2018 15" A1990 Macbook -defective CD3215C00?

[BevanJ]

Hi all,

Just one of the ports on my Macbook just died whilst flashing an Arduino.

There is no 5V usb power out of the port when plugging in a peripheral, and the computer will not charge through the port either. The other 3x USB-C ports all behave normally when supplied with power, or when delivering power.

Another thing of note is that every now and then, the computer will spit a warning that one of the USB ports is drawing too much power, even with nothing plugged in.

I've got a copy of the schematics for this board. From what I can see, both U3200 and U3100 are fed with 5V for peripherals from the PP5V_S4_X_USBC bus. I'm guessing that through some sort of internal MOSFET switching in the U3100 and U3200, this is put out to the USB port?

Assuming the above is correct, the PP5V_S4_X_USBC 5v bus is fine - given the other port works fine.

Given that it's just one port that has been knocked out, should I change the CD3215C00 associated with it?

I gather that all the CD3215 chips must be operating for the computer to charge through any of the USB-C ports.

I've got a descent binocular microscope at home, along with a really good iron (HAKKO, with micropencil), and Quick 861DW hot air station - I usually use this kit for SMD circuit development and design, but never thought I would be doing surgery on my macbook.

I've just ordered some CD3215C00 chips and a reballing stencil off aliexpress.

Can anyone tell me if the XA(U3100) or the XB(U3200) port is the port on the LH side of the computer, closest to the screen?

Cheers for your help and advice,

Bevan

[mon2]

Re: 2018 15" A1990 Macbook -defective CD3215C00?

Hi.

1) do not worry about the actual USB power delivery controllers from TI (ACE) on the logic board at this time. Yes, each ACE controller must be communicating with the host before the logic board will dial up the 20V power delivery contract to power this box. That is, you may be ok at this time.

2) Instead, review the one-time fuses (white in color) near each of these ACE controllers. Meter in DC volts mode. Check the voltage on each fuse - one both sides.

The fuses of interest are F3000 and F3001. See attached.

One fuse is for the left side of the logic board; other fuse is for the right side. You should check both.

The power rails you need to check are :

PP20V_USBC_XA_VBUS

PP20V_USBC_XB_VBUS

What are the readings for both - again - check both sides as one of these fuses is likely blown.

Post your results. Pretty sure you can source these fuses from Arrow.com - that is where we recently did the same for a blown backlight fuse.

0603 SMD size 32V @ 6A

If indeed you confirm that one of these fuses is blown, we may be able to drop ship them to you from Arrow with free shipping.

[BevanJ]

Re: 2018 15" A1990 Macbook -defective CD3215C00?

Thanks Mon2,

Are those fuses still used for supplying 5V power out of a USB port though? From how they are arranged feeding MOSFETs Q3100 and Q3200, it appears they are only for switching the 20V supply onto the port, and in turn for charge protection?

I can definitely get my hands on 0603 32V 6A fuses, if I don't already have them at home. I'm ordering stuff from RS frequently anyway.

PPDCIN-G3H feeds the battery charge circuitry.
PP20V_USBC_XA(orB)_VBUS is directly wired to the USB-C port.

For normal 5V output operation, is PP5V_S4_X_USBC switched onto PP20V_USBC_XB_VBUS through the CD3215 (linking PP_5V0 and VBUS)?

I'll give the fuses a check when I get home from work this evening.

Bevan

[mon2]

Re: 2018 15" A1990 Macbook -defective CD3215C00?

This is USB Type C power delivery stuff.

Your external Apple Type C power adapter is the provider.

It will power up, by default @ 5VDC and spit out a list of USB Power delivery contracts to the host logic board. The ACE controller will then select the most suitable contract from this list which for your box will be 20V @ 3A or 5A.

Then this voltage rail is used to power the logic board.

For your Arduino or similar consumer device, you will not have this contract negotiation going in so you will remain with 5VDC on Vbus of the USB Type A / Type C connector.

Using tools from ST or Cypress (now Infineon) - you can monitor the PD contract hand shakes and negotiations between your external power adapter and this logic board. You can also order some fair cost (often < $ 20 USD) Type C power meters which will display the final power contract using an OLED display, etc. It is a cool industry built upon what Apple has created and others have fixed.

Review J3300 which will be a molex style mating connector which hosts a dual USB Type C paddle board. It is this section that passes on or receives the voltages and USB signals from the outside world.

Respectively, there is JB500 is the other USB logic board that mates with another dual Type C paddle board.

There are some great youtube videos on how to teardown this box so that you can rebuild without losing screws

Aliexpress sells spare USB Type C paddle boards, etc.

From memory, believe that you should have visibility to the white SMD fuses without the teardown but may be wrong. Based on our first post, suspecting that one of these fuses is blown.

[BevanJ]

Re: 2018 15" A1990 Macbook -defective CD3215C00?

Cheers Mon2,

I'm fairly familiar with USB-C and how the PD handshake takes place.

I've got a USB-C monitor/voltmeter/ammeter coming.

"For your Arduino or similar consumer device, you will not have this contract negotiation going in so you will remain with 5VDC on Vbus of the USB Type A / Type C connector."

I get this - this 5V is also missing if I plug in a USB Type A device through an adapter on this port. HOWEVER, from the schematic, it doesn't look like this 5V comes through F3000 or F3001?

I'm looking at it in board view and one of the SMD fuses is on the upper side of the PCB. So, the question is - is the back USB-C port XA or XB? Hopefully it's the one I can measure the fuse of without a logic board removal.

Bevan

[mon2]

Re: 2018 15" A1990 Macbook -defective CD3215C00?

TBH, not sure which connector is which, physically. My database only had the schematic (no board view for the A1990). Will investigate further after some Netflix viewing

In tracing the paddle boards on the schematic, they show only the 20v rails as the power to the USB connectors.

The same rail is 5vdc (default) unless the power delivery handshake kicks it upto 20vdc on Vbus.

Will not hurt to review the voltage on the fuse that is accessible.

[BevanJ]

Re: 2018 15" A1990 Macbook -defective CD3215C00?

Well,

I'll be damned. I pulled the logicboard to measure F3000 and F3001. Both have continuity, and are intact.

Swapped the two paddle boards for further fault elimination, and refitted the motherboard.

And... all 4x USB-C ports are now working normally?!

And yes, I did do an SMC, PRAM, etc reset before this whole exercise.

Is there a CD3215 fault state that requires a complete power cycle to unlatch? I'm genuinely puzzled by this. Here's hoping the port keeps going.

How frustrating.

Bevan

[xytrax]

Re: 2018 15" A1990 Macbook -defective CD3215C00?

Hi everyone, I have an A1990 board but it's 820-01326 and I don't have PP3V3_S5 and I can't find whoever creates it in the schematic either. I had a previous failure in U3200 and when I replaced it it worked for a few days and then it turned off, checking voltages I found myself with PP0V9_SSD in short I solved it but goodbye to PPV3_S5. I hope you can guide me on this, thanks