Chromebook USB-C no longer recognized USB devices after shorting

[theflyingsquirrel]

I have a Samsung chromebook Pro and one of the USB-C port have something inside and was shorting. The particular thing look like a loose soldering debris. It was at a bit off middle so I think it was shorting the vbus and data pin. I ended up desoldering the whole usbc port and removed it. Upon removoing the usbc port, the chromebook boot again. I can charge with the remaining usbc port (now I found out that when I charge with higher power charger, it does not work), but I cannot get it to recognize any USB stick I plug in. May be the handshake is gone? In the pictures, you can still see the USBC port, but it is already removed now. I replaced the PI3USB30532ZLE chip already but it is exactly the same. Anyone familiar with USB handshake have some insight into what is wrong? And what could have blown when the USBC port is shorting?

[theflyingsquirrel]

I finally have it figured out!
I traced the VBUS line from the USBC port and looking at the schematic from another laptop, I found the USB PD and Type-C current-limited power switch X5PT4 from the picture which is similar to NX20P5090UK from the schematic.
The V out pin from this chip is directly connected to the VBUS pin on USBC port.
This is a bga chip so I had to remove it in order to get access to the pin. With the USB stick inserted, I measured the Vin pin and 5V is present. I then measured the EN pin and it is 3.3V. This chip is active high so that means the gate is closed and Vout should have 5V. But so far I got 0V so this chip must have been fried.
I took the same chip from a scrap board and reballed it. It is a bit challenging since the grids are small, but I eventually got it done. After replacing the chip, with the USB stick plugged in, I measured VCC pin at 5V!
I then booted up the machine properly and it can finally detect the USB stick!
I then test the PD charger and VCC charges at 20V instead of getting stuck at 5V like before. I think this USBC port is fixed.
However, when I test the other port, the Vcc is still at 0V. I think the same chip from the other side is also fried. But now I know what to do!

So to sum it up, so far, I replaced:
- USBC port
- the 2 MOSFET chips for both CC line
- the 2 TVS diode for both CC line
- STM32 USBC controller chip
- PI3U chip
- USB PD and Type-C current-limited power switch X5PT4

And that is for one USBC port. The other side still get Vcc at 0V with USB stick plugged in.

Thanks mon2 for all your help. Without you, I do not even have any idea of where to start. You gave me method to test and debug the circuit and direction of where to look.

[mon2]

Best to share clear pics from all angles to confirm this USB mux has been soldered in properly.

[theflyingsquirrel]

Best to share clear pics from all angles to confirm this USB mux has been soldered in properly.

I did not solder in a new USBC port yet.
But the other USBC that is connected with flex cable to this PCB is not detecting anything.
I was thinking the USBC are probably connected to the same hub.
I added a picture with the flex cable.

[mon2]

I mean the PI3USB30532ZLE chip. This part is used to steer either the TOP side of the USB-C blade or the BOTTOM side of the USB-C to the logic board's USB host controller. That is, only one side will be active at any time. If this part is not soldered in properly then the USB-C port will appear defective.

[theflyingsquirrel]

I mean the PI3USB30532ZLE chip. This part is used to steer either the TOP side of the USB-C blade or the BOTTOM side of the USB-C to the logic board's USB host controller. That is, only one side will be active at any time. If this part is not soldered in properly then the USB-C port will appear defective.

I see. OK I will take more picture of it when I get to the board.

[theflyingsquirrel]

I mean the PI3USB30532ZLE chip. This part is used to steer either the TOP side of the USB-C blade or the BOTTOM side of the USB-C to the logic board's USB host controller. That is, only one side will be active at any time. If this part is not soldered in properly then the USB-C port will appear defective.

I've uploaded more pictures. It seems fine.

[mon2]

Can you check the attached?

Is this just flux?

[theflyingsquirrel]

Can you check the attached?

Is this just flux?

