Menu
What's new

Lightburn Testing, etc.

Bill Arnold

Bill Arnold

1974
Staff member
Messages
8,633
Location
Thomasville, GA
Since I've been reading about Lightburn from a few of you, I decided to download the test version to try out a couple of things. I don't plan to use it to run my laser system; bCNC is working fine for that and I run that wireless. I'm mainly interested in how it handles photographs. It appears to be a snap for processing a photo, so I'm playing with it a bit more.

I ran into something that I haven't figured out yet, though. I did one photo, processed it, created the gcode, downloaded it to bCNC and ran it. The photo ran as expected, although I need to use better material for the final run. I read up on the settings information a bit more, then processed another photo. The second one caught me off guard because it printed upside down and mirrored! Any idea about the reversal thing?

In the process of running the second photo, I discovered I needed to tweak the driver for my X axis (long axis, two motors). I had set all the drivers on my J Tech BOB to 0.50V initially, then ran the X axis driver up to 0.80V. Everything I had run looked fine until I ran this photo today. Because of its transitions from dark to light and back, there were frequent accelerations, occasionally causing a blip in the motors. Depending on which way the axis was moving when this happened, the gantry would shift either right or left. So, after seeing that happening, I adjusted the X axis driver to 1.0V and ran the same gcode again. The results are below.


Bobbie_Test-a.jpg
 
Last edited:
You probably have it sorted, but I did have pausing type issues before if I was trying to print something too fast with the last. Doesn't sound like it would be an issue with your stuff since you've printed these out correctly before.

The issue I had was that the laser code over ran the buffers on the controller, so it would pause and have issues trying to catch up to the data stream or lose some movement data.

As far as the upside down and mirrored thing, It might be something to do with how you have the origin setup. A lot of the k40 machines home to the back left corner, but are set to have the origin in the front left. Might want to just check those out a bit.
 
The upside down and mirrored thing was caused by setting the machine home setting to the left rear. With it set to left front, the image is the correct orientation. I don't know how I got it out of whack there, but I've been exploring Lightburn quite a bit to see if that's what I want to use for images.

The photo I posted is the most complex thing I've tried to burn so far, with all of the shading. I made the image small - 70mm wide by 87mm high - to see the results quicker. Even at that, there were more than 325,000 lines in the gcode file. I heard a pause about 10% into the run, then it ran fine after that. There was never a pause at the points where you see the jumps in the top image, just the motors dragging. On the second run, I still heard a pause about 10% into it, but there was no disturbance in the burn.
 
I have never heard any pauses while testing out photos. The photos that I have run, have had up and about 10 to 20 million lines of code. The photos I have tested are four inches high, bye about an inch and a half wide, and it would be going back and forth until the picture was done. Usually around one to two hours, depending on the number of lines per inch. Where I ran into problems with the motors, was after the last test, and it returned to home rather quickly, and screwed things up.
 
I changed my photo to 600lpi to see how it would generate. I have the image sized at 2.75 by 3.5 inches. The gcode file grew to over 3.5 million lines and 49MB in size. I'll do a test run with it in the morning.
 
How wide is your beam, I think 600lpi is a bit over kill.

I think my k40 is at best .2 mm.

600lpi would = .021 mm line width? Pretty sure the diode lasers have a bigger dot size than that.
 
I haven't been to my shop yet today - appointment with our financial adviser this morning.

On the LPI thingie, the J Tech 2.8 diode laser is specified at a typical 0.01" (0.254mm) focused dot size, with a possible minimum 0.006" (0.152mm). So, if I'm interpreting correctly, the LPI range would be 100 to 166. Lightburn defaults to 0.100mm line spacing or 254 LPI, which will provide overlap of the lines for better detail, I suppose. In any case, 600LPI does seem to be lots of overkill.

An LPI of 600=0.042mm (25.4/600) per line, which is 36 times greater resolution than the 2.8 diode is capable of producing.

In any case, the fun continues after I take LOML to lunch at Ruby Tuesday. :rolleyes:
 
Bill Arnold said:
An LPI of 600=0.042mm (25.4/600) per line, which is 36 times greater resolution than the 2.8 diode is capable of producing.
Click to expand...

I knew you'd be better at the math, LOL.

Not to mention that 600lpi would take a fair amount of time to run.

Just got my dust boot reprinted with a mount for a cheapo 6watt laser I got off of ebay. My Jtech 2.8 quit having the power to be able to even give the wood a suntan.

I'll be posting on that once I get it done, but my brother is coming into town for a week, so I have a feeling I won't be getting much shop time in for a while.
 
Dan Noren said:
... The photos I have tested are four inches high, bye about an inch and a half wide, ... Usually around one to two hours, depending on the number of lines per inch. ...
Click to expand...

Brent Dowell said:
... Not to mention that 600lpi would take a fair amount of time to run. ...
Click to expand...

Dan's photo is 6 square inches. He posted it takes 1 to 2 hours to run at, apparently, 600 LPI.

The photo I tested is 9.625 sq in. At the Lightburn defaults, it processed at 254 LPI. It ran in 15 minutes.
 
Last edited:
UPDATE:

I ran the same file again and watched the bCNC screen as it started. The progress bar at the bottom showed it loaded 9999 lines of the 325000+ the file contains. When it got to line 9999, there was a pause, then the bar reset showing all 325000+ lines loaded and it finished printing.

Also wanted to see what happened if I doubled the resolution to 0.050mm line width. The file jumped to 650000+ lines. I loaded it and the progress bar again showed 9999 lines initially. When it got to line 9999, it paused and shut down as if finished.

Sounds like it's time for a conversation on the bCNC or Lightburn forum!
 
Last edited:
Dan Noren said:
Now there is where we differ Bill, I use lightburn to generate and save the gcode, and I use ugs to run the gcode.
Click to expand...

I corrected my comment in post #16. I use bCNC to send gcode, as I stated in my opening post. I'm evaluating Lightburn as a vehicle to generate gcode for photos.
 
You must log in or register to reply here.
Top