Clarification on encoder index pulse for Rigid Tapping

Ok I have setup a mill with a servo for the spindle and I was using the servo built in encoder for the ZAB pulses fed into a Mesa 7i96 for tapping with a one to one belt to the spindle. This has been fine.

Now I want to drive the spindle with a reduction of about 4:1 this of course will not work using the servo encoder.

Can I still use the AB from the servo and put a sensor on the spindle with one pulse per rev.

There is a disc on the spindle already with 8 slots. I am thinking I cannot use that unless I make it a 1 slot disc.

Edit: If I am not mistaken the servo encoder might have a lag in the AB with regard to the Z from the spindle disc

That is how mine is setup and it works fine. I used the AB from the servo and installed a plate with one slot on the spindle for the Z.

See here

JohnnyCNC wrote:

That is how mine is setup and it works fine. I used the AB from the servo and installed a plate with one slot on the spindle for the Z.

See here

Yes. I have done it that way on several machines but always had the ZAB from the same slotted disc but was not sure about the servo not being in direct sync with the Z sensor.

Edit: I put the same answer to the linked post as well (by mistake)

I had the same concern about the sync between AB & Z but it seems not to matter.

If you have a relatively high resolution encoder, a separate
non-synchronized index pulse should work fine as long as it
has a consistent leading edge (the index logic will be triggered
on the edge of the index pulse) There may be a count or so of
uncertainty but if the encoder resolution is high enough, this
should not matter.

Another issue with a wide index pulse is that it will occur in differing
places depending on the direction. This does not affect LinuxCNC
tapping or threading however since synchronization to index
is always done in the same direction.

Thanks Peter. That makes perfect sense. My encoder is 2048 and my Z slot is rather wide.

PCW wrote:

If you have a relatively high resolution encoder, a separate
non-synchronized index pulse should work fine as long as it
has a consistent leading edge (the index logic will be triggered
on the edge of the index pulse) There may be a count or so of
uncertainty but if the encoder resolution is high enough, this
should not matter.

Another issue with a wide index pulse is that it will occur in differing
places depending on the direction. This does not affect LinuxCNC
tapping or threading however since synchronization to index
is always done in the same direction.

Peter Thanks for the feedback
I am wanting to use the AB pulse from the encoder in the servo motor for the spindle and wanting to using a simple narrow slotted disc on the actual spindle as the motor and spindle are not 1:1

So would a narrow slot say 5mm hole in disc be fine or would a longer slot be better?

I think most anything would work as long as its wide enough to let the
full light beam through. Because Index is edge triggered, a wide slot is not harmful.