RE: more R,P,Y axis conventions (rotary)

Hi Ballendo,

A little <snip> and more below. :>)

ballendo@... wrote:

>
> Alan,
>
> (snips and more below)
>
>
> Step/Dir S/W usually has a "bit flip" available to the jog key(s) > Dir.
> BTW, the "shift" keys (as used by MaxNC) could be a good choice
> for "additional linear axis" COMMON keys. This would give us 6 axis
> jog capability from a std. keyboard (cursors,PgUp/Dn,2 sets of +/-
> ,and shift keys)
>

Yes, for the step/dir version, I'll need bit flips to replace the phase
data. I used the "shift" keys for a while (easy to "see" by BIOS call)
for a while, but for other then the X axis, I had (have) trouble
remembering which way something moves! Grabbing the keyboard interrupt
and eating my COMMON keys replaces that trick.

I replaced my incremental move keys '-' & '+' with '<' & '>' (like G91,
except we don't get g-code yet). This frees up the '+' & '-' keys for
the 4th axis. I set up an increment size, in inches or degrees (see
"NEW" below), and then move that amount when the '<' or '>' keys are
pressed. A sixth you say?

I have a "main" where the user can select an axis, move incrementally or
absolutely, (the "current axis") or enter jog mode (+ other commands).

I enter "jog" mode from a command menu (well, just a 'J' key for now),
to prevent the operator from accidental moves. In jog, the COMMON keys
move the four (4) axis, or the panel encoder moves the "current" axis.
Slight paradigm shift. In main it's "current axis". in jog, it's "press
the key", and the encoder (jog mode 2?) operates on the "current axis"
also.

> >(I've been testing B) to indicate (the little pointer on the
> >circumference) the right direction on the table!
>
> B rotates around Y. So the plane is XZ (involved axes). Positive Y is
> BEHIND the sherline. So you look at the rotation from behind the mill!
> And maybe "THROUGH" the rotary table. (what?! Yours isn't selectively
> invisible?!<g>) I paid extra on all my machines for "selective
> invisibility". Pretty expensive option, but it sure helps setups!
>

Oh well, gonna SKIP that option!

>
> >Both, huh? Well, if I command 2.00" on Y, I get + movement (tool to
> >rear of the part, yes). But currently, if I command 200.0 deg on
> >the B, the table rotates ccw to 160 degrees (according to the
> >indicator).
> >Which I don't think I want! But then the "tool" went to 200 deg!?
> >OUCH! Which do I want?
>
> Are you in absolute (G90) or incremental (G91)?
>

From "main" (as above) an 'E' command (or menu selection?) allows you
to type in a new absolute (like a G90) location. I'm trying something I
think is NEW here, If "current axis" is XYZ, then I want inches. if
ABC, then I want degrees. neat?!

> >Would I really use G02 and G03 on the table? It would be nice to be
> >able to get it to "rotate a cut".
>
> ON the rotary table or WITH the rotary table?
> (XY arcs with B vs. XB "arcs". curves, really.)

Maybe I'm thinking of XB? Ya know, to cut that crankshaft plate?

>
> Andrew Werby can prob'ly show you what happens when you combine two
> rotating motions. Try(from memory): http://www.computersculpture.com
>
> And certain types of Cam could make use of this type of move. Or
> lettering a surface?

I can see the lettering as XY (where Y addresses the table). Probably
save multiple concatenated rotary tables for later!

>
> >Configurations are a GOOD thing, I just don't want to have to change
> >the phase outputs or feed rate calibrations.
>
> Multiple configs' can also allow "scaling" by setting FALSE step
> sizes. And allow Ellipses by setting unequal step sizes! On controls
> which don't "support" either of these! (Kevin Carrol clued me in to
> this a few years back. He wrote the original code for Stepster.)

I'll keep this in mind, but probably not for this jog program. I DO
want to do a FULL coordinated move g-code control program. That was
started about 18 months ago (for CY-545's), from the other end!

> Hope this helps.
>
> Ballendo

I think things are coming along!

tnx,

Alan KM6VV