Re: home and limit switch in series

["Darrell Gehlsen" <darrell-at-machinemaster.com>]

Maybe I am missing something, but if all of the switches are in series for
NC switches and connected to one input. When homing, do it one axis at a
time always looking at the same input. You could home as many axes as you
can control as long as you back out of the switch before calling it home.
If you some how end up with a switch tripped and powered down, to get out of
Estop, you would first Override Limits and jog off the switch. Once that is
done, home as usual.
Concerning limit switches, the program knows where it is and has "Soft"
limits built in so if a switch gets tripped the program should be able to
tell which axis tripped the switch. If a switch trip is detected when no
axis is near a limit then that must be an Estop.
That makes 2 switches for each axis and what ever you think you need for
Estop switches all on one input.
Did I miss anything?
Darrell

----- Original Message -----
From: Doug Fortune <pentam-at-home.com>
To: Multiple recipients of list <emc-at-nist.gov>
Sent: Tuesday, March 28, 2000 10:38 PM
Subject: Re: home and limit switch in series


>
> Joel Jacobs wrote:
>
> > Seems kinda complicated to have so many switches for home/limit.
>
> Three switches per axis.
>
> > Wouldn't this work?
> >
> > Just two NC switches in series or two NO switches in parallel for each
> > axis - only three pport inputs used - one for each axis.  The same
switches
> > could be used for home and limit.  Each axis moves at HOMEING_VEL in the
> > direction of HOMING_POLARITY until it hits a switch.   The axis then
would
> > decelerate to a stop and slowly back up until it's off the switch.  It
could
> > optionally continue until an index pulse is recieved from an encoder or
a
> > hall switch on the lead screw.  This position would become 'machine
zero'.
> > Then move to HOME_OFFSET at HOMING_VELOCITY and reset axis position to
HOME.
> >
> > Using this method EMC would have to know which way it was moving when it
hit
> > a limit.  If it hits a limit maybe it should stop and back off the limit
> > before entering Estop.  That would reduce the possibility of the machine
> > being turned off with a limit switch activated.
>
> Exactly.    But the scheme fails if the machine is turned off (on purpose
or by
> accident)  if one or more of the axes has tripped (opened) the HL  switch.
Then
> upon awakening,  further pulses to Home in the homeward direction will
make it
> overtravel and grind against the stops with no feedback and possibly
damage
> something.   Remember, upon awakening, you don't know which of the two
> (home or limit) positions you are at.  (If it had tripped the limit
switch, travelling
> slowly in the homeward direction would untrip the limit switch, giving the
hint
> that you were now at the far end, and could now rapidly Home.)
>
> Since it is nice to have one (or more) emergency stop buttons arrayed
around
> the machine, my suggestion was to series wire it (them) with
Home-Overtravel
> switches (ie overtravel switches only on the Home end) on every axis,
coupled
> with a "try-to-home-each-axis-sequentially"  (ie non-simultaneously)
algorithm
> to guarantee no indeterminate states during Homing.   That way you get a
lot of
> use out of only one input line to the computer, including the emergency
stop
> function.
>
> This results in 4 axis Home/Limit capability with emergency stop and
> reliable cold start capability with only 5 inputs.
>
> Future excercise: Figure out how to retain the above functionality for
> 5 axis with only 5 inputs!
>
> Doug Fortune
> pentam-at-home.com
>
>
>