OpenPnP with LinuxCNC via linuxcncrsh?

plhsystems wrote:

I think I like your previous idea of having a switch permanently made when the axes is not at home and someway down. That way, you could use two switches (one for each side) and have the centre point where neither switch is made and hence home. So homing would be simply:

• See which switch is made
• Move that side so that it goes up
• Stop when the switch in no longer made
  
  
  And as long as the switches are at the same height, finding the true centre point wouldn't be too hard, just halve the distance from when the switch on one side releases to when the switch on the other side makes.

That's not quite what I meant. If you did that then you would have no limit switch unless you add 2 more. I was winging it when I machined my head base, I added 2 threaded holes for inductive proximity switches positioned as limits. I planned on doing what I said first, now I plan on adding the pot

tommylight wrote:

1 switch + home_offset will home it anywhere, including in the middle.
As for precission, it uses a stepper with a pulley directly driving heads, so even a cheap microswitch will be more precise than the drive parts.
Using an optical switch and a 0.9 degree motor seems better for accuracy, as there is no need for a lot of torque, and this is for very small parts.
 

Ya wanna explain that process to me? I know how homing and LinuxCNC works but I'd like you to explain the process to me with "1 switch + home offset"  so I can explain where it fails and maybe you'll understand.

I wan't too concerned about limit switches, as long as you can find home and know the what maximum travel is then you should be good, but if you machined the head already then pots are probably the best way to go.

That should work. I was thinking 2 switches but you do only need 1, if the switch is made, move one direction until it releases and if not, move the other way until it activates.

Having read the homing section of the manual, it looks like what we are trying to achieve is standard, i.e. the axis travels in one direction until the switch is made and reverses until the switch releases, exactly what we need

 

And by what I can gather from this pic

 

The home switch is ignored after homing anyway, please correct me if I am wrong

Home switch is only valid during the actual homing procedure.

For a setup like the one I pictured several posts back, that's actually pretty easy to do. You just mount an inductive proximity switch on top of the base behind one of the gear racks. The switch will be active about 40% of the range which is probably fine.  That is a standard homing procedure, you set it up to travel towards the side that activates the switch, after it makes it backs off. If it's made when homing is started it only travels in the back-off direction.

That only works if you're setup mechanically allows it. Personally I'll be doing something like a 5-turn potentiometer and using the  HOME_ABSOLUTE_ENCODER = 2. The nice thing about this is it actually is an "absolute encoder" So it adds closed loop feedback for little coin.

Actually what I think I'll do is use one of these "high-torque" pots as the lower pulley shaft threaded in from the back.   www.digikey.com/en/products/detail/tt-el.../8146R5KL-25/4780742

Not much modification needed.

 

They're quite expensive, I think I'll stick with the switch

A $35 pot is not expensive considering the cost of everything else. I'm only using the "expensive" pot because it's serving as an axle for the lower pulley. The same type of pot in a lower duty version mounted freely to the stepper or lower pulley is alot less expensive.

But to each their own, this is what I'm doing.