______________ ________________ ____ _____________ __ ________
/_ __/ ____/ / / ____/ ____/ __ \/ __ \/_ __/ ____/ |/ / / _/ _/
/ / / __/ / / / __/ / / / / / / /_/ / / / / __/ | / / / / /
/ / / /___/ /___/ /___/ /___/ /_/ / _, _/ / / / /___ / | _/ /_/ /
/_/ /_____/_____/_____/\____/\____/_/ |_| /_/ /_____//_/|_| /___/___/
A rewrite of the Telecortex project in NodeJS which controls APA102/SK9822 LED strips over OPC
Video of Steamed Hams on a previous version of TeleCortex
Our Balena image running on a Raspberry Pi 3 can control 4 channels of 300 SK9822 pixels pixels with UDP over on-board WiFi at about 200 FPS.
We've created a docker image that sets up your Raspberry Pi 3B+ as an Open Pixel Control server with 4 SPI ports. The docker image is designed to run on BalenaOS. This is the recommended way of using TeleCortex if you don't need to do development on the server. Otherwise, there's a lot of messing around required.
Sign up for BalenaCloud (it's free!) and set up an application. Use its wizard to make an image with your network details.
Or you can download and configure your own image as long as it runs docker. Head over to https://www.balena.io/os/#download and get yourself a BalenaOS image.
Once the Pi(s) is running you can push an image to it by following the instructions below
With a fresh install of raspbian lite, get your raspberry pi configured with the correct locale using raspi-config
. Optionally, while you're there you may wish to enable SSH, set the correct WiFi country, change the root password, and set a unique hostname.
(this can be done by editing the file on the SD card, or while the pi is on) As root, add the lines
dtparam=spi=on # this one is not necessary if you've already enabled it in raspi-config
dtoverlay=spi1-2cs
to /boot/config.txt
and reboot.
For more options (e.g. 5 SPI devices) and pinouts see /boot/overlays/README
.
To test, ls /dev | grep spidev
should show
spidev0.0
spidev0.1
spidev1.0
spidev1.1
# Fresh pi needs update
sudo apt-get update && sudo apt-get upgrade
# Install essential tools
sudo apt-get install vim git zip python
Different versions of Node support different CPU architectures.
v10
is LTS until 2020 and supportsarm64
,armv6l
andarmv7l
v11
went EOL in 2019 and supportsarm64
,armv6l
andarmv7l
v12
and above only supportsarm64
andarmv7l
This means that the version of node you should install depends on your architecture and whether you need security patches.
Some versions of raspbian / dietpi for your pi's architecture will install node v10+ out of the box, some won't. The way to find out is
apt info nodejs | grep 'Version'
We need this version to be greater than or equal to 10. If it's not, you'll need to use option B or C, depending on your architecture, if it is, you can just install it with
sudo apt install nodejs -y
Note: sometimes, package managers will link the node binary to nodejs
, not node, you could do something like
ln -s /usr/bin/node $(which nodejs)
You can determine your CPU architecture with:
uname -m
If if your architecture is armv7 or later (RPi Model >2), you want to install v12
or later
If if your architecture is armv6 (RPi Model 1, Zero), you want to install v10
.
This script will install the latest node for a particular version.
NODE_VERSION='v10' # or whatever
ARCH=$(uname -m)
NODE=$(curl https://nodejs.org/dist/index.tab | grep v10 | grep linux-${ARCH} | cut -d $'\t' -f 1 | head -n 1)
curl "https://nodejs.org/dist/${NODE}/node-${NODE}-linux-${ARCH}.tar.gz" -o "node-${NODE}.tar.gz"
tar -C /usr/local -zxf "node-${NODE}.tar.gz"
You may need these build tools.
sudo apt-get install gcc g++ make cmake
This will add the nodesource repository to apt
# install node 11
curl -sL https://deb.nodesource.com/setup_11.x | sudo -E bash -
sudo apt-get install -y nodejs
node -v
npm -v
git clone https://github.com/Laserphile/JS-Telecortex-2-Server
cd JS-Telecortex-2-Server
See: these instructions
npm install
npm run build
npm link
This will only work as a PBX server on OSX. You need /dev/spidev...
for pi-spi
to work.
