While I agree this is a 'Bad Idea(tm)' - there are ways that are safer than others. And I sorta agree that a wireless throttle for a longboard would be more rad (and one could argue safer) than a long, wired controller of 'one size fits all' length and prone to folding. Annnnnyway...
Use a complicated radio protocol.
That may seem counter intuitive, but the ISM bands which these things will be using are really noisy (especially around electromagnets
) and you'll want a failsafe that's rather specific (and encrypted if used in 'groups' (a gaggle?) or sold at scale). This takes up a bit more computation to perform but are done in hardware 'for free' in most chips like TI's chipcon, Digikeys Xbee, and the ilk...and you can even use good ole TCP/IP with a microcontroller like the MSP430 w/ a chipcon radio built in for a less than the cost of a hamburger.
When 'over complicating' it with a MSP430 you can also do things like smooth the throttle signal to offset noise, a crappy quality potentiometer, or smooth a signal from an 'on/off' button into something usable as a 'throttle', and even prevent wheelies (or hold them). It's just some extra code to add 'later', once you have a working throttle.
Additionally, you'll want more features eventually (haptic feedback, battery level, etc), other things to talk to (garage door opener, other cyclists, etc) and ways to prioritize packets. Things you won't be able to do as easily with a basic 'point to point, serial replacement' protocol.
If you really want to do this, here are some things you should buy:
- a usb->serial (FTDI cable) if you don't have one already.
- A pair of 'Xbee S2 radios' or you can get two msp430s+radio from TI's website for free.
- A 'throttle input' whether that be a thumb throttle you stole from something or a bare bones switch or potentiometer.
- the connectors between devices (usually a 2.54mm or .1" spaced pin on the microcontroller side).
high level steps of how to build it:
- establish a basic 'hello world' example that blinks lights wirelessly (or wired, if you still want that).
- Next add a few lines to make it blink at different intervals.
- Once that works, read up on PWM. Sleep. Read it again.
- Dump some code in to handle the PWM math.
- Make it spin a motor the ESC's PWM input on your bench.
- Celebrate!
- Think about and implement some safety failsafes and limits for a 'test run'.
- Test on the bike/board. Don't die.
- Celebrate! Go faster.
- Iterate on that idea over and over.
Should you go this route, I'll happily offer any help depending on what step you are at.
Edit - Also you might wish to consider Bluetooth LE - as integration into phones is a bit less hassle than rooting and using a harness on your Android phone. Again, this is more of a consideration if the end goal is mass sales.
