#help-with-circuitpython

1150 messages · Page 2 of 2 (latest)

stone briar
#

So where is says **services: Service, entry: ScanEntry

how does it differentiate between the end of *services and the beginning of entry? (Sorry. My python noobieness is showing)

Nvm. Just found entry on keyword only arguments

gray cargo
# 
>>> def print_all_1(*args, extra):
...     for arg in args:
...         print(arg)
...     print("extra", extra)
...     
>>> 
>>> print_all_1(2, 4, 6)
Traceback (most recent call last):
  File "<python-input-7>", line 1, in <module>
    print_all_1(2, 4, 6)
    ~~~~~~~~~~~^^^^^^^^^
TypeError: print_all_1() missing 1 required keyword-only argument: 'extra'
>>> print_all_1(2, 4, 6, extra=20)
2
4
6
extra 20
stone briar
#

if *args is used then the second argument becomes key-word only

gray cargo
#

you can give a default value: extra=999, and then you don't have to supply the arg

stone briar
cosmic swift
#

Interesting appearance on the Adafruit front page...

image.png
#

Is that in reference to just setting up the board with CP? Or is Adafruit going to start promoting Firefox for all web based tools?

solar orchid
placid granite
#

Hi, I just started with circuitpython on a rp2040. To execute boot.py I need to restart the board, right? Atm I'm doing this by pushing the tiny little reset button. Is there something like machine.reset() or another way to reboot the board?

abstract wedge
#

microcontroller.reset()

placid granite
#

Is there a way to automatically import reset from microcontroller for repl cli?

abstract wedge
#

you could create a r.py file that imports and calls it, and then do import r whenever you want to reset?

placid granite
#

sound like an idea...

wooden dragon
#

(not in safe mode though)

sullen lake
molten marten
#

What errors do you get?

pearl rapids
solar orchid
solar orchid
# sullen lake Has anyone used the audiospeed module (https://docs.circuitpython.org/en/latest/...

Here's an example I just tested on a Pico running 10.2.1 :

import time, board, audiobusio, audiocore
import audiomixer
import audiospeed

SAMPLE_RATE = 22050
CHANNEL_COUNT = 1
BUFFER_SIZE = 2048

i2s_bck_pin, i2s_lck_pin, i2s_dat_pin = board.GP20, board.GP21, board.GP22

audio = audiobusio.I2SOut(bit_clock=i2s_bck_pin, word_select=i2s_lck_pin, data=i2s_dat_pin)
mixer = audiomixer.Mixer(voice_count=1, sample_rate=SAMPLE_RATE, channel_count=CHANNEL_COUNT,
                         bits_per_sample=16, samples_signed=True, buffer_size=BUFFER_SIZE) 
audio.play(mixer)

wav1 = audiocore.WaveFile("amen1_22k_s16.wav") 
wav1adj = audiospeed.SpeedChanger(wav1, rate=0.5)

mixer.voice[0].play(wav1adj, loop=True)

while True:
    print('hi', mixer.voice[0].playing, time.monotonic())
    time.sleep(0.5)

And there does seem to be a bug in SpeedChanger if no mixer is used. Looks like in my refactor to pull it out of WaveFile and into it's own package, I messed up non-mixer playing. (filed issue https://github.com/adafruit/circuitpython/issues/11020)

cosmic swift
# solar orchid Perhaps because Firefox finally supports WebSerial? Seems also like a brand dea...

