#circuitpython-dev

1 messages ยท Page 387 of 1

idle owl
#

@devout jolt Doing the jerk thing and tagging you. Do you have any experience with the UDA1334A I2S DAC?

#

See the details above.

devout jolt
idle owl
#

Any idea if I can wire a 2-wire speaker to the breadboard pins and have it work?

devout jolt
idle owl
devout jolt
idle owl
#

Exactly what I was about to link.

#

I have this at my other office in the US. But not here in Canada with me.

#

So I'll need to order it here.

#

Thanks so much for the confirmation!

ember iris
manic glacierBOT
slender iron
#

@gilded cradle are you able to do the review above?

blissful pollen
slender iron
#

thanks @blissful pollen

#

(dan and jeff are on vacation)

manic glacierBOT
blissful pollen
#

No problem, I vaguely recalled they were going to be away. Feel free to ping me if you need reviews I can do my best to look

slender iron
#

thanks!

manic glacierBOT
slender iron
gilded cradle
blissful pollen
#

No problem!

gilded cradle
#

I was dealing with a flooding issue in the building here.

blissful pollen
#

ouch that doesn't sound like fun

gilded cradle
#

No, it could have been worse. Just ruined a stack of paper, but carpeting needs drying.

blissful pollen
#

yea at least you were there to notice, its when you're out and it happens things go real bad

gilded cradle
#

Yeah

manic glacierBOT
manic glacierBOT
#
nathanielbutts
[adafruit/circuitpython] New comment on issue #5266: audiobusio\.I2SOut: audio distorted "crackly" playing MP3

I believe that was the problem. Inspecting the file showed it to be 320kbit/s and 44.1khz, so basically .wav like. I used xcfa to move it to 64kbit/s and the pops are gone. It sounds kind of muted and scratchy, but that has probably to do more with manipulating the file too much. I'll play around with ripping at the correct rates and try again.

Thanks for looking at it!

manic glacierBOT
slender iron
#

found 250 bytes for samd21 by sorting the text section at link time

#

๐Ÿ• time

ember iris
#

You're on a roll!

atomic summit
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
capellini
[adafruit/circuitpython] New comment on issue #5183: Convert all circuitpython modules to use MP\_REGISTER\_MODULE

I think that I've now converted all modules in shared bindings that have a *_MODULE defined in py/circuitpy_mpconfig.h. I didn't see a module listed in MICROPY_PORT_BUILTIN_MODULES_STRONG_LINKS for the following modules in shared bindings, so didn't use MP_REGISTER_MODULE:

  • multiterminal
  • paralleldisplay
  • nvm (submodule of microcontroller)
  • os (listed in MICROPY_PORT_BUILTIN_MODULE_WEAK_LINKS and alt name listed in MICROPY_PORT_BUILTIN_MODULE_ALT_NAMES and I don't quite...
manic glacierBOT
#
capellini
[adafruit/circuitpython] New comment on issue #5248: Add \_\_name\_\_ to all \`board\` modules

@PracticalMetal We want add a __name__ attribute and assign it a q-string value of 'board', since that's the name of the module. So that, at the serial terminal, instead of getting a blank name for the module:

>>> import board
>>> board
<module ''>

we would get 'board' instead:

>>> import board
>>> board  # the next line prints 'board' instead of a blank string
<module 'board'>

Inside the pins.c file for each board (`ports//boards//pins...

manic glacierBOT
manic glacierBOT
manic glacierBOT
lone sandalBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
tannewt
[adafruit/circuitpython] New comment on issue #5183: Convert all circuitpython modules to use MP\_REGISTER\_MODULE

espidf, rp2pio and samd are port specific modules so I'd do those as a group.

The other four are micropython modules that we reorganized into shared-bindings. os and time are in this group too. I think they can be made strong links. MicroPython has uos and utime too which is why they have strong and weak links.

I'm not sure multiterminal is used anywhere. We used to use it on ESP8266 for webrepl but I'm not sure we'd go that route when we support ESP32 again.

`paralle...

idle owl
#

So I have this UDA1334A hooked up to a Pico. First thing I tried was the tone code from the guide. It "works"... but it doesn't seem to actually play the tone. As in, there is clearly an attempt at transmitting audio, there is a sort of static silence being played to the headphones at the appropriate interval, but it isn't a tone. Tried MP3 code, and it worked fine. So I'm a little confused. It's standard tone code, though it has a "volume" variable, but I increased that and it did not resolve the issue.

#

I remember the tone volume being included because the audio coming from this board if you don't have external volume control was painfully loud.

#

(I included a warning on that guide page to let the code run before putting the headphones on so you knew how loud it would be before having them in/on your ears.)

#

Tweaking things isn't fixing it, and frankly, I have no idea what I'm doing with it so I'm simply changing values to see what happens. Really, I've been comparing it to the Circuit Playground library tone code, and altering things that differ. It's not helping.

tidal kiln
#

@idle owl is this unique to pico?

idle owl
#

Valid question. I have not hooked it up to anything else this time around.

#

I can try another board.

blissful pollen
#

I'm trying to think why the MP3 would play but the tone wouldn't. What wouldn't be getting set that lets one work but not the other

idle owl
#

Grabbing some other board.

blissful pollen
#

can you post (or have a link?) to the tone code?

idle owl
#

I do, it's in a guide. One moment.

#

The major change I made was increasing the volume variable.

tidal kiln
#

hmmm. looks like have the max98357a amp, but not the uda1334. so can't help test. sry.

idle owl
idle owl
#

I'm wondering if not all pins are created equal on Pico.

blissful pollen
#

the audio code is different on the two as well. As far as I know basically all the pico pins do everything

idle owl
#

I have a script that returns valid pin combos. In theory if they're invalid, it errors (it did that initially on the Metro). It did not error on the Pico.

idle owl
#

Oi...........

#

The pin combo script runs forever. Yeah, I don't think pin combo matters on Pico.

#

This is the combo I used, it is returned by the script Bit clock pin: board.GP0 Word select pin: board.GP1 Data pin: board.GP2

#

Hmph.

#

ยฏ_(ใƒ„)_/ยฏ

#

I'm out of ideas.

#

Last ditch, I'll try it on Feather RP2040, see if it's a Pico or RP2040 issue?

#

If I have a soldered one anyway.

blissful pollen
#

Hmm so exact code runs just replacing the sine wave part playing the MP3. I'm not really an I2S expert and trying to recall when I worked on some RP2040 audio stuff

idle owl
#

Do not have a soldered one. ๐Ÿ˜•

#

Kinda like them not being soldered because STEMMA, but I guess it's time to sacrifice one of them.

blissful pollen
#

I am finding I end up soldering them and if I want them someplace where I don't need it buying another one. Didn't say it was an economical plan

idle owl
#

I already added one to my cart ๐Ÿ˜†

blissful pollen
#

I don't have the I2S decoder so can't help there sorry

idle owl
#

I'll at least solder up a Feather and test it.

#

@tidal kiln It still might help to hook that MAX up to a Pico and see if it plays a tone over the speaker. Would at least determine if it's I2S or the UDA.

tidal kiln
#

already wiring...

idle owl
#

Well played.

#

BRB, soldering

tidal kiln
#

yah. it'll take me a few.

#

@idle owl generally working. getting good tone. but it's not consistent. goes randomly loud/quiet on different loops.

# 
import time
import array
import math
import audiocore
import board
import audiobusio

I2S_CLK = board.GP0
I2S_WS = board.GP1
I2S_DATA = board.GP2

audio = audiobusio.I2SOut(I2S_CLK, I2S_WS, I2S_DATA)

tone_volume = 0.1  # Increase this to increase the volume of the tone.
frequency = 440  # Set this to the Hz of the tone you want to generate.
length = 8000 // frequency
sine_wave = array.array("h", [0] * length)
for i in range(length):
    sine_wave[i] = int((math.sin(math.pi * 2 * i / length)) * tone_volume * (2 ** 15 -1))
sine_wave_sample = audiocore.RawSample(sine_wave)

while True:
    audio.play(sine_wave_sample, loop=True)
    time.sleep(1)
    audio.stop()
    time.sleep(1)
idle owl
#

Well what the what.....

#

Let me try on this Feather.

#

This is a new one. ValueError: Bit clock and word select must be sequential pins

#

swaps pins around

blissful pollen
#

oh i wonder if its using PIO

idle owl
#

No tone.

tidal kiln
#

@idle owl also:

Adafruit CircuitPython 6.3.0 on 2021-06-01; Raspberry Pi Pico with rp2040
>>>
blissful pollen
#

The PIO to write some stuff it has to be sequential pins

idle owl
#

Ohhh

#

Let me try 6.

#

Been running 7.

#

Hah!

#

Tone!

#

does the bug filing time dance

#

Wow that was a frustrating one.

#

I don't even know what I'm supposed to be filing here. ๐Ÿ˜„

idle owl
#

@blissful pollen Do you have an idea of what I should be reporting here?

#

"Tone playback broken on 7x on RP2040" seems vague.

#

Or maybe it isn't.

blissful pollen
#

I don't think it is that vague. I'd mention the sequential pin error message and that MP3s work

tidal kiln
#

@idle owl which 7.x? let me see if i can recreate same failure.

idle owl
idle owl
blissful pollen
#

If I have a chance later I can take a look at the code reading this while in a meeting that's just going on and on ๐Ÿ™‚

idle owl
#

Let me load 6 on the Pico and see if I get tones.

tidal kiln
#

@idle owl silence on 7.x. can got back to 6.3.0 and it works again.

idle owl
#

Silence on Pico ๐Ÿ˜ฆ

#

on 6x

#

With your code, so I eliminated the changes I made

ember iris
#

is that problem still present on the absolute latest build? I feel like a couple of timing related changes were just merged for the pico last week

idle owl
#

Checking.

idle owl
ember iris
#

Sweet bug away!

#

It just means the release candidate is more likely to be better!

idle owl
#

I'm not ok with the fact that I can't get a tone on 6x though

#

And cater could.

#

I mean, I've hard reset the board between because you have to unplug it to get back into the bootloader.

#

I could nuke it I guess and try again.

blissful pollen
#

The fact you had a weird issue on the Feather RP2040 is there too. May be something in the RP2040 I2S code

idle owl
#

Right, but there was a tone on 6x on the Feather.

#

Blergh.

