Getting back to LinuxCNC plasma table

[acourtjester]

One thing is you can get all the speed you want with the servos, it far exceeds the need for plasma cutting.

[N2 Welding]

One thing is you can get all the speed you want with the servos, it far exceeds the need for plasma cutting.

With that said is it a stretch to assume it would be a waste of money for someone on a tight budget to use servo motors?

[acourtjester]

That is correct there are many stepper motor table working just fine, all the other tables I built had steppers. Using the servos on my part was more of an interest of seeing the difference then the need. Being retired and a rabid hobbyist I was able to do the switch, with no schedule to meet I can play with different projects, and share the experiences with others.

[SegoMan DeSigns]

One thing is you can get all the speed you want with the servos, it far exceeds the need for plasma cutting.

With that said is it a stretch to assume it would be a waste of money for someone on a tight budget to use servo motors?

The Clearpath servos Tom & I use require no driver or encoders as they come built in so they are not that far out of line for cost. I have 3+ years on my NEMA 34's and they are still doing fine moving my 400lb gantry.

[cutnweld]

not sure how you can move a table too fast. I run 1200 ipm on my table, and with the linux touchoff speed as an added bonus it is not a laser but it does help close the gap on speed

[acourtjester]

Ok new question, switching back to Plasma on my table. I am using QTPlasmaC GUI and the Probe Test works as it should with moving to pierce height. But when running aa G-code (dry run) I don't think it moves to pierce height just to cut height then moving on the cut path. I have used different LinuxCNC Post Processors in SheetCam but looking at the G-code I don't see the probe command G38. What post processor should I use with an Ohmic sensor??
I have used other post processors with an ohmic sensor but not with LinuxCNC so I cannot use them as they use a G31 command for the probe.

[cutnweld]

The linux cnc plasmac post processor is the one I use. There is some "Ohmic3" code written by Rodw that gets put in someplace, if you pm him or caretech they would be able to point you in the right direction.

[acourtjester]

Thanks I will check on it.

[Rodw]

Qtplasmac does not use G38 probing. Instead it does everything automatically when it receives a spindle on command by using external offsets. I am a bit rusty with plasmac, There are some heights that need to be respected (safe height, probe height etc - entered in the GUI) It would be best to ask on the Linuxcnc forum.

I gave Les a hand to write the plasmac post processor. There are some settings at the top of the file to change diffeent behaviour. BY default it lets Sheetcam control the tool library but most people export their tools from Sheetcam and import them into Plasmac as then you can tweak things on the fly as plasmac then just uses the Sheetcam tool number to bring up the cut settings.

If you have got ohmic probing, I think there is a GUI setting to enable it. There are no gcodes. You just need a signal that goes true when the probe hits the material, plasmac will then back off slowly until contact is lost. Its good to have a float switch too becasue then you can probe a lot faster as the overtravel is taken up on the slide.

[acourtjester]

Thanks Rod, There are so many things with QTPlasmaC and LinuxCNC. Here is a routing ( Axis GUI) I use for my router bit height that used G38. I thought it would be the same for Plasma.
G38.2 z -2 f20
g92 Z0.0
g1 z0.8
g38.2 z -1 f10
g92 Z0.0
g1 Z 0.5
g92 z0.75
M05
M30

I just tried to find it in the manual, but did not now. This is similar to what I had found and used with other software for setting pierce height in the G-code.
It actually be in there when I do run a cutting operation, I just did not see it in the dry run. At the start I was using a floating head switch and it cut fine. I have added the ohmic sensing, so the simple thing is just run testing to see how things go.

[Rodw]

Connect your ohmic probe to plasmac.ohmic-probe
There ares some settings for probing on th parameters tab,
You must have ohmic enabled on the front screen

There are ots of things about qtplasmac which are unusual. Good luck with it.

[acourtjester]

Thank Rod I'm sure I have it hooked up correctly as both the ohmic test and probe test work fine. The ohmic led comes on when the button is pushed and I touch the torch tip to the work surface. The probe test moves down touches the metal surface and then re-tracks to the Z DRO 0.160.