Yeah I was thinking a brand deal as well since clicking on it takes you to a page on Firefox's site about using it with Adafruit. I wouldn't mind a bit of transparancy on Adafruit's part to explain this a little. Are they recommending Firefox now because it is actually better than other choices (more reliabile, more features, more something else)? Or are they simply in a marketing deal with Firefox?

sick cairn
#

I imagine it’s because Firefox is FOSS and independent. But a clear comparison table would be cool

sick cairn
#

I wish the chart went back further so I could see Netscape Navigator

cosmic swift
sick cairn
#

Mosaic was never my favorite, no. But Netscape had tabs before anyone else

molten marten
#

I think InternetWorks was the first browser with tabs

#

Bought by aol

#

I only know this cos I was watching an LGR video yesterday hahaha

sullen lake
#

@solar orchid I just got back and saw your message. StreetChicken.wav played at half speed is a banger. Thanks!

dusky shard
molten marten
#

Oh hmm might be able to import the keyboard driver for the repl with that, could be handy

dusky shard
#

I don't know if anyone knows the internals of circup here but I have a couple of libraries I wrote and import in my code.py that in turn import a few Adafruit std libraries. When I either place my libraries in /lib or compile them to mpy and leave at top level circup can't find the dependencies to add the Adafruit libraries to lib/. If I leave them at the top level and uncompiled to mpy circup finds the dependencies and installs them. I'm pretty sure circup either can see into mpy files to find recursive Adafruit libraries to add or must have its own list somewhere of what depends on what.

solar orchid
wooden dragon
#

I am a little surprised that circup doesn't find the dependencies in a .py file in /lib, but it's been a long time since I had my hands in that code

#

by the way @solar orchid I keep delaying the patch you submitted to discotool because I wanted to take time to read through the code in detail, in part because I don't have access to a MacOS 26 compatible machine (or Windows 11 for that matter which is also broken it seems)

solar orchid
gray cargo
#

If discotool is broken then Adafruit_Board_Toolkit probably is too.

wooden dragon
#

I don't know ? It's the code that gets the drives and pairs them with the serial ports by diving in the USB tree, but Adafruit_Board_Toolkit just does the serial ports with the pyserial inherited code, right ?

#

I mean it might still break, but not for the same reason

gray cargo
#

the port listing is done in an OS-specific way, with code I copied from you. I think I did the part of the windows code myself.

wooden dragon
#

yeah it's the code we took from pyserial/list_ports (and submitted fixes to the pyserial repo years ago)

gray cargo
#

ah, I forgot that

wooden dragon
#

but that's not related to the issue with 26 that was fixed here by todbot

dusky shard
solar orchid
dusky shard
sick cairn
#

I'm having trouble with some "object with buffer protocol required" errors. Is it true that bytes, bytearray, and array all support the buffer protocol in CircuitPython?

gray cargo
#

The usual reason for that error is that None was accidentally passed

#

you could put a print statement in and see what the arg actually is

placid granite
#

I'm trying to get a usb_hid led only usb descriptor right. I'm trying on rp2040. Does anybody know of some example code? Atm, I have a keyboard with no inputs, but only an output for two leds (msg waiting & ring)...

sick cairn
placid granite
sick cairn
solar orchid
crystal shadow
#

Need some help locating a very old resistive touch screen driver. I have a Adafruit TFT FeatherWing - 3.5" 480x320 Touchscreen for Feathers PID: 3651 which uses the discontinued STMPE811 I would like to use this display but now it looks like i am out of luck any ideas would be appreciated

molten marten
#

Try the stmpe610 driver, some indication says it should work

royal lance
#

Hi!
I cannot run CircuitPython on my ESP32-S3 board. This is ESP32-S3-Tiny by Waveshare, but instead of 512k/4M it has 8M/8M RAM/Flash.

  1. vanilla adafruit-circuitpython-waveshare_esp32_s3_tiny-en_US-10.2.1.bin gives a loop in bootloader:
ESP-ROM:esp32s3-20210327
Build:Mar 27 2021
rst:0xc (RTC_SW_CPU_RST),boot:0x8 (SPI_FAST_FLASH_BOOT)
Saved PC:0x403830b1
SPIWP:0xee
mode:DIO, clock div:1
load:0x3fce2820,len:0x1c4
load:0x403c8700,len:0xb24
load:0x403cb700,len:0x27b4
entry 0x403c886c

  1. With rebuit tinyuf2, copying the UF2 image stucks with I/O errors in dmesg.
    After reset there is the same loop with just more logs from tinyuf2.

