CircuitPython in 2022

As every year, it’s time for a yearly summary and plans of all things I do that are related to CircuitPython.

The Projects

This year I made a record number of keyboards, and also greatly improved my keyboard library, uKeeb. Some features were added, because CircuitPython now supports them: CapsLock LED, disabled disk and console, NKRO and BOOT modes, keypad module. Others I added as new keyboards needed them: proper hold/tap with rollover timeout, layer switching improvements, mouse support, LED animations.

My FluffBug robot is now at hardware version 5, but software-wise there was no progress. I got the per-leg inverse kinematics done, and I should write the walking code, but I can’t find the motivation.

The MIDI Ocarina is now at version 3. I gave up on the barometric pressure sensor for detecting blowing, as its recovery time was poor, and instead switched to a PDM microphone using audiobusio. This works great. I still need to write the MIDI code for it.

I tried the new camera code, but I couldn’t get it to work with the camera breakouts that I have. Either I’m using wrong pins, or something broke in the IDF version I used. I have no energy to investigate this further.

I also gave up on making the uGame console work with newer versions of CircuitPython – the changes in interrupt handling and time module related to adding sleep made the SAMD21 port work so erratically, that I simply gave up. I did fix the PewPew code by avoiding using the time module altogether.

There is also a but in the PWM code on SAMD that makes it hang completely in a loop waiting for a register flag that never gets set. I got that far debugging it, but why it never gets set is a mystery to me, and nobody seems to be interested in fixing it. I worked around it in Stage library by simply setting the backlight pin high, not using PWM at all, but I got bitten by the same bug recently in the ocarina code.

The Future

This year I want to finish FluffBug at least to the point where it walks and uses the leg sensors and the distance sensors. I wish I could get the camera to work with it as well. I might also try a gesture sensor, and an accelerometer.

I received a nice ESP32-S2 development board with a display from MakerFabs, and I want to get the display working with displayio. Unfortunately it’s a 16-bit parallel, with non-consecutive data pins, so I would need to work on the relevant module first. No idea how much work it would require.

There is a cool watch project using a SAML22 microcontroller, which should be very similar to SAMD21. The author mentioned that he wants to get a port of CircuitPython working on it. I’m very interested in this, because that chip also has an LCD-driving circuitry, which would allow me to make even cheaper version of the PewPew console, with a custom LCD screen.

I might be making another handheld game console, with a 3.2” screen and ESP32-S2 as the brains, but I’m not entirely sure yet. I’m doing research on how viable it would be, and collecting parts.

Conclusion

This year brought some cool developments around USB and keyboards, but also a lot of frustration with keeping my existing projects working despite new bugs. The SAMD family seems to now get very little attention, and it suffered greatly with the addition of sleep support. Not being able to find and fix those bugs is very frustrating, and drains energy I could use on projects instead. The new boards are ridiculously powerful, but considerably more complex to use in projects, unless using a ready development board. I think I will focus on the Lolin S2 Mini boards for my projects this year, however, while I mentioned some things I hope to work on, I am aware that I have very little energy left for this. Hopefully my health and the overall global situation will improve.