#

Is my response to this.

tidal kiln
#

audio.show()?

idle owl
#

Probably.

blissful pollen
#

I2S on the RP2040 is a PIO program

idle owl
#

That clarifies very little for me ๐Ÿ˜„

#

I don't have a huge grasp on RP2040 interworkings.

#

Which is to say... no grasp.

blissful pollen
#

lol sorry. the PIO is their cool assembly like language to rapidly write input/output basically as fast as the processor clock speed

#

From quickly looking at the PIO program the timing pins are both "sideset" pins, and while the name isn't important the way they are written two means they have to be sequential. I believe that is for BCLK and WSEL

#

But this is where I know very little about I2S ๐Ÿ™‚

idle owl
#

Incidental that I made them that way.

#

I didn't know ahead of time, I simply wanted the pins to be in order in the constructor.

manic glacierBOT
#
kattni
[adafruit/circuitpython] Issue opened: #5283 Sinewave tone playback on RP2040 no longer working as of CP7

CircuitPython version

Adafruit CircuitPython 7.0.0-beta.0-183-gb80ffee10 on 2021-09-01; Raspberry Pi Pico with rp2040
(but also the 7.0.0-beta.0 release)

Code/REPL

import time
import array
import math
import audiocore
import board
import audiobusio

I2S_CLK = board.GP0
I2S_WS = board.GP1
I2S_DATA = board.GP2

audio = audiobusio.I2SOut(I2S_CLK, I2S_WS, I2S_DATA)

tone_volume = 0.1  # Increase this to increase the volume of the tone.
frequency ...
idle owl
#

@tidal kiln @blissful pollen ^^

manic glacierBOT
main furnace
#

On the MAX i2s chip I get bad results if I leave the Gain pin floating. Volume fluctuates randomly.

idle owl
#

The UDA doesn't appear to have gain.

main furnace
#

I have one UDA plugged into the breadboard just waiting for me to wire it up.

idle owl
#

@main furnace If you have a Pico and you're up for testing, I wouldn't mind a sanity check on the tone code. Carter was testing with the MAX I2S chip, not the UDA. It's also absolutely fine to decline.

manic glacierBOT
idle owl
#

@tidal kiln @blissful pollen WAV file playback doesn't work either on Pico running 7 absolute latest. Getting silence. ๐Ÿ˜ฆ So could it also be audiocore-related?

tidal kiln
#

works on 6.3.0?

idle owl
#

Checking.

#

@tidal kiln Not for me.

#

But I couldn't get tone working on Pico for me either. ๐Ÿ˜•

tidal kiln
#

oh. you didn't get 6.3.0 tone to work? you haven't been able to get anything i2s audio?

blissful pollen
#

I'd bet something audiocore related for the RP2040. I don't think I have any I2S devices around though to test it though

idle owl
#

over I2S.

tidal kiln
#

but no tone on pico using 6.3.0?

idle owl
#

Correct.

tidal kiln
#

weird. that worked for me.

idle owl
#

I know.

#

That's why I pinged you again. ๐Ÿ˜„

#

the souuuuund..... of silence.

tidal kiln
#

so dunno if it's a UDA vs. MAX thing? or something else?

idle owl
#

That is accurate. But it also might be a red herring. If audiocore is borked for RP2040.

#

I will update the bug.

tidal kiln
#

yah. given we aren't 100% same hardware. it could be (insert answer here).

idle owl
#

Exactly.

#

Bleh.

manic glacierBOT
slender iron
blissful pollen
slender iron
#

thank you!

manic glacierBOT
idle owl
#

Fascinating, I did a print when it was supposed to be playing, I thought it was doing the static during the playback. Blergh.

#

@main furnace That's you?

main furnace
#

On CP6 the .wav files would eventually stop producing sound.

blissful pollen
#

Taking a quick look at the code I am wondering if it is something with setting the frequency, it has to change the PIO playback frequency. It could explain why some clips seem to play

main furnace
#

yeah, I'm 'rsbohn' on github.

idle owl
#

Can you try my wav file? And can you post yours for me to try?

main furnace
#

haha I didn't realize that that song was StreetChicken.

idle owl
#

It and its name is a John Park special. ๐Ÿ˜„

main furnace
#

๐Ÿ˜„ I have this 'drama.wav' 22050 samples/sec.

#

I think I downloaded it from this channel, actually.

idle owl
#

Drama was a JP special as well!

#

I got so tired of the coin sound and the laugh.wav, that I begged for something new. So he put together a few clips for me.

main furnace
#

JP is awesome.

#

Both .wav files playing CP7 (8/17):: RPi Pico::UDA1334A.

idle owl
#

I'm starting to suspect my Pico. ๐Ÿ˜•

manic glacierBOT
main furnace
#

I will thank you for giving me the incentive to wire up the UDA breakout. Thanks!

idle owl
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #5283: Sinewave tone playback on RP2040 no longer working as of CP7

If you have an analog amp (TPA2012, etc., or try on the MacroPad), does PWMAudioOut work?

It appears there are conflicting test results of RawSample vs WaveFile.

I had trouble with I2SOut due to miswiring sometimes.

I did WaveFile testing with the test program in #5196, adapted to the SAMD or RP2040 board I was testing on (I2SOut vs PWMAudioOut, and whether or not there was SPEAKER_ENABLE).

tulip sleet
#

@idle owl try making the length longer (multply the length by say, 8, at least), and also take out the loop=True

#

But if that code works on the MacroPad with PWMAudioOut, I am confused (see how similar it is to the library code)

idle owl
idle owl
#

With a pop.

tulip sleet
#

Can you do PWMAudioOut on the RP2040 to an analog amp?

idle owl
#

Which I think is the click.

#

@tulip sleet Like this? ```py
import board
import audiomp3
import audiopwmio

audio = audiopwmio.PWMAudioOut(board.GP0)

decoder = audiomp3.MP3Decoder(open("slow.mp3", "rb"))

audio.play(decoder)
while audio.playing:
pass

print("Done playing!")```

#

to the PAM8302?

tulip sleet
#

the low-level code doesn't know it's from Wave or Raw or PWM vs I2s.

#

try a wav file, not MP3

idle owl
#

So try that code with WAV or tone?

#

ok

#

I still have the PAM on the breadboard

tulip sleet
#

try it both with Wave and tone

idle owl
#

checking wiring.

#

Wiring verified.

#

Checking tone.

tulip sleet
#

does the same wav file play on the macropad?

idle owl
#

Checking.

#

Tone playback is just a squeal too

tulip sleet
#

the squeal is due to the amplifier amplifying noise, I think

#

you have ground connected to the - input and GP0 to + input?

idle owl
#

Yes.

#

Sorry, trying to get this wav file onto the MacroPad, copying is taking forever.

#

Yes it plays.

#

Save wav file.

tulip sleet
#

macropad uses GP16 to the speaker amp; try GP16 on the Pico.

idle owl
#

(I admit I was using Finder initially, and it wasn't working.)

#

Ok.

tulip sleet
#

same program? or you are using the macropad library?

idle owl
#

Same program

#

Not MacroPad lib

#

GP16 is still squeal.

#

For tone.

#

Let me try wav.

#

Same.

blissful pollen
#

I only quickly glanced at the I2S code but any chance Dan that something is not being set in the PIO program frequency correctly for tones? could explain why mp3s work or anything "standard"

tulip sleet
#

I2SOut was working for me earlier, so I don't understand why it isn't now. I didn't change that part, I was changing the DMA

idle owl
#

What release did you change DMA

#

Should I try the version before that of 7?

tulip sleet
#

I tried wav files with many different sampling frequencies

#

alpha.6 I think had the fix

idle owl
#

Let me try an earlier alpha.

#

Alpha 4 the file works.

tulip sleet
#

yes, alpha.6 has #5097

idle owl
#

It's scratchy, obvs.

#

Well, it played once anyway. ๐Ÿ™„

tulip sleet
#

yeah, it had multiple problems

idle owl
#

Right.

#

But proof of concept.

#

going back to latest.

#

(โ•ฏยฐโ–กยฐ๏ผ‰โ•ฏ๏ธต โ”ปโ”โ”ป

#

IT'S WORKING.

#

Trying tone.

tulip sleet
#

power cycle in between if not working

idle owl
#

Tone playback is squeal and click.

#

Not a tone.

#

That's with PWM

tulip sleet
#

but is a tone on MacroPad?

idle owl
#

Trying one more time

#

No.

#

It's a click and a fshhhh silence.

#

Without the MacroPad lib.

#

Running beta.0

#

not the tip of main.

tulip sleet
#

alright, I will try this Monday or Tuesday

#

unless Scott or someone else wants to step in; we will fix it

slender iron
#

I'm planning on doing an rc with it as a known issue

idle owl
#

Scott was going to take a look during AaE. It was blocking me on a guide, but Limor unblocked me with other stuff.

#

Also that.

#

I'll create the I2S template as though it's all working, and then we'll add it to the guide with Pico once it's fixed, basically. In the meantime, I will do other things.

#

@tulip sleet Thanks for the help. I'm still not above blaming my Pico.

#

Limor figured you'd take a look when you got back.

tulip sleet
#

sure; it's broken on beta.0? or is it rc.0 that you are testing, or alpha.6?

idle owl
#

beta.0, and the latest build off of S3, which is a built of beta.0 with a number after it.

#

Tried with both.

tulip sleet
#

ok, that's helpful; something is weird, but I won't stew about it

slender iron
#

it's not urgent ๐Ÿ™‚

idle owl
#

๐Ÿ˜„ See to it that you don't! You can sort it soon enough.

tulip sleet
#

bye!

idle owl
#

Later!

manic glacierBOT
#
tannewt
[adafruit/circuitpython] New tag created: 7\.0\.0\-rc\.0
orchid basinBOT
main furnace
#

When I load the latest build (from S3) the Pi Pico stops showing up as the E: drive (no drive at all). When I go back to a build from 8/17 it works again.