  2. Much smaller (2.8M) adafruit-circuitpython-waveshare_esp32_s3_tiny-en_US-9.1.4.uf2 has the same issues

  3. Images for unexpectedmaker_tinys3 (quite generic board with same S3-N8R8 chip) don't work either

  4. I've built my own build-waveshare_esp32_s3_tiny_n8r8/firmware.uf2 which behaves the same

  5. MicroPython ESP32_GENERIC_S3-20260406-v1.28.0.uf2 works just fine -- copied fast, then boots in next second

  6. Tried writing the app directly: esptool write-flash -z 0x10000 build-waveshare_esp32_s3_tiny_n8r8/circuitpython-firmware.bin -- all the same

Could anyone suggest steps to diagnose or fix this?

gray cargo
royal lance
molten marten
#

So a question, on my picocalc nothing appears on screen if I dont tell it to, however if my program crashes i get the output on the screen, is this because I setup the screen in my program and circuitpython keeps that even if the program crashes?

royal lance
gray cargo
royal lance
#

oh, I tried to look PSRAM stuff in datasheet. Turns out S3, S3-WR0OM, S3-PIC0 are different chip families.
N8R8 may refer both to WR0OM and PIC0 (may be more, but I stopped here). I have PIC0 apparently.
Fortunately, in all three R2 is always Quad SPI, and R8/R16 are always Octal SPI

gray cargo
royal lance
#

but all this stuff looks like a matrix to me:

  • we have variety of boards with specific wiring
  • we have variety of chips with same footprint, but not very compatible firmware-wise

I'm new here, so I will not insist on rewriting build scripts and web pages

#

nice, 7 boards have the same error (actually, I copy-pasted qio from one of them)

$ git grep -l 'IDF_TARGET = esp32s3' | xargs grep -l 'CIRCUITPY_ESP_PSRAM_SIZE = [^2]' | xargs grep  'PSRAM_MODE = qio'
boards/circuitart_zero_s3/mpconfigboard.mk:CIRCUITPY_ESP_PSRAM_MODE = qio
boards/m5stack_cores3/mpconfigboard.mk:CIRCUITPY_ESP_PSRAM_MODE = qio
boards/m5stack_cores3_se/mpconfigboard.mk:CIRCUITPY_ESP_PSRAM_MODE = qio
boards/unexpectedmaker_feathers3/mpconfigboard.mk:CIRCUITPY_ESP_PSRAM_MODE = qio
boards/unexpectedmaker_nanos3/mpconfigboard.mk:CIRCUITPY_ESP_PSRAM_MODE = qio
boards/unexpectedmaker_pros3/mpconfigboard.mk:CIRCUITPY_ESP_PSRAM_MODE = qio
boards/unexpectedmaker_tinys3/mpconfigboard.mk:CIRCUITPY_ESP_PSRAM_MODE = qio
gray cargo
molten marten
solar orchid
molten marten
#

Right right, is there a specifically approved method for initialising the display and then specifically showing REPL or console?

shrewd meadow
molten marten
#

That is.. super convenient actually

#

thanks!

shrewd meadow
#

you're welcome.

solar orchid
dusky shard
#

No end of new things one learns in this channel! I'm already using this new trick on a Featherwing.

dusky shard
#

On further testing and looking at the library sources I think this happens automagically when I import the library adafruit_hx8357 based on BusDisplay. That is the import of the library sets the display.root_group to displayio.CIRCUITPYTHON_TERMINAL.

I can see that resetting display.root_group back to the above might be useful when a program exited that was using the display so it could show REPL output again.

stone briar
#

I dropped the nicenano circuitpython into this controller and it drops out of bootloader mode. when I put it back into bootloader mode its still called NICENANO what could i try to get this thing back to square 1?

viral flume
stone briar
#

Pro Micro nRF52840

#

I reflashed it with

adafruit-nrfutil --verbose dfu serial --package nice_nano_bootloader-0.6.0_s140_6.1.1.zip -p COM3 -b 115200 --singlebank --touch 1200

And got this 'success message'

image.png
#

But when I drop in the circuitpython uf2. it doesnt work. maybe the controller is borked.

