Re: Belt tensioners on toothed belts

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "ballendo" <ballendo@...>
wrote:

>
> Tom,
>
> Ron's reply reminds me of something I've seen in older torch

drives...

>
> The position feedback is de-coupled from the drive. IOW, applied

to

> your description: The timing belts are NOT being used for drive,

but

> only to determine position.
>
> I've seen quite a few that had a "hidden" (means non-obvious; at
> first glance it looks like th emotor is driving the rack))

friction

> drive on the long beam, and usually rack gear and resolver/encoder
> for feedback. Could your belts be feedback only? (IMO that makes

your

> prior description make more sense, as then the tensioning spring

is

> only there to keep the belt tight enough to the feedback
> resolver/encoder, and flex wouldn't matter...)
>
> Just some thoughts, I'm curious now<g>
>
> Ballendo
>
>

No, nothing that sophisticated. The encoders are on the motors.
Toothed pully on motor is belt connected to gantry drive shaft (rack
on both sides) with another toothed pully. The idler arm pivot
point is closer over the motor than the drive shaft pully but not
directly over it. The arm is held against the belt using a radial
force type coil spring. Very little spring tension is applied ,at
least not enough to counter the different forces, so the belt flexes
a lot. since the pivot point is not centered over the midpoint of
the belt span it flexes different distances based on the direction
of travel. In this particular case the bolt holding the pivot arm
had worked loose and it was actually binding as the belt moved
towards the motor. The motors were strong enough to overcome the
drag but the overload fianlly took out the axis servo drive
(Rutex). I won't go into the electronics design other than to say
there was only one large fuse for all of the drives.

I think a rigid idler would probably be a better appoach ( and
probably be fewer moving parts).