#

Same result with ItsyBitsy RP2040 (older 8/30 build works as expected).

slender iron
#

tries it

#

@main furnace I don't get a drive either. ๐Ÿ˜ฆ

#

I should have tested it before making an RC

main furnace
#

oops

slender iron
#

thanks for catching it ๐Ÿ™‚

#

fbb works from yesterday

manic glacierBOT
jaunty juniper
#

but only on windows apparently ?

lone axle
#

I think Scott is on linux and mentioned that he doesn't see the drive either.

jaunty juniper
#

ah, I see it on my mac, not on windows

orchid basinBOT
jaunty juniper
#

oh fascinating that's an actual file name: /media/CIRCUITPY/.TemporaryItems/folders.501/TemporaryItems/(A Document Being Saved By BBEdit)

#

ah no it's a directory name

slender iron
lone axle
#

I need make some projects with my pico's. I haven't used them much except for testing a few things and it throws me off every time that there is no reset button

slender iron
#

updated tinyusb to even newer ๐Ÿ™‚

#

thach has been fixing bugs found with the usb compliance test

lone axle
#

fwiw I do get a CIRCUITPY drive with Adafruit CircuitPython 7.0.0-rc.0 on 2021-09-02; Raspberry Pi Pico with rp2040

slender iron
#

huh, interesting

#

could be flaky

#

I saw it once after linux reset the device

#

popped right up with the build I just posted

manic glacierBOT
main furnace
#

firmware.uf2 -- now the drive appears

slender iron
#

awesome! Just made a PR for that update

#

will need to test other ports tomorrow

#

before I do rc.1

lone axle
#

Ah, yeah I ran microcontroller.reset() and something does seem to be off the drive eventually showed up but took like 20-30 seconds when normally it's only a few seconds max.

#

I will check that firmware you posted real quick as well.

slender iron
#

it updates tinyusb further. thach made fixes for samd and rp2040 since the previous update

lone axle
#

Yep no issues with that one. Drive re-appears within a few seconds like normal.

slender iron
#

great!

manic glacierBOT
jaunty juniper
main furnace
#

Very much improved if you add a reset button.

manic glacierBOT
manic glacierBOT
#
vkichline
[adafruit/circuitpython] New comment on issue #3999: Support ESP32\-NOW

ESP-NOW is a great remote control protocol. It can achieve much longer range (with much lower bandwidth) than WiFi and needs no network access, making it resistant to many issues that affect portable control systems, such as security.
I would use this protocol today if it were available to write a wireless multi-servo control, using an M5 for the UI/HID and an ESP32-S2 as the receiver/controller.

spiral elk
#

STM32F4 seems fine with CIRCUITPY drive visibility on rc0 FWIW

manic glacierBOT
#
hathach
[adafruit/circuitpython] Pull request review submitted: #5286 Update tinyusb for samd and rp2040 improvements

Both SAM51 and Rp2040 passed all the Compliance Test (chapter 9, HID, MSC) as expected :+1: .

However, SAMD21 (tested on M0 metro) has issue with following chapter9 test cases:

  • TD 9.16 Enumeration Test where the host repeatedly issue bus reset + enumeration for 150 times and
  • sometime TD9.9 Halt Endpoint Test.

Note: SAMD21 pass that same tests on my simple example, it is rather similar to a previous issue where setup packet is missed on samd21 as mentioned in this note https:...

manic glacierBOT
manic glacierBOT
atomic summit
#

Does anyone know how to get the frozen modules updated to the latest commit hashes? Specifically the Neopixel library?

#

It's pointing to b2f21e6 but the latest is 0f4661c that has a fix in it that prevents me from using 7.0 with my boards.

#

Can I just checkout the newer commit hash in my local repo and then do a PR against that to get it updated?

#

Or does someone "official" have to do that?

manic glacierBOT
atomic summit
#

Well I did it... let's see if I get in trouble ๐Ÿ™‚

manic glacierBOT
#
mocleiri
[adafruit/circuitpython] New comment on issue #4671: Native Tensorflow Lite support

I found out that the circuitpython stm32 port doesn't support C++ which is a problem but I think it can be worked around.

Would it be possible to intake this micropython commit to add c++ support to stm32?

https://github.com/micropython/micropython/commit/97960dc7deb7a0e691fca5944402cd03386b744b

The other option is to build the c++ parts externally and link them into the firmware. I'm already doing this for the tensorflow bits but it could be expanded to also include the micropython...

manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
lone sandalBOT
manic glacierBOT
manic glacierBOT
ember iris
#

@tidal kiln for the sgp40--would you like me to add an example for lower memory use loosely following the example I outlined, or would that cause more confusion with the sensor usage overall?

manic glacierBOT
tidal kiln
ember iris
tidal kiln
#

i think existing example is ok

#

as you mention, the sensor is sort of meaningless without the VOC algorithm

#

but if someone wanted to be simple/low memory, they'd just never call measure_index and only use raw

#

since you only import/create an instance if measure_index is called

manic glacierBOT
tidal kiln
#

you'd want something like your other example if things were kept separate

#

if VOCAlgorithm weren't part of the SGP40 class

ember iris
#

That makes sense. Thanks for helping with the pull request and being a second set of eyes. I had been dug deep in the sgp40 docs and just missed a lot because of it

tidal kiln
#

since you'd need an example showing how to bring both in, create, and use together

#

yah. np on the review. wasn't able to actually vet the algorithm itself. but it seems like it's working and providing useful results for your use case. so you were the tester for that ๐Ÿ™‚

ember iris
#

The algorithm is nuts. It's basically two phases (one for the heating element being turned on and one for the heating element fully heated) that pipe into a 24 hour table of, 'Right I'm seeing this much of a change, I'm not use to it, and I only respond to VOCs so lets bump the score' with 'Ok I've gotten use to this, lets go to a score of 100'

#

To understand if that's a smart approach I'd need to understand the physics of their metal oxide layer, which I might one day look up, but for now I'll trust sensirion

ember iris
#

Should I be testing the release candidate, or should I be testing the absolute newest?

tidal kiln
#

for CP firmware?

ember iris
#

Yup! I'm updating most of my boards and am curious what's the most useful build to use right now

tidal kiln
#

sort of depends

#

in general, use latest release, and if you find any issues, bring them up

#

absolute newest if you are helping test fixes for such issues

ember iris
#

Good to know! Thank you!

tidal kiln
#

"latest release" could either be stable (like 6.3.0) or alpha/beta for next version (like 7.0.0 beta 0)

#

"absolute newest" is whatever latest commit is. very much a moving target.

#

so there can be like 3 flavors

#
  • stable
  • unstable
  • absolute newest
#

there can be cases, with newer hardware, where unstable is the only option. shouldn't be that way forever. just until a stable release is made.

#

macropad is like that right now

ember iris
#

That makes sense. So for the process of moving a release candidate to a stable release--do the absolute newest changes get omitted unless one of the changes solves a major bug that was present in the release candidate? And in the absence of a major bug, is the release candidate the version that gets rolled into the stable release?

tidal kiln
#

that's generally it. but scott/dan would better know those details.

ember iris
#

ok I think I get it. Thank you for the explanation, I appreciate it!

onyx hinge
#

Rc.0 had a pretty serious bug for rp2040 so that may influence your choice too

ember iris
#

Since I'm getting more comfortable with circuit python, I'm starting to be more willing to use unstable versions in hopes that I'll get better at bug hunting

#

And the projects are just small senor arrays for around the home, so stability isn't that critical

manic glacierBOT
#
tannewt
[adafruit/circuitpython] New tag created: 7\.0\.0\-rc\.1
orchid basinBOT
onyx hinge
#

That's a lot of new boards.. all since rc 0?

stuck elbow
#

they tend to cluster

jaunty juniper
#

Scott stopped the auto update for rc 0 because of the issue with rp2040

#

so it's the new boards since the beta

manic glacierBOT
onyx hinge
#

Thanks everyone for rc.1! I did bring the hmi board so I may even try it out on vacation

atomic summit
#

Haha, I just spent 6 hours re-flashing all of my FeatherS2Neo stock with rc0! Oh well... everyone can update their own boards when they get them!

manic glacierBOT
orchid basinBOT
manic glacierBOT
main furnace
atomic summit
stuck elbow
#

this is a slippery slope

#

you could wait a year and put 8.0 on them

atomic summit
orchid basinBOT
orchid basinBOT
onyx hinge
#

if I understand correctly .. any use of while will be flagged by a future pylint? that'll be, umm, controversial.

manic glacierBOT
jaunty juniper
onyx hinge
#

oh phew won't be enabled by default

jaunty juniper
#

they call that "extensions" or plugins ?

onyx hinge
#

I hope there's enough value in having a more consistent style in circuitpython code, vs just mindlessly changing code until it causes the linter to stop complaining. it's a trade-off

#

based on past experience I have about a 25% chance of breaking code I'm authoring by fixing pylint diagnostics, which is frustrating.

jaunty juniper
#

that's a real danger yeah, especially when they add new stuff and like me you accidentally update pylint in your venv

grave ledge
#

Some questions regarding the new board definition for Lolin S2 Mini: Since the board has no factory-installed NeoPixel, I'm guessing that NeoPixel should be removed from mpconfigboard.h & pins.c? Unexpected Maker creates pin assignments (e.g. An, Dn, DACn) for his boards for py code compatibility. The Lolin S2 Mini has labels only for SoC GPIO numbers. How to handle this?

spiral elk
#

GPIO only is also pretty normal. Since the silkscreen and the supplied decals only list the GPIO numbers I'd suggest leaving them at that.

#

In addition, with regard to the neopixel assignments, they're not on optimal pins even if you wanted them to remain. A lot of folks will only be using the outer rows to plug them into D1 devices I suspect.

grave ledge
#

I'm leaning in this direction too. It's nice to see I2C, SPI, etc. enumerated at dir(board, but people would need to look up the GPIO labels in that case. Probably best to be minimally specific; its normal to change pin assignments in code when changing boards.

spiral elk
#

Given the lack of silkscreen guidance or even one of those fancy infographics, having to look up the pins is probably inevitable.

grave ledge
jaunty juniper
#

no it can't

spiral elk
#

It might be worthwhile to look at the Micropython pins.h for that board to see what they've done there. Its somewhere to be found on the wemos github.

jaunty juniper
#

the board doesn't have dedicated SPI or I2C pins on the silkscreen, neopixel or neopixel power pins, all that should be removed

#

(from mpconfig.h too)

spiral elk
#

The maker does define specific pins for SPI/I2C in their micropython port though

#

(first-party-written)

grave ledge
jaunty juniper
#

is that information available in doc ?

spiral elk
#

Those are more sensible pins too, consistent with the D1-mini-subset of the header

#

They match the D1 mini's pins for those functions

#

Probably justification to leave the pin definitions in, since there's already an established ecosystem of add-on boards using those pins.

grave ledge
#

esp32-s2 datasheet says I2C & SPI2/3 on any GPIO pins. D1 mini 'D' pins could be labeled, but shields are inconsistent for their use, at least with I2C often overlapping with other uses (some shields have pin use pads).

spiral elk
#

IMO, they should be left in, but changed to the micropython assignments.