#

Nvm...I just reflashed it again and this time it worked. 🤷‍♂️

dense olive
#

I'm a little confused about differences in reading I2C data, depending whether an I2C Controller is reading, or an I2C Target is reading.

When an I2C Controller writes to an I2C Target, the I2C Target doesn't need to know how many bytes need to be read: https://docs.circuitpython.org/en/latest/shared-bindings/i2ctarget/index.html#i2ctarget.I2CTargetRequest.read & https://github.com/adafruit/circuitpython/issues/10423#issuecomment-2974082922. The I2C protocol takes care of signaling when the message is complete.

But when an I2CTarget writes to an I2C Controller, there doesn't seem to be an API mechanism other than reading until a buffer is filled, or until a specified slice of that buffer is filled: https://docs.circuitpython.org/en/latest/shared-bindings/busio/index.html#busio.I2C.readfrom_into. The result is that multiple messages can get concatenated into a single read, and message boundaries are lost. Shouldn't there be a mechanism for an I2C Controller to detect the end of the message and stop reading?

storm spire
dense olive
#

Thanks, I'll read more on the specs. I have a different application, but it occurred to me that perhaps a sensor might send an (unknown length) sequence of values... but there's no way for a Controller to figure that out (currently?). I control both ends in my application, so I could add a layer of protocol to send the length first, but I was hoping to keep it simple.

storm spire
dense olive
#

Yes, in my case, the Target moves on to the next message, but the Controller keeps reading as if it's the same message. I'll add a data length and make it a two-part read.

#

I could also flip it so that there are multiple Controllers writing and one Target reading, but that seems like it's asking for trouble with collisions and throughput degradation and general robustness.

storm spire
#

targets that implement a “register” model for their application protocol often implement an auto-increment procedure for extended reads

storm spire
storm spire
sick cairn
dense olive
#

I think a two-step read will do it... Controller reads a length, then fills a buffer/slice of that length with the next read. Then the message is done. This works with espressif Controller, and I will test with raspberrypi and samd. I have a CRC already. I2C Target only runs on samd and raspberrypi.

dense olive
placid granite
placid granite
# solar orchid Are you intending to follow an existing HID usage or your own? Here's an examp...

Well, I'm at the beginning of learning and do not really know what is the best way to reach my goal: make a simple led show up in Linux under /sys/class/leds - like the leds on a keyboard, but just a led.
I thought I'd start looking at the descriptor used for a keyboard, then get rid of the keyboard part to just have the led part left over.
I do not need to do fancy stuff to the led, just switch it on and off.
My led for testing is the blue led on the rp2040 board. If I get a hold on how this all works, I'd love to make a rgb led work.

wanton basalt
#

Having digital test muxes to see when the controller is having a line at high Z is a life saver. Unfortunately most consumer products don't have those features exposed :/

dense olive
#

I realized that I do need multi-controller since my multis have to be espressif (I2C Target has to be either raspberrypi or samd). But in testing multi-controller more thoroughly with a raspberrypi I2C Target... two controllers works pretty well, but start increasing the number of controllers, the data sizes, or the frequency of messages and things go bad. And then recovery of multiple MCUs becomes an issue. So I've ruled out using I2C for my current project. Also ruled out SPI for now since SPI Target (available on M4 only) has a fatal issue. That leaves UART if I wan to keep costs low, but UART needs a multi-level hierarchy of forwarding due to limits in the number of UARTs per MCU.

sick cairn
#

What are you trying to do??

dense olive
#

I'm trying to move a lot of data (frequent and large messages) from many espressif devices to a single other device, preferably espressif or raspberrypi, for further processing. The higher-cost solutions involve adding hardware like Ethernet or CAN (but CAN also breaks down with high bus utilization / multiple senders of frequent large messages).

