Thanks Roland,
What I've found is that once the socket has reached that midpoint between flats 
it sticks there.  Rounding off the edges of the inside didn't fix it but did 
help.  I can even shut off the air so I can pull it down slowly against the 
return spring pressure and turn the spindle carefully by hand and I can still 
make it stick in that one area.

The problem with turning the spindle is that the socket spins freely in the 
wrench so even the slightest bit of friction once it touches the nut causes it 
to turn with the spindle.  I tried 1RPM all the way up to 30RPM and it didn't 
matter.  The rounding of the nut and socket means 50% of the time it works 
perfectly and the other 50% sticks.

Also if my set screw is too loose then on the way up the socket pulls out.  And 
the set screw loosens after about 10 cylces.  Hardened metal against hardened 
metal doesn't work well.  A softer screw didn't work much better.  Way to much 
rattle on the socket/driver connection.

I never thought I'd have this much trouble...
John


> -----Original Message-----
> From: Roland Jollivet [mailto:roland.jolli...@gmail.com]
> Sent: November-22-21 4:01 AM
> To: Enhanced Machine Controller (EMC)
> Subject: [Emc-users] Spindle positioning.
> 
> You probably don't need more than 1 bar on the lifting air cylinder, up or
> down. No need to force it down.
> So let it down, then if you can give the spindle a 1RPM burst to let the
> nut slip over.
> And have a switch on driver Z to know if it's down over the nut.
> 
> You could also stop the spindle and descend the driver as the spindle
> reaches it's last 5RPM? before stopping.
> 
> Roland
> 
> 
> 
> On Mon, 22 Nov 2021 at 10:03, John Dammeyer <jo...@autoartisans.com> wrote:
> 
> >
> > > From: dave engvall [mailto:dengv...@charter.net]
> > > On 11/21/21 9:24 PM, John Dammeyer wrote:
> > > > Is there a way to tell the spindle to turn until it finds the index
> > and stop so it's always stopping at the exact same spot?  Like
> > > decelerate to 0.5 rev per second  (30 RPM) or slower and stop when the
> > index happens?  With either an M5 or the button on the
> > > user interface?
> > > >
> > > If I were? doing that? today I would use 3 sensors. IOW a guard sensor
> > > either side of index, rotate quickly to detection of guard slow way down
> > > and search for the central sensor. How you know which is the shortest
> > > path is left as an exercise for the "student". ;-) Or maybe one simply
> > > defaults to searching in one direction.
> > >
> > > If you don't need orientation then a stepper motor and a gear would seem
> > > to be as simplistic as one can get.
> > >
> > > Dave
> >
> > One still has to line up the socket to the hex head of the drawbar.  And
> > since it can be at any position relative to the spindle the sensing has to
> > happen against the hex head.  Rotate spindle until hex head is at a known
> > position.  Then rotate socket from and indexed position until it matches
> > the socket and then send it down.  This could all happen in under 100mS and
> > be almost transparent to an end user.
> >
> > There are times where I wish I had BMT-30 cone type tooling or just only
> > TTS.
> > John
> >
> >
> >
> >
> > _______________________________________________
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users
> >
> 
> _______________________________________________
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users



_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to