This project is a complete package for controlling radio AC outlets with Node and Python. All documentation is in the respective directory, see below.
- Run
npm installin theappfolder. - npm start (for development)
This can also be installed as a node dependency.
This project is for wireless (433MHz) AC outlets controlled by the Raspberry Pi. I've wanted to control AC outlets for a while now. All home automation solutions were far to expensive for me just tinkering. I've seen similiar projects but most use an Arduino to send the radio signals. I wanted to use my pre existing temperature logger web server and add control to it. Thus this project was born.
Thoughts: I previously made a power tail that wired directly to the Raspberry Pi which worked fine for controlling 1 outlet however it was clunky, not very portable, and most likley a fire hazard.
The wireless outlets are already IEEE certified and encapsulated. So there is no worry of anything going wrong or someone getting electrocuted. The only trade of is now I don't know the state of the AC outlets. Communcation is only 1 way. I was thinking about just making a cronjob that runs every 15 minutes to ensure that the state is correct most of the time...
Usage: python ac_outlet_control.py <type> <outlet> <state>
For example: python ac_oulet_control.py zap 1 1 will turn outlet 1 on.
- Raspberry pi
- AC outlets (These are much better than ByeBye ones I used previously)
- 433 MHz transmitter/receiver
- Diligent Analog discover
First step is capturing the waveform. I had an USB o-scope from college that I used however I've seen people use adacuity and their sound card. Either one will work.
I first hooked up the 433 Mhz receiver to my Diligent o-scope. Wiring it as you you assume, 5v to 5v, Ground to Ground, and I hooked the positive scope probe to the data pin and the negative scope probe to ground. I couldn't easily trigger on rising edge or falling edge due to the amount of noise, NYC isn't the greatest place to test signal sniffing, even tried putting it in the microwave, a faraday cage, to decrease noise. I ended up having to just hold the remote button and make the time range big enough to capture a few signal periods. This was the result:
From this I was able to figure out that the signal was made up of two signals: A Long high followed by a short low and a Short High followed by a long low. Using the o-scope software I was able to measure the wave peroid. After measuring the wave's peroid and identifying the pattern I created a int array, a binary representation of the wave form, where a 1 was the Long High follower by short low and an 0 was a Short high followed by a long low.
From that I was able to write a program that trasnlates the array into a series of High/Low voltages send to the transmitter that can turn on or off the AC outlets.
After looking at several different waveforms I was able to decode the signals:
SIGNAL EXPLAINATION:
[0,1,1,0,1,1,1,0,1,1,0,0,1,1,1,0,1,0,0,1] [1,0,0,1,0]
↑ ↑ ↑ ↑ ↑ ↑
A B C D E F
A - Base Signal
B - State bit. 1=on 0=off
C/D - Remote outlet identifier
E - Channel. The ac remote has two channels. 1-3 / 4-6
F - Parity bit.
Easey peasey! Now I can control everything, up to 6 outlets.
The problem now is that I wanted to use a python script to run this however Python doesn't runn fast enough to access μs delays. Thats where wiring pi came in. A great c library for low level gpio access for the raspberry pi that also has a Python wrapper. Once I installed that I created a the python script in about 30 minutes and I was good to go!
Wiringpi
git clone git://git.drogon.net/wiringPi
If you have already used the clone operation for the first time, then
cd wiringPi
git pull origin
Will fetch an updated version then you can re-run the build script below.
To build/install there is a new simplified script:
cd wiringPi
./build
Python dev tools
sudo apt-get install python-dev python-pip
Python wiringpi2 wrapper
sudo pip install wiringpi2