Yes those are just flux. I tried cleaning with alcohol and brush, but it is remain there. It is not solder for sure.
I noticed when I plug in some charger, the charging is intermittent. But some lower rate charger seems fine.
I ordered new USBC ports. I will try to put it in and when I received it and see if the other port act differently.
Do you think some capacitor or resistor is shot? Or worse, some trace are burnt?

[theflyingsquirrel]

I used a thermal camera to check when it is on. I find that this chip got really hot.
STM32F051C8U6
I searched and it is MCU 32-bit ARM Cortex M0 RISC 64KB Flash 2.5V/3.3V 48-Pin UFQFPN EP Frame.
Anyone know what it does?
May be I will try to replace it and see if it will work.

[mon2]

This is a Cortex M0+ microcontroller that will have internal firmware. Unless you can extract the original firmware, the replacement part will not function in this case because the replacement will be a blank part.

[theflyingsquirrel]

This is a Cortex M0+ microcontroller that will have internal firmware. Unless you can extract the original firmware, the replacement part will not function in this case because the replacement will be a blank part.

Can I strip it from another same model board? Would the content be unique to each board?

[petehall347]

what is this solder blob ?

[theflyingsquirrel]

It was from picture I took before. That part has been fixed already.
The stm32 got to 69 degree C fairly quickly. I am thinking it is damaged from usbc port shorting.
I removed the PI3 chip before I used the thermal camera thinking that is the damaged part.

[mon2]

IMHO, cannot see them using unique code inside of the microcontroller but not speaking from experience. I do know that the microcontroller has been compromised so there are methods (not usually available to the public) to extract code out of even 'protected' microcontrollers. As a start, yes, recommend to carefully flux and remove the parts and swap with a known good donor part. The part in the pic appears to be QFN or BGA from a quick glance so you may have to reball it if it is a BGA package.

Update - it is QFN so you are in good shape to use flux + patience to remove and replace.

[theflyingsquirrel]

IMHO, cannot see them using unique code inside of the microcontroller but not speaking from experience. I do know that the microcontroller has been compromised so there are methods (not usually available to the public) to extract code out of even 'protected' microcontrollers. As a start, yes, recommend to carefully flux and remove the parts and swap with a known good donor part. The part in the pic appears to be QFN or BGA from a quick glance so you may have to reball it if it is a BGA package.

Update - it is QFN so you are in good shape to use flux + patience to remove and replace.

Thanks. Luckily, it is qfn. I will try swapping with another one from other board to see if it works.😁

[theflyingsquirrel]

I replaced the STM32 and there is no change in performance. The chip is not heating up fast anymore but USBc port still does not detect my USB stick. Kind of disappointed.

[mon2]

Post pics of the reworked areas. Test the USB port pins using DIODE mode. No power to the board when testing in diode mode. Red meter probe to ground; black meter probe to the pin to test.

[petehall347]

could it be something silly like it needs enabling .

[theflyingsquirrel]

Post pics of the reworked areas. Test the USB port pins using DIODE mode. No power to the board when testing in diode mode. Red meter probe to ground; black meter probe to the pin to test.

I will take some picture of the rework area when I get a chance to open it again.
I used a USB C tester to test when charging and it seems that the charging voltage is almost 20V while the battery is rated for 7.5V. I am not sure if that is normal. I will attach a picture of the test result. There is 0V in data line.
I tried to measure the USBC port with diode mode, but the pin is too small and some pin are inaccessible to measure. I will have to get an expansion board to plug into the usbc port to measure.

could it be something silly like it needs enabling .

It is not likely as I have a few machine of the same model and none of them need enabling.

[mon2]

@ ~20v, the power delivery section / controller / firmware is working on your logic board. Do not disturb these sections. Next, continue to check if the same ~20v travels onto the battery charger IC power path.

Share again the markings on your battery pack. What is the voltage marking on the official battery pack?

On power delivery communication, only the CCx (ie. CC1, CC2) pins are active for the 1-wire BMC communication between the PD controller and the PD external power source / adapter. Respectively, the data lines on the same USB connector could be disconnected and PD can still function.