I have a power chair that I’m slowly converting to a robot and wanting to use the control system that is built into the unit, that said I’m trying to understand the programmer input that connects to the joystick module for programming the unit ( speed, braking , etc. )
The module is a P&G Pilot Plus and believe it or not it might as well be the latest nuclear invention of the time as far as protected info from any source.
The manufactures covet there technology with great secrecy to the general public.
The joystick has 3 pins to connect to for the serial input/output on the chair, 2 of those pins are directly connected to the battery 1 goes to the internal circuit board with small battery wires +/- also going to the circuit board , the plug is of a dual nature in that it is also the charge socket.
Not sure of your “serial” input from the joystick. Should be two analog inputs, “pots”. The pots need a voltage across them, and they give a rail-to-rail voltage change. Can you verify this? I wouldn’t think it would be serial, as you’ve suggested.
If you have a DMM, you should be able to “read” the voltage from the two pots to verify.
The solution to this type of interface is to feed it two analog signals (PWM and a cap filter, or a D/A), and simulate the pot voltage inputs.
If it really is “serial”, then you need to “snoop” the interface signals, and see if you can read any sort of “packet” data out of it. Could be SPI, or I2C I suppose, but I’d doubt it. What makes you think it’s serial?
The joystick that you are referring to is NOT analog , it is a proprietary serial single wire and after much experimenting i opted to buy a Sabre Tooth H-bridge and end the headaches.
There is a way to interface the physical joystick part by making a circuit to emulate the stick either -5 to +5 or -12 to +12 , the stick is a hall effect device and is hard to get the electric circuit to emulate it.
The circuit has to be perfectly regulated to be exactly at the center points just to get the chair to initialize.
You also have to remove the brakes from the motors , which is not hard.
After doing a pile of research on the same thing, the wheel chair joysticks are usual double halls effect per axis. Done that way for redundant safety interlock.
IMO fastest way to get the thing moving for the least $$ would be to trace the circuit backwards from the motor outputs and find the motor controller chips on the board, logic probe the pins or figure out what chip is it and find the data sheet, then just use a arduino to feed the motor controller ICs and still have the option to use another output on the arduino to unlock the brakes, usually power to the brake to disengage
