SMD Model F controller w/ RP2040

[jak123]

I'm wondering if anyone has taken a SMD Model F Controller and tried to pair it with a Sparkfun Pro Micro RP2040 controller. The Pro Micro RP2040 appears to be pin-compatible in all the ways that matter for this project. It would be nice to be able to run more complex configurations of QMK (with more layers and more advanced features enabled) that are tricky to fit into the traditional ATmega32U4 based Pro Micro. It would also be nice to have the additional I2C channel that is available on the RP2040 available for rotary encoders, screens, and other peripheral hardware.

I realize that getting the QMK build working on a RP2040 is a lot more complex than just a re-compile, given that parts of the cap-sense specific code is written in assembly and is intimately tied to the ATmega32. That's part of why I'm wondering if anyone else has built such a beast, or if anyone is working on such an upgrade.

[pandrew]

I was considering this, it should be possible, but I'm not working on it right now.
A couple things to note:

• Hardware wise, for solenoid support you would have to cut the "+5V" trace that goes to the solenoid header, and connect it to the RAW/VBUS pin. The rest of the board should work at 3.3V. The threshold offsets may or may not need to be adjusted.
• It may not make sense to do this. Rico's working on an RP2040-based controller: https://geekhack.org/index.php?topic=117555.0
  It is currently only the small form-factor one used in Ellipse projects, but it makes more sense to just create 3.96mm pitch variants of the Leyden Jar. When the TH and SMD Model F controllers were created, JLCPCB would not solder USB connectors for you, and atmega32u4 chip stock was very unreliable, and this was before the pandemic. Now JLCPCB solders USB ports fine, and I don't think i've seen RP2040 out of stock a single time, so the separate pro micro design doesn't make sense anymore.

[jak123]

As is evident by the fact I completely missed that the RP2040 is a 3.3v controller, I'm much more of a software person than a hardware one. My hardware experience is "just enough to be dangerous." Grafting a Pro Micro RP2040 onto the existing controller board seemed like a great way to skip ahead straight to the software bits.

I agree that a version of the Leyden Jar looks like an interesting path for production boards. I'm super interested in how it (or something similar) might play out for the Brand New Beamspring. It's still tempting to try to put something together here and now, stick it in a spare F, and see what I can get working. I finished my months-long "dream board" restore of an F AT just a few weeks ago, and now that I've been playing around with firmware configurations, the limitations of the original Pro Micro have become a bit more clear. Plus, I had really wanted to put some encoders on it.

[Ellipse]

Yes, Rico has developed a working RP2040-based controller (both PCB hardware and Vial firmware) that will hopefully replace the ATmega-based xwhatsit controllers for the new Model F and Beam Spring projects, among others' projects I am sure.

I have tested a prototype Leyden Jar controller on the Round 2 beam spring prototype and it works well in my typing tests. Rico has successfully tested the controller on a new Model F as well.