Nomad
Version One
Software
Problems
Too fast - while steering and throttle worked well the car itself didn't like moving at low speeds. Instead it was too fast for easy testing.
Short Battery Life - the main power pack was 7.2v and didn't contain much charge - therefore the car could only be tested for a short period of time before it was necessary to recharge for 30mins.
Direction - as GPS was being used to decide the direction that the vehicle was facing it was necessary for the car to move before this direction could be found. A digital compass on the i2c interface could be used to overcome this problem (however calibration might be an issue).
Messy - the construction was rushed resulting in a messy construction.
Version Two
Upgrade - Bought a r/c hovercraft from Maplin for £24.99 - 3 motors, one for lift and the other two for left and right. No servos instead it moves like a tank. Didn't take long before I'd taken it apart, the decoder chip is PT8A97BPE and after a bit of searching I managed to find the data sheet. As the left and right motors are either on or off there is no need for a PWM signal. Instead I identified the pins from the decoder chip to the individual motors and found that by applying 3.3v the motors turn (I guess it goes through some more circuitry). i therefore attached some wires to the correct pins on the decoder chip, also added a ground wire and wired this up to the I/O pads on the waysmall.
Todo
Check that flicking the state of the GPIO is sufficient to start and stop the motors.
Add GPS
Check range of radio modems and add them
Sort out power for all the components
Perhaps increase power supply for lift motor
Add i2c digital compass
Write simple waypoint following software.