CP2102 based UART not functioning properly

[trevor229]

Hello, first I want to take a second to thank you all for the amazing work you guys have done here, incredible work!

I am attempting to replicate the schematic shown in the diagram with some generic level shifter modules with no luck. The issue I am having is that when the low voltage side is disconnected, the high voltage side floats up to B+ at around 20v. Once the low voltage side is connected (TX pin) it drops to 3.6v and any data send via UART stays at that voltage level.

This has happened with 2 brand new boards and any battery I tried. I suspect these level shifters aren't the proper device to use considering they are designed for open drain signalling such as I2C. What circuit was used in the video? I believe I would have better luck with something similar.

Also FWIW, I am using a Silicon Labs CP2102 based USB to UART adapter and it seems to be functioning properly, outputting the proper sequence via logic analyzer capture.

[mnh-jansson]

This is the setup i have currently. I use a step-up converter for the high level voltage instead of VBAT+ becuase i was testing communication with VBAT+ lower than 9V. Is the USB to UART adapter you are using 3.3V?

[trevor229]

Thanks for the reply. It has solder pads to select between 3.3v and 5v but I have left it in the 5v configuration. I wouldn't be too surprised if my level shifters are just garbage regardless.

[mnh-jansson]

Change it over to 3.3V and try again

[trevor229]

That seems to be producing better results. I am getting about 13.6v out of the HV side towards J2, and J1 now does something instead of staying still. Still no actual data response yet. Here is what I got while scoping and also a logic capture.

[mnh-jansson]

Could you confirm that m.idle() sets J2 to 0V and m.high() sets it to 13V?

Also, what battery are you testing?

[mnh-jansson]

Added note on the 3.3V in the README.

[trevor229]

I can confirm that J2 does follow both m.idle() and m.high() when issued. Idle drops to 0v and high rises to ~13.6v

I verified this on an XC6.0 dated Jan 14 2021 with a OneKey symbol. The indicators light up and stay illuminated while the interface is hooked up to it.

[EmperorArthur]

@mnh-jansson Why does the wiring diagram have DTR connected to RXI? That's not normal, and can easily be the cause of the problems. Especially since some drivers and boards will continually drive it high or low.

Standard practice to leave DTR disconnected, and disable hardware flow control.

[mnh-jansson]

@EmperorArthur It doesnt, DTR is connected to TX. This was to workaround an issue we found with fake FTDI chips not responding to break condition. We use break condition to keep J2 low in idle, otherwise the battery will go into "dumb charge mode"(legacy charge signal). I wrote a little note about this, but i guess i can revert the picture to avoid confusion

[EmperorArthur]

@mnh-jansson Thanks. That's a pretty clear note too.

@trevor229 Can you post a picture of the adapter you're using. Also, if you disconnect RXI from your board, does the battery response change? Good boards can handle the 13+V input, but cheap ones might do funky things.*

* Technically, it could let the magic smoke out of your PC, but that's highly unlikely.

[trevor229]

Here's my adapter and setup. I went ahead and ordered a genuine FTDI FT232 based adapter that will be here tomorrow to try out, always good to have another anyways and this one uses a jumper for 3.3/5v instead of solder. FWIW I am running on Linux Mint 22.1

[Spud2233]

I am having the same issue, I have a similar cheep aliexpress level shifter and a CP2102 serial adapter.

[mnh-jansson]

I suspect the break condition is the issue here. If anyone with the CP2102 can confirm if the m.high() and m.idle() function work as expected, this would help alot in troubleshooting this.

[Spud2233]

I retested my setup using the internal UART of a raspberry pi 3B and got the same result so I threw my logic analyzer on.

Pi 3B internal UART (GPIO14&15) with J1 connected to RX & logic analyzer.

Pi 3B internal UART (GPIO14&15) with J1 connected to the logic analyzer only.

USB CP2102 adaptor with J1 connected to RX and logic analyzer USB CP2102 adaptor with J1 connected to logic analyzer only

From what i could see it looks to be an issue with the RX side rather than the TX as both the CP2102 and raspberry pi serial interfaces hold the RX pin at ~1.6V-1.8V while at idle.