Re: [CAD_CAM_EDM_DRO] as simple as 1,2,3...

Bill,

Yes, in a way. I was saying earlier in response to your post that it is
a simple enough thing to get encoder bits into a PC. Once in the PC
(without an "encoder chip"), you watch the two lines and "step" through
a software state machine. 4 states, every state '4' you identify as a
cw or ccw pulse (assuming std. quadrature, not 4X).

Trouble is, you have to be on top of the bits, watching the bits for
changes, which signify a change to the next state.

I don't recall, were you wanting to write the threading code? Did you
mention BASIC or 'C'?

How slow do you run the spindle? And how many steps/rev. on the
encoder? A US Digital encoder chip could take some of the load off the
PC (I've been meaning to get some). In fact, Camtronics has the DRO
boards, which would give you a "look" at how hard it would be to "keep
up".

This approach would be making the "carriage axis" of the lathe track the
spindle? 200 steps to rev. the carriage axis, would equal one rev of
spindle (single pitch thread, same as carriage thread)? And you
therefore need a "software gearbox", to simulate the back gears used for
threading.

Just a thought, what if you put a 200 step (line) encoder on the
spindle, and used that encoder to step circuit in FILES (Jon's?), and
drove your carriage axis (assuming 200 steps/rev)? You could then run
it at threading speed to judge if it could keep up? THIS WOULD ONLY
ALLOW you to "match" the carriage lead screw thread, not too useful, but
would be a minor proof. Actually, you'd probably want 1024 or 2048 etc.
steps, so there would be some good even divisions possible. And the BIG
question, could your PC keep up with 2048 steps/rev of the spindle?

Simplified, the PC needs to keep up with maybe 2048 steps at your
threading spindle RPM.

Hope this isn't too confusing the way I explained/examined it!

I might be persuaded to put program in the FILES to read the JOYSTICK
port and spin a number on the screen. Bear (bare?) in mind that it takes
time to write to screen as well! Jack might want to try, if he hasn't
already determined the steps (lines) of his encoder.

ACTUALLY, it's simple enough for anyone to do in BASIC!

Alan (rambling mode off)


BillDarby wrote:

>
> This sounds like the solution to the question that was put earlier. IE: Is there any simple way to feed the output of an encoder
> into your computer? Is it an answer for threading?
>
> Bill Darby
>
> Alan Marconett KM6VV wrote:
>
> > Jack,
> >
> > You could:
> >
> > 1. Connect A & B phases to Joystick's sw 1 and 2.
> >
> > 2. Use a short program to read sw bits.
> >
> > 3. Write program to Count the "steps" (walk through all 4 state
> > transitions of A & B).
> >
> > 4. Run program, start at known shaft position, turn slowly until 1
> > complete revolution. Or do N revolutions, and divide result by N.
> >
> > 5. (extra credit) Connect the index to 2nd joystick switch 1, and an
> > edge of it's signal to "gate" one full (or N full) revolutions of the
> > shaft.
> >
> > Alan (just did that)
> >
> > jmw@... wrote:
> > >
> > > Is there a way to determine encoder line count for those among us who
> > > happen to be temporarily down on their luck and find themselves scope-
> > > less? The manufacturer disavows all knowledge. I assume one spins the
> > > encoder at a known rate, but then what? This unit has A, B, Index,
> > > ~A, ~B, etc pinout.
> > >
> > > Maybe the thing could be connected to a PC somehow?
> > >
> > > Thanks..
> > >
> > > --Jack