Re: Auto tool changers

Alan,

The "model" is a homing routine. Of the type used by T-tech. No
switches. Run the stepper "enough" to be "sure" you are TO/past the
next tool position, and then reverse and STALL the stepper against
the detent, as Bill described. K.I.S.S. Then stop the pulse train,
and do not idle the motor. (see the P.s. for a "better" idea, IMO.)

Or have two switch circuits (this uses two input bits). One is a home
switch, but it's really just a "flag" that we've just passed slot 1
(of the three). You know this because the "other" switch circuit has
three activated positions, corresponding to the tool positions.(I'm
assuming that the "six" tools come from THREE turret positions, by
utilising end or side approach) When BOTH switches activate you are
at the "base" tool position. BTW, these switches don't need to be any
kind of accurate. You just want a way to find and verify "starting"
slot. Then you want a "blip", as you pass each slot.

With this you can easily implement "shortest path" tool selection. If
we're sitting at slot 1 (we just initialised, which means we drove
until we got the AND of the 2 switches, then reversed and are now
sitting against the pawl at slot 1). If tool 3 is called we go one
way. If tool 2 is called we go the other. The "blips" from the
switches keep us "in the know". One direction the blip comes BEFORE
thepawl ,the other it comes aFTER the pawl. And there's no electrical
reason not to expand the idea to 6 slots (if the mechanicals can be
made to fit)

BTW, since this is a lathe, be sure the action of the tool turns the
turret against the pawl. This makes the holding force less important,
as the pawl will take the cutting load.

Hope this helps.

Ballendo

P.S. Thinking about this a little further, we only need ONE switch
circuit. Just make the "home"/slot 1 flag or detent BIGGER than the
others. This way, slot one is indicated when the bit blip stays active
"longer" than at the other "index" blips. Simple software to
implement this, IMO.

P.P.S. I'd consider a double acting air cylinder to provide the
motive force. Fast. Very fast. And unaffected by dirt/debris;
compared to the worm. small size means not much air/co2 to operate.
(that Co2 reference comes from the turbocarver dental burr
investigations :-)

--- In CAD_CAM_EDM_DRO@y..., Alan Marconett KM6VV <KM6VV@a...> wrote:
> Hi Drew,
>
> OK, I saw the slots, and I now see three holes on the
circumference, Are
> these the ones? Yes, three photo interrupters could give a direct
> readout of the positions. the "edge" encoder I mentioned was to be
used
> as an index. The PIC can "find" the index on startup, and keep
track
> of the 6 positions. I understand that ENCO already has this type of
> encoder setup on the spindle (I think I've got one coming!).
>
> Bill tells me that the turret "drops back" into the detent, so I'm
> guessing that a "ratchet" effect is in play here, and one would only
> have to insure enough steps to get a little past the next detent
> position, and then allow the mechanism to drop back.
>
> I suspect that the original DC motor was just shut off, and the
detent
> took over, therefore, the stepper implementation should do the same.
>
> Instead of a PIC, a Cybernetic Microsystems CY545 stepper controller
> could be "programmed" to the the job. A little more costly then a
PIC,
> 'tho. (I've just been talking to them again).
>
> Interesting project! I wish there was a (simple) way to "model" the
> mechanism, so that the driver/code could be worked on. The worm
gear
> data would have to be known, and other details of the mechanism, or
the
> actual hardware.
>
> Alan KM6VV
>
>
> Drew Rogge wrote:
> >
> > The six positions consist of three slots and three holes.
> >
> > It might be more interesting to do somekind of BCD encoding
> > or something which would allow the PIC to know what position
> > the turret is currently in. That way you wouldn't have to
> > always start the turret in the same position when the PIC
> > powers up. Or I guess you could have somekind of a home switch
> > the PIC could read.
> >
> > I'm curious how one would prevent the stepper from continously
> > pushing back on the "stop" too hard and/or for too long.
> >
> > Drew
> >
> > Alan Marconett KM6VV wrote:
> > >
> > > Hi Bill,
> > >
> > > Perhaps if a simple encoder disk is added to the turret. 6
slots (one
> > > to the edge), would allow two photo interrupter sw's to report
the
> > > position of the turret. A PIC could then drive a stepper (even
a simple
> > > unipolar driver), and do the "rotate past, drop back to lock"
sequence
> > > that you've said is required to change the tools. I have a
Sears/EMCO
> > > 5" lathe (no CNC... yet), would this fit on my lathe? How
much? I'd be
> > > interested in a swap of some kind to get a mechanism to work
with (kinda
> > > putting the cart before the CNC horse!).
> > >
> > > It only looks like three tool positions in the pix?
> > >
> > > Alan KM6VV
> > >
> > > wanliker@a... wrote:
> > > >
> > > > In a message dated 2/21/2002 11:09:26 PM Mountain Standard
Time,
> > > > johnhe-uk@s... writes:
> > > >
> > > > > I'm thinking of three of four
> > > > > tools. I've looked but not found any prices on the internet
> > > >
> > > > If you go to the following URL, there are pictures of the
guts of an EMCO
> > > > COMPACT 5 CNC/PC TURRET.
> > > >
> > > > http://www.grifftek.com/c5cnc/c5cnc.html
> > > >
> >
> > --
> > Drew Rogge
> > drew@p...