I am building a system that will control aiming of a directional antenna. I need to be able to control the direction and scan rate of the antenna. I have had much success in doing this in the past with a geared stepper motor, and basically using the counted step approach with a sensor for a home location. In an effort to make the system smaller, more agile, and per-manufatured parts. I was looking at hobby style servo motors and using a mounting system such as http://www.servocity.com/html/spg5485a-45_360o_rotation.html. I have limited experience with servos, and have never been able to successfully control there speed of rotation. Is there a way to easily control a servo's rotation speed?
I also welcome your input on Stepper vs Servo? The Requirements: Smoother Precise Scanning (.25 deg resolution or better), Variable Rate of Scan from 1min to 2 hour for full rotation, Rotating 3lbs of gear and cable.
Is there a way to easily control a servo's rotation speed?
It's much like controlling the speed of a stepper. If you want to move a servo from say the 0 degree position to a 90 degree position, you get maximum speed by just writing on servo write command to 90 degrees and the speed will be the maximum the servo can mechanically do. However to go slower you just make smaller servo write commands and pace the time between increments the position commands until the final 90 degree position is reached. So to summarize you can with software, slow a servo down, but it's maximum speed is limited by the servo motor/mechanics. I think the same can be said of stepper motor controlling, no?
The Requirements: Smoother Precise Scanning (.25 deg resolution or better)
In my tinkering a standard servo is limited to about .4 deg resolution.
I don't know if there is such a thing as a 'standard' servo. Specific models have different resolution capabilities, the better servos will list such in their datasheets, the majority will be silent on the subject. One thing for sure to get the best resolution possible of a given servo never use the servo.write(degrees) command but rather use the servo.writeMicroseconds() command. That give you 1000 steps of resolution (1msec to 2 msec) from a software perspective, but what a specific servo can utilize from that is a function of the quality of the specific servo used, there are some very good ones if chosen for that property.
From what I've seen most standard (aka inexpensive servos) have a built in ~5us deadband to reduce hunting. This can be changed in some digital servos. A long time back I made the below setup using a bamboo skewer as a long pointer so I could see the small servo movements as the servo control signal was changed in 1us increments. Someone should be able to easily duplicate this to check the resolution and position repeatability of their particular servos.
zoomkat:
From what I've seen most standard (aka inexpensive servos) have a built in ~5us deadband to reduce hunting. This can be changed in some digital servos. A long time back I made the below setup using a bamboo skewer as a long pointer so I could see the small servo movements as the servo control signal was changed in 1us increments. Someone should be able to easily duplicate this to check the resolution and position repeatability of their particular servos.
Back in my R/C flying days the quality of a servo's centering ability was a big selling point. If say you have a servo being commanded to say at 90 degrees and mark it's exact position with a pointer attached, and then command it to say 180 degrees then 0 degrees and finally back to 90 degrees, how close is it from the reference mark you made earlier. In the instrumentation world we would call this 'repeatability' but the R/C guys used 'centering' as usually servos were set up where the servos were in their mid travel range when flying straight and level, except the throttle servo of course.
