Drag-Knife Plugin

[GraphicMan]

I believe this is why I had some problems in cutting small letters
Please find the attached Image, Notice the transition between retract depth and cut depth while making a swivel action

Should not be the swivel made all at the retract depth and when done Z goes to cut depth and continue the path?
Not sure what wrong I could have done, I have also attached the CM file


 

[Dragonfly]

And you can change the direction and start point; http://www.cambam.info/doc/plus/cad/EditPolyline.htm

It is very important to make all closed curves to be in uniform direction because it determines the direction of the pivoting arc.
I wrote you a message, Eddy

[EddyCurrent]

And you can change the direction and start point; http://www.cambam.info/doc/plus/cad/EditPolyline.htm

It is very important to make all closed curves to be in uniform direction because it determines the direction of the pivoting arc.
I wrote you a message, Eddy

Sorry Ivan, I just replied now.

So are you saying the plugin should change the direction of all polylines to be the same ?

[Bubba]

So are you saying the plugin should change the direction of all polylines to be the same ?
*****************
Yes, As far as know it would help. Looking at it from my vantage point.

[lloydsp]

Lacking that, the plugin would require to predict which direction to move (start-point to end-point, or otherwise end-point to start-point).  In the event that it had to move from end-to-start, there is some work that would require to be done on the geometry calculations, in order to get the arc to curve the "right way", because the value stored in the arc's definition is relative to the direction of 'travel' of the arc.

It's not absolutely necessary that all the arcs and polylines progress in the same direction, but lacking that, Eddy will have to do some internal work to fix the direction issues.

Lloyd

[Dragonfly]

So are you saying the plugin should change the direction of all polylines to be the same ?

Yes.
This simplifies the movement and next segment direction analysis.
I didn't put much effort but in practice I am checking the direction, trying to change it, and there were cases when it wouldn't change. So I included a second try and then simply the code exits if direction change fails again. No idea why this is happening. Besides all initial polylines must be closed and the new ones - open due to the cross overlap of start/end segments to ensure the material is fully cut.
I am using a commercial Roland type blade holder and blade. The holder's adjustable foot (nut) has a flat lower surface and is adjusted so that the blade protrudes a tiny bit just enough to cut the foil and leave a mark on the backing. The tension is adjusted too just enough to make the blade pierce the foil and the foot is in constant contact sliding over the foil surface and this is how a constant depth of cut is achieved on a not absolutely leveled surface. So I am not using blade lifting at sharp corners and my generated blade paths are flat, all points at z=0.

[EddyCurrent]

New version added to first post.

1. Added radio buttons to set direction of selected polylines.

Picture shows the results.

Please note this does not change direction of the Polylines in the source geometry. Inside the plugin a copy is made of the source geometry and it is the direction of the copy that is changed.

[airnocker]

Well if it ain't one thing it's another, ya know?

Did my test with the new drag knife plug-in (not the very latest version, but the one before that as I'm running 098N).

My first engraving generation completed but although I went all the way through on most of the material in the set and because I'm using 1/4" acrylic as a cutting base and it's  probably not the flattest thing I didn't get full depth cutting in all the items BUT MOST OF THEM.

Then my lifting tape being a few years old didn't lift the letters off the backing and weed nicely like it used to, so I picked up the boxes and letters I could manually and moved them onto the white acrylic.  So the boxes and the first "Gawly" are from the first cut trial.

The "Gaw" letters to the very right are from the second test cut and there I had a weird, new experience.  I increased the DOC from .008" to .010" and regen'd the g-code then replace the new code into Mach3.  It acted like it generated the new tool paths.  I moved the X axis zero over to new material and started the second cut.  It did the "Gaw" letters just fine then moved over to do the "CNC" but oddly raised the bit an inch or so higher, then "cut" the "CNC" letters in the air.  It raised the Z higher after each letter.  w-e-i-r-d.

All that aside, you can see from the boxes (the first is 1" square and each one is half the size of the preceding box) that there are lead-in artifacts on each one and the smaller the box gets to worse you can see the "correction" move.

The .9" letters look pretty darn good but you can still see the lead-in or start artifact and some dips on the serifs from the swivel arcs.

As the letters in the photos are the same size and font (Times New Roman), the "CNC" is without any plugin from several years ago and the boxes and "Gawly" stuff are from today with the plugin.

The .cb file I used is the one posted hear earlier in the thread.

[airnocker]

This is probably a more useful A-B comparison of the two "CNC" letter versions.  The Drag Knife-12c.JPG was taken with a tripod I'm fairly certain, while the CNC2.JPG was a hand-held 1/15 second exposure.

Also, the only parameter I defined with the plug-in was knife offset.  I left all other parameters at their defaults.

Doing more tests....

[airnocker]

Finally had better success with lifting the cut letters from the backing and transferring to the white acrylic.

Apologies for the poor camera focus.

[GraphicMan]

@EddyCurrent

Did you check out the swivel action issue?
please let me know
thanks

[EddyCurrent]

@EddyCurrent

Did you check out the swivel action issue?
please let me know
thanks

We might have to wait until Bill gets back to it.

[GraphicMan]

Eddy can you check the plugin again
The cut path is in +ve Z not negative (cuts made in the air)
please recheck

[EddyCurrent]

It says in Bill's text

"A positive value for RetractDepth will be set as a -Z value."

So have you set that correctly ?

[GraphicMan]

It says in Bill's text

"A positive value for RetractDepth will be set as a -Z value."

So have you set that correctly ?

I know and I have doubled checked that and compared to the original script form
The script works properly It does convert any value to -ve Z (cutting down in the material)
The plugin does not, check the generated path
even though when trying to input a -ve value the plugin shows a warring massage that only +ve values should be used, but the -ve is still there and when I run the plugin it produce the path properly as -ve value for the Z, so i believe this is a bug, check it yourself