour .

Thank you for your help.

[mon2]

Re: MacBook Pro A1990 doest turn on

4. PP3v3_S5 is 1.3V

Meter in resistance mode (2k).

Red meter lead on PP3v3_S5; Black meter lead to ground.

What is the resistance to ground for this power rail ?

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hey,

Unfortunately, I tried to measure P3v3MAIN_PGOOD and by mistake, I think that I shorted P5VP3V3_VREG3 to PPBUS_HS_3V3GH_T ( C7653 and C7650), As a result the USB C meter still shows 5V but the current is changing between 320mA to 470mA. does it mean that I shorted U7650 ( PMIC )?
I need to solve this issue first

[mon2]

Re: MacBook Pro A1990 doest turn on

For the 20V power delivery contract to be selected, each USB controller needs to be communicating.

See here on a logic flow that should be applicable to your logic board:

https://logi.wiki/index.php/CD3215_Bootup_Sequence

and

https://logi.wiki/index.php/CD3217_a...er_on_Sequence

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Thanks, Mon2,

About the short that I did, Do you think it's solvable by replacing the U7650 ( PMIC TPS51980A) or the short may kill the board?
For now, I am looking to buy TPS51980A from a reliable website and not from AliExpress.

Thanks again,

Nitzan

[mon2]

Re: MacBook Pro A1990 doest turn on

If you suspect the operation of U7650, confirm the voltage present @ L7660 with the adapter connected.

What is the voltage reading (PP3V3_G3H_T) ?

Also confirm the voltage @ P5VP3V3_VREG3 .

Just measure it carefully.

Note that BOTH of these rails are @ 3v3 so likely there is no harm done here and your U7650 is working fine.

On the replacement note, not every shop in China is evil. Yes for sure there are bad seeds over there and know this from personal dealings but they are rather easy to spot most of the time. Work with a high rated shop on Aliexpress and you will be fine. We have done lots of purchasing for our OEM designs and repairs with success.

If you truly need to source this regulator, can find you a good vendor that we have used for the same part. Most of these shops are family run businesses and are not bad people.

First post the voltage readings for the above.

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Thanks Mon2 ,

I will get thinner probes for this measure and keep you update.
Two questions and one request :
1. what do you mean by , "with the adapter connected" ? do you mean with USB C meter connected ?
2. The PPBUS_HS_3V3GH_T was actually 12.3V when i measured it and not 3.3v, and i think its correct because its PPBUS . After that i shorted the signals as i mentioned the voltage on PPBUS_HS_3V3GH_T start to fluctuate from 12.3V to lower values as 7V and even lower.
3. please send me a link to the recommended AliExpress store .

And once again , thank you very much . you are so helpful .

Nitzan

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hey Mon2,

I did double check ( search in the web) on the correct value which suppose to be on PPBUS_HS_3V3GH_T, and its 12.3v as i said. So as i said , my first measure before the short was 12.3v and now its fluctuate.
its mean that i did short between 12.3v rail to 3.3v rail.
Nitzan

[mon2]

Re: MacBook Pro A1990 doest turn on

Hi.

1) Yes, you are correct. My error, the PPBUS_HS_3V3G3H_T is indeed as you are reporting. Schematic notes to be ~13v1. If this voltage rail is fluctuating, try the following first -

Remove all power to the logic board. Meter in resistance mode (2k is ok).

Red meter probe onto C7650 (your favorite part); Black meter probe to ground (shield is ok).

What is the resistance reading ?

It is very possible that the brief shorting event has caused some internal damage to this regulator but post your resistance on this rail.

2) The USB C 'ACE' controllers communicate to the host over the I2C bus. Let us confirm there is no issue with the I2C communication lines.

Leave the meter in the resistance mode. No power to the logic board.

Check & post the resistance to ground for each of the following parts:

R5246
R5200
R5201

It is important that there is no fault on these lines.

These 'ACE controllers' communicate with the external Type C power adapter over the CC1 & CC2 lines. Each such controller offers these lines.

Check the resistance to ground for each controller using the meter. Make a note of each pair of values for each controller.

Also consider to use the DIODE mode on your multimeter. With the DIODE mode (still with NO power to the logic board), place the RED Meter probe (yes RED meter probe) to ground.