If you want to push to your pi without going through the pipeline. Make sure you also "enable local mode" on BalenaCloud if its a cloud image.
-
npm install --global --production --unsafe-perm balena-cli
-
Get the hostname of the device you want to push to with
sudo balena local scan
-
build and push to the device with
balena push <hostname>
-
Be patient, the output will just stop for about an hour with no loading screen.
-
You'll know the push completed when you get to
[Build] [main] Successfully tagged local_image_main:latest [Info] Streaming device logs...
To view the output of the main container, you can do
balena ssh <hostname>
- then inside that session, do
balena container ls
to get a list of running containers. the app is likely the latest created container which means is necessary for the next step to work - Attach to the output of the main container with
balena attach $(balena container ls -lq)
npm start:nodemon
npm start:dev
get a list of command line options with
npm start -- -h
spi
--spi-clockSpeed SPI Clock speed to use for all devices
[number] [default: 3000000]
--spi-mode SPI Data mode to use for all devices
[number] [default: 0]
--spi-channels SPI Channel information (as JSON)
[string] [default: "{"0": {"bus": 0, "device": 0}, "1": {"bus": 0, "device":
1}, "2": {"bus": 1, "device": 0}, "3": {"bus": 1, "device": 1}}"]
pbx
--pbx-ports PBX Serial port definitions for js-pixelblaze-expander
[string] [default: "{"0": {"name": "/dev/ttyS0", "options": {"channels": {
"0": { "capacity": 300 }, "1": { "capacity": 300 }, "2": { "capacity": 300 },
"3": { "capacity": 300 }, "4": { "capacity": 300 }, "5": { "capacity": 300
}}}}}"]
--pbx-channels PBX channel definitions for js-pixelblaze-expander
[string] [default: "{"0": { "port": 0, "channel": 0 }, "1": { "port": 0,
"channel": 1 }, "2": { "port": 0, "channel": 2 }, "3": { "port": 0, "channel":
3 }, "4": { "port": 0, "channel": 4 }, "5": { "port": 0, "channel": 5 }}"]
Options:
--version Show version number [boolean]
-d, --devType Type of device used
[choices: "PBX", "SPI"] [default: "SPI"]
-p, --port port used to listen for OPC commands
[number] [default: 42069]
-b, --brightness global brightness value (from 0.0 to 1.0)
[number] [default: 1]
-t, --transportProtocol OSI Transport Layer protocol with which the server
will listen [choices: "TCP", "UDP"] [default: "UDP"]
-m, --middlewareProtocol Protocol used to translate colours before sending to
device
[choices: "colours2sk9822", "colours2ws2811", "colours2ws2812", "colours2rgb"]
[default: "colours2sk9822"]
-h, --help Show help [boolean]
so, if you wanted to use a different serial port, devType and middleware, you could do:
npm start -- src/main.js -d PBX -m colours2rgb --pbx-ports '{"0": {"name": "/dev/tty.usbserial-AD025M69", "options": {"channels": { "0": { "capacity": 300 }, "1": { "capacity": 300 }, "2": { "capacity": 300 }, "3": { "capacity": 300 }, "4": { "capacity": 300 }, "5": { "capacity": 300 }}}}}'
If everything is working, you should see something like this:
🛰 UDP Server listening on port: 0.0.0.0:42069
Before setting up an OPC client such as JS-Telecortex-2-Client, you can test that everything is working by running this command in a new terminal window on the pi:
printf "\x00\x00\x00\x03\x00\xff\x00" >/dev/udp/127.0.0.1/42069
printf "\x01\x00\x00\x03\x00\xff\x00" >/dev/udp/127.0.0.1/42069
printf "\x02\x00\x00\x03\x00\xff\x00" >/dev/udp/127.0.0.1/42069
printf "\x03\x00\x00\x03\x00\xff\x00" >/dev/udp/127.0.0.1/42069
# |ch-|cmd|cmd_len|--colors--|
# ^ ^ ^ ^
# ch: 3 ' | | |
# cmd: 0 (8bit RGB) | |
# cmd_len (uint16BE): 3B |
# colours: 1 green pixel /
This will send a UDP OPC message to localhost port 42069 that sets the first pixel on each channel (0-3) to green. For more info, Check out the OPC protocol details