Hi Joe,
It was an interesting problem to solve given that X-plane 11 responds as though the autopilot is trimming when using the trim switch. Also the sim is loosely modelled on a Rockwell Shrike Commander which has a electric trim indicator, so I really don't care about the physical position of the wheel. If I do decide to include absolute position my preference would be to use a Hall sensor. I just hate pots! The motor is a carriage motor out of an old DEC LA210 printer that I happened to have lying around.
The Arduino code is written as a state machine that operates in manual, electric and autopilot modes. A PI(D) controller controls the speed and direction depending on the mode and trim change required.
Manual mode calls for a speed of 0 RPM with an Ki of 0 to simulate mechanical drag and the encoder pulses are sent back to AM to command the sim to pitch up or down. Electric mode uses the AM Game controller interface to read the yoke switch and send a state change to the Arduino, which in turn drives the motor and the encoder pulses command the sim. At present when the autopilot servos are engaged and the autopilot is trimming up or down, an appropriate state change is sent to the Arduino to run the motor but NOT send back the encoder pulses. When the autopilot has finished trimming it reverts to manual mode so that it can be overridden. If you hit the trim switch in autopilot mode the sim disengages the autopilot as expected and it reverts to manual mode.
I have just finished fiddling around with the autopilot constants and the Brunner FFB Yoke settings and it is working reasonably well. There is a bit too much gain in the PI(D) controller but it is not worth trying to tune those until I have the mass/inertia of the trim wheel and pulleys in the system. I pass those parameters to the Arduino via AM and messageport so I don't have to keep flashing the Arduino during tuning. Hence this thread.
Cheers
Rob.
