Hey there,

Just noticed your project doesn’t have a license file. Thought I’d mention that adding a license could be really beneficial, especially for a cool project like this. A license helps protect your work from being copied or misused. It also makes it clear how others can use, modify, or distribute your design.

I wasn't stepping on your work. It's truly what I've been looking to play with. I have lots of boards, sensors, but this finally pulls it altogether.

Thanks so much for your work on this!

I'm printing the files as we speak so I can start to tinker!

Hi,

First of all, really cool project, thank you for sharing this!

Reading the schematic and the datasheet for the TCA4307 a potential issue arose, the current modules use a single row of pogo connectors for power and communication.
Those connectors aren't staggered and don't ensure a consistent order of connection/disconnection.

I'm trying to come up with a simple solution to quite my worries and wanted to ask if you had issues hot swapping the modules?
Power issues? Communication issues or lockups?

Hey! Noticed that the keycap STL is linked and listed in the instructables article, however it is nowhere to be found within the repo

Hi. Can you please update CAD/STL sections with .step files? i would like to make a my own macropads and modules based on your design

hi this is so cool of you @sb-ocr! ok disclaimer im writing this in my brake rn I will revise this later the example likely is faulty currently

tldr.

perhaps this helps to decode the module nr. in the interrupt pulses fast enough.

import board
import array
import adafruit_pioasm
import rp2pio

count prof = adafruit_pioasm.assemble( """
.program counter

mov ~y y          ; initial prep for addpulse

reset:            ; this block initiates grace time and waits till pin is low
set x 11111           ; set retry counter or grace period max 5bit
wait 0 pin 0          ; wait till pin is low

findedge:         ; this tries to find pulse or exits
jmp pin addpulse      ; when pin high add one to pulse count
jmp x— exit           ; else exit when grace period is zero 
jmp findedge          ; else retry

addpulse:         ; this adds one to pulse counter
jmp y—                ; adds one since y is inversed in exit
jmp reset             ; reset grace period and test for next pulse

exit:             ; this writes pulse count to isr
jmp y! reset          ; when y zero alias no pulses detected reset
mov ~y y              ; inverses y/pulse count so isr is filled with real count
mov isr y             ; depletes y to zero and fills isr
mov ~y y              ; inverses y to prep it for addpulse
push noblock          ; push isr and do not wait for empty tx fifo
"""

countpio = rp2pio.StateMachine(
    program    = countprog,
    frequency  = 0,  # 125MHz pio clock speed
    jmp_pin    = board.A0,  # interrupt line pin 
)

pulsecount = array.array( 'L', [32] )

while True:
    while countpio.in_waiting() < 1:  # wait until at least one count is available
        pass
    while countpio.in_waiting() > 0:  # read all words in fifo
        countpio.readinto(pulsecount)  # read from rx fifo
        print(pulsecount[0])
    countpio.clear_rxfifo()  # reset rx fifo to make room for new counts 
   

limits since one instruction takes 8ns at 125MHz

11111bin/32dec *8ns = 256ns is the grace period until pulse count is reset
2 *8ns = 16ns is the minimum high time for one pulse
i dont know the minimum low time of a pulse but perhaps 8ns duno
assumes a default low state of the interrupt line
tx fifo contains up to 4 pulse counts then counts are lost
dwell time for one pulse count until the next is 7 *8ns = 56ns

more info

this is circuit python. with rp2pio and adafruit_pioasm libraries from adafruit you can init a pio of the rp2040 to count pulses until no pulse is detected for 256ns (but this is adjustable). After one count the pio needs 56ns then it’s ready to count again (perhaps this can be helped by using two pios). So no simultaneous inputs are possible like this. I think. anyways feel free to close this at anytime this is just a thought

import board
import adafruit_pioasm
import rp2pio

pwminprog = adafruit_pioasm.assemble( """
.program PwmIn

; algorithm:

; loop:
;    reset the 'timer'
;    loop: 
;       decrement timer
;       test for falling edge 
;    record timer value as pulse width (actually, (0xFFFFFFFF - x)*2/125MHz is the pulse width)
;    loop:
;       test for rising edge
;       decrement timer
;    record the timer value as period (actually, (0xFFFFFFFF - x)*2/125MHz is the period)

