Isnt a tiller pilot overkill (not to mention inefficient) for this setup?
I have connected remote
control servos, some $30 units can do 40kg/cm torque and draw only a small fraction or 0.3 amps of power being much higher efficiency. Then I have magnetoresistive sensors which can be mounted anywhere on the boat, or just use gps
feedback if moving fast enough or.. a wind vane
sensor I built as well. It looks like a mini vane except it just turns a potentiometer and you can set the wind
angle with it. The advantage is it can cancel oscillating movements rather than amplify as we all know in some sea states with a wind vane
when the boat rolls it makes the vane move and causes the boat to turn! It ends up correcting itself bbut the boat doesn't steer the optimal course.
Work is still in progress, but I did have some success in wellington harbor a few months ago. So basic conversion cost to drive a vane from compass
would be under $100 total and far better logic and feedback control than any tiller pilot. I have source code on github for automatically computing the PID loop gains.