its been a very long time - really i could be wrong - but iirc, no. for that i think you would use the DFU bootloader on the panel (which is just USB, albeit A-A in this case) or JTAG on the pcb.
^
(to be ultra-clear, this is totally unrelated to norns or rasbpi.)

i’m talking about something that makes your USB port look like a serial COM port. like something with FTDI in it. or the adafruit thing i linked above. or a teensy/avr/etc with appropriate firmware implementing USB CDC device class.

or of course just an actual rs232 port. (all you need is rx, tx and ground) - but in this (probably rare) case do check the voltage levels - really old systems could be putting out crazy stuff like 5vc, 12v… 25v?

1 Like

Yeah - I think that’s the cable I have - which I was using to get debug with meadowphysics - connected to the pins on the back of the module.

Aha… I’ve forgotten already after a month. That’s just for debug, not programming/flashing.

Edit - turns out I also have an FTDI-friend board which should work as well.

2 Likes

righto - yes, same thing you would use to get debug prints off the UART from a libavr32 monome device. (trilogy/ansible/TT - aleph has a CDC device port builtin.)

bleep bloop all the docs
https://elinux.org/RPi_Serial_Connection

highly recommend familiarizing w/ this if you are hacking on rpi at all really - it will be needed sooner or later.

Thanks. I’ve just bought the Adafruit cable you mentioned. It seems easier indeed. The only issue is that I was not planning to use the UART gpios (14 and 15) when I designed my board so I used them for connecting the three push-buttons :sweat: Too bad, otherwise it would have been easy to solder a UART breakout…

Edit: I’ve updated my board to include a UART header

Good News Everyone!

The framebuffer select thing @zebra did is working good for me with my ssd1322 oled. The 2.0 dev branch is updated with that change now.

So… this means no more having to change screen.c for fb1.

Instead…

matron.sh would be updated to

./build/ws-wrapper/ws-wrapper ws://*:5555 ./build/matron/matron -f/dev/fb1

and /etc/systemd/system/norns-matron.service should also be updated to add

-f/dev/fb1
5 Likes

hrmf, this makes me think we should just change matron.sh to start/restart the service, putting all magic numbers in one place.

anyways, matron.sh and start.sh should only be needed for debugging, if at all. they are relics of an earlier time when the services didn’t exist; we should probably remove them but there are now a ton of posts on here recommending their use… so…

Sounds good to me. Then keep matron.sh around for bacwards compatability?

1 Like

20 characters of finished DIY-project.

Edit:
And he/she plays nice with good old 64.

31 Likes

With norns 2.0 fast approaching(today maybe). Anything in particular one should think of before upgrading? Can you (theoretically) use update.sh and then fix the button/encoder overlays + battery status removal + framebuffer stuff afterwards? That would be lovely.

Depending on how the kernel upgrade goes, that might need commenting out too. I was on 1.xx and lost all audio drivers. YMMV

1 Like

I’m out most of the day today, but i can try to work up my upgrade notes later tonight

  • as mentioned above you’ll probably want to comment out the kernel update in update.sh
  • Battery status thing Is no longer an issue.
  • framebuffer is now assignable so there’s a new way to get that working.
3 Likes

This is all great stuff. Should be a breeze then. If it’s out befare my bedtime today I’ll give it a shot and report my results.

I’m testing the “new” UART connection, it works perfectly :slight_smile:

screen /dev/cu.usbserial* 115200

Ready for a new install from scratch…

2 Likes

Does anyone know how much adaptation would be required to run this on an Asus Tinkerboard? I have one lying around.

that’s what I did and everything went fine

ok, I sFTPed the update into home/we/update, checked update.sh for any kernel update info(found none?) and tried to run update.sh. Got the following error:

The version.txt file is for some reason missing. weird stuff.

The version.txt file normally just contains something like 190314. You could try manually creating the file…

yeah, that was not the issue. had my head on a bit wrong. But yeah, 2.0.0 is working, only edits I had to do was get the USB audio working and the new frame buffer fix from @okyeron. I’m having some network issues, any tips on how to troubleshoot? Cannot seem to connect to wifi in the system/wifi settings.
I can still ssh into the RPi, but Maiden is not functioning anymore either(but says it’s running when running journalctl -u norns-maiden.service -b in terminal.

Apart from that:
Really painless update process actually. hopefully I’ll have some more time to test later today.

One of the major parts of the wifi/networking improvements was switching to NetworkManager. For the most part matron is now simply shelling out to various nmcli calls so my first thought would be to check for missing packages. I’d try installing the package (which are commented out) here:

What does the output from journalctl -u norns-maiden.service include? Are you trying to access maiden via an mDNS name like norns.local or IP address? If using an mDNS name I’d try the IP address just to rule out mDNS…

If you can get into maiden but still running into problems with the on device wifi menu then I’d be curious to know if there are any errors reported in the matron repl within maiden when you navigate to the device wifi menu and try stuff.

After a restart of the RPi enc2 and enc3 wont work and it stopped connecting to my network. Will need to connect to a monitor and get a keyboard to debug this later on.