#

And ditch the neopixel or make it a -Doption on the build

#

But choose a pair of pins that are more accessible, possibly the one that their one-neopixel shield uses.

#

Oh never mind that last, that uses the same one as the micropython-assigned SDA...

jaunty juniper
#

CIRCUITPY_STATUS_LED_POWER and NOPIXEL_POWER should be removed anyway

spiral elk
#

Yar

#

Oh, the new neopixel shield uses D4, which corresponds to gpio 16 on the S2

#

that would be a potential candidate for board.NEOPIXEL

grave ledge
spiral elk
#

That's not their first neopixel shield, there was an earlier one hardwired to D2 which is not very useful.

grave ledge
#

Note, CP build fails when using the latest esp-idf pulled with esp-idf/install.sh

#

Yes, their original NeoPixel shield appears EOL (only clones still exist)

spiral elk
#

are you sure? I just updated and built a couple hours ago.

#

I'm gonna probably order a few more boards, gonna take a look through shields too and at least order a couple of the neopixel ones.

grave ledge
#

tried again with adafruit/circuitpython main, using esp-2021r1-8.4.0 from .espressif/tools/xtensa-esp32s2-elf (not the older esp-2020r3-8.4.0)

grave ledge
jaunty juniper
grave ledge
#

okay, clear now thanks.

onyx hinge
#

I'm trying to figure out how espressif's parallel lcd library pin names map onto circuitpython ... espressif has "cs", "wr" and "rs" pins. circuitpython has "chip_select", "write", "read", and "reset" pins. "read" never seems to be used (or is just always driven high / true)

orchid basinBOT
manic glacierBOT
onyx hinge
#

hum there are already two versions of the HMI DevKit ... mine is silkscreened "V1.1" and the versions are ??? and 2. The only schematic in the repo is for "1.1".

#

๐Ÿค”

orchid basinBOT
onyx hinge
#

huh my HMI devkit does NOT power on with the USB C-C cable that I use on my phone... it's fine with an A-C cable I also had around.

#

I verified the C-C cable works between computer & phone. so hm.

manic glacierBOT
manic glacierBOT
ember iris
onyx hinge
#

There is no usb pd negotiation ic but I assumed you could get 5v 500ma regardless. Anywayi have a working cable and that's what counts

ember iris
#

huh. No idea. Chalk it up to phones are weird I suppose?

manic glacierBOT
#
mrdalgaard
[adafruit/circuitpython] New comment on issue #5233: nRF52840 \- cannot find rfm9x if battery powered

I just tested this on another SPI device (BMP390) to see if this is a general issue.

I used the following bit of code:

import time
import board
import adafruit_bmp3xx
import digitalio
spi = board.SPI()
cs = digitalio.DigitalInOut(board.RX)
bmp = adafruit_bmp3xx.BMP3XX_SPI(spi, cs)

while True:
    print(bmp.pressure)
    time.sleep(0.5)

It works fine on USB power, but on battery it also fails:

>>> supervisor.get_previous_traceback()
'Traceback (most recen...
orchid basinBOT
manic glacierBOT
manic glacierBOT
turbid radish
#

?serverinfo

digital shoreBOT
#
adafruit
Owner

adafruit#3230

Channel Categories

8

Text Channels

59

Voice Channels

6

Members

30712

Roles

35

ID: 327254708534116352 | Server Created ยท
manic glacierBOT
stuck elbow
#

it looks like the failures are due to earlier commits, that somehow slipped past

#

what should be done in such case?

jaunty juniper
#

is the pylint version used set to a fixed version ?

stuck elbow
#

no idea

#

I guess I will make a separate commit that fixes it first

lone axle
# jaunty juniper is the pylint version used set to a fixed version ?

I think the pylint that runs on the library code is fixed version 2.7.1 currently, from here: https://github.com/adafruit/Adafruit_CircuitPython_ImageLoad/blob/071d6e8f7d2c17ff5bcc18d0160d7e6b8074aa5d/.pre-commit-config.yaml#L21 but when it gets run on the examples from here: https://github.com/adafruit/Adafruit_CircuitPython_ImageLoad/blob/071d6e8f7d2c17ff5bcc18d0160d7e6b8074aa5d/.pre-commit-config.yaml#L33 it's executing just pylint in the shell which is using the one installed from here: https://github.com/adafruit/Adafruit_CircuitPython_ImageLoad/blob/071d6e8f7d2c17ff5bcc18d0160d7e6b8074aa5d/.github/workflows/build.yml#L47 without specifying version and thus taking the newest current release.

stuck elbow
#

I just went ahead and fixed all the warnings and failing tests

#

@lone axle I would greatly appreciate if you could take a look at it

lone axle
#

Will do after I finish up with work for the day. about another hour and a half or so.

stuck elbow
#

weekend soon!

manic glacierBOT
onyx hinge
#

Later we can move the standard i2c and spi buses to this macro

#

Instead of specifying them in multiple places (mpcofgig and pins)

jaunty juniper
#

yeah, I'm looking into doing that, but that will wait after 7.0.0

manic glacierBOT
#
mrdalgaard
[adafruit/circuitpython] New comment on issue #5233: nRF52840 \- cannot find rfm9x if battery powered

This issue might be related:
https://devzone.nordicsemi.com/f/nordic-q-a/74754/nrf52840-application-won-t-completely-run-without-power-applied-to-vbus
Since the nRF chip has VBUS directly connected to it, it might actually change behaviour based on whether USB power is connected.

This could also be related, but have not tested with a capacitor yet:
https://devzone.nordicsemi.com/f/nordic-q-a/65475/spim-3-mosi-must-be-tied-to-ground-via-capacitor-to-work/268076#268076

SPIM3 does look ...

jaunty juniper
#

@onyx hinge would this be a reasonable way to do it ? (and add the same for SPI, UART, etc.) - BOARD_I2C is true if there are default I2C pins in mpconfigboard.h

#if BOARD_I2C
#define BOARD_MODULE_DICT_DEFAULT_I2C \
    { MP_ROM_QSTR(MP_QSTR_SDA), MP_ROM_PTR(DEFAULT_I2C_BUS_SDA) }, \
    { MP_ROM_QSTR(MP_QSTR_SCL), MP_ROM_PTR(DEFAULT_I2C_BUS_SCL) }, \
    { MP_ROM_QSTR(MP_QSTR_I2C), MP_ROM_PTR(&board_i2c_obj) },
#endif

#define CIRCUITPYTHON_BOARD_DICT_STANDARD_ITEMS \
    { MP_ROM_QSTR(MP_QSTR___name__), MP_ROM_QSTR(MP_QSTR_board) }, \
    { MP_ROM_QSTR(MP_QSTR_board_id), MP_ROM_PTR(&board_module_id_obj) }, \
    BOARD_MODULE_DICT_DEFAULT_I2C
bitter bronze
#

Where can I find some code examples of CIRCUIT Python for the NeoTrinkey? I have reported that I get a 404 error when I try to follow the link on the learning page.

onyx hinge
#

@jaunty juniper close, I think.. you need a #else from #if BOARD_I2C that defines BOARD_MODULE_DICT_DEFAULT_I2C as expanding to nothing. Other than that, I think the above is on the right track.

#

you have to delete the lines from the individual pins files too of course, and I don't know if BOARD_I2C is a define that exists already

jaunty juniper
#

it is, I noticed that when I looked into implementing a second I2C singleton on the QT PY 2040 a long time ago

onyx hinge
#

@jaunty juniper I think that's on the right track, yup

manic glacierBOT
manic glacierBOT
#
kamtom480
[adafruit/circuitpython] Pull request opened: #5300 Fix incorrect use of allocate\_memory\(\)

next_code_allocation and prev_traceback_allocation are in the IMMOVABLE group. See: https://github.com/adafruit/circuitpython/blob/7ce3f256509ba64b63ea926922d22854f8fb8ac7/supervisor/shared/memory.c#L39. However, calls to the allocate_memory() function are with the argument movable = true. This causes an assert: https://github.com/adafruit/circuitpython/blob/7ce3f256509ba64b63ea926922d22854f8fb8ac7/supervisor/shared/memory.c#L168

#
jerryneedell
[adafruit/circuitpython] New comment on issue #5233: nRF52840 \- cannot find rfm9x if battery powered

Interesting -- the first link does appear to be the same issue

amaul
amaul 4 months ago
SPI signals are obviously coming through with VBUS powered.  With VBUS disconnected the only signal that I have a heartbeat on is Chip Select. MISO, MOSI and CLK are low and do not even trip a trigger acquisition.  Signal frequency doesn't really match the working configuration, but I supposed that would be expected.  Logic level and voltage level looks correct.  I'll need to check with the BSP on...
manic glacierBOT
#
mrdalgaard
[adafruit/circuitpython] New comment on issue #5233: nRF52840 \- cannot find rfm9x if battery powered

An additional observation -- If I power it up via USB, then disconnect the USB and leave it on battery power, it continues to operate normally. The issue only occurs if it is powered up or RESET when in battery power.

Yes, and since the person in the first link does not seem to be using CircuitPython, it would seem to be a problem with the hardware, or the bootloader, which is in common.

manic glacierBOT
manic glacierBOT
lone axle
# bitter bronze Where can I find some code examples of CIRCUIT Python for the NeoTrinkey? I hav...