sick cairn
#

SPI would typically be the simplest option. What is this about needing multiple UART masters? RS422 supports multi-drop

#

You could still use a CS pin like SPI does to select which device should be communicating

near dagger
#

Hihi! I know this might not be the place to ask it but i'm pretty sure it is as its most likely a CiruitPhython thing. So i have a Matrix Portal M4 (Display is the 64x32 RGB LED Matrix - 2.5mm pitch) and i did all the setup, it has the files it should have and its connected to WIFI but I've hit a wall and i am unsure on what to do next, I've read the setup pages on the website but i am still unsure on what to do next, if someone could tell me or guide me in the right direction i would apricate it! again i apoligize if this is not the right place to ask this

dense olive
# sick cairn SPI would typically be the simplest option. What is this about needing multiple ...

Yes, SPI would be nice, but Target is only supported on M4 and has an open issue that is blocking its use in all but the simplest cases. UART is fine, but with limited UARTs per MCU, this yields a multi-level forwarding hierarchy. RS-422 is one driver, multiple receivers ...and is it supported in CircuitPython? RS-485 is supported and supports multiple drivers, I did look at that some time ago and dismissed it for reasons I don't remember, perhaps similar concerns about robustness under heavy utilization, or the caveat about best-case timing on the rs485_dir pin. I could test it out and see.

#

@near dagger Can you link to the tutorial you're using, and also describe more what is working so far, and what is not?

solar orchid
# placid granite Well, I'm at the beginning of learning and do not really know what is the best w...

If you want to emulate the LEDs on a keyboard (which only have on/off state, no brightness), then you can use the built-in HID keyboard emulation that CircuitPython is already doing and just intercepting those HID reports. That would look something like this in your code.py:

import time
import usb_hid
import adafruit_hid
keyboard_hid = adafruit_hid.find_device(usb_hid.devices, usage_page=1, usage=6)

print("waiting for LED commands from computer")
while True:
    out_report = keyboard_hid.get_last_received_report()  # from computer
    if out_report:
        print(["%02x" % x for x in out_report])
        time.sleep(0.5)

No need for a boot.py.

In Linux, you would trigger it with commands like this:

echo "1" | sudo tee -a  /sys/class/leds/input41::scrolllock/brightness
echo "0" | sudo tee -a  /sys/class/leds/input41::scrolllock/brightness

and "input41" will change depending on when and where you plug in your CircuitPython device.

near dagger
#

and it seems to be working? like its conected to wifi and everything and its all set up, i just dont know what to do next, like if theres more to set up

dense olive
near dagger
#

Ok thanks!

#

It just didn’t seem to be displaying anything when I was running code in circuit python!

#

That’s what the issue was I’m pretty sure, sorry I completely forgot about that XD

dense olive
#

yeah, I think that guide is just for set up, other guides have more interesting visual examples

near dagger
#

Ok, thanks!

dense olive
#

For UART with RS-485 https://docs.circuitpython.org/en/latest/shared-bindings/busio/index.html#busio.UART, what do I do with the rs485_dir output pins? I don't find any CircuitPython examples. Do I need external transceivers, termination resistors, bias resistors? This may get complicated pretty fast, and I wonder about robustness with many drivers. Almost $7 for a transceiver, I may be better off shelling out for Ethernet (or inexpensive ESP32SPI co-processors).

sick cairn
#

(Could make a switch with two mosfets if you’re really worried about cost)

dense olive
#

The busio.UART module seems to support RS-485 to some degree, but it's not clear to me what it does and doesn't do, or even how to hook it up. It runs with one driver, but if I add a second, the second's writes are never read. But then I almost certainly am not handling the rs485_dir pins properly.

