Upgrading Liaocheng KT-1060 (M2 Nano) to 32-bit GRBL for Lapida Business

Background: I am reviving a Liaocheng KT-1060 CO2 laser (1000x600mm) for a Lapida (tombstone) engraving business in Cebu, Philippines. The original LiHuiYu M2 Nano board is confirmed dead after bench testing; it fails to register via lsusb or dmesg on my Debian (ThinkPad T480) setup.

The Plan:

  • Hardware: Replacing the dead M2 Nano with a 32-bit GRBL-compatible board (e.g., Monport K40+ or MKS DLC32).
  • Repairs: Restoring rat-damaged wiring with 20 AWG stranded wire and replacing brittle air assist tubing.
  • Software: Switching to LightBurn on Linux to utilize advanced dithering (Jarvis/Stucki) for high-detail granite portraits.

Goals:

  1. Achieve native LightBurn support on Debian without proprietary dongles.
  2. Leverage 32-bit processing for smoother, higher-speed stone etching.

Looking forward to any tips from the community on optimizing LightBurn settings for polished black granite!

I replaced my Ruida with a DLC32 board. Got it running OK, but functionality of the pulse was not there, that had to be generated with other hardware.. Swapped it back after a few weeks and pretty much forgot about it.

Sounds like you have a project ahead of you. I’ve seen some rat infested cars/boats, I’d hate to have to rewire any of them.

What kind of laser power supply (lps) or other supplies are in the machine?

Might want to post some photos… lots of diy types here. :wink:


32 bit machines do have a true 16 bit hardware pwm generation, of course it’s adjustments can only be in whole steps.

If you have a 256 or 1 byte, your 100W machine (idealistically) would give you a step of 100/256 = 0.390625 watts/step. Conversely a 16 bit machine gives you 100/65536 for a step of 0.001525879 watts/step.

A 1% step changes the pwm enough to change to change the pwm output by 1%. For an 8 bit machine a 1% change is 2.55 steps in the control, but you can’t have fractions, so each step is 2 or 3, somebody has to truncate it or round it up before it can be used.

Same 1% change with 16 bits results in a change of 655.36 steps to make that 1% step in pwm.

Also these are the number of based on using the co2 across it’s complete range, which usually starts around 10% power. That takes 10% off one end. You can’t really use 0, so that’s off the table. Unlikely you’ll use fraction in the power percentage. All percentage end up as integers, as far as the machines I’ve worked with.

Also, co2 glass tube machines are not linear, they are in a class called negative resistance devices.

I really doubt you’ll notice a difference, control is there to a certain extent, but the difference of even a few watts would be difficult if not impossible to detect.

Don’t be crushed when it’s not really as adjustable as people hope. Most materials don’t have a range for this to be a real benefit… IMHO.

:grinning_cat:

Hello, thanks for your response, and apologies for the late reply - had to visit the site and take photos. The current M2 board is dead - needs replacement - I need a board that supports LightBurn


You will need to find some documentation for the original board. Although it can be done by chasing the signals down to ensure you have it configured correctly.

Lots of companies make K40 upgrade boards that allow grbl to operate and Lightburn also.

:grinning_cat:

1000w in, 60w out? Is that for real?

Definately not Energy Star compliant!