Are there specific examples that you're looking for? This page has an example for how to work with the Neopixels: https://learn.adafruit.com/adafruit-neo-trinkey/neopixel-blink and this page the capaticitive touch input: https://learn.adafruit.com/adafruit-neo-trinkey/capacitive-touch-neopixel-brightness For most intents and purposes this device is just a standard CircuitPython device so anything from https://learn.adafruit.com/circuitpython-essentials would be relavent as well, although some things would be difficult to hook up to that device in particular since it doesn't have broken out IO pins available.

bitter bronze
#

@lone axle Great. Thanks for the pointers. I also asked on help-with-circuit-python and received a pointer.

lone axle
bitter bronze
#

@lone axle I had not seen that page nor seen it when I was looking. Are there similar pages for other products and if so where are they located. This would be very useful information that could be published on some of the other pages. Thanks again.

lone axle
#

but I don't think every product has the link on that downloads page. I usually use this trick to find them. Load up one of the products that does have the page and then edit the URL in the address bar. You can change out the 4 digit numerical product code to be whatever product you want. If you search up a product on the main adafruit site and then check in the URL you can see it's product code.

https://learn.adafruit.com/products/3333/guides
https://learn.adafruit.com/products/[put product code here]/guides
bitter bronze
#

@lone axle I am not sure how I missed the "New Guides" heading on the Learn page. Sometimes I get tooo focused and run into the forest/trees problem. Again, your insight into how to find things should be more prominently identified. Perhaps it would be useful for Phil or Lady to occasionally mention this during Show and Tell.

bitter bronze
#

@lone axleI have often wished for a "Guide for the Impatient". When I am focused on getting something up and running, I would like to have the bare bones steps to show how to boot the device, download the UF2, pointer to the code, pointer to the Guide, subgraph of code projects. I hate to get bogged down in text and I would really like to have a MindManager type diagram for each product with visual clues. Perhaps someone on the team already has some Python GUI building expertise and code that could lay the groundwork for this capability to use more visual navigation. Users could chose to use the textual, visual or combination for navigation. I would be glad to test and perhaps contribute.

lone axle
#

That would be pretty neat for sure. The main guide for each device should contain all of that information but it does tend to be a bit heavier on the text than it sounds like you're looking for

#

Essentially it would be like Device "Quick Start" pages

#

or "Cheat Sheet" for the device, though I try not to use that term when possible since it implies breaking the rules sort of.

manic glacierBOT
orchid basinBOT
manic glacierBOT
#
maholli
[adafruit/circuitpython] Pull request opened: #5302 updating pycubed firmware for mainboardv05

PyCubed users:

  1. Note pin changes between mainboardv04 to mainboardv05!
  2. Remember that pycubed's low-level boot counter is not present on the circuitpython main branch. This is because a truly low-level boot counter implementation requires modifications to circuitpython's main.c file.
    • To implement the PyCubed boot counter for yourself, add the following to /main.c
      #include "shared-bindings/nvm/ByteArray.h" // PYCUBED BOOT COUNTER
      Pri...
manic glacierBOT
onyx hinge
devout jolt
#

hi CircuitPython wizards, I have a native library question: is there a way to have a native module called in the background regularly by the supervisor without modifying supervisor/shared/tick.c? I notice this is how keypad does it, but I also see supervisor/shared/background_callback.c being used by a few other modules. What is the differences between the two approaches?

crimson ferry
#

in rc.1, is there a meaning to NeoPixels (violet) and LED flashing once per second after startup? The Blinka shows up in the REPL on FunHouse, for example, then it goes into this pattern. Was running beta.0 fine.

#

(same behavior on all esp32-s2, PyPortal was fine with the upgrade)

crimson ferry
#

rc.0 is the same behavior as rc.1

jaunty juniper
#

hmmm I see the normal green blink every 5s after code has done running, or solid white (a little blinding still on Funhouse !) when in the REPL

manic glacierBOT
manic glacierBOT
crimson ferry
#

the rc.1 issue seems to be in my boot.py, probably something to do with alarm.sleep_memory

low sentinel
#

yo, anyone know of a circuitpython repo on github that runs Circuitpython tests (e.g., on push)?
I have CPython but seems like it would be really nice to test with even the linux build of circuitpython.

#

just looking for an example to follow. I spose I can check out circuitpython, build the linux port and run tests but seems like a hassle someone else would have figured out already ๐Ÿ˜…

hollow token
#

I haven't seen something like that at all

#

I really want to code coverage and unit test my CirPy project(s) but I don't see the capacity without a "fake"/"soft" board

#

I say this without any insult but I don't think these kinds of testing are on a feature map (yet)

#

I'd like them to be but I don't think (aka this is my opinion and not a fact) this kind of feature will be in any time soon

low sentinel
#

Personally I donโ€™t really care about coverage as a quality indicator. Just interested in turning a PR red when it makes circuitpython not work ๐Ÿ˜‰
Will have to spend a little time and see if thereโ€™s a relatively straightforward path. Probably have to implement some subset of pytest, thinking about it a little more ๐Ÿคฆ

hollow token
#

coverage is pretty useful not for quality but making sure the quantity of code is "correct"

#

It can be used for a number of different stats, quality isn't a great one

#

With pytest, if you run into pexpect, let me know, I'm one of the maintainers over there :)

#

(pytest uses pexpect for self testing, etc)

#

@low sentinel But what does "not work" mean here? Not build or break functionality/bugs that people might have been using previously?

hollow token
manic glacierBOT
low sentinel
#

@hollow token thereโ€™s theory about what coverage does but I have regularly seen it (1) reduce velocity, (2) increase hacks to reduce lines of code, (3) cause false tests that execute branches but assert nothing (I.e., it definitely does not indicate correctness from my perspective). Itโ€™s all anecdotal but yeah, not a fan & itโ€™s okay that others like it!

โ€œNot workโ€ was chosen intentionally as vague. I want to test a circuitpython/native interface that might behave differently in major revisions of circuitpython. Maybe Idk what youโ€™re asking about, I want to test my code on the target runtime. Seems like a normal thing to want to do? ๐Ÿ˜ƒ
Thanks, Iโ€™ll ping you if I run into troubles when/if I get a little time to do this.

hollow token
#

So you want actual tests and testing that CirPy builds isn't going to cut it, aka unit tests

#

and yes, unit tests are pretty normal and coverage is a tool that can be used as with any tool, incorrectly

manic glacierBOT
atomic summit
#

I ported the MicroBit 5x5 font over and write a font/scrolling text class - I need to work on the font kerning because the font doesn't include any width data, so going to need to parse the font and add some extra meta data, but I'm pretty happy with this!

manic glacierBOT
manic glacierBOT
#
sommersoft
[adafruit/circuitpython] Pull request opened: #5304 Include Frozen Libraries In build\_board\_info\.py

Fixes #4672

Tested with the following:

(.venv) ~/Dev/circuitpython/tools:$> python
Python 3.8.2 (default, May 20 2020, 16:07:42) 
[GCC 7.5.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import json
>>> import pathlib
>>> from build_board_info import get_board_mapping
>>> fpath = pathlib.Path().home()
>>> with open(fpath / "Dev" / "frozen_board_mapping.txt", "w") as f:
...     json.dump(get_board_mapping(), f, indent=2)
... 
>>>
...
manic glacierBOT
lone axle
jaunty juniper
low sentinel
#

@lone axle thanks these are great resources! Between this and redm Iโ€™ll be squared away.

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #5283: Sinewave tone playback on RP2040 no longer working as of CP7

[deleted and reposted - previous post had several errors]

The UDA1334A breakout cannot drive a speaker: it is line-level output meant to drive an amp. I tried a UDA1334A into a TPA2012 analog amp, playing WAV files and I am hearing output. Are you testing the UDA1334A directly to a speaker, or with an amp in between?

I am not hearing the sine wave test at the top, though; I'm working on that.

BTW, I am hearing more scratchiness on I2SOut than PWMAudioOut, but that is a different issue.

orchid basinBOT
orchid basinBOT
crimson ferry
#

why would boot.py execution behavior be affected by the content of code.py?

stuck elbow
#

fiesystem corruption?

crimson ferry
#

I started with a fresh storage.erase_filesystem()

jaunty juniper
#

what's the issue ?

crimson ferry
#

I can get it to happen consistently, but no minimal example (yet?)

#

if I have a boot.py (content of boot.py doesn't seem to matter?) in the release candidate (doesn't happen in beta), AND I have my normal code.py on device, The device, FunHouse in this case, will start up, I'll see the Blinka icon on the display, nothing else prints, then it goes into flashing purple and the LED every second. Needs safe mode and deletion or renaming boot.py to recover.

#

But a trivial code.py doesn't seem to trigger the behavior.

#

My code runs fine in the release candidate, only if there is no boot.py

jaunty juniper
#

so it would be your normal code causes a reset loop if there is a boot.py ?

crimson ferry
#

as far as I can tell, it should never be getting to my code.py, since boot.py doesn't even execute (e.g., a boot.py with just a print statement

#

just the Blinka icon and black screen

#

but then it goes black totally (no Blinka) when it starts the lighting pattern

#

I just can't fathom how the existince of anything in a code.py would affect whether boot.py runs successfully

#

any esp32-s2 that I've tested does this, on rc only, not beta

#

PyPortal does not have this behavior

jaunty juniper
#

to me it looks like code.py causes a reset loop

crimson ferry
jaunty juniper
#

you might not see anything on the screen because it happens before the screen is refreshed

crimson ferry
#

I'll try a delay

#

with this boot.py, same bahavior:py print("hello from boot.py") while True: print(".", end="")

jaunty juniper
#

yeah I just tried and the presence of a boot.py causes a boot loop

#

(on a funhouse in this case)

crimson ferry
#

but not always, certain trivial boot.py + certain trivial code.py run just fine

#

ah, interesting ...empty+empty

jaunty juniper
#

I am too lazy to test with an install without tinyuf2

crimson ferry
#

haha

jaunty juniper
#

but it's always a good test to do

stuck elbow
#

I don't think you can print in boot.py?

jaunty juniper
#

it goes to boot_out.txt

stuck elbow
#

there is no usb serial yet...

#

ah

crimson ferry
#

only built-in displays

#

trying it w/o TinyUF2...

#

empty code.py + empty boot.py is fine without TinyUF2, I'll file an issue

#

thanks for the help in narrowing it down

manic glacierBOT
#
anecdata
[adafruit/circuitpython] Issue opened: #5305 Presence of boot\.py in many cases will cause a boot loop on ESP32\-S2 with TinyUF2 \(rc, not beta\)

CircuitPython version

Adafruit CircuitPython 7.0.0-rc.1 on 2021-09-02; Adafruit FunHouse with ESP32S2

Code/REPL

None

Behavior

On an ESP32-S2 device (Fun House in this case) without TinyUF2, device behaves as expected.

