Air Scribe with PlasmaC

phillc54 wrote: Edit: You could still use M5 for the end of both as that will stop all spindles.

Yes, I was going to say once a spindle was set, the $n on F5 was optional.

rodw wrote:

thefabricator03 wrote: This is where I dont really understand, If we are using multiple spindles does that mean that the G-code commands for the plasma torch wont apply if we have spindle 1 selected? Do we have to define spindle 0 as the plasma torch or is it already set up?

Well my take on this is that each spindle has its own spindle.n.on signal and we use spindle 0 to trigger the torch (using spindle.0.on)
We are currently treating our machines as single spindle machines. But I think if you did gcode like this (and I'm no expert with gcode)

Code:

(set a spindle speed on all spindles to keep Linuxcnc happy) S100 $0 S100 $1 (Cut something) T1 M6 (select plasma tool - x, y offsets = 0,0) M3 $0 (use plasma) ...( cut some stuff) M5 $0 Go x1,y1 (go somewhere else) M3 $0 (use plasma and enable plasmac via spindle.0.on) ...( cut some more stuff) M5 $0 (Engrave something) T2 M6 (select scribe and set offsets and enable scribe via spindle.1.on) M3 $1 (use engraver) .... (engrave some stuff) M5 $1 G0 x2,y2 (go somewhere else) M3 $1 (use engraver) .... (engrave some more stuff) M5 $1

When you see how clean that code is, it does not really make sense to muddy it by adding a second signal to lower the torch. Sure that might happen but it should be done behind the scenes they way Plasmac does it. Plus you could have a whole row of tools up there. I did see Grotius post a pic with a drill bolted onto his gantry once.

Ok I understand now, So what do I need to set up to give it a go? I take it I need to write some kind of code for the post processor in sheetcam to spit out the correct G-code.

Do we need to define the Spindle0 as the plasma torch in PlasmaC? or hal/ini? And spindle1 as air scribe.

Also where do I set up the parameters for spindle1, As in turn on certain outputs to activate scribe relays, I take it that that is done in the Hal? I would like to set up that the scribe lower setting is turned on then after X amount of time the scribe it turned on then motion starts. I am assuming this also done in the Hal.

Well its all automatically supported and stays out of the way until its used as I showed so there is nothing really to do other than hook spindle.1.on to your output.

Code will not run until spindle.1.at-speed is set true. I would connect a spindle.1.on to whatever solenoid needs to come on first and use a delay component to trigger the next one. You may need to chain to a second delay to give the scribe to get to the surface before triggering spindle.1.at-speed.. Let me know how you want it to work and I can be a bit more specific.

thefabricator03 wrote: I would like to set up that the scribe lower setting is turned on then after X amount of time the scribe it turned on then motion starts. I am assuming this also done in the Hal.

Maybe sometihng like this.

rodw wrote:

thefabricator03 wrote: I would like to set up that the scribe lower setting is turned on then after X amount of time the scribe it turned on then motion starts. I am assuming this also done in the Hal.

Maybe sometihng like this.

Code:

loadrt timedelay addf timedelay.0 servo-thread setp timedelay.0.delay-on 0.5 net spindle-1-on <= spindle.1.on net spindle-1-on => my_7i76e-lower-output net spindle-1-on => timedelay.0.in net delayout <= timedelay.0.out net delayout => my_7i76e-scribeon-output net delayout => spindle.1.at-speed

Thanks Rod, That looks like it should do what I need.

Now the fun part, Modifying the post processor. I now realise that the code I posted earlier is for Mach3 not Linuxcnc,

Would I be better off trying to modify my CandCNC post for this task?

Stefan, I had thought once we know how we are going to do it, I might have a go at adding it to my POST seeing you are using it.

It should just be changes to the onPenDown and onPenUp methods as thats where the M3 and M5's are issued.

I did not get very far today. I went up to the shed to play with the ohmic sensing and decided I needed to upgrade to the latest version of master branch and after that, homing is not working on the X axis. Its just ignoring the home switch and plowing right through until it hits a limit switch. I think that must be a regression in master branch as all the signals were correct. If you email me the POST you are using I might get a chance to look at it over the weekend.