Updated to Master Branch, Stuck in E-Stop

Just updated everything to 2.9.0 so I could pull the Plasmac updates more easily. I ran the plasmac configurator.py and updated my .ini file.

When I start linuxcnc and verify that it's running 2.9.0-pre0, axis starts up but I cannot toggle out of e-stop.

If I change my config to use Gmoccapy, I can get out of e-stop and everything appears to be working (I have motion).

I had a working configuration before I updated.

Thanks

Bit of a long shot but try changing your estop in hal to this.
I think this is identical to what you have, just formatted differently. I vaguely remember having this problem in axis back with an early 2.8 build and gmoccappy was OK.

Does the plasmac sim do the same thing?
If you run a non plasmac axis sim, do you have the same fault?

I tried the change and it has the same behavior.

The plasmac axis sim works fine, as does non plasma axis sims.

Perhaps it's something with my axis files? Seems strange that gmoccapy works and axis doesn't.

Edit: Looking at the pin "iocontrol.0.emc-enable-in" it's reading FALSE. Shouldn't this be TRUE due to the setp I have in my HAL? Perhaps it's getting set to 0 somewhere else?

newsig is an obsolete command Maybe thats it. See
linuxcnc.org/docs/devel/html/man/man1/halcmd.1.html

I would compare the hal file for a sim against your config.

Is this a gantry machine? If so, its set up as wrong for 2.8/2.9

and you seem to be mixing stuff up in these sections.

Why does joint 2 connect to stepgens 3 & 4?

I'd be tempted to run pncconf and get a working config and then cut and paste the stepgen stuff into your hal file as there were significant changes between 2.7 and 2.8/2.9 But if you opened ran your original 2.7 config, the config files should have been converted to the new format

This is a gantry machine, but I have not separated out the two Y motors into individual joints to take advantage of the update. I have only ever used 2.8, and it was working really well with this config. Motor Y2 was just slaved (with a negative scale value) to Y1.

Edit: Joint 2 is for my Z, not Y. My machine is really a XYYZZ

This example I did might help if you wanted to use the joint axis stuff.

forum.linuxcnc.org/47-hal-examples/34072...ranch-gantry-example

The spindle needs a slight change due to multispindle which came after I did this. Its easy enough to cut and paste the stepgen stuff.

phillc54 wrote:

Are you saying that if you change axis to gmoccapy in the ini file you posted then it works?

Do you get any error messages in the terminal when you try running axis?

Yes, if I comment out the Axis line and uncomment gmoccapy, it works as in I can move my machine around. Nothing plasma related is loading but I assume that's because I'm missing some python files in my dir possibly? I couldn't find a plasmac_gmoccapy.py like the plasma_axis.py.

No errors are shown when loading Axis.



As a test I started a config I have for the same machine (same hal file, and mostly same ini), but when I use it as a router (no plasmac stuff). Still on 2.9.0, it comes out of e-stop just fine. It appears to be something specific to plasmac+Axis?

I'm going to try starting from scratch, converting the functioning sim over.

Edit: Starting with the Axis sim (/home/bryan/linuxcnc-dev/configs/by_machine/plasmac/axis), I converted everything to a working machine. Copied the exact same machine.hal file (called plasmaworks.hal) and remade the symbolic links to the /plasmac/ folder. Totally stumped why this worked when all of my hal and ini files are the same. Going to have to take a closer look at the differences in the other files.

Machine is out of e-stop and moving around and all of the plasmac related stuff (tabs and panels) seem to be functioning. Will attempt a cut later today.