Re: [CAD_CAM_EDM_DRO] Limit and Homing switches

Hi Alan,

Comments are scattered in your message.

Alan Marconett KM6VV wrote:

>
> Hi Drew,
>
> You could be right, but as long as you just need the index to happen,
> and not STAY valid, then it should be OK, IMO.

If the index isn't latched after the step is issued then I'm afraid that
the controller may not see it.

> I'm not totally clear on
> this, but essentially you can't "use" the uSteps as INCREASED
> resolution, just "S M O O T H E R" motion.
> Are you sure you need the index to stay on (not just the event)? And
> also, how many encoder "steps" do you have?

The encoders I'm using are 512 counts per rev. So that's 2048 in
quadrature.

> That has to figure in.
> Isn't that the STEPS you command with? So I'm thinking you command 5X
> (encoder) steps, and it moves. So you can stop on one encoder step,
> right? Or am I all wet? The microstepping should be "inside" these
> steps? No? Wouldn't you get an encoder step that is 5 microsteps
> wide? I may not have this right, so we need to get clarification on
> this!

The G3x0's aren't what I would call a micro stepping drive. In the G320
and
G340 in 1X mode the servo turns one quadrature step for each pulse on
the
"Step" input. So in my case that would be 1/2048 revs per step. If I
instead
jumper the G340 to say 5X mode then the servo turns 5 quadrature steps
for
each pulse on the step input from the controller. This would be 5/2048
revs
per step or about 410 steps per rev. If as Mariss mentioned in another
message,
the index pulse is one line wide and there are, in my case, 512 lines
per rev
then it's possible that none of my 410 steps will stop at a point where
the index pulse is asserted. In 10X mode it takes only ~205 controller
steps
per rev and I'm less likely to see the index. I guess if the controller
is
polling for the index often enough then it may not miss the index. But
if for some reason it's not looking when the index goes by then I'm
SOL.

>
> I think there is more to a G340 then I understand. Yeah, I KNOW there
> is! That's why I want to try servos... although I have LOVED playing
> with stepper motors sense first seeing them in about '77! They spin up
> pretty fast you say? Humm...
>
> Alan KM6VV
>
> Drew Rogge wrote:
> >
> > I'll put the scope on the thing and see if it wiggles. The problem that
> > I foresee with the X5 is that for every step the software issues, the
> > motor is going to turn 5 quadrature counts worth of rotation. If the
> > index pulse only occurs every 4 counts of rotation then it's possible
> > that index pulse could happen in the middle of those 5 counts and not
> > be asserted when the step is done. Things would probably be worse if I
> > was doint X10. I guess I'll just have to try it and see what happens.
> >
> > I think you'll find the servo stuff a lot of fun. It's neat to see them
> > spin up pretty fast.
> >
> > Drew
> >
>

--
Drew Rogge
drew@...