Another plasma component...

Phill

Success! After getting my mechanics.hal right the probing-test is working! The one thing I did notice is that it did not set the plate surface to 0 on the DRO ( I expected to see the DRO at .090 when the torch was at pierce height). I ran quite a number of trials and they all turned out the same on a floppy piece of 2mm HR STL on a really chewed up bed. I could pass a .089" dowel under torch every time and a .093" dowel wouldn't pass under the torch. It is too late here to cut a sheet of parts to get some scrap for line tests so I will need to wait until the morning. I will get the line cut macros written tonight. I guess the good news at this point is that I have relatively clean examples of .ini and .hal files to use a 7i76E and THCad with the plasmac branch.

John

islander261 wrote:

Phill
Success!
John

Very exciting guys. I just have not had the time to play for so long. Really want to get my ohmic sensing working this weekend.

Phill

I am sure that those from the external THC camp will say why bother. My present system uses mostly standard hal components to do THC and in the past when I worked with EO branch I would set the G54 Z to zero when I probed to work piece surface. I did this simply with G92 Z 0.000 in my probing routine. It is very comforting to see the correct pierce height and then the correct cutting height displayed on the DRO. My current configuration doesn't update the DRO when the THC is active so you see the initial cut height for the full cut. When I was using the EO branch I had the DRO updating during the cut. It was nice to have because it gave you a sense what the material was doing (warping). Actually as I recall the EO branch "automatically" tracks the Z movement so that all is required is to zero the G54 ( work piece top surface) Z when probing is complete. Yes the work piece does warp and that causes changes in the work piece (G54) Z 0. One never changes the machine (G53) Z0 when cutting, this is set at homing and remains fixed. Because we update the G54 Z 0 every pierce the warpage doesn't matter because we are constantly correcting it. I usually limit my IHS skip distance to less than 2" ( more like 1.25") on the thin material I use.

I was able to probe from random Z heights with no problems. The default configuration is saving and reloading. Still no luck adding to the materials list, not a big problem as I will most likely edit it to the materials and conditions I use.

Thanks again for all the work. I hope to get a few line test cuts in tomorrow.

John

phillc54 wrote:

This has been fixed, the GUI's set it correctly. You can leave the setp plasmac.mode 2 in your HAL file if you wish.

Thank you for this fix I will update and try it out.

I'm not sure what is going on here, islander261 is also having this problem. I have tried many times here and cannot get it fail.
I will keep trying.

Let me know if there is anything I can do on my side to help diagnose this problem. The only thing I can think of that may cause an issue is I changed the name of the machine in the INI file. This caused a new "machine-name.cfg" and "machine-name.mat" file to be generated. linuxcnc does seem to load changes made directly to the file but sving doesn't happen.

I have tried touchoff at 1mm in a metric config and 0.1" in a imperial config without any problems. In the sims I just leave touchoff at Z0 and don't touchoff for any tests. The component itself does not use the touchoff height at all.

This morning I managed to figure out why these errors were occuring (it was my fault). My machine table only uses a single motor for the y-axis (joint_1) but also has a rotary 4th axis (joint_3). When I converted the config over from sim I missed this line in plasmac.hal:

net p_comp:axis-position        joint.3.pos-fb                  =>  plasmac.axis-z-position

which was reading my rotary position as the z-position and causing confusing errors. To fis the issue I simply changed it to read:

net p_comp:axis-position        joint.2.pos-fb                  =>  plasmac.axis-z-position

but I was wondering if it would not be better if the line was:

net p_comp:axis-position        axis.z.pos-cmd              =>  plasmac.axis-z-position

my brief check in hal config told me that these should be equivalent but I dont have enough Linuxcnc experience to be sure...

Do you still want to see my config I should be able to upload it later today (machine computer doesn't have internet)?

Thanks for your help
Glenn