TPS25750EVM: USB-PD-CHG-EVM-01 - MacBook Pro M1 Not Charging

Ryan,

Are you able to capture the I2Cm bus lines in and I2C bus Analyzer (I use a Saleae Logic Pro 8)?  

Can you capture a PD trace (I use a Total Phase PD analyzer)?

I will see what I can do to find a M1 MacBook Air to test with.  I don't have one easily accessible.

I do have a Saleae Logic 4 that I can use so Yes I can supply a capture of the I2Cm bus when plugged into a M1 Macbook Pro.

I do not have anything to capture a PD Trace unfortunately.

Do you think we could make this work so I could use the Saleae Logic 4 to capture the USB-PD transaction on the M1?

github.com/.../hla-usb-pd

The Saleae logic pd decoder is not that useful.  Let's start with the I2CM trace.  You can just attach the sal file from you capture so that I can look at it.

I can interpret the I2C events to see what is happening

Here is the I2CM data capture when plugging the USB-PD-CHG-EVM-01 into a 2021 MacBook Pro M1 13".

The gap in the middle is where I pulled the USB-C cable out and flipped it's orientation.

I'm using a genuine Apple USBC cable that came with the new MacBook M1.

You can see it just keeps unsuccessfully negotiating something so Sink or Source never happens.

USB PD is working fine with other devices and older Mac's.

https://drive.google.com/file/d/1iv7Kty46Bwzix2b667eXeOgIehb1NQJJ/view?usp=sharing

** I have the SCL & SDA Labeled backwards in that data capture FYI so swap them.

Thanks!

Ryan,

I am just guessing, but this looks like there is a hard reset happening in the PD traffic so the part is continually trying to reconnect.

I also noticed that you have a very significant power glitch happening in this capture:

Can you attach connect 2 analog traces to VIN_3V3 on your system and send me the results?

That spike reaches 4V which is outside of the operating range, so you will have to identify and fix the source.

Hey Chuck, these were taken on the USB-PD-CHG-EVM-01 you guys are selling us. The only thing I have done is supply a 4 cell battery pack and your current firmware you supplied me with. 

So fixing the 3.3v source would mean the EVM design could be bad correct? 

So where does that leave us with the TPS25750 properly charging the newer M1 MacBook Laptops?

What's the next step to trying to solve this? 

Ryan,

I will reach out to the board designer of the PD-CHG-EVM and see if he has any idea where this power spike is coming from.

I am also trying to get a hold of a M1 MacBook to test with on my side.

I can probably get you the USB PD data captures of the MacBook M1 and its Apple charger if that would be of any help. 

Also I took the voltage readings on the Jumper 7 pins on the EVM Charging board if that could be a reason for the spike.

Ryan,

If you can get a PD capture that will help quite a bit.  It will tell me what is attempting to happen on the type C side.

I will work locally on the voltage spike.  It should not be happening.

I'm happy to report that out of the blue the USB-PD-CHG-EVM is now charing my Macbook Pro M1 laptop now.

I also found out by accidentally touching the 3.3v LDO that's it's bad and running at 160 Fahrenheit so it must have went bad for some reason. The getting 5.04v on the input pin and 4.99v on the output pin when it should be supplying 3.3v.

So even though the LDO's output should be 3.3v it's 4.99v and it's allowing my Macbook M1 to charge so I'm happy.

I will swap the 3.3v LDO with a new one and see what happens then but I'm hoping the issue with the Macbook was the bad LDO.

I'll report back.

Ryan,

I'm glad you've resolved your issue.  Please reach out if you have any other issues.

Regards,

Chuck

I replaced the bad 3.3v LDO for a new one and now have 3.3v output again.

Now the Macbook Pro M1 will not charge when plugged into the USB-PD-CHG-EVM.

All other USB C devices I plug in work just fine.

Any idea why having the 3.3v LDO output 4.99v is allowing the Macbook Pro M1 charging to work every time and never when the output is 3.3v as intended?

If I add a USB C dongle to my Macbook Pro M1 and plug the USB-PD-CHG-EVM into the the Dongle it charges the laptop but the Macbook says it's connected to a 52w input vs when the LDO was outputting 4.99v and connected to the Macbook just fine it would say it's connected to a 60w adapter. So I guess the USB c Dongle is lowering it's power output level to leave some power to supply it's USB ports maybe?

Ryan,

My biggest concern is that the other LDO going to 5V has potentially damaged the TPS25750.

I have a very similar Dongle to that one.  It reduces it's PDO output by it's internal power consumption.  It uses 8W of power for it's internal circuitry and provides 52W to the downstream port.

Regrads

Chuck

OK, just for more info on this issue I can confirm that I tried a original Apple Thunderbolt .8 meter cable between the EVM board and the M1 Macbook and the Mac would not recognize the charger.

I tried a new USB-C Belkin Cable and the Macbook does not recognize the EVM.

I tried 3 different USBC hubs and the ANKER did work but only supplied the Macbook M1 with 45w of the 60w.

The Hyper Plus Macbook Dongle did not recognize the EVM charger. And I tried a Belkin 4 port USB C hub with USB-PD pass through but it did not recognize the EVM when plugged into it.

So I'm sticking with the HooToo Shuttle USB-C dongle for now which does recognize the EVM and supplies the M1 Mac with 54W of power.

After more research I found that Apple has blocked a certian amount of USB-C hubs from connecting and powering the M1 Macbooks due to a previous issue they had where Macbooks would end up bricked from certian USB hubs with USB-PD pass though.

I can't find any info as to which USB hubs are blocked from connecting and which are not or how they even tell which is which if it's not by the vendor ID.

Here is a thread on the Apple Developer forum where their recent Monterey Mac OS update has stopped some USB-C devices from charging or being noticed at all.

https://developer.apple.com/forums/thread/683409

With the current firmware you supplied me with after the LDO went bad and was simply passing through 4.99v the Macbook M1 did recognize the EVM consistently and it showed up as a 60w Power Source and it did charge the M1 Macbook from empty to full so I know the firmware will work but there must be something off.

All I can do for now is try to get a USB-C PD communication trace log so we can get an idea of what is going on. I'll supply that data as soon as I can find a logging device unless you guy figure that out first.

Only thing not working now is USB BC1.2 protocol.

What is the current timeline for getting the BC1.2 Protocol working?

What is the timeline for getting the online GUI updated now?

Ryan,

I am going to do some testing on the BC1.2 on my side early next week.

I will also look into the Mac issue on our side.

I am working internally to get a GUI timeline, but I do not have one right now.