While having a servo open to measure the pot output voltage, I went ahead and modified it for continious rotation. It was interesting to note that there is a wider pwm control band than I had expected. With the servo centered at 130 (mini ssc II servo controller), the variable pwm band seem to extend from ~117 to ~143 for the servo (standard type servo). This might make for some interesting setups like, making a tracked ROV powered by servos and steered with a joystick.
Wasn’t sure if you were asking, but the servos have an RPM of about 60 when modified for cont. rot.
What I was noticing was that I could get ~10 discrete speeds between 0 and max rpm. This might roughly equate to 0, 6, 12, 18, 24, 30, 36, 42, 48, 54, and 60 rpm. Might be useful as a poor man’s variable speed gear head motor. I was actually expecting a much tighter speed control range.
The speed response from a modified servo is not linear, it’s exponential. It will reach full speed when the pulses it receives are still relatively close to the 1.5mS stop value.
zoomkat was this a hitec servo you modified?
I was thinking about the speed thing you observed and if the control loop in the servo was just using the P part of PID (i.e. Position error) to determine servo power then it would make sense that you would have a small range of signal input that would correspond to different rotational speeds. If the control loop used the I (i.e. Integral) term however the power applied would grow grow fairly quickly as the accumulated error should quickly dominate the position error.
The servo I modified is a Tower Hobbies std TS-53. There is no error “windup”, so it appears to operate from the position error only.
Note: The tower hobbies TS-51 was a Hitec HS-300. This may be a hitec servo…