Limit switches modified

As mentioned yesterday, it looked possible to modify / rearrange the existing, rather nice looking limit switches and trigger ramps. Sure enough, it was fairly straightforward.

I only had to machine (narrow) one of the X axis ramps in the end. The Y axis ramps are a different design (possibly Omron?) and are the correct width to only trigger one of the switches. I also had to reposition the +L ramp on the X axis to give a little more distance between it and the dead stop, and shift it back 8mm so that it had a switch to itself - that was  a matter of drilling and tapping 2 new M4 holes in the underside of the table. The limit switch trigger ramp was hardened, so I annealed it, machined it down in width from 10mm to 5mm, then crudely rehardened it. It's not exactly critical.

Then the Y axis, which simply required one of the ramps to be shifted up 8mm to its own switch locus. I sensibly snapped off an M4 tap before replacing the ramp in its original position and instead moving the other one.

Then the switches had to be rewired. I now have 3 of the 4 switches in use. In each switch assembly there is a +L, -L and dead stop switch and these are wired in parallel (limits) or series (dead stops) for both axes. Next, connect the switches to the controller inside the console and configure the homing etc parameters.

NB: My reading / interpretation of the Chinglish instruction manual was clearly wrong. The +L and -L inputs are actually the dead stops and the X0, Y0, Z0 etc are the home switch inputs. So I had to decide between rewiring all the switches again or moving the ramps so that they are encountered in the opposite sequence. The latter was simpler and took little time. I may have lost about 10mm of motion but that's nothing to fret about for now.

Set up a DTI on the table to check I have the correct scaling on the X and Y axes. It's a crappy 1" thing but accurate enough to tell me that I have a scaling error. The servo drives can be configured to register 2 or 4 positions per line on the encoder ie to count every up and down edge on the A and B inputs. So if I understand it, for a 1000PPR (line) encoder, it requires either 2000 or 4000 steps per turn and I'd entered 1000, being naive. I'm pretty sure the default for the drives is 2 edges per line, in which case my controller seting should be 2000, not the current 1000. Certainly, when I ask for 25.4mm, it only moves 1/2" so I have a 2:1 scaling error. Yes, a value of 2000 in the "electronic gear numerator" parameters for the axes seems to correspond to the default settings in the driver. Clearly different conventions apply in China (Newkye) and Hungary (CNCdrives).