Thanks for the reply and discussion. I think this can make your product even better.
I am using a ComMarker Omni X system, with the 150mm lens. In the USB stick distributed with the machine, there was no configuration file. As recommended by ComMarker, I had to do the lens calibration myself, using the Lightburn program “calibrate galvo lens”. This worked fine, and I was able to correct for scale, bulge, skew and trapezoid.
In this note, I would like an additional correction parameter for angle non-linearity, having to do with the galvo angle and the distance on the build plate being slightly non-linear, scaling not as theta but tan(theta). This is a non-linear correction to x from x (going as x^3), not as a correction to x from y, as in bulge, skew and trapezoid.
Here is a photo of the test calibrator I made, which is similar to the one in Lightburn but with more distances (1cm spacings) and labeled to make things easier to measure. Note, I designed the round holes to be cone shaped into the plastic so that it is easy to insert the caliper points and measure the distances more accurately with a caliper. (I’ve check for no bias here.) When I measure the spacing at the center, I find 9.9 ± 0.05 mm, whereas at the edge I find 10.6 ± 0.05 mm. You can even see by eye that the spacing is greater on the outside.
Because I am trying to use Lightburn for laser machining of a flex circuit board, I would like to calibrate with this additional correction, which seems to be a simple correction from distance proportional to theta to tan(theta), for both the x and y axis. I have also figured out a way to use my calibrator to iterate on the Lightburn “calibrate galvo lens” feature to do the lens correction even better, using this precision measurement with calipers.
In the Lightburn program, I looked at the “prefs” file where all the calibration information is stored, and the only reference was to the “height = 150”. I have checked by changing this value that it tells the program the size of the bed. It seems the program does not know the lens height. But maybe all that is needed is a rescaling of the internal galvo theta from theta to tan(theta), or maybe with a scaling factor that changes the dominate theta^3 term in its Taylor expansion. I am assuming the calibration program gives the 150mm size of the bed, after which it can compute something like the scan speed.
I hope you can help here. I think with this correction and caliper calibration one can get quite accurate positioning of a pattern, opening up new applications for systems that are not too costly. I can share info about details (files, purchase list etc) if that is interesting to you.
Thanks
John
