Another plasma component...

Phill

Good news, I set the debounce delay to 5 like my system runs and much better behavior. Only one bouncing episode and that was caused by probing into a puddle on the sheet. So 13 tries and 12 successful lines cut! Torch pulse works as it should to get preflow going to clear water off sheet.

I also tested enabling and disabling THC enable from Gcode and it works just fine using a motion.digital-out-xx hal pin. I tried to link everything up in a post_gui_hal file but no go so I had to add a connection line to plasmac_panel.hal. I must be loading the GUI wrong for this to happen but I ran out of time for testing.

Well I am going to have to get a PP going now as want to test with some real work.

Thank you

John

John,

Attached is plasmac.comp with some changes to probing.
I am not 100% sure if the math for the heights is correct, so you should do some test probes first to check that.
I have followed your lead and pushed down a further 0.005" after initial contact, I hope this solves the problems.
If it works ok you could try reducing your debounce value as much as possible.

To install it:
copy it to your home folder
$ . path/to/scripts/rip-environment
$ halcompile --install plasmac.comp

If it doesn't work and you want to use the original plasmac then:
open and resave the original plasmac at your-repo/src/hal/components/plasmac.comp
$ make


Cheers, Phill.

I have split the Gmoccapy run/config panel into two so there is now an option of having a single panel with two tabs on the left hand side or two tabs that go behind the preview like John has done.
I have also fixed the broken tooltips.

Cheers, Phill.

Phill

I fetched this mornings update for the branch. Thanks for moving the tabs. Please see attached screen shoot on my new screen.

Problem. I went to cut a sheet of parts and make video and no joy. Back to same old probing problem, torch goes down and touchs sheet, Ohmic Probe LED comes on, Torch just bounces up and down +/-.005" while Ohmic-Probe LED intermittently changes state. Same response when trying to run a program and when doing a test probe. What do I need to check for? I think the last plasmac.comp I was using yesterday was dated 03 April.

John

Phill

A quick update. Probing with the float switch works as it should. Is this because with the float switch the movement is so much greater(.005" vs .115")?

John

John,
I am not quite sure what is happening. Did you try the plasmac comp I posted above . It is different to the one in the branch.

Cheers, Phill.

Phill

There is/was no attachment at this end.

I am a very poor coder but it looks to me the difference is that the float switch logic is looking for the float switch offset to not be exceeded and there is no such offset for ohmic probing. I can see how the travel can easily exceed zero_target with no sheet deflection and a bit of debounce.

I know why this wasn't a problem before. I was doing the line test cuts on the skeleton of sheet that I had already cut productions parts out of. So I was testing on a pre warped very flexible piece of material.

John

There is/was no attachment at this end.

Sorry John, I must have pressed the go button before my poor old brain had caught up...

Cheers, Phill.

islander261 wrote:
Please see attached screen shoot on my new screen

John,
I see you show the arc voltage as an integer. Is this normal practice for plasma?

Cheers, Phill

Phill

I just want a display that isn't jumping around too much.

I am different than most users in that I mostly cut one material and thickness. So if we use the HT rule that you need to hold your cut distance to better than +/-.005" for good cut quality. This is about +/- 6 arcvolts on my system (.070"/74v=.000810811in/v). If you are using standard Duramax consumables you are closer to +/-11v. I find on my system using a THCad card for arc voltage measurement that I get about .1 to .2 scaled volts noise whether I am cutting or not. So for me not seeing any fractions is ok. On the system I built I can only set the arcvoltage to a 1v resolution.

John