Re: Limit - Home switches

Hi PJ,
How the switches are wired is often a function of the
software. "Big" machines will have one input per switch.
Small
machines using a PC parallel port have limited inputs so series
wiring is often necessary. Software must handle these conditions.

Safest: multiple inputs
Cheapest: single input

Depending on the software, a home switch and limit switch can be the
same thing at the "zero" end.

The zeroing of a "big" machine (3 axis for example) would
first
start by zeroing the Z axis. This allows the spindle to (presumably)
be clear of any fixtures that might be mounted to the table when the
X & Y are zeroed. Once the Z has been zeroed, both X & Y can start
their zeroing process at the same time.
Zeroing of an axis goes as follows:

1) Is axis currently on limit switch now? Yes then go to step 4 else
go to step 2.

2) Slowly start moving the axis towards the HOME position and start
looking for switch closure. Keep looking until closure is found and
then go to step 3. If switch has not been found after a certain
parameterized number of pulses (which is usually MAX travel) error
out.

3) Stop the axis. Go to step 4

4) Start it going in the reverse direction (even slower than before).
Go to step 5.

5) Wait until the axis comes off the limit switch. When it does, go
to step 6

6) Now if you have encoders with a Z pulse go to step 7, otherwise,
you can call this HOME. Because of switch hysterisis (especially
with mechanical switches), the position it comes off the switch may
be a little different each time you come off the switch. You can
also elect to now move the axis forward an additional fixed number
of steps to give you clearance from the switch, and call that HOME.
Goto step 8

7) Systems with encoder and a Z pulse: Now that you've come off
the
switch, start looking for a Z pulse. Once you find the Z pulse, then
ZERO all counters relating to this axis, then software tells the
motion profiler that it should be at ZERO, but that its' actual
position is at a negative position (this negative number is also a
parameter, and can be up to half an inch from the point the Z pulse
was found). Seeing the error, the profiler will automatically move
the axis to the ZERO point, and now you can call this HOME. The
extra half inch gives you a little margin for ZERO away from the
switch. Additionally, the point that the Z pulse is found should be
about half a motor revolution away from where the axis came off the
switch. The switch is moved during machine assembly so that this is
the case. This makes it so that the switch hysterisis is always
within a definable window. If you come off the switch, and find the Z
pulse BEFORE getting out of the window, error out. Else Go to 8

8) Now that you've found come off the switch and have defined
HOME,
if software now detects a switch closure, this is an error, stop all
motion. If software is functioning well, you will not need to
connect the home switches to an E-STOP condition. If you don't
trust
the software then somehow connect the switch to E-STOP as you see fit.

Jeff
www.xylotex.com


--- In CAD_CAM_EDM_DRO@yahoogroups.com, "P. J. Hicks"
<hickspj467@a...> wrote:

> I would like to add limit and Home switches to my Sherline. I am

trying to make this machine as close to a 'real' or 'big' machine as
reasonably possible to act as my learning platform. I have down
loaded several drw of how limit and home switches are wired and they
seem to conflict. One shows them all in series activating an E-Stop.
Others show individual axis switches wired back to individual pins on
the controller.

>
> What are the crucial differences and why? Which way works most

reliably and is safest? Which way can use the least number of
switches to get homing and limits?

>
> PJH
>
> [Non-text portions of this message have been removed]