#

I'm more concerned about complexity (hardware, but also software and error recovery) than cost. If busio doesn't naturally handle multi-driver, then I'm more likely to add a network interface than to write complex low-level software specifically for RS-485. I don't even know if RS-485 will handle my overall project even if I do get the fundamentals working. This is really a networking application, I was just hoping to trim the per-MCU cost by using built-in comms.

crystal shadow
molten marten
#

Nice!

sick cairn
dense olive
#

maybe I'm using the wrong terminology, I need multiple MCUs writing, and one MCU reading from all of them

sick cairn
#

Ok. Does it have to be simultaneous or can they take turns on the same bus?

dense olive
#

they can take turns, but then there's complexity with how to coordinate that... think of multiple writers generating messages at random (but frequent) times. they could buffer multiple messages until it's their turn

sick cairn
dense olive
#

well, the writers do randomly have messages to write... but they could be arranged into regular intervals

sick cairn
#

Each get their own timeslot?

dense olive
#

there's some waste to time multiplexing though

sick cairn
#

Yes

dense olive
#

I thought that was what you were suggesting

sick cairn
#

Or they each get their own opportunity

#

How much data and how often?

dense olive
#

writer #2 may accumulate a lot more messages than writer #5, for example... then the time slot has to accommodate the busiest writer

#

think of the messages as random lengths of within a range of a say 100 to 1000 bytes, they may come in very fast, very slow, or something in between... some writers may routinely generate much more data than others

#

I could use a CS-type pin, but that would take up a lot of pins, similarly trying to multiplex... any kind of back-and-forth synchronization communication is prone to error and needed recovery (or data loss)... without a simple scheme, I'm probably better off using standard networking

sick cairn
#

Hard to design without solid requirements, but ok. How about chaining the CS signal so each device tells the next one when it’s complete?

dense olive
#

maybe, unless a node goes down

sick cairn
#

You really want automatic arbitration? You could assert with a shared signal and do exponential + random backoff for conflict. Receive the sent data and make sure it matches what you intended in case two devices assert simultaneously. Send a CRC checksum at the end of the packet for integrity

#

But I’m not sure this extra complexity gets you extra performance

#

If you need more bandwidth, use a higher bitrate

dense olive
#

Yes, I'll have a CRC (unless using TCP, then it's probably redundant). Multi-controller I2C with I2CTarget would have automatic (built-in) arbitration but it's too finicky for my heavy bus utilization. RS-485 (if I understand it correctly) with multiple transmitters would have automatic (built-in) arbitration, not sure if it could handle my number of transmitters, and message sizes and frequencies. I'm confident basic UART would work, though it makes for complex topology and message forwarding. I guess I'm trying to avoid inventing something new if there's something off-the-shelf with robustness built in. So I've been exploring SPI, I2C, and UART as possible alternatives to higher cost-per-node networking. But I do make a trade-off between the complexity of something more custom, vs the cost of something more standard.

#

Thanks for the suggestions and for listening as I sort through the issues.

sick cairn
#

I’d love to help more, but it’s really unclear what your requirements are. You’ve expressed concern over conflicts, but I’m not sure if the issue is throughput or latency

#

And then there’s the issue of robustness. What should happen in the event of a single-node failure?

#

How often do you expect this to occur?

#

If you can nail down some requirements, we can design a system to meet your needs

  • What is the worst-case for packet size, frequency, and number of nodes?
  • How important is it to receive these in a timely manner?
cobalt heath
#

Hi! I'm working on a CP project where I need to time software processing in my pico2 from the moment the radio chip receives preamble to the moment a message was transmitted. The goal is to time the software processing time as accurately as possible.

I understand CP is designed to prioritise user friendliness over hardware access so hardware interrupts are hidden intentionally by design. My work is meant to be integrated to a larger CP project, which is why I can't just move to C or MP for direct access to interrupts ^^;

My question is:

  • What would be the "circuitpythonic" way of doing this? If I use asyncio for software interrupt, for example, is it possible to characterize the few ns of software delay?
  • If I want to read hardware interrupt anyway, how would I achieve this in CP? I know a bit of C so I don't mind having to frankenstein something together.