[Rodw]

Sounds like it should work. I can't help. I sold my table 12 months ago

[kn612]

Ok new question, switching back to Plasma on my table. I am using QTPlasmaC GUI and the Probe Test works as it should with moving to pierce height. But when running aa G-code (dry run) I don't think it moves to pierce height just to cut height then moving on the cut path. I have used different LinuxCNC Post Processors in SheetCam but looking at the G-code I don't see the probe command G38. What post processor should I use with an Ohmic sensor??
I have used other post processors with an ohmic sensor but not with LinuxCNC so I cannot use them as they use a G31 command for the probe.

My torch does a dry run at cut height instead of pierce height too

[cutnweld]

Yup mine too. just tried it

[acourtjester]

Thanks guys sounds like that is normal
Something else I just found out when running the config for setting up the GUI there is a mode selection 0, 1, and 2. If you select 2 it has less buttons for the features 0 has them all.

[kn612]

Yessir, modes tie to your plasma unit and thc. Depending on how you get arc ok signal and arc voltage. Mode 2 would be for something such as a proma thc. I run mode 1 w/ thcad and a miller 625 that has cnc option.

[caretech]

Acourtjester -- see the excellent QtPlasmaC docs here:

http://linuxcnc.org/docs/devel/html/pla ... asma:modes

Mode 0, 1, and 2 described in detail there. Lots of other very useful info on that page too. It's worth a read.

[acourtjester]

I am having an odd problem with QTPlasmaC, I am trying to cut some metal (about time). All things seem to work but the Arc OK signal is not functioning. The torch fires and tries 3 times and stops, I can see the Arc Ok in signal in Hal Show each time the torch fires and the signals are show in the Hal file. I changed the mode from 2 to 0 no change!! So it seems I am missing something. Attached are screen shots of the Hal file.

[kn612]

Have you tried mode 1? Arc ok input is used with modes 1&2.

Pierce height low enough to transfer? Good ground connection? I haven't had to use hal show yet, I'm pretty limited on help on that one.

[acourtjester]

It was on 2 and I tried 0, I will try 1 in the morning. Hal show is need as you select the signal you want and the display shows the name and a red ball. when the signal is active it turns yellow. I would activate Hal Show and slide it to the side and watch it when the torch fires.
Thanks for the reply

[acourtjester]

It was on 2 and I tried 0, I will try 1 in the morning. Hal show is need as you select the signal you want and the display shows the name and a red ball. when the signal is active it turns yellow. I would activate Hal Show and slide it to the side and watch it when the torch fires. the torch stays on long enough to blow through the 3/16" as 25 amps. It should pierce and drop to cut height then torch out and move to the next cut start. The torch will fire 3 time and error out no Arc OK
Thanks for the reply

[acourtjester]

Well I as still fighting the glitch in the software, it still will not function with the plasma on. I have a center punch routine I use to mark the drill holes before cutting the parts out. It would just light the torch and stop, like no Arc OK signal. But as above it is getting into the software. I unclicked the Auto Volts button and it then ran that routine. Video below, but now it will not run the cut G-code and gets an error about the material list not finding it's spects. With plasma off it will run the G-code completely without errors. I had run a plasma test cut sometime back just fine. I had a PC die and started over with a new PC and it looks like the version I loaded has a glitch. I have done a complete new installation on a SSD with Book Worm Debian 12. Maybe installing it on the new table PC will correct the problem.

[kn612]

Tom, I saw you were getting a joint 2 limit switch error as well on the pierce over on the Linux forum. Would you be getting some noise when the arc starts causing that issue? Can you do a hal show and watch your joint 2 limit switch pin and see if it comes on when the arc starts?

If you are going to take the time and do a fresh install I recommend mx Linux:

https://forum.linuxcnc.org/9-installing ... -new-users

This has always worked very well for me on both a Toshiba laptop and a Dell desktop. I tried bookworm and kept getting a joint following error and with the same exact config with mx Linux on the same hardware it works, zero issues. Did a fresh install last month, following those directions it yielded me qtplasmac version 2.10.

[acourtjester]

Thanks for the link for the install, it seems the guy put some time into it, and I need the help with them (little experiences with Linux).
Interesting comment at the end I guess some heated words was pasted with another member. I have a few SSDs so I can do an install on one of them using this procedure. I think the setup on the table now is Dedian 11 bullseye, when using it for a router with Axis things worked fine (no plasma noise). On the other PC may have been a different version and did not have plasma problems??