Servo Motor Deadband Width

Hi all,

I have found a servo motor with a deadband width of 1µs and a pulse width range of 1400µs. From this, I calculated the pulse width per degree to be 7.73µs (180 degrees rotation). The deadband is +/-0.5µs, right?

So, the accuracy (theoretically) is 0.13 degrees? Also, the error of the motor is 6%? I understand it may be higher and it is unlikely to be that accurate in practice.

I would really just like some clarification as to what I am doing is right so any guidance or help would be greatly appreciated.

Kind regards,

Marc

The first thing to note is that most servos don't move exactly through 180 degrees - most have significantly less movement and some may have a bit more.

Also the resolution and repeatability of the servo depends primarily on the quality of the feedback potentiometer within it.

The best way to determine the characteristics of a particular servo is by writing a short program and measuring the effect.

...R

I had a project where I was concerned about the accuracy of the servos I was using.
I was making a hexapod with 18 standard servos.
As Robin2 mentioned, I measured the servos.
I purchased 20+ servos and mounted each one in a jig with a protractor and measured the servo response at multiple points across its range.
There was some variance among the individual servos.
I considered building a matrix in code to compensate for that variance but then chose to ignore it. The deviation was small enough that ti did not realistically affect my project.

Now you have piqued my curiosity. What is your project?

As best as I remember the typical hobby servos I've tested had a movement dead band of ~5µs. Below is a simple setup I used with a bamboo skewer attached to the servo horn and the skewer tip at a line on a piece of paper so I could see each servo increment movement. I made some code that made it easy to send 1µs command increments to the servo and detect when the servo actually moved.

resolution1.jpg

resolution1.jpg

Hi all,

I would like a servo/stepper motor to turn a valve on/off and I am unsure which motor to use. I have calculated the theoretical accuracy of a servo motor and if correct, is 0.13 degrees with an error of 6%.

This is the servo motor:

I calculated the following:
Deadband width = 1µs
Pulse range = 1400µs
Pulse width per degree = 7.73µs/degree
Accuracy = 0.5/7.73 = +/- 0.064 degrees
Error % = +/-6.4%
Is the accuracy of the final angle? E.g. 90 degrees will realistically be 85-95 degrees (approx)?

Is this correct? I understand in practical terms it will not meet this. Can this be improved upon using a stepper motor?

Any help is greatly appreciated,

Marc

This seems to me identical to the question in another Thread you started and in which you got Replies but never responded. Double Posting is against the rules of the Forum as it wastes everyone's time.

Not only that, but I strongly suspect you are still in search of a solution to this even older question - that makes for Treble Posting.

Please click Report to Moderator and ask to have them all merged into one Thread so we have all the info in one place.

...R

"Can this be improved upon using a stepper motor?"

Maybe, I think stepper motors are used in some 1200 dpi printers.

On or Off- how can a couple of degrees make any difference?
In general, Stepper motors are more precise which is why you have them in printers and DVD drives.

Hi,
How many degrees do you need to turn the valve, OPEN to Close?

If you use a stepper, you will need to use limit switches to indicate fully OPEN and fully CLOSED and a feedback device to tell you the position of the valve.

What is your OVERALL application, you may be OVER THINKING the issues.

Tom... :slight_smile:

TomGeorge:
If you use a stepper, you will need to use limit switches to indicate fully OPEN and fully CLOSED and a feedback device to tell you the position of the valve.

In case of a strong enough valve that blocks when fully open or fully closed, and a not overly strong stepper: turn the stepper in the close direction for at least as many steps as it takes from fully open to fully closed. Then you're guaranteed to have fully closed the valve. That's your reference position.

Of course also works with the valve fully open as reference, which one is best depends on the overall situation. Maybe OP will come back this time to clarify?

TomGeorge:
How many degrees do you need to turn the valve, OPEN to Close?

This stuff has all been covered in the OP's other Threads.

...R

"Is the accuracy of the final angle? E.g. 90 degrees will realistically be 85-95 degrees (approx)?"

My tinkering with typical standard size hobby servos indicates they are capable of maybe .5 deg. repeatable positioning.

@mh95ncl

TOPIC MERGED.

Could you take a few moments to Learn How To Use The Forum.
Other general help and troubleshooting advice can be found here.
It will help you get the best out of the forum.

Robin2:
This stuff has all been covered in the OP's other Threads.

...R

Not my problem. literally.
If OP starts a new thread, then lets see the facts that have been established in the first post of the new thread.
Citing previous threads is fine, but if the OP wants advice lets state ALL the facts first in the new thread.
Tom... :o :o :o

TomGeorge:
Not my problem. literally.

Unlike me, you must not object to asking (or answering) questions that have already been dealt with.

...R