Re: [CAD_CAM_EDM_DRO] servo runaway prevention

ccs@... wrote:

> > Encoder cables falling off is a really bad thing.
>
> I guess you are right that a following error would only be generated
> if the machine were commanding a move at the time the encoder failed.
>
> It strikes me that some sort of backup 'sanity check' in the amp
> itself would be good. This might take the form of a second low-res
> encoder. If pulses are seen when command velocity is zero, error out
> and shut down. And amp with a tach might have something to check
> consistency between the two - if one signals motion and the other does
> not, something bad is happening.

In the Allen-Bradley 7320 CNC control that I started with, the tach
signal was sent back to the encoder card. This card compared the
tach signal with the encoder signal. If the velocity of the tach signal
was ever 10% greater than a velocity derived from the encoder counts
coming in, it would cause an E-stop. In the days of encoders with
light bulbs in them, I suspect this could happen every month or 2.

Jon