Servo Pulse Duration Question

[WBegg]

While fiddling with a new damper today, I noticed that the servo (MG90S) only moves every 10% increase when % speed is changed in 5% intervals. I can understand this, and write it off to the resolution of the feedback circuit of the servo (tried 2 different servos).

So for "gits and shiggles", I hooked the output of the HM servo circuit to my oscilloscope, and noticed something a bit strange. While the output was quite linear, there is a "hump" from 10% to 0% as the graph below shows. I had the limits set from 1000 to 2000, and it hit 2000 uS at 5%, then dropped back down to 1950 at 0% (about in line with the 10% output).

Any reason for this? Just curious. The Damper doesn't move when going from 0 to 5% anyway.

 

[Bryan Mayland]

I noticed that the servo (MG90S) only moves every 10% increase when % speed is changed in 5% intervals.

This is by design. What you're seeing is the "servo write holdoff" which is HeaterMeter waiting up to 10 seconds to move the servo if the move is small. This prevents it from moving back and forth between two durations that are so close that the servo just twitches and keeps the motor active. It will move, you just have to wait.

I would check your data again on the 0% output being 1950 and 5% being 2000. Is it possible you're seeing the holdoff in effect and the output isn't what you're expecting it to be yet? It would take special code to make something like what you're seeing happen and it is just (max-min)*percent + min. I might not have noticed in testing if this was occurring but it seemed to look right every time I've tested it. There can also be a large amount of jitter in the PWM signal which might taint the results.

I'll put it on my list of things to check next time I have the whole thing hooked up to the scope though, just in case.

 

[WBegg]

Thanks Brian.

Yeah I noticed the "servo write holdoff" when doing the test, as it took some time to read the triggered pulse on the oscilloscope.

I'll run the test a few more times with this in mind, to check out the variance I was getting before with the 0-5% readings.

Like I said, the servo moves as prescribed, no issue with the operation, just an academic curiosity.

Cheers,
W