Reducing Drive Backlash

On Friday, April 26, 1996 at 3:00:00 AM UTC-4, Steve Scampini wrote:
Hello Everyone,
Does anyone know if the following is a *practical* (and in fact,
*practiced*) way of generating a low backlash drive:

Take two nominally identical stepper motors with gear drive reducers
(with the same reduction ratio). Hook the shafts together.
Run them in *opposite* directions "for awhile" until the backlash is
wound out of the combination. Stop this torture. Now that the backlash
is would out, step both motors in the *same* direction.

What do the mechanically inclined (or even horizontal) think of this
scheme? Potential gotchas?

TIA,

Steve Scampini

This method will make matters worse as the system will be grossly over constrained. There are many ways to build essentially zero backlash systems. They can then (if necessary) be driven in only one direction.

http://www.richardfisher.com

On Thursday, 14 January 2016 08:27:59 UTC-5, Helpful person wrote:
On Friday, April 26, 1996 at 3:00:00 AM UTC-4, Steve Scampini wrote:
Hello Everyone,
Does anyone know if the following is a *practical* (and in fact,
*practiced*) way of generating a low backlash drive:

Take two nominally identical stepper motors with gear drive reducers
(with the same reduction ratio). Hook the shafts together.
Run them in *opposite* directions "for awhile" until the backlash is
wound out of the combination. Stop this torture. Now that the backlash
is would out, step both motors in the *same* direction.

What do the mechanically inclined (or even horizontal) think of this
scheme? Potential gotchas?

TIA,

Steve Scampini

This method will make matters worse as the system will be grossly over constrained. There are many ways to build essentially zero backlash systems. They can then (if necessary) be driven in only one direction.

http://www.richardfisher.com

Celestron's original dual AC motors more or less did this.

On Thursday, January 14, 2016 at 5:22:51 PM UTC-5, RichA wrote:

Celestron's original dual AC motors more or less did this.

Doesn't make it a good method.

http://www.richardfisher.com