STM32F411 (LQFP64) and RP2040 Modules for QMK and VIA Keyboard Software Development
Already have an account? Log in.
To make the experience fit your profile, pick a username and tell us what interests you.
Hardware: Micro, Keys, etc...
Micros: STM32F411RETx & RP2040
Software Support: QMK/VIA
Switches: MX type hot-swappable
Keys: 98 Keys + Encoder Switch (includes 10 Dedicated Macro Keys)
Rotary Encoder: Volume / RGB.
Tools:
Other references:
keyboard-layout-editor.com
Will publish Kicad files and Code after validation...
stm32_64pin_dev_board.kicad_pcb_bom.csvCurrent BOM (may change)...Comma-Separated Values - 2.54 kB - 12/26/2024 at 14:39 |
|
Testing the new RP2040 Module in usevia.app
Note, encoder in now on the right, it replaces the pause keyboard button as it is never used...
Assembled boards arrived today (from PCBWay). A quick test reveals I didn't screw up as they worked perfectly, including the RGB.
The 2040 having more available GPIO's simplifies the design process and with the driver:vendor option RGB is a breeze and the flashing is easy with drag and drop, definitely the way to go for your first keyboard project.
One thing for certain, without a microscope or camera with zoom, hand soldering these tiny devices would be a major challenge for anyone and not recommended for your average hobbyist or old engineers like me.
The cost of 5 boards (3 completely assembled and 2 empty), including courier amounted to $87.34. That's about $22 per assembled board which is as good as it gets... I should have got all the boards assembled, I will in future.
Next Step:
It would be tempting to just add the electrons from the dev board to the main keyboard PCB and call it a day, but as I now have 3 perfectly good RP2040 Keyboard Modules, I think I go with a basic keyboard PCB, that is a board with only the switches, diodes, RGB leds and the 30 pin connector. This also has the advantage of being able to repurpose it for use with other modules.
There is no difficulty in hand assembling the bare-bones board as the diodes, hot-swappable connectors and RGB are easy enough to hand solder, thereby reducing the overall cost.
Working on a new version of the bare-bones board
Improvements:
Reasoning:
The bare-bones keyboard PCB is intended for development along with either the STM32 or RP2040 development modules. My main aim was to simplify development and limit potential problems by separating development into two parts, then combine to create the final keyboard.
Nothing is wasted, as I can combine any module with a bare-bone board to create a working keyboard.
I will update this log when I complete the new board...
Place holder for additional images
Expands the 1.27" connector to 2.54" for ease of testing (the 1.27" connector is too small to safely probe). The expander board also includes an optional 30 pin FFC / FPC ribbon connector as an optional method to connect future boards. All Kicad files will be published after validating operation.
Added an RP2040 Development Board to project.
This module is for testing RP2040 based keyboards and supports all the available GPIO's. It includes four RGB leds for general testing of RGB code but can drive a led strip or the keyboard RGB via the 3 pin connector.
A 30 pin 1.27" pitch connector connects to the test keyboard consisting of 99 MX Switches, 99 Matrix Diodes, 99 RGB leds and no processor... it is used for testing both STM32F411 and RP2040 modules.
Note, the 1.27"connector is too small for real world testing, so I have added an expansion/conversion board to convert from 1.27" to 2.74" and also include a 0.1" ribbon cable connector as an alternate connection method.
In an ideal world, the keyboard connector would have the same pin-out for both STM32F411 and RP2040.
Revision 2.0 of the module removes RGB test leds as it intended to be used with the Basic Keyboard (no smarts/processor)...
Echo Keyboards
As I'm working on several keyboards, I grouped them under the brand name Echo Keyboards...
Models are:
News:
Finally worked out how to use keyboard-layout-editor.com correctly to generate the required code, I was just copying the raw data instead of downloading the json file which made life more difficult...
This link solved most of my problems: https://www.caniusevia.com/docs/layouts
I have created a new git repo for the current code: here
Received two prototype (populated) boards in post. Preliminary test went well, everything appear correct, including the code for the new 6 x 17 matrix layout as opposed to the 10 x10 matrix used in my original Blackpill keyboard.
I failed to notice that QMK supports 25xx-series for SPI (and 24xx-series for IC2) so the current test EEPROM footprint needs changing.
As noted earlier, the new dev board are designed to use a 6 x 17 matrix and the QMK keyboard files require updates to reflect this. After updates to the code, the keyboard generate the correct codes but the layout in usevia.app needs a little work, the Enter key is not in the correct position ++
Update: 21, December
Corrected STM32F411 code, now all key in the correct place, ISO/ANSI switching works, everything else works with usevia including settings RGB... but awaiting expaned board to test RGB.
Added RP2040 version (awaiting boards from PCBWay)... are we there yet?
Added Expander Board (awaiting boards from JLCPCB)... Arerived, perfect...
STM32 Keyboard Development Module (adding more to options board)...
Added optional program header, connections for MISO, MOSI, SCK and SS
Michael O'Toole
Ironpark
deʃhipu