Black meter probe onto the CC1 line of the first 'ACE' controller. Record the reading. Repeat for the CC2 line and so on for each of these controllers.

Looking for blown out parts that could cause the halt of the 20V contract negotiations.

3) The replacement 'ACE' controllers are available here:

https://www.aliexpress.com/item/3296...rch-mainSearch

(same supplier whom we have used many times in the past)

The part listed in the schematic is the 'CD3215A' - be sure to share your logic board details and reference this part. Request that the part be 100% new and original - which is true from this vendor. There are other shops that may discount the part a bit but will ship a refurbished part. They pull them off logic boards and reball them. In our opinion - how did they test the pulled part ? Recommend a brand new part where it is possible. There are variations of this sacred part so be sure to remain with what is recommended in the schematic.


We have also purchased the same TPS51980A parts from the a few times over the past 2-3 years:

https://www.aliexpress.com/item/4000...19e9314fHSpcHB

It is a good idea to pick up a few of these as they are very low cost (in Asia).

Also buy a few spares for the replacement Type C controller ('ACE controller') to fix the faulty USB port that is preventing the 20V power delivery contract.

1. what do you mean by , "with the adapter connected" ? do you mean with USB C meter connected ?

Adapter = your Type C power adapter. Yes, also recommend the type C meter that will show the voltage & current draw which is a great way to gauge the operation of the logic board. Watch many of the Louis videos on youtube for inspiration / guidance. Review @Piernov's Logiwiki and also the Repairwiki for more details on logic board repairs.

On the Type C meter - they are not all created equal. We have a few in the building and almost thought we had a dead to the world logic board with 0mA current draw. The meter was not as good as an earlier purchase which showed us that we were drawing ~14mA (better than being totally dead) for an ipad repair. For now it may not be critical but just a FYI.

Use plenty of flux and hot air with low air pressure (so other parts do not go flying). Never force the surface mount part OFF the logic board. When the solder is melted, a gentle nudge of the tweezer will make it move. THEN you are ready to assist to lift the part but again very gently. Otherwise, the PCB pads can be ripped off the board which makes it a larger mess to repair.

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Wow ,
So many goodies, thanks.
I need to bring my thinner probes for those jobs.
will do it and keep you update.

thanks a lot

Nitzan

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hey Mon2 ,
As I promised here are the results :
the resistance on C7650 is 126.4K

The resistance on R5246,R5200 and R5201is above 212k those resistors are connected to the same net PP1V8_SLPS2R, so no short.

I checked CC1 and CC2 in resistor mode and in diode mode for U3200 and U3100, the reason for that is that originally a little bit of water touched the USB C port of those chips.

U3200 :
CC1 :
resistance: 610k
diode mode: 0.824v
CC2:
resistance: 612k
diode mode: 0.824v


U3100 :
CC1 :
resistance: 613k
diode mode: 0.844v
CC2:
resistance: 611k
diode mode: 0.861v

So what is the next steps?

Thanks,

Nitzan

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hey Mon2,

I also forgot to mention that before the short accident ( C7650 ),
U3200 gave me three out of four LDO rails.
3.3v , 1.8 x2 were exist ,except from the 1.1V LDO rail who was missing.

Nitzan

[mon2]

Re: MacBook Pro A1990 doest turn on

Hi. The readings look ok at this time.

When you mate the external power adapter, do you see power cycling of the 5V rail ? The first post notes 0.4A but then does this power rail cycle down to 0A and repeat ?

Do you see the same power up cycling with the next USB port ?

Do you know which CD3215 is faulty (U3100 or U3200) ?

Does one of the CD3215 get hotter than the other with the external power adapter ?

Be careful with the removal / insertion of the external adapter and try to remove with a straight in / out action. This is to prevent any cross shorting of the Type C contacts which can cause damage to the CC1 / CC2 lines as they short to the 20V rail but for now, you are not @ 20V. The more current designs of the controllers are tolerant to such high voltages without damage.

