Benbox Laser Machines/CO2 Lasers/3D Printers

I have grbl-1.1f installed on a couple of machines and today isolated a problem with the firmware. If a limit switch is asserted when  the CPU comes out of reset, the HOME function does not work. I ended up manually moving the rail and resetting the CPU.

I am etching some material that reacts to the laser at low power. My problem is the first pixel on every line, from both directions, is much darker than the rest of the line. I am not sure of the exact length but it looks like single pixel under the microscope.

I have a core-xy etcher that has been in use for over two years. Recently started having odd behavior when performing homing operations. About 50% of the time it works as expected. Triggers the limit switch on the Y rail, backs off a bit, drives the carriage along the X rail till the limit switch trips, and backs off 5 mm. The other 50% it does a small surge and stops wait an "error<-1>" message on the console.

Firmware is grbl-1.1f.

Can replicate in a terminal emulator by typing $h.

I have been experimenting with burning images in scratch paper. Scratch paper consists of 4 layers, paper, clay, crayon, ink. Normal use is using a scratch tool remove the lop layer of India Ink exposing the colored layer below. Some users on a Facebook page tried using a laser with fairly good results. I think most use a CO2 at high speed and very low power at a 600 DPI resolution.

I tried with a 2.5W diode laser. The trick is to apply enough energy to burn the ink but not touch the crayon. Got acceptable results when I filtered the black pixels to be 1 out of 4 with a speed of approx 3400mm/min and a power of 90 out of 255 on the thinner brand of paper and a power of 150 on the thicker paper (ink seems to be different).

I suspect that part of the problem is I do all burns at a 0.1mm resolution and that the grid scheme effectively changed that to something close to 0.20 in X and 0.14 in Y. I also means that the laser is only driven On for one pixel (0.1mm). I need to do the math and tests but suspect that the laser is not fully ON before being turned OFF. Question is how to quantify and control that process. What test patterns and materials can be used? Is there a difference between running that laser as a raster and firing while in motion vs coming to a stop and pulsing the laser.

There are two goals in this quest. First is to making the underlying colors more vivid. Second is to learn how mathematical transforms (filters) affect the end result. After laser etching for almost 4 years (mostly slate & 2 tone wood) I feel like an absolute beginner. I don't know the art, math, or physics necessary to get repeatable results. If there is any interest we could create a category here specific to this technique.

I purchased 4 different brands/type of paper. Have only experimented with two. Both were "rainbow" crayon layer. Attached is picture of first success. BTW - looks much better than the picture

My 2.5W diode had been performing poorly. Replaced it with a 3.5W that has a much larger forcus point. Been doing mostly cutting so switched back to 2.5W. Todaty I wanted to do an etch and did the test burns on cardboard. Came out horrible. Any horizontal lines that were less than 3 pixels wide wide either very faded on just not there. Put the 3.5 back in and repeated test. Again not very good. Did some test cutting and both diodes behaved as expected. Did some test etching and both misbehaved.

This is with grbl-1.1f with laser mode enabled. Looked at the gcode and the speeds are not changing when doing the horizontal line so I don't think it is related to acceleration. 

Wondering if grbl with laser mode on always mucks with the laser power.

Current thoungths:

1. Disable laser mode.
2. Change the gcode generator such that it only emits a Fxxx on a change in speed.
3. Lower both the power and the speed.

Any other?

Attached is the image that I am trying to etch. If anyone does etch please post you results. On my fixture I am burning at a 0.1mm resolution so this image is 110x110mm.

Couple of years ago I built an etcher with a rotary table for etching wine glasses. It has a spring loaded chuck that hold the glass in a vertical orientation. The X axis is the rotation and the Y is vertical. Adjusting the Z (distance from the glass) is manual via a slide mechanism. Last week I added a Z leadscrew and motor. I purchased some short distance (200mm) non-contact sensors and would like to follow the contour of the glass.

Current thinking is to run a distance calibration pass and have the host software do the math to compensate for distance and speed.

I can use either Arduino Nano with GRBL or a STM32 with home-brew firmware. Both have pros and cons. Possible implementations:

1. Add code to STM32 base
2. Change to GRBL-2.0 for ARM.
3. Trim unused code from GRBL-1.1 (ie ARCs, info messages, status reports)
4. Switch boards to an ESP32 package.

Switching to GRBL 2.0 for ARM appears to be the preferred long term solution while the ESP32 with radio link sounds interesting. There is an active ESP32 development effort that I am monitoring.

I prefer 3 for quick testing as the distance sensor has examples for Arduino. Any thoughts on how to get 2K code space from GRBL?

After three years of playing with diode lasers I am finally getting around to determining what is the ideal distance from lens to work. I know it depends on the lens. Originally I was at 150mm, went to 120 for a long time, currently at 100mm. My 2.5W has been under-performing lately and getting a good focal point is difficult.

I have been advising new members to read Ralph's excellent document on calibrating your diode laser using BCL. After searching and some PM's with Ralph it appears I must have been mistaken. He does have a series of video tutorials that cover the topics but not a document.

However, I recall printing one. This was during the early days of BCL development when a number of us volunteered to be  beta testers. As I normally don't use Windows on my home systems, I setup 2 dual boot laptops, one with Windows 7 and another with 10. Those environments are long gone.  If anyone has the docs please post or email.

In the meantime, watch the videos!!!.