Sc + Mach3: Rotary Plasma, Square Tube Issues - Help Needed

diesel · Post by **diesel** » Wed Nov 09, 2016 11:03 pm

Hello:

I'm seeking clarification on the Sc rotary plugin and its interaction with Mach3.

I've watched WMF's and Les' YouTube videos many times and am still not able to sort out how to make it all work properly

I'm running a CandCNC Ethercut with full 4.5/5-axis support (no need to unplug anything to make the rotary axis operate).

The Z-axis is giving me grief. Depending on the post processor I try, the Z either hits the upper limit switch or drives the torch right into the workpiece after touch off. I haven't tried actually cutting anything yet since I would like to verify motion first. In order to do dry runs, I have the DTHC in manual/disabled mode.

Tom Caudle suggested on the Plasma Spider forum that I need a different Mach3 profile for the Sc rotary plugin in order to properly cut square/rectangular tubing. Tom is suggesting that there is some interaction between the Sc post processor and Mach3 that needs to occur and doesn't appear to be occurring.

I'm completely unaware how to create and setup a new Mach3 profile.
I currently have 3 profiles and paid for licensing for each Sheetcam plugin when I got my brand new table just over a year ago: flat table plasma (appears to be working fine), rotary plasma and router (never even used it)

Today I've upgraded from Sc TNG 6.0.15 to 6.0.19 and the behavior persists.

The other day, the Z-axis was working fine when doing rotary cutting motion tests. The issue I was having then was with "cutting" geometry on the apex of the square tubing radius - the motion would "cut" half a circle on one side of the pipe and then "cut" the other side of the circle on the complete opposite side of the pipe.

Here's a video of that somewhat successful motion test:
https://youtu.be/ELypQafY1Lo

I probably have at least 30 hours invested in research and experimentation trying to make it operate properly. Any advice that someone could provide to get me back on the right path would be appreciated greatly.

Thanks,

Matt

Table specs:
LDR Motion Systems 5x10 - new Sept. 2015
Custom all aluminum gantry with a lot of Z-axis travel and added X-axis travel. (No need to remove or remount torch for rotary. To switch from flat to rotary, in the Mach3 flat profile I just use a macro that homes all axes and drives the X-axis off the edge of the table and it self-centers the torch over the rotary)
Scribe
Stepper motor based CandCNC Ethercut with DTHC-IV (running latest firmware)
Mach3
Sheetcam TNG 6.0.19

Post by **Les Newell** » Thu Nov 10, 2016 12:14 am

That motion is really odd. Just a thought - in Mach make sure your rotary axis is defined as a linear axis, not rotary. Mach's rotary implementation does not play well with SheetCam.

diesel · Post by **diesel** » Thu Nov 10, 2016 3:04 pm

Thanks for your reply, Les!

Is that the IJ mode setting?
I have to admit that I am not a fan of Mach3's GUI.

Cheers,

Matt

Post by **Les Newell** » Thu Nov 10, 2016 3:12 pm

In Config->general config make sure 'A axis is angular' (or whatever axis your rotary axis is) is NOT checked.

With angular mode enabled Mach 'rolls over' some moves, taking the shortest path rather than what it has been commanded to do.

diesel · Post by **diesel** » Thu Nov 10, 2016 3:25 pm

Excellent - thank you!
I will look at that ASAP.

Cheers,

Matt

diesel · Post by **diesel** » Thu Nov 10, 2016 9:12 pm

Hi Les:

In Mach3 A, B, C axes are unchecked/linear.
My table uses A for the rotary.
I'm at a loss...this must be something very simple.

Cheers,

Matt

Post by **Les Newell** » Fri Nov 11, 2016 1:25 pm

Sorry it took me a while to reply. Could you try the development version of SheetCam. Your licenses will work with that version. What post are you using?

diesel · Post by **diesel** » Sun Nov 13, 2016 5:59 pm

Hi Les:

Thanks for getting back to me.

I can certainly try a development version of Sheetcam.
I'm currently running 6.0.19.

I was finally able to get the Z to follow the contour of the tube. To make this happen, I ended up restoring the original post that shipped with my table and the rotary axis. All motion tests I completed looked OK...perhaps the torch height was a bit close.

With all of that in mind, I decided to try a test operation on 4" square tube (0.25" wall) and DTHC-IV active. Almost immediately after the pierce, the torch height was about 1" above the tube. Methinks I may have to disable DTHC for cutting tube.

Attached is a picture showing a test part that was created with the THC active which cause the torch height to be much too high for the internal geometry.

Thanks for your help!
Further tweaking is needed, but I'm happy that I was able to get it to do something productive.

I also will need to tweak my CAD drawings.
I'm not liking the odd geometry at the corners.

Cheers,

Matt

Post by **Les Newell** » Mon Nov 14, 2016 2:53 pm

diesel wrote:I was finally able to get the Z to follow the contour of the tube. To make this happen, I ended up restoring the original post that shipped with my table and the rotary axis. All motion tests I completed looked OK...perhaps the torch height was a bit close.

That is good news. The cut height being a bit low could be your switch offset.

I decided to try a test operation on 4" square tube (0.25" wall) and DTHC-IV active. Almost immediately after the pierce, the torch height was about 1" above the tube.

That sounds like your DTHC voltage setting may be incorrect. Does it work on flat sheet?

I also will need to tweak my CAD drawings. I'm not liking the odd geometry at the corners.

Yes, it can be tricky getting the corners right.

diesel · Post by **diesel** » Mon Nov 14, 2016 10:17 pm

Les Newell wrote:That is good news. The cut height being a bit low could be your switch offset.

OK, so you bring up a great point...one that has confused me a bit.
Mach3 has a switch offset and so does Sheetcam.
Do I calculate the offset in Mach3, set it in Mach3 and then also set that same value in Sheetcam?

Les Newell wrote:That sounds like your DTHC voltage setting may be incorrect. Does it work on flat sheet?

Yes, I haven't had any issues with the flat sheet in a while...and my issues were due to inexperience. I still need to do some voltage testing with square tube as it is a whole different beast.

Les Newell wrote:Yes, it can be tricky getting the corners right.

One would think that solid modeling (Solidworks) would help with this.
I need to keep on experimenting!

Thanks for your help!

Cheers,

Matt

Post by **Les Newell** » Tue Nov 15, 2016 11:51 am

diesel wrote:Mach3 has a switch offset and so does Sheetcam.
Do I calculate the offset in Mach3, set it in Mach3 and then also set that same value in Sheetcam?

It depends on how the post is configured. If the post generates the reference moves you need to set it in the post. If referencing is done as a canned cycle in Mach you need to set the switch offset in Mach.

Yes, I haven't had any issues with the flat sheet in a while...and my issues were due to inexperience. I still need to do some voltage testing with square tube as it is a whole different beast.

I would have expected the THC voltages to be very similar though you may get torch dive on the corners of rectangular tube if your axes can't move quickly enough. The rotary and Y axes have to move pretty fast when cutting around corners.

One would think that solid modeling (Solidworks) would help with this.

Ideally you want to be able to 'unwrap' your solid model. That way the geometry should be pretty much perfect. I have a suspicion unwrapping is part of an extra cost module for SolidWorks.