 
.wrap_target
    mov x ~NULL         ; start with the value 0xFFFFFFFF

timer:
    jmp x-- test        ; count down
    jmp timerstop       ; timer has reached 0, stop count down
test:
    jmp pin timer       ; test if the pin is still 1, if so, continue counting down
timerstop:              ; pulse is over (or timer has reached 0)
    mov ISR x           ; move the value in x to the ISR: the pulsewidth (0xFFFFFFFF-pulsewidth)
    push noblock        ; push the ISR into the RX FIFO

timer2:
    jmp pin timerstop2  ; if the pin has become 1, the period is over, stop count down
    jmp x-- timer2      ; if not: count down
timerstop2:
    mov ISR x           ; move the value in x to the ISR: the period (0xFFFFFFFF-period)
    push noblock        ; push the ISR into the RX FIFO

.wrap
"""


pwminpio = rp2pio.StateMachine(
    program          = pwminprog,
    frequency       = 0,
    jmp_pin           = board.A0,
    in_shift_right  = False,
)

lowtime = array.array( 'L', [32] )
hightime = array.array( 'L', [32] )

while True:
    pwminpio.clear_rxfifo()
    while pwminpio.in_waiting() < 2:
        pass
    pwminpio.readinto(hightime)
    print(hightime[0])
    pwminpio.readinto(lowtime)
    print(hightime[0])

even more info

im not too good at programming and quite new to this in general. anyways
I saw your video and your issue with missing inputs due to pulling inefficiencies. you proposed to encode the module nr. in the interrupt pulses. I recently tried to read a pwm signal of a phillips hue lamp and had the same issue with circuit python being to slow. I solved the issue by reading the signal via a pio of the rp2040. hope this is relevant to you since it is all circuit python it shloud be possible to integrate without a rewrite in c or so

Hey! Can I ask you to attach schematic files of the boards for the Core and Modules, as was the case for the first version of the macropad?

Hey! I have currently finished building the main macro-pad module. I have put the firmware code on the board, and I have checked my soldering a few times now. I cannot get the RGB to light up (The OLED works fine), and the board is constantly inputting strokes into my computer. I am not sure what is causing the issue. Here is my testing keymap in main.py:

keyboard.keymap = [ [ KC.A, KC.B, KC.C, KC.D, KC.E, KC.F, KC.G, KC.H, KC.I, KC.J, KC.K, KC.L, ] # Layer 1 ]

Here is my output in a random text editor when this is saved to the board: (Note I am trying to push buttons, hold buttons etc, to try and get this to stop, which is where the new/random other letters start appearing)

ddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddiddjdklefdghlghbeaddlfcdgaddjkldefghdadibcddgaddjklefgdhadibdcdddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddgaijklefghadddddddddddddddlbcjijklefghaddddddddddlbcjijklefghaddddddddddddddddddddddddlbcgaklefghaijjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjbcdiijklefghaddddddddddddddlbcjijklefghadlbcgaijklefghadddddddddddddddddddddddddddddlbcgaijklefghadlbcgaklefghaijjjjjjjjjjjjjjjjjjjjjjjjjjjabcdgjklefghaiabgijklefghadddddddddddddddddddddddddddddddlbcjijklefghaddddddddddddddddddddddddddddddddddlj

And here is my soldering work:

I'm not sure if you have encountered this issue, however I would like to resolve this issue. If you need any more files, code snippets, or log files let me know.

-Demon

@sb-ocr You wouldn't believe it, I did a school Design & Tech major works on this concept a few years back, you've implemented pretty much everything I ideated right down to the custom pogo pin connectors, only I did a base mounting board design. Sharing some of my old files here as they might be of interest! Obviously it's not super technical as it had to fit within particular marking guidelines and cover particular topics, but I'm sure you'll enjoy the blender render promo video that looks remarkably like the one in your youtube vid! If you ever decide to do a version 3, I'd love to have a chat, I've got other projects on at the moment but still would love to make this a reality for the live production / DJ space! Love your work!

MIDIFIER Design Folio.pdf

Just a quick question on the status of the irq handling issue. As suggested, I am holding off a build.

Hi,
First of all, great job with this project :D i really want to finish the build and use it.

I've solder the components on boad and uploaded the firmware.
When it starts up i get this error.

I've commented the ocreebModule import line and the lines where its used just to check if i can get anything running, but after i did that i've got this error :(

I've put the code on an empty XIAO RP2040 and i got the same error.
Im new in this hobby and i dont know if i need anything connected to that pinouts to run the code or not.

Here is my board.