On the same device with TinyUF2, starting clean from a fresh storage.erase_filesystem(), with either empty boot.py + empty code.py, or some (but not all) combinations of trivial and non-trivial boot.py + code.py, the ...

manic glacierBOT
#
gingle
[adafruit/circuitpython] Issue opened: #5306 Bitbang SPI looses least significant bit

CircuitPython version

Adafruit CircuitPython 6.1.0 on 2021-01-21; Adafruit QT Py M0 with samd21e18 also
Adafruit CircuitPython 7.0.0-beta.0-161-g65753a1c2 on 2021-09-01; with samd51j19

Code/REPL

# 09/03/2021
# looses both least significant bits
# tested on QT Py and Metro_M4_Express

import board
import digitalio
import adafruit_bitbangio as bitbangio


nEN = digitalio.DigitalInOut(board.A1)    
nEN.direction = digitalio.Direction.OUTPUT

spi =...
manic glacierBOT
#
aivarannamaa
[adafruit/circuitpython] Issue opened: #5307 Raw REPL greeting missing in CP 7\.0 rc1

CircuitPython version

7.0rc1 for RPi Pico

Code/REPL

# not related to code

Behavior

In MicroPython up to latest release and in CircuitPython up to (at least) 6.3.0, entering the raw REPL gives fixed "greeting": raw REPL; CTRL-B to exit>'. In CP 7.0 rc1 for Raspbery Pi Pico, this greeting is lost and Thonny IDE can't read the confirmation about entering raw REPL anymore.

I haven't tested, but I suspect all tools depending on pyboard.py (eg. rs...

manic glacierBOT
atomic summit
ornate breach
#

Nice!

atomic summit
#

Now I plan to ship some cool demo code on the board, including using the ntp stuff from @jaunty juniper as well as adding some extra glyphs and maybe some digital signal displays based on a pin state...

#

I might actually work on this on my stream tomorrow - hacking code realtime in CPY is super sweet!!!

ornate breach
#

It really is ๐Ÿ™‚

manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
aivarannamaa
[adafruit/circuitpython] New comment on issue #5307: Raw REPL marker missing in CP 7\.0 rc1

can you help us plan how to make our possible to localize it in the future such as in circuit python 8?

The best I can think of is prefixing the standard marker with translated explanation, so that Ctrl+A would produce something like this:

NB! You have entered the mode which is meant for automatic tools, not for human users. Press Ctrl+B to get back to friendly shell
raw REPL; CTRL-B to exit
>

Current Thonny version would then output the first line as if it was regular ...

#
dhalbert
[adafruit/circuitpython] New comment on issue #5283: Sinewave tone playback on RP2040 no longer working as of CP7

I have a fix that restores I2S output, by not trying to reuse the buffer supplied by RawSample as an output buffer. However, it is intermittently buzzy, which can sometimes only be fixed by a power-cycle. I hear this same buzziness in 6.3.0, so I think that was an existing problem.

The current RP2040 code does not do anything special to expand mono samples to stereo. I2S output is always stereo. By contrast, the SAMD implementation sets a special "mono" mode of the SAMD peripheral, which...

crimson ferry
# jaunty juniper empty boot.py, no code.py, boot loop

I think it's more complicated... if I start clean, empty boot & code will work repeatedly. It seems like there's some unexpected state being retained in some cases, and once the error triggers, it will keep triggering across hard resets until boot.py is removed. It would almost have to be something in flash.

manic glacierBOT
#
anecdata
[adafruit/circuitpython] New comment on issue #5305: Presence of boot\.py in many cases will cause a boot loop on ESP32\-S2 with TinyUF2 \(rc, not beta\)

I think it's more complicated... if I start clean, empty boot & code will work repeatedly. It seems like there's some unexpected state being retained in some cases, and once the error triggers, it will keep triggering across hard resets until boot.py is removed. It would almost have to be something in flash?

Another behavior I hadn't seen before: Some non-trivial boot.py and non-trivial code.py combinations will not trigger the looping, but the code.py will not run. boot.py output:...

warm verge
#

Hi, not sure if I'm being dumb or not. I have a feather S2 with a oled shield. When I plug into my macbook via usb-c the feather seems to randomly reboot and the drive disappears and then re-appears (as if I unplugged it and re-plugged it). Is this something that is happening for a reason? The screen says "Code stopped by auto-reload. soft reboot output:".

#

I'm new to micropython but a very experienced python dev.

stuck elbow
#

does this also happen without the shield?

#

the drive shouldn't disappear

warm verge
#

ah, this might actually be a watchdog running in the code.

stuck elbow
#

circuitpython does a soft reboot when it detects a write to the drive, but that shouldn't make the drive disappear

warm verge
#

What is the "Code stopped by auto-reload" bit? Is that part of the bootloader?

#

Actually, the drive seems stable now. No idea why.

#

and as I say that, it drops. I'm going to remove the watchdog code and see if that stops the drive disappearing.

stuck elbow
#

that's the soft-reboot after writing to the disk

warm verge
#

Thanks for the help @stuck elbow

manic glacierBOT
#
gingle
[adafruit/circuitpython] New comment on issue #5306: Bitbang SPI phase=1 lowest bit is wrong

Hi Dan,

I first noticed this as I was driving a DDS chip. Using busio the DDS did not set to quite the right frequency. The LSBit is generally not a big impact but for the DDS it was the LSbit in the msb of the frequency set and set the frequency about 10% low over what it should have been. I was using the equivalent of the metro_m4_express. I then did a Busio version which worked OK with the same polarity/phase settings. I then tried the bitbang version on the QT PY which is the waveform th...

#
gingle
[adafruit/circuitpython] New comment on issue #5306: Bitbang SPI phase=1 lowest bit is wrong

Hi Dan,

I am not well versed in Github unfortunately. One more thing to master. I posted the waveforms and programs on the Adafruit Forum under 'SPI Busio vs Bitbangio' on sept 3. It is under gingle in Adafruit Circuitpython and Micropython. The DDS was an AD9832.

Could you send any replies to @.@.>? I am doing most of my work at home and have my files on my home computer.

Thanks,
Glen

From: Dan Halbert @.***>
Sent: Tuesday, September 7, 2021 9:36 AM
To: adafruit/c...

slender iron
#

@tulip sleet free now

thorny jay
#

Skipping the meeting without notes... ๐Ÿ˜ฆ

slender iron
#

<@&356864093652516868> Meeting is starting in 5 minutes or so. (Was delayed 24 hours for Labor Day holiday here in the US.)

turbid radish
#

Link to the doc?

slender iron
ember iris
#

Going to lurk/afk through the meeting

tulip sleet
#

@turbid radish you can ctrl-Z the expansion of your name to undo it

#

i have to do that all the time

idle owl
#

I hit escape before moving on from my name.

turbid radish
#

Thanks

tulip sleet
#

that is better! I have to retype my name if i ctrl-z, but it does not expand the second time

turbid radish
#

Definitely, thanks Scott

lone axle
idle owl
#

Oh snap!

#

Entirely ๐Ÿ˜„

lone axle
onyx hinge
#

๐Ÿ‘‹ missing the meeting due to being a tourist

idle owl
onyx hinge
modern wing
idle owl
#

@crimson ferry Are you on mic today?

crimson ferry
#

sorry

idle owl
#

No worries ๐Ÿ™‚

trim elm
#

Oh yeah massive hug report to Justin for all the help

#

Probably saved me 30 hours of work at this point

pearl notch
#

A group anti-hug for weed whackers and leaf blowers

modern wing
idle owl
#

Sounds good.

jaunty juniper
lone axle
jaunty juniper
#

it lists latest stable and unstable if more recent

#

(like the download page)

idle owl
#

@jaunty juniper That sounds really useful.

crimson ferry
#

where does the board_id come from? it doesn't seem to be in the ports/*/boards files, but it seems to match the binary/UF2 filename?

slender iron
#

it's the folder name

crimson ferry
#

oh ๐Ÿคฆ

modern wing
#

Thanks!

errant grail
#

Thanks everyone.

jaunty juniper
#

thanks everyone toebeans

manic glacierBOT
slender iron
#

Here is the notes document for Mondayโ€™s CircuitPython Weekly meeting. It is at the normal time of 11am Pacific / 2pm Eastern here on Discord. Everyone is encouraged to attend! Please add your hug reports and status updates even if youโ€™ll be attending the meeting - itโ€™s super helpful! If you are unable to attend but would still like to include updates, feel free to include them in the notes and weโ€™ll read them off during the meeting. Hope to see you there! <@&356864093652516868> https://docs.google.com/document/d/1FFbgAbScOMYtfsaJNmNGlxLPucVKkRWj6YuQ2r4KEdk/edit?usp=sharing

Google Docs
CircuitPython Weekly for September 13th, 2021

CircuitPython Weekly for September 13th, 2021 Welcome to the CircuitPython Weekly meeting notes! Feel free to add your Hug Reports and Status Updates early. During the meeting, we go through them as a round robin sorted by username. If you canโ€™t make the meeting and would still like to partici...

slender iron
manic glacierBOT
#
water5
[adafruit/circuitpython] Issue opened: #5309 ATSAMD build error \(audio\_dma\.c\)

CircuitPython version

Adafruit CircuitPython github (2021-09-08), with ItsyBitsy M4 Express / Mini SAM M4.

Code/REPL

make BOARD=itsybitsy_m4_express
or
make BOARD=mini_sam_m4

Behavior

audio_dma.c: In function 'audio_evsys_handler':
audio_dma.c:420:31: error: implicit declaration of function 'dma_write_back_descriptor' [-Werror=implicit-function-declaration]
420 | (DmacDescriptor *)dma_write_back_descriptor(dma->dma_channel)->DES...

#
dhalbert
[adafruit/circuitpython] Pull request opened: #5310 Fix RP2040 I2S: always copy to output buffer

Make RP2040 I2S work again.

  • Don't attempt to use the sample buffer as the DMA buffer if no conversion is needed. Instead, always copy the sample buffer into the output buffer.
  • Move the I2S PIO restart slightly to avoid starting before an error check.
  • Add explanatory comment about getting mono->stereo conversion for free.

The quality of the I2S audio is not great: there are buzzes and noise, which happen erratically. A power cycle or reset can improve things temporarily. However,...

idle owl
#

@tulip sleet Can I please get builds for the Pico and the Feather RP2040 from you? Or I can wait until tomorrow when the checks have completed. Up to you.

#

Builds from #5310.

tulip sleet
manic glacierBOT
tulip sleet
idle owl
manic glacierBOT
#
eat-sleep-code
[adafruit/circuitpython] Issue opened: #5311 Mini Remote Control: No signal sent is valid?

@tannewt

I am tinkering with parts from an old AdaBox -- a Circuit Playground Express and Mini Remote Control [Product ID: 389].

I installed CircuitPython 6.3.0 on the Circuit Playground Express. I also created a lib folder on the CPE and dropped the adafruit_irremote.mpy from Bundle for Version 6.x in it.