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hey Mon2,

I opened the Board View and the schematic of MB: 820-01041A.
From the board view U3200 is the Master chip which connected directly to the SPI ROM, and U3100 is the slave which connected thorough UART port to the Master in order to connect the ROM.
In the schematic U3100 is the salve and U3200 is the Master.
What do you think?
Nitzan

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hey ,

As I mentioned before,
I was stuck on 5V with current draw of 380mA to 400mA no boot looping.
I think that the faulty CD3215 is U3200 , because the Master / Slave thing that I wrote .
About the replacement , I have access to BGA soldering machine, I just ask my self if it will not hurt the Touch ID chips and components.

Nitzan

[mon2]

Re: MacBook Pro A1990 doest turn on

1) You actually have 4 'ACE' controllers on this logic board. Can you confirm the same checks on the additional 2 * 'ACE' controllers @ UB300 & UB400 ?

These will be on the opposite side of the logic board and near the USB connector paddle boards.

On this note, do you see the same current draw across the use of the other USB Type C ports ?

2) Also interested in the resistance reading to ground from my post # 2.

Meter in resistance mode (2k).

Red meter lead on PP3v3_S5; Black meter lead to ground.

What is the resistance to ground for this power rail ?

This rail is very low and there must be a reason for the pull down of the voltage.

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hey Mon2,

1. , yes all the 4 USB C ports were stuck on 5V with current draw of 380mA to 400mA , no boot looping.
2. The resistance for PP3V3_S5 is 19.23K.

So what do you think?

Nitzan

[mon2]

Re: MacBook Pro A1990 doest turn on

What are the diode and resistance readings for the CC1 & CC2 lines of these 2 type c ports ?

The resistance of the power rail is good but the voltage is still low.

[noz_1974]

Re: MacBook Pro A1990 doest turn on

Hello Mon2,

For U3100 and U3200 I already wrote.
If you mean about UB300 & UB400 , I will disassemble the board I get back to you ( didn't want to do it till I get the replacement parts, and didn't want to peel off the black cover ). So , did you mean the reading for UB300 and UB400 ?

once again thanks a lot

Nitzan

U3200 :
CC1 :
resistance: 610k
diode mode: 0.824v
CC2:
resistance: 612k
diode mode: 0.824v

U3100 :
CC1 :
resistance: 613k
diode mode: 0.844v
CC2:
resistance: 611k
diode mode: 0.861v

[mon2]

Re: MacBook Pro A1990 doest turn on

Do you see the 1v1 (BMC power) rail on U3100 ? This is very important rail needed for the power delivery negotiations using the CC1 / CC2 lines.

If U3200 is missing this voltage, remove all power and check the resistance to ground of this off rail @ C3204. If the resistance is very low then consider to remove C3204 off the logic board and check the resistance again for this rail.

It will be a good idea to source more of these ACE controllers from Aliexpress for replacement.

Use high temp tape to protect the surrounding SMD parts so they do not go flying. Use lots of flux - dab it all over the top surface of this part and it will start to flow as it heats up. High heat and low air pressure works for us. Increase air pressure if needed but in small amounts.

Use a tweezer to assist you - gently nudge the IC at the corners to see if the solder is melted. Gentle nudge is enough to confirm this and keep applying heat in circles to hit all balls under this device. Then flip it over once the part is swimming in the melted solder.

Clean up the pads with more flux and solder braid with a high heat soldering iron and flat blade tip. Lots of youtube videos on this topic. If the solder braid gets stuck - STOP!!! else the BGA pad will be ripped off the PCB layout and then it is often game over. The braid acts like a heat sink so do use the proper tools. Ideal if you have a local shop that works with this stuff like a good cell phone repair place.

Forgot to blab about soldering on the fresh part...you will use low temp (aka solder with lead) to tin over the balls that you just cleaned up. This is with a blob of melted solder hovering over each pad. You do not actually touch the pad surface with the tip of the soldering iron but run over the pad surface to coat it. Then solder braid with flux again to remove the bumpy surface. Next, use alcohol (Costco pharmacy stuff is great) and clean up the residue / flux so you can view each shiny pad surface. Lots of videos on youtube on the process.

Be sure you note the location of pin #1 of this IC for the replacement.

Apply lots of flux before getting started on the fresh replacement.

Then apply at about the same location as the original. Hot air again and when all is well, the part will sink into position and self align.

Use the tweezers again to nudge the corner, gently. This is to 'self-align' the part onto the pads. 'Make it dance' as Louis says in his videos. Then remove the hot air and allow the logic board to cool down to test again.