Another plasma component...

Phill,

I dont really understand G-code, I use Sheetcam for all of my G-code generation,

I have been using Rod's post processor and code snippets for my g-code,

Hopefully we can get PlasmaC to a stage where Les from sheetcam can write us a new post processor.

I still have not gotten around to testing the pierce only function , I will test it out on a piece of scrap tomorrow if I get time.

I don't have all of the answers but I think the multispindle features would support most of this.
If plasmac always recognised spindle.0 as being the plasma cutter, a user might define spindle.1 to be an engraver and perhaps spindle.2 as a drill and so on.

Then you just need to select the tool with the toolchanger commands to set the offsets correctly.

So the pin spindle1.on would turn on the engraver and spindle2.on the drill etc.

EDIT: then plasmac does not really need to support any extra features, it would be up to the user.

rodw wrote:

I don't have all of the answers but I think the multispindle features would support most of this.
If plasmac always recognised spindle.0 as being the plasma cutter, a user might define spindle.1 to be an engraver and perhaps spindle.2 as a drill and so on.

Then you just need to select the tool with the toolchanger commands to set the offsets correctly.

So the pin spindle1.on would turn on the engraver and spindle2.on the drill etc.

EDIT: then plasmac does not really need to support any extra features, it would be up to the user.

Does PlasmaC already support multispindle feature? What your saying makes sense to me, I was thinking how I could add a additional axis later if needed - e.g. rotary axis.

thefabricator03 wrote:

Does PlasmaC already support multispindle feature? What your saying makes sense to me, I was thinking how I could add a additional axis later if needed - e.g. rotary axis.

Yes Linuxcnc supports multispindles from V2.8
linuxcnc.org/docs/devel/html/gcode/m-code.html#mcode:m3-m4-m5
linuxcnc.org/docs/devel/html/man/man9/mo....html#SPINDLE%20PINS

The rotary axis is super simple too. I've played with that and out a 5th stepper controller into the spaceship just for that.You just add another (rotary) joint in your ini and hal files and in your ini make it a XYYZA machine.

Anyway, in conjunction with a toolchanger, I think this is probably the best cleanest and most Linuxcnc like method. Maybe Andy who wrote the multi-spindle stuff might have some ideas.

So will I need to set up any air scribe specific code?

Or can I create a tool that is for the scribe and create a sheetcam post that sets up a tool change to the scribe tool on another spindle?

thefabricator03 wrote:

So will I need to set up any air scribe specific code?

Or can I create a tool that is for the scribe and create a sheetcam post that sets up a tool change to the scribe tool on another spindle?

I'd have to think about it a bit but I think it should be able to be managed in Sheetcam like the center punch tool is now. Move scribing to a new layer and do another operation with the tool. Probably just need to make sure that the onPenDown function selects the correct spindle on an M3.

It will be interesting to hear what Phill says about the tool changing.

If you do a Google Search for Sheetcam engraving, it would be interesting to see what is suggested. I will say this is so different to Mach3, so much cleaner and using standard Gcode, no custom M codes.

Time for some more pictures,

The ugly side of owning a water table, Cleaning it out,



351 cleats in just over 4 hours, 100% yield, Nothing better than being able to walk away from the machine while it cuts and know that it wont break down. Cut all cleats with no operator intervention. Just a testament to how good PlasmaC really is!

thefabricator03 wrote:

Time for some more pictures,
351 cleats in just over 4 hours, 100% yield, Nothing better than being able to walk away from the machine while it cuts and know that it wont break down. Cut all cleats with no operator intervention. Just a testament to how good PlasmaC really is!

Great work. So how does that error rate compare with previous pre-plasmac experience?

rodw wrote:

Great work. So how does that error rate compare with previous pre-plasmac experience?

With my CandCNC controller, I had a lot more errors compared to PlasmaC , When I first purchased the controller I could not get the simple probe to work properly, I was constantly getting in contact with CandCNC but they did not know what was going on. It would not recognize a probe input, The z axis would sense the top of material with the ohmic but the axis would continue to move downwards causing the drive to fault. I managed a work around changing the post processor but I never felt I could trust the system.

Its so reassuring with PlasmaC that when I leave the machine it will continue to work as it should. If anything does go wrong I have the knowledge of building the controller myself and the great team on this forum for help.