I then placed this code into a code.py file on it:

import board
import pulseio
import adafruit_irremote

pulseIn = pulseio.PulseIn(board.IR_RX, maxlen=120, idle_state=...
idle owl
#

@tulip sleet Tried Pico since that was first available. WAV playback worked immediately. I'm still getting silence for tones. ๐Ÿ˜• Going to try the other Pico I brought with me just to try it, but .. eh.

manic glacierBOT
tulip sleet
#

@idle owl this is the test program I used for a sine sample. I separated the cycle size and the length so I could vary one without affecting the other.

๐Ÿ“Ž sinei2s.py
manic glacierBOT
idle owl
#

With UDA and headphones.

#

Hmph.

#

Ok. I haven't heard crunchy yet. But I believe you on that one.

#

@tulip sleet Do you want me to test on Feather as well?

tulip sleet
idle owl
#

Ok. Sounds good.

tulip sleet
#

what is the diff between your tone program and mine??

idle owl
#

I'll send you mine.

#

give me a sec, it's in a bunch of other commented out programs because I was going back and forth so quickly testing things the other day.

tulip sleet
#

i have a zillion small programs: they are all in the .zip in the PR

idle owl
#

@tulip sleet ^^

#

The I2S pins are obvs different.

#

But that's all I changed in yours.

#

And yours works.

tulip sleet
#

one is loop=True and one isn't. I will try in a few minutes: just about to eat a quick meal

idle owl
#

Ok. Take your time!

tulip sleet
# idle owl Ok. Take your time!

since the sample you created is not looping, it is quite short. Try changing the * 8 to * 80. It makes a short audible blip tone for me

idle owl
#

Ok checking.

tulip sleet
#

at the same time, you may need to change the length in the for loop to 8000 // frequency, to generate just one cycle that's repeated over and over

idle owl
#

There's a faint something, and if I change the frequency, that faint something changes pitch. But it's not a tone.

tulip sleet
#

I"m not sure why the * 8 was added in the first place

idle owl
#

Ok trying the second thing you suggested.

tulip sleet
#

anyway, my test program is essentially the same thing. cycle is 8000 / frequency (I used float division for better accuracy).

idle owl
#

There it is.

#

Take out the * xx and it plays a tone.

#

Weird though

#

If I simply take it out of the length variable, it is still silence.

tulip sleet
#

the original tone code in, say, the CPy lib, does not have the *8, and it has loop=True, right?

idle owl
#

If I change length inside the for loop to (8000 // frequency), it plays a tone.

#

They're the same thing!

tulip sleet
#

right, because otherwise that value is wrong, and will not generate a sine wave of the corret frequency

idle owl
#

ohhh

#

I don't entirely follow, but I've seen weirder, so I'll trust you.

manic glacierBOT
#
tannewt
[adafruit/circuitpython] Pull request opened: #5312 Selectively build boards based on changed files

This uses a step output from the test job to set the build matrix
for the board build jobs. The built boards depends on which files
were changed.

  • Changes contained within ports/*/boards/ will build each board.
  • Changes contained within ports/* will build all boards from the
    port.
  • All other file changes will build all boards.

All boards will be build for pushes in adafruit/circuitpython as
well.

A side-effect is that we no longer need to explicitly list the
boards to b...

tulip sleet
#

that's why mine separates the two. cycle is the length of a single sine wave cycle, up and down and up to zero again. length is how long the whole thing is

idle owl
#

Ok, so it has to have the * 80

#

AND change length in the for loop.

#

To play the tone.

tulip sleet
#

IF you don't have loop=True

idle owl
#

I was about to flip a table over here trying to get the tone back.

#

Hmm ok.

#

Point.

tulip sleet
#

it is playing the whole sample, just once, without loop=True. So the longer length is, the longer it will play

idle owl
#

Works with loop=True and the rest doing the thing.

#

Much clickier this way.

tulip sleet
#

e.g. from CPy library: self._sample.play(self._sine_wave_sample, loop=True)

idle owl
#

Right.

#

I don't know why my code ended up this way. It's gone through so many iterations, and I'm not even sure where I got it from.

#

Odd about the clickiness tho.

#

Maybe more time for it to happen.

#

Or my perception of it is wrong.

tulip sleet
#

the i2s quality is still not good, though. if you let it play many times it may turn rough

idle owl
#

Fair enough.

#

We tout it as higher quality! ๐Ÿ˜†

#

I guess I'll leave that bit out of the template for now.

#

(Which means forever because I can't imagine remembering to go back and change that if we resolve that.)

tulip sleet
#

it's higher quality except on RP2040

idle owl
#

Ahhhhhh

#

Blergh. Ok.

#

Have to make that section a template section then.

tulip sleet
#

it should be pretty fine on the other boards

idle owl
#

So it says that for other boards but not RP2040.

#

Fair enough.

#

Now I kind of want the CircuitPython version numbers to include -grungy instead of -dirty when altered. ๐Ÿ˜„

manic glacierBOT
#
urish
[adafruit/circuitpython] New comment on issue #5212: USB output only starts after main\.py finishes \(RP2040\)

Can this simulator be run headless? I'd love to use it for library CI to track memory use.

Yes, I made a small demo repo for you it runs the simulator from the command line inside a Node.js progress:

https://github.com/wokwi/rp2040js-circuitpython

It's just the bare minimum that you need to get the Pi Pico CircuitPython build running in your terminal - if you need any help customizing it to run on the CI, I'd love to help!

idle owl
#

@tulip sleet Ok, I'll reply to the PR.

tulip sleet
#

thanks for the testing; we do have to figure out what's wrong with the RP2040 I2S. The way it's implemented is very different from other ports.

manic glacierBOT
idle owl
#

@tulip sleet Merge it?

tulip sleet
#

@slender iron do you want to review #5310?

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] Issue opened: #5313 RP2040 I2SOut sounds bad

RP2040 I2SOut can have many clicks and buzzes, and other artifacts. The quality can vary between successive plays of the identical audio. A power cycle or reset can make it sound better, but only for a while.

The PIO I2S implementation is free-running, at the same clock rate as the DMA writing. But it appears to get out of sync or bad data is introduced somehow.

See #5283 and #5310 for further comments and test programs.

slender iron
#

Just left for a walk. Iโ€™m ok if you want to merge. Otherwise I can look when I get back. Must have come in after I got through my email

tulip sleet
#

tnks, I think we'll go ahead for now, since it's a quick fix for RC

manic glacierBOT
tulip sleet
#

@idle owl I am updating the frozen libraries to catch up for 7.0.0 final, but are there any upcoming library changes I should wait for?

idle owl
tulip sleet
#

great, thanks!

manic glacierBOT
slender iron
#

@tulip sleet looks good

manic glacierBOT
orchid basinBOT
manic glacierBOT
crimson ferry
#

This has probably been asked and answered somewhere, but what do you do to make the (esp32s2) CP image smaller so that a DEBUG build will make successfully?

onyx hinge
#

@tulip sleet I'd like to get the fix for thonny in 7.0.0, I'll work on it tomorrow while I'm at adafruit

#

If it's not too late

tulip sleet
manic glacierBOT
spiral elk
#

I suspect that the REPL marker is partially to blame with some of the problems I'm having with the Lolin S2 Mini. (After Thonny fails to connect to it, it goes into crash mode and I can only reproduce this with Thonny, I can connect to repl just fine with screen, putty, etc. without crashing it)

manic glacierBOT
#
maholli
[adafruit/circuitpython] New comment on pull request #5302: updating pycubed firmware for mainboardv05

How about making a new board def for the new revision? That way the old one will still work. You can add the boot counter behind a flag and then merge it in if you want to be on main.

@tannewt are seperate board defs your preferred method of handling major board revisions? Part of me yearns for a cleaner solution than that, especially since folks are wanting firmware for multiple D51 flash sizes (J19 and J20, for example). Maybe solution is discrete board defs but easily managed ...

onyx hinge
tulip sleet
#

he did a revert, but Scott would like something better in the long run, yes. But that doesn't have to be for 7.0.0

manic glacierBOT
manic glacierBOT
manic glacierBOT
orchid basinBOT
slender iron
crimson ferry
#

thanks, I set up to try that, and will, after I find and catch the chipmunk that got into my house

slender iron
#

thanks!

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #5313: RP2040 I2SOut sounds bad

I did further testing, and I don't believe this is a software problem. It was some kind of electrical issue with the MAX98357A and how I wired it. When I connected a Saleae up the problem went away. When I removed the Saleae ground return, the problem came back. I connected Vin and GND more directly, and that fixed the problem. Reducing the MAX98357A gain might also help.

crimson ferry
#

๐Ÿฟ๏ธ issue: Closed.
DEBUG=1 make issue: โœ…

#

but , ๐Ÿคฆ , got the funhouse all set up to test and realized it doesn't have the debug pins exposed, so I have to test this on a device without a built-in display

waxen pike
#

Howdy! I searched, and didnt find what I needed, so perhaps I'm looking for the wrong thing, or using the wrong terms

I'm currently working with a MatrixPortal M4 w/ the 32x64 matrix display. I'm having a hard time finding a way to display simple symbols, even common unicode, for like arrows. I'm assuming this is due to the bdf and/or the terminalio.FONT font not having these characters in them.

So, are there any symbol / glyph bdf fonts? Or do I need to figure out tilegrids?

manic glacierBOT
#
anecdata
[adafruit/circuitpython] New comment on issue #4984: ESP32\-S2 DEBUG=1 builds failing

Reducing wifi buffers was another suggestion to workaround this, but I couldn't find where / how to do that, so turned off modules so that it built. Just as a stake in the sand, I was able to build DEBUG=1 on esp32-s2 7.0.0-rc.1 with the following changes:

CIRCUITPY_AUDIOCORE ?= 0
CIRCUITPY_AUDIOBUSIO ?= 0
CIRCUITPY_AUDIOBUSIO_I2SOUT ?= 0
CIRCUITPY_AUDIOMIXER ?= 0
CIRCUITPY_CANIO ?= 0
CIRCUITPY_DUALBANK ?= 0
CIRCUITPY_IMAGECAPTURE ?= 0
CIRCUITPY_ROTARYIO ?= 0
CIRCUITPY_PS2IO...
#
anecdata
[adafruit/circuitpython] New comment on issue #5305: Presence of boot\.py in many cases will cause a boot loop on ESP32\-S2 with TinyUF2 \(rc, not beta\)

Switched to Cucumber RIS since Funhouse doesn't have debug pins exposed.

