@scientiffic Without magnetometer yaw value will always drift.

Yaw value cannot be determined from accelerometer readings. Yaw value is integrated gyroscope readings which are rate of change. Integration is never perfect and this will cause yaw to slowly drift.

Apparently this was caused by accidentally writing 0x0 to AXP192_DCDC13_LDO23_CONTROL which turned off DCDC1 which provides power to ESP32 chip. The behaviour I was seeing was normal boot and then the firmware turning DCDC1 off.

I was able to revive the M5StickC by connecting G0 to GROUND and randomly doing short and long power button presses while trying to erase flash. After few tries erase flash command suddenly succeeded.

Debugging this further, I am able to get some serial output by first pressing the power button for ten seconds and then press it for one second.

After the short press I can see the boot. up sequence and then it goes quiet again. Is this the old problem of M5StickC not booting if battery is too empty?

I was in process of writing a C driver for the AXP912 chip.

Everything was going fine until I started working on the 0x33 register ie. AXP192_CHARGE_CONTROL1. I wrote the value 0xd0 ie. 0b11010000 which is supposed to mean:

0x80 : 0b10000000 -> enable charging
0x40 : 0b01000000 -> target voltage 4.2V
0x10 : 0b00010000 -> 15% end charge current
0x00 : 0b00010000 -> 100mA charge current

0xd0 : 0b11010000

After running the program M5StickC stopped responding in the usb port. USB device can be found but the connection times out.

-cut-
$ make erase_flash
Toolchain path: /Users/tuupola/.espressif/tools/xtensa-esp32-elf/esp32-2019r1-8.2.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc
Toolchain version: esp32-2019r1
Compiler version: 8.2.0
Python requirements from /Users/tuupola/esp/esp-idf/requirements.txt are satisfied.
Erasing entire flash...
esptool.py v2.8-dev
Serial port /dev/cu.usbserial-XXXXXXXXXX
Connecting......................................_____

A fatal error occurred: Failed to connect to ESP32: Timed out waiting for packet header
make: *** [erase_flash] Error 2
-cut-

I can connect to the usb port but there is no output.

-cut-
$ screen /dev/cu.usbserial-XXXXXXXXXX 115200
-cut-

I though maybe this is one of the intermittent error M5Sticks have and tried to flash another unit. After flashing same problem.

I tried updating the USB controller firmware. Update was successfull but did not help the problem. Also tried to connect G0 and GND when programming. Did not help.

So my guess is somehow AXP192 can be configured so that it does not allow either the ESP32 or serial port start.

Question is how to recover?

@heybin said in Add framebuffer library to UIFlow:

sorry, These days I am busy migrating micropython to the master from lobo micropython,so adding this feature may a couple of weeks later

Does this mean M5Stack project is moving away from Loboris for and using the original vanilla instead?

Replying to myself here. Went to eat, came back and 0x56 is found on the bus. Although something weird is going on. Setting motor speed and reading the encoder value works only if also the wheel has USB attached. I guess the unit needs the internal battery to be properly charged to function.

How is the M5 Wheel / Balance supposed to work? Do I need to connect it via the Grove interface or are the bottom pogo pins enough? I have attached the wheel to a M5Stack Fire. When scanning the I2C bus only MPU6500 and IP5306 are found.

>>> i2c = I2C(scl=Pin(22), sda=Pin(21), freq=400000)
>>> i2c.scan()
[104, 117]

If I turn the wheel off the I2C scan times out and triggers the task watchdog. So it seems the wheel is atleast connected to the i2c bus.

>>> i2c.scan()
Task watchdog got triggered. The following tasks did not reset the watchdog in time:
 - mp_task (CPU 1)
Tasks currently running:
CPU 0: IDLE
CPU 1: IDLE