Another Plasma Install

Hello All

I have finally gotten around to getting hardware installed in housings and cables made to retro fit my plasma table. The table is a much modified Torchmate 2 using a Torchmate branded Flashcut controller. I use my table on a daily basis for production work so I will not be able to actually connect to the physical machine until I am sure that I have a fully working control system. The table is a gantry type with dual drives on the Y axis.

What I have working now. I am running LCNC on an old Dell Optiflex 745 SFF box (free at the dump!). My hardware interfaces are a Mesa 7i76e card and a Mesa THCad10 card. The 7i76e is connected to the native Dell NIC. The operating system is LDME-2 and has been updated several times. I am running LCNC 2.8.0 PRE from the master branch to get gantry support. For the GUI I am running the plasma version of gmoccapy 2.1.6 (at least that is what displays at the top of the GUI). It was quite a journey getting all these parts playing together, much trial and more error to get to this point. I am not really looking forward to trying to update any of this. The only "bug" I can find at this time is that I cannot start the halscope from the setup page of gmoccapy. At this time I have the GUI displaying the arc voltage, ArcOk LED, Float Sw LED from external hardware. I can fire the torch (for testing only) using the torch button. I made the HAL connections for doing this in the plasma.hal file, the main HAL file contains all of my hardware definitions. I am not sure if this is the correct way to do this and welcome any advice on the "correct" way.

I live in the US so I use archaic units of measure. I have the main gmoccapy working in inches with exception of the THC functions. Is there a thread about how to change the units for the THC functions? I couldn't find one or didn't understand that one applied to this problem. Please help with any direction. In the future I want move most of the THC controls to a different tab or page that is not visible (taking up space) on the main page and add new controls. I find that I rarely change the THC settings ( I cut the same material >90% of the time) during routine use. Even when changing material the only parameter that is routinely changed is the arc voltage to compensate for cutting speed changes. The response of my present THC is so slow that I rarely change the pierce delay.

My present table has no limit or homing switches. I do use soft limits to keep the torch over the material (I almost always use full sheets of the same size). The gantry does rack over time from the turn on and off jitter from the stepper drivers. I usually place the material to be cut on the table and align it to the table using the laser cross hairs on the torch. I then manually set the machine and then the part origins to the sheet. I want to be able to use combined limit/homing switches to keep the gantry square and prevent running off the end of the drive racks when jogging. I will post more on this this later. I am still confused about how all this works with getting the joints available for jogging prior to homing and the axis afterwards. This is one of those topics that I thought I had clear in my head from reading other posts until I am face with actually doing it.

I will be replacing the current Z axis drive on the table with a new one. This is where things are going to get complicated and non standard. The Z axis will have the following switches: Z+ limit, float switch for Z- limit (broken torch heads are expensive), torch crash switch. The plate surface will be sensed by conductive contact between the torch shield and the top of the plate, this is a requirement as my present system uses this and it works very well on the thin plate I work with. The relays and optocoupler for this are located in the control cabinet. I think that I want to use the Z+ limit as a combined limit/homing switch, machine Z 0 will be about 1/2" below the switch in Z+ (up) travel. The float switch will act as the Z- limit. I am not sure whether it will just do a feedhold function or do both a feedhold and axis disable (estop?). The torch crash switch will be activated when the torch magnetic mount is broken off of the Z stage, this will cause a feedhold (stop all motion and shut off torch) to be asserted. I intend on using a modified G38.1 command or subroutine to find the plate surface and set part Z0. This is complicated by the fact that the conductive switch can be falsely triggered by slag build up or water between the shield and nozzle (Hypertherm torch). This requires testing for a shorted condition as a preamble or part of the G38.1 probing block. My present Z control system is external to the main RS274 machine controller so I don't use G38.1 probing. Any comments on this? I will post photos in about a week. I want to have all of this tested and working before installation on my table.

I am looking for guide or forum topic the explains how all of the various files work together to make a gmoccapy installation work. I am also looking for the correct tools to modify these files. I tried to install a version of a GUI builder once and never really figured out if I was successful.

John

Moderators

I posted in the wrong area. Please move this to the GMOCCAPY area. TIA

John

The first picture I see with the error. the only time I have seen it is when I ran the wrong Kernel and that is with the newer uspace 2.8, did not have that error with 2.7.6 or what ever it was, it ran fine with both kernel, until I update Uspace to the newest one and run it with the wrong kernel.

what supply voltage are you using on TB-1

Hello

Making progress here. I have managed to change the THC portions of the GUI to display in inches with the correct ranges.

Remember to make backup copies of working files before you break them. I modified the plasma.py file with the text editor to have the correct ranges for each control. I renamed the existing plasma.pyc file to force some action that I don't know about. Then to fix the units display from mm to in I opened the plasma.glade file by right clicking on it and selected the suggested Open with Glade menu choice. In the upper right pane I clicked on window1 and the gui parts appeared in the main pane. Selecting each mm instance in turn I was able to edit the text in the lower right pane when selected. I save the file when finished. I then renamed the plasma.var file to force something I don't understand to happen so the new plasma.glade is used. Learning a lot by reading forum posts and a lot of trial and error here. Sorry about posting such a small compressed file but the forum editor wouldn't allow me to post a file with the .py extension.

John

Thanks for sharing this with your config files. I am in the process of building a scratch built plasma table. I have just got the control box working but the wiring is incomplete. I have all but built the gantry. The Z has 4 sensors and will also carry proximity sensors for the Y limits.

I will try and follow your config conventions as I put mine together. Unfortunately, I will have to undo your changes to support your archaic measurement system. Many years ago, my brain supported this system at the kernal level but now following a software upgrade some years ago, it simply refuses to acknowledge it even exists! Fortunately there is a workaround as it can still recall some key conversion factors which allowed me to deal with the strange rack I used for driving the y axis. It has M6 mounting holes spaced on a 5.75" pitch. What a mess!

This is one of those topics that I thought I had clear in my head from reading other posts until I am face with actually doing it.

This sounds like the story of my life right now with this project!