Another plasma component...

the optional use of m62/m63 does allow for functions emulating hypertherms truehole to a degree (no it won't vary the torch current or switch gases but it does stop the torch early while motion continues )

HI
Finally, I have some time to play with PlasmaC component it looks like a great piece of software.
I have a question, on this configuration, you are getting the feed from the configuration GUI, not from the postprocessor
I can clearly see the pros of that, however, I was wondering if we can send the feed and the feed rate from postprocessor using analog Out connected to plasmaC feed input, I use SheetCAM and I like to use rules like reduce speed on small circles, etc. so if I set my post-processor to out an analog out just before every Feed,.

do you think it can work?

thanks for all the great work at PlasmaC component

grijalvap wrote:

HI
Finally, I have some time to play with PlasmaC component it looks like a great piece of software.
I have a question, on this configuration, you are getting the feed from the configuration GUI, not from the postprocessor
I can clearly see the pros of that, however, I was wondering if we can send the feed and the feed rate from postprocessor using analog Out connected to plasmaC feed input, I use SheetCAM and I like to use rules like reduce speed on small circles, etc. so if I set my post-processor to out an analog out just before every Feed,.

do you think it can work?

thanks for all the great work at PlasmaC component

Pedro, great you've joined the plasmac revolution! The feedrate can be set from gcode Per chapter 6.1 in the docs
linuxcnc.org/docs/devel/html/plasma/plas...l#_material_handling
Chapter 6.2 following outlines how to slow for holes.

There are a couple of post processors for plasmac on this thread. There may be a couple of updates so read it all and get the latest ones.
forum.linuxcnc.org/plasma-laser/36568-po...-and-plasmac?start=0

on about page 9, I've shared some screen dumps of sheetcam code snippets for slowing down for holes.
Les at Sheetcam has offered to write a postprocessor for sheetcam and I'm waiting to be sure the final feature set is stable to involve him.. There is some current discussion about allowing enabling and disabling the torch between am M3 and M5. With your experience, it would be good if you could share your thoughts on this after reviewing the last couple of pages here.

Pedro

Welcome back!

I am still quite a few releases behind the current PlasmaC standard. I hope to bring myself up to the current release point in the next week. As others have mentioned you can set the feed rate from either the materials list or Gcode. I have done both with success. Please feel free to email with any questions. Even though it is of no use to me (I have a TD power supply) PlasmaC can benefit from incorporating your HT RS-485 code.

John

phillc54 wrote:

I have deferred work on the arc addition to the GCode parser for a while as there are problems if any words use named or numbered parameters or expressions. I will possibly release one that will just report an error if it cannot work it out but to do it 'properly' is a major job.

Phill, if you come back to this, could you make sure there is something that can disable this parsing? Somehow, it does not feel right to build features that modify the Gcode as a user might not want to enable this type of behaviour.

what do you want to do the to / with the gcode?

what would plasmac be looking for?

or wanting to do?

So I don’t know if this belongs here or in a new post. If it needs to be moved let me know and I will delete it and start a thread.

Im not a Linux programming genius, hell I could barely manage getting it installed. I don’t know if what I’m wanting can be added to the plasmaC or would need to be its own thing. The steel I cut is all high dollar stuff. Yesterday I cut up a 11”x23” sheet of stainless that cost me $100. So a mess up can be really costly and go down hill fast. Now I can’t cut it real fast as I’m running a 30xp and sitting at around 20”/min. I’m using sheetcam to nest and save it as a number attached to that sheet so I can load it back up and add more parts later and know what’s already been cut out. My issue is finding the exact starting corner. I have been using a square to get it square to the table and then bringing the torch down and moving over till I just barely touch the edge. Then set this as -.375 in g54 for x and y. This gives my just enough to clear the edge of the sheet. I do this because I set the edge of the parts at the edge of the sheet in sheet cam. This has eliminated the fudge factor guessing for me.

Yesterday I was wondering why could I not have a touch off button. The torch has ohmic so it could use that. Back when I ran cnc machines the new hass mills would load up a touch off tool. You would manually eyeball it over the corner and hit the touch off button. It would come down in the Z and find the top. Then move over in the X any Y a programmed amount and then come down a set amount and move over till it touches the side. It would set that as X and then retract and do the same for Y. It would then enter these numbers into the G54 offset.

If I could somehow build this function into my machine I would be tickled pink. We won’t bring up the other feature that would put it over the top. Well why not just mention it at least. Another machine I ran would touch off in 2 places on an edge and rotate the axis in the machine so the x and y would like up parallel to the sheet/part.

Robert,

I have been working on a post-processor for Fusion360 to use with plasmac and asked about making some changes in the g-code parser to handle the speed change when cutting circles as well as adjusting the kerf width due to cutting slower, and even extending the cut past the end of the circle with the torch off. Phill decided this was probably still better handled in the post-processor. There are a number of things that make this difficult to do from a "finished" g-code file as compared to doing it in a post-processor. And, it appears he has found a new issue we would have to overcome with the current g-code parser to implement it as well.

Steven