Hi all,
I have an interesting project idea on the boil. It's planned for go-live in mid 2015 so iv'e given myself enough lead time so do all the research and testing properly.
In a nutshell - I want to build 120 individually controllable wireless LED modules
The idea is that cast on a stage will be wearing these modules, each module is individually addressable(Probably by an ID which is written in along with the program). From the lighting desk(Sending DMX 512, probably via Artnet tot he transmitter module which will translate ArtNet to my own signal) we can dim the LED on any one of the modules to produce patterns etc on the stage.
The DMX,ArtNet side of things isn't really worrying me right now, the bit that i'm stuck on is the RF hardware. I've played with the nrf24l01's and thought that they would be perfect but from what I can tell they require some sort of 'bonding' process kinda like Bluetooth, so I cant just have a simple 'broadcast' system.. Appears there is a max of 6 per network?
I'm not worried to much about data integrity or the modules ACK'ing to the transmitter etc.
I had planned on sending about 10 updates per second to each of the 120 units, so by my thinking that's 3 bytes per module per update(Address byte,PWM level byte and maybe a checksum byte) times by 10 updates per second over 120 units = 3600Bytes per second
I haven't been able to find any reference info anywhere, lot of other projects do a many transmitters to one receiver type thing(For data loggin or security etc) but what i'm doing seems to be the reverse and therefor difficult to find any information on what others have done before.
The Jaycar modules are far too slow for what you need.
They are at best limited to 4800 bps and they need considerable overhead and specialised coding (usually manchester coding) to operate reliably.
OK for 1 to 1 links, but not 1 to 120.
I think (I haven't tested this module) you'll better choose a module like this one
It's a 2.4GHz module, which allow buad of 250k to 2M, which should be enough.
For address I would use DIP-switches, since then you can change the address, without having to modify any of the code (and who would want to upload 120 different programs just for the address that's changed..)
The LED dimming through PWM as you said and you're done, I'm not sure, but this may even be done with an ATtiny84 I think (11 pins available)
3 for shift register (piso) to read DIP-switch, 1 for LED pwm and I think 3 for SPI to the transceiver module, makes it perfect for ATtiny84.
Steen: that module is the one I've been working with and best I can tell is limited to a basic network of 6 devices, although i'm not certain that this is correct.. can anyone clarify?
mauried: you say the jaycar device is too slow as i'd only get 4800bps yet I need only 3600. Do you have nay examples of the code that would be required to drive these units? Why would a 1 to 1 vs a 1 to 120 be any different in my situation?
About the number of ANT network node (ANT is the name of the protocol implemented by nrf24l01 chip), 6 are the pipes. A pipe is as ort of subchannels id(s) the devices uses to receive/send messages in a user transparent way. The pipes share the same RF frequency. A pipe can be seen as a usb endpoint. If you need more than 6 devices, you can use the mesh capabilities of nrf24l01+ chip. Multiple devices can be arranged on tree architecture in a user transparent way. The addressing is based on a 40 bit identifier.
I have done some testing with these module, also for a DMX/Lighting related project.
I have found that the ones with the on board antenna have limited range of a few meters or so at 250kbps
What would probably be a better idea is to have a NRF24L01+ with the LNA amplifier and SMA antenna
You could probably get away with using the dead cheap "antenna on board" modules for the on-stage users
If you are working totally simplex, as in one TX multi RX then there isnt a limit to the number of RX as far as I know, you could set them up all on one channel but index them seperately on different bytes in different packets
I have written some very nice 'Auto-pairing' software which the transmitter does a scan for the quietest wireless channel to use, then tells all the receivers what channel to jump to for data reception - works a treat !!