Hi,
Craig, my switches are all proximity sensors... are there advantages to using mechanical switches?
For Home switches I believe there are advantages, for Limits, no.
For a Limit switch all that you really care about is that the switch operates when the axis is adjacent the switch. It does not especially matter if it operates +-0.1mm
nor does it particularly matter when the switch deactivates when you back up.
Home switches are different. Recall that Mach references the machine coordinate of an axis to zero (or Home Offset if so programmed) not when the switch activates but when it deactivates.
The axis will advance towards the Home switch until the switch activates and then decelerate to a stop. Then the axis will back up and the machine coordinate zeroed as the switch deactivates.
There is a premium here on exactly when a switch activates and when it deactivates. Snap action micro switches are very good at repeatably activating at exactly the same spot again and again.
The Omron switches I use have a precisely defined and repeatable hysteresis, ie a given travel before the switch deactivates.
With proximity switches and to a certain extent optical switches there is a minor variation not only when the switch operates but most importantly when the switch decativates when the axis reverses.
Its for this reason that I prefer snap action microswitches for Home switches.
Having told you the reason that I like mechanical switches there will be many to say the contrary and with experience to back the claim. I suspect in truth both solutions are adequate.
If you really want accurate homing then you'd use Index Homing. Given that I have servos on all my axes and the the servo drives have an auxiliary encoder output including an Index signal
I could invoke Index Homing easily. It would require an extra input to Mach for each axis of course, so four currently and five coming.
I made my own breakout board for my new mill. It has an eight wire plug for each of the six axes, four wires for differential Step/Dir signalling, one for an Enable, one for a Reset,
one for a servo Alarm and the last one is a 0V Return. I could I suppose redesign to have say a 10 wire plug which would allow at least one extra wire for Index Homing....but really do I need it?
I can Home or Reference my linear axes to within 0.02mm just with snap action microswitches....is that not good enough? I could with Index Homing do better, to within 1um....but is that
really a big advantage of just skiting? Once I fully develop my fourth and fifth axes I may have to revisit this decision, I really need to Home or Reference my rotary axes within say 15 arc secs,
and I suspect that Index Homing may be the only way to get that sort of accuracy. If that were the case then I would retrofit just the fourth and fifth axis with Index encoders rather than redesign
my breakout board.
Craig