I also did some backreading in the server and found this thread #help-with-circuitpython message where @abstract wedge (sorry for the ping!) mentioned "C module in CP" as a workaround, which currently seems like the best solution to me. But I'd love to get some inputs from someone more experienced than I am ^^; ty in advance!!

abstract wedge
#

the pulsein module saves precise timings of pin changes

#

it uses internal timers

cobalt heath
sick cairn
#

You’re not going to get 1us accuracy in Python. I think the best you could hope for is 100us

#

But you could get it with the PIO

#

If the radio tells you when you receive a preamble, that is

cobalt heath
# sick cairn How about you time the pins with the PIO?

ty for responding! This was super helpful!
I looked into it bc I wasn't super familiar, but PIO is deterministic, right... So uncertainty would be the interaction between PIO and python? I was curious about how you'd estimated that number :o

sick cairn
#

Manually triggering the garbage collector might reduce your measurement jitter

cobalt heath
sick cairn
#

Hmm. Maybe 100us was too pessimistic. Try it and see

#

Google claims the function overhead in Circuitpython is only a couple us

near dagger
# near dagger Ok, thanks!

I followed all the tutorials and everything and I don’t know why but it still won’t display anything, am I missing something? Genuinely I’ve tried everything I can and I’ve asked different people and I’m so lost and confused at this point

ancient crown
near dagger
#

I haven’t and I’ll do that, thank you

#

Everything seemed to be set up and working as like the green light was flashing on the back and everything else seemed to be working as the guide said

ancient crown
#

that'll replace circuitpython, but is easy to go back once the test is done

#

just copy back the circuitpython firmware UF2

#

and remember that UF2's are copied to the BOOT folder (not CIRCUITPY)

near dagger
#

Ok

#

Follow up question, is there a way to completely wipe the board and start from square one?

ancient crown
near dagger
#

Danke!

#

Hopefully that’s a last resort if the other stuff doesn’t work XD

ancient crown
#

if the simpletest UF2 works, then we can take a look at why it's not working in circuitpython

#

it should just be a code issue if the simpletest UF2 works

near dagger
#

I might not have stuff that I needed downloaded tbh, I have a feeling I missed something

ancient crown
#

download the UF2? or download firmware to the matrix portal?

near dagger
#

I’ll have to double check

#

All I know is I got it connected to WiFi and that’s the furthest I got, nothing else was working

ancient crown
near dagger
#

Danke! I will take a look when I get home, thanks again

#

I’m sorry for being slightly dumb with all of this, I’m like brand ish new to coding and stuff

ancient crown
#

no worries. asking here is perfect. it'll be best if you can ask again when you're with the hardware.

#

there'll be some stuff worth knowing that's just generic circuitpython stuff

#

and then stuff for getting the matrix to work

#

it can be a lot of new stuff to deal with initially

near dagger
#

Would I need a hub 75?

#

Wait nvm sorry

ancient crown
#

you don't need anything else. just that and the matrxi portal.

near dagger
#

I have the 64x32 RGB LED Matrix - 2.5mm pitch, idk if that’s the same

#

yes it is

#

Ohhhh, product ID, that’s what PID means

ancient crown
#

"hub 75" is an unfortunately overused term. it's often applied to any LED matrix panel thing and implies there's a standard. but there can be lots of variations in these panels. but if you got the LED matrix panel from adafruit, it should work.

#

yep, sry, PID = product ID

#

adafruit shop URLs are pretty simple - they include the product ID,

#

just wanted to verify what panel you have

#

software generally must be done correctly depending on the panel

#

that's why there's more than one simpletest UF2 for example

#

they are panel specific

near dagger
#

Ohhh ok, that might be my issue

near dagger
ancient crown
#

the cool thing about the UF2 is it does not require anything else in terms of code

#

it's stand alone

#

just download and copy to BOOT folder

near dagger
#

Ok

#

I also have the gif folder with that file and Arduino all set up with the proper libraries and everything

#

So it might be the UF2 thing