My Newb Plasma Table Attempted Build

The circle from Toms code you posted here doesn't look circular to me. It is shown with the start point at the top when it is actually at the right side and it then seems that it is a bit short in the Y axis. Can you measure it to see.

phillc54 wrote: The circle from Toms code you posted here doesn't look circular to me. It is shown with the start point at the top when it is actually at the right side and it then seems that it is a bit short in the Y axis. Can you measure it to see.

It's just the cell phone camera. The edge was a little beveled and that's making it look that way.

Try these two circles, both 2" dia. You may need to change the feed rate as they are 10ipm, the same as Toms.

You can probably use you marker pen if you don't want to cut as it seems to show the same issues.

phillc54 wrote: Try these two circles, both 2" dia. You may need to change the feed rate as they are 10ipm, the same as Toms.

You can probably use you marker pen if you don't want to cut as it seems to show the same issues.

Ran them both. They appear to be ok...forgive the stray marker outside of the main path, I kind of have to apply a little pressure to the marker after probing to get it to write.

dvn4life1972 wrote: I again used the wizard, to make a square this time, and the error appeared.

On that square there can be some weaves clearly seen, and that points to mechanical issues, but the strange thing is they are present only on half of the cut.

I think those are caused by the acceleration being a bit too high again. I need to go back thru my notes, PCW had walked me thru lowering that before the PC changeout...but this config is from just before that was done. It doesn't show up on the other 10ga scrap I was using, but it really shows on this 18ga thinner material.

Aside from that, the same error is present. I'll get that acceleration lowered...it's an issue when cornering.

I've spend all day trying to do a from-fresh config using pncconf. Now the machine will not run correctly unless the velocity and accel are set to one specific setting. The reason I am even attempting to just do a brand new configuration and rebuild the machine in linuxcnc is because nothing else is working with the existing one, so why not? Everyone has pretty much resorted to just stating it must be something physically on the machine, even though it was working fine even with the laptop when the issue began to occur. Now it has a desktop PC, linuxcnc 2.8, etc etc and it will cut the test pieces perfectly with simply gcode not produced by a common pp.

I understand that we may be at a point where nobody has any other ideas or may even be tired of trying, but if that's where you are please let me know so I can start obtaining what I need to move to something different altogether.

Thanks

Why would you do a fresh config?

If it was working and then quit without any software changes
that suggests a hardware issue (whether mechanical or electrical)

PCW wrote: Why would you do a fresh config?

If it was working and then quit without any software changes
that suggests a hardware issue (whether mechanical or electrical)

I'm fully willing to replace some hardware, but what? The drivers and motors?

What I mean about pncconf, is that if the base configuration has relatively
tight following error limits and you don't get following errors, then the basic
configuration is not contributing to the problem, You may of course need to
tweak the acceleration and velocity numbers, but this should be done with
"Calibrate" menus not by re-running pncconf

Can you try swapping just the X,Y step connections?
( this does require homing to be disabled! )