Re: [CAD_CAM_EDM_DRO] Re: Closed-loop steppers, was: breakout boards plus - hard choice

scyvt wrote:

> Jon
>
> Isn't it the general practice for encoders to be mounted directly to the servos?
> What was your reason for leadscrew mounting?
>

This eliminates any cyclical errors from the timing belt and the
sprockets. If the sprockets are bored off-center, for instance,
then the position would have a sinusoidal error that repeats
every motor or leadscrew rotation. The other reason is the
drive assemblies I adapted to the Bridgeport were already made
that way. See the 3rd picture here :
http://jelinux.pico-systems.com/CNCconv.html
for a look at these drive assemblies. The red thing is the
Sumtak encoders I installed in place of the resolvers it was
originally made with.

> In the case of the Pico universal stepper controller, does it matter where in the drive chain the
> encoder is placed?

As long as there is minimal backlash, no. The belts are kept
VERY tight with a roller on an eccentric, so there's no backlash
there. If the leadscrew is low-backlash, then a linear encoder,
hopefully with high resolution, should be fine.

This machine uses the PPMC, which is a related controller for
analog velocity servos.

Jon