  1. erased flash
  2. installed TinyUF2 combined
  3. installed CP rc.1

Initial test with no boot.py and default code.py gives this in the debug console on startup:

# ...
I (904) cpu_start: Pro cpu up.
I (1318) spiram: SPI SRAM memory test OK

Creating an empty boot.py gave an ambiguous result on hard reset, repeatedly. No "Hello World!, but it's possible that tio didn't reconnect in time. It ...

lone axle
# waxen pike Howdy! I searched, and didnt find what I needed, so perhaps I'm looking for the ...

I think you are correct in your assumption that the default font does not include many of the special character symbols (I think only the ones available on a standard keyboard using only default keys and shift).

I don't know specifically about unicode symbols but Jeff did port a copy of an early version of font awesome for use with circuitpython. Here is an example that shows how to use it: https://github.com/adafruit/Adafruit_CircuitPython_Bitmap_Font/blob/main/examples/bitmap_font_label_forkawesome.py perhaps there are symbols within that font that will suit your needs.

waxen pike
manic glacierBOT
waxen pike
onyx hinge
#

@tulip sleet is this the non-updated version of USB descriptors? bitmap_keyboard = usb_hid.Device( report_descriptor = bitmap_keyboard_descriptor, usage_page = 0x1, usage = 0x6, in_report_length = 16, out_report_length = 1, report_id_index = BITMAP_KEYBOARD_DESCRIPTOR_REPORT_ID, )

#

Is the fix report_ids = [BITMAP_KEYBOARD_DESCRIPTOR_REPORT_ID] and remove report_id_index?

manic glacierBOT
tulip sleet
onyx hinge
#

@tulip sleet thanks, looks like my page in that usb hid device (at least) didn't get updated

tulip sleet
#

i did update my HID page in the guide

onyx hinge
#

the only device I have with me will probably run into trouble with this esp32s2 boot.py bug that has been reported so I don't want to experiment right now

#

and also some stuff on my personal blog, which someone mentioned here on discord the other day I think

manic glacierBOT
#
jepler
[adafruit/circuitpython] New review comment on pull request #5315: Echo control characters used for REPL modes

This character is defined as "enquiry". Some terminal programs respond to it, for instance the rxvt terminal on Linux will respond with the binary string \033[1?2c, explained as

"ESC [ ? 1 ; 2 c" (``I am a VT100 with Advanced Video Option'')

The other characters you used have some kind of meaning in ASCII (and, by inheritance, as unicode/utf-8) but ^E is the one that I think may cause a problem with "real, in the world" terminals.

onyx hinge
#

did anybody get a handle on this esp32-s2 problem yet? I wonder if scott & I didn't fully get that gc bug fixed or something.

manic glacierBOT
tulip sleet
manic glacierBOT
#
microDev1
[adafruit/circuitpython] Pull request review submitted: #5315 Echo control characters used for REPL modes

This doesn't completely resolve #5307.

While working on this issue I noticed that the raw repl marker was missing because pyexec wasn't included in the qstr sources and thus translate() returned NULL.

Initially, #5308 contained a fix for this issue by adding pexec to qstr source but later I changed the PR to reverting localization when the concern about localization affecting the use of the message as a marker was raised.

onyx hinge
#

aha that explains why the strings didn't appear, I guess. I would have told you I tested it but clearly that would have been a lie.

orchid basinBOT
manic glacierBOT
crimson ferry
#

what's the generic way to print to some serial from within core code for debugging purposes?

tulip sleet
#

mp_printf(&mp_plat_print, "x: %d\n", x);

#

but on ESP32-S2 you can use the ESP_LOGI() macro

crimson ferry
#

I tried the esp-idf logging, but it didn't seem to work in main.c (I could have easily put it in a bad spot)

tulip sleet
#

you have to do ESP_EARLY_LOGI() or something like that

#

look for EARLY

crimson ferry
#

ah, right, will try both

#

thanks

waxen pike
manic glacierBOT
#
tannewt
[adafruit/circuitpython] Issue opened: #5317 Translate message in pyexec and use invisible characters to convey serial state to tools

Follow up from #5307. CP 7 broke Thonny and other serial based tools because it relies on certain messages. Translating these messages breaks the tools. Instead, we should add invisible characters that these tools can used to determine serial link state (human repl, raw repl vs paste mode). These characters will remove the dependence of these tools on the messages that may change from version to version.

cc @aivarannamaa

onyx hinge
#

@slender iron from here I could just feel your pain at being practical, but given time we'll be able to get to the better place you're aiming for ๐Ÿ™‚

slender iron
#

you know me too well

onyx hinge
#

nah, but i know you a good amount

manic glacierBOT
#
mrdalgaard
[adafruit/circuitpython] Issue opened: #5318 nRF: deep sleep increases power usage

CircuitPython version

Adafruit CircuitPython 7.0.0-rc.1 on 2021-09-02; Adafruit Feather nRF52840 Express with nRF52840

Code/REPL

import time, alarm
time.sleep(15)
time_alarm = alarm.time.TimeAlarm(monotonic_time=time.monotonic() + 30)
alarm.exit_and_deep_sleep_until_alarms(time_alarm)

Behavior

After startup my cheap multimeter measures around 0.5mA for the first 15 seconds, and over twice that for the last 15 seconds (in deep sleep)
Light sle...

crimson ferry
#

"bisect" means something like: black-box testing the 106 S3 UF2s between the most recent known good-build and the oldest known issue-build, binary-search-like, to find which merge introduced the issue?

tulip sleet
#

you could use the S3 UF2's if you have a known good one, but I clean them up every month or two

crimson ferry
#

beta.0 was fine, rc.0 was not ...I'll do it manually since the builds are already there and I have to load & run each anyway to see if boot.py works

manic glacierBOT
slender iron
#

BLE central debugging

manic glacierBOT
jaunty juniper
#

to be fair, you only need to test approximately log2(106) builds though

crimson ferry
#

right, the range of 106 ...actually only 53, I was also counting the .bin files ...actually picked the first bad one on the first round, but had to run 5 more to confirm

manic glacierBOT
manic glacierBOT
idle owl
#

(โ•ฏยฐโ–กยฐ๏ผ‰โ•ฏ๏ธต โ”ปโ”โ”ป

#

@tulip sleet So I switched back to my way original tone code. Which loops. But I didn't need to make any other changes, including to the volume variable, and I have tone playback.

#

So I guess you fixed it, is my point, but you fixed it more than I think we thought you did. Or maybe you ended up dragging my bastardised code back to this original version in the end anyway.

manic glacierBOT
idle owl
#

Hmm. Where is this I2S error coming from? ```ValueError: Bit clock and word select must be sequential pins

orchid basinBOT
#
durapensa
[adafruit/circuitpython-org] New comment on issue #752: New "Install UF2 Bootloader" section in download/board\.html?

I made a start on this and found that many boards are named differently in the respective ports/esp32s2/boards directories:

TINYUF2                           CIRCUITPYTHON        
adafruit_feather_esp32s2          adafruit_feather_esp32s2_nopsram 
adafruit_feather_esp32s2_tft      adafruit_feather_esp32s2_tftback_nopsram                         
adafruit_funhouse_esp32s2         adafruit_funhouse                
adafruit_magtag_29gray            adafruit_magtag_2.9_grayscale    ...
blissful pollen
idle owl
#

Ok, that's what I'm testing on, and I changed the pins to force the error. But wasn't sure if it was RP2040-specific.

#

Thanks.

blissful pollen
#

Just double checked and yeah it is in the RP2040 port and not in the samd port when i checked

idle owl
#

Have to include an extra section in the template to add a warning for RP2040 for pin selection.

blissful pollen
#

Figured it was to add to a guide so wanted to double check myself for my own sanity too ๐Ÿ™‚

idle owl
#

Fair enough! Appreciate the thoroughness. The templates are even worse than simply adding it to a guide because they're going to go in every board guide eventually. So tweaky stuff like this makes them vaguely more frustrating/complicated. But the end result of a tailored page for each board is going to be really nice to have.

blissful pollen
#

ha now i'm really glad I double checked! But I can imagine when it varies just slightly its frustrating at times

idle owl
#

Yeah, some of these templates have a bunch of sections like "If this is for a SAMD51 board, include the following, otherwise.." etc. Some templates are quick and smooth. Others, I'm finding, not so much.

#

@blissful pollen Clear enough? The bit clock and word select pins must be on consecutive pins! They can be on any pins you like, but they must be in consecutive order, for example, A0 for bit clock and A1 for word select.

orchid basinBOT
#
Neradoc
[adafruit/circuitpython-org] New comment on issue #752: New "Install UF2 Bootloader" section in download/board\.html?

The way that section works is by reading this file referenced as site.data.bootloaders in the template.
https://github.com/adafruit/circuitpython-org/blob/master/_data/bootloaders.json

It contains the information on which port has a bootloader, and the conversion between the circuitpython board ID and the bootloader board ID (bootloader_id in the template). As far as I know it's entirely manually made (including the version number). But note that it currently contains 0 information on t...

blissful pollen
#

Reads clear to me

idle owl
#

Thanks!

orchid basinBOT
#
durapensa
[adafruit/circuitpython-org] New comment on issue #752: New "Install UF2 Bootloader" section in download/board\.html?

Ahh, thank you. I edited that file for lolin_s2_mini, also adding esp32s2 to bootloaders, for testing but overlooked the possibility of bootloader_id being different from the board's key in boards.

In that case, these boards should be an easy add:

adafruit_feather_esp32s2          adafruit_feather_esp32s2_nopsram 
adafruit_feather_esp32s2_tft      adafruit_feather_esp32s2_tftback_nopsram                         
adafruit_funhouse_esp32s2         adafruit_funhouse       ...
manic glacierBOT
idle owl
#

Wait, can you write code in CircuitPython, that expects input in the serial console? Like pausing a WAV file until you press enter in the serial output..? Or is that a Python thing.

crimson ferry
tulip sleet
#

there's no pattern - it's not automatic

#

e.g. aws s3 rm --recursive s3://adafruit-circuit-python/bin --exclude '*' --include '*20210822*

jaunty juniper
#

oh the mpy-cross S3 might need a cleanup too

tulip sleet
#

i'll do that, definitely, thanks

crimson ferry
#

@idle owl ```py

input("<<<")
<<<xyz
'xyz'```

idle owl
#

I just managed to find an example I did using input() ๐Ÿ˜„

crimson ferry
#

lol, I had never done it before, but have had it in the back of my mind