K40 Laser Conversion

I picked up a Chinese K40 laser cutter a few years back. I had a little play with it and decided the software was abysmal, the controller was horrible and it was grossly unsafe. I stashed it on the ‘forgotten projects’ shelf and left it there.

Now, since our hackspace is still closed due to Covid-19, I’m really missing our nice Epilog laser cutter, and I have all kinds of projects on the go that really need a few small pieces making. I could outsource them – but where’s the fun in that?

So, it really is time to break out the K40 ‘blue box’ and get it up to scratch.

A look Outside

Here it is in all its, ahem, glory.

That orange window is just coloured plastic. I have no idea if it specifically absorbs CO2 laser wavelengths, but I’m going to replace this with window glass, with a partially reflective film coating. I know for sure, this will block any glancing laser reflections. CO2 lasers cannot penetrate glass, only mark it.

Laser Bed

Laser bed view, with optical end-stop held for testing

The ‘bed’ was removed for to allow measuring the maximum axis travel. There is no Z-height adjustment, so focusing the laser is not possible. I will need to construct my own bed from either expanded metal mesh or honeycomb material. I will be able to make the bed move vertically and in a level manner. It is important the bed is level with the axes, to ensure the laser operating distance is consistent as the axes move.

Quite a bit of mechanical work to do there, and not much space to do it in.

Control Electronics & PSU

Laser power supply. Super dangerous this.

Ugh, where to begin? Too much mixing of line AC and low voltage going on. The earth connection from the laser power supply is screwed to the painted enclosure panel. So not actually connected to it. There is no earth bond to other enclosure sections, or the metal lids.

No interlocks on the access doors, so you can merrily poke your fingers in there whilst it’s running. No shielding over the line input to the 20kV power supply. No strain relief, no grommets where high voltage lines go out through a hole in the metal, etc. This is only a brief summary, it is worse than this, the more you look, the worse it gets.

After stripping off the FPC connectors for the optical end-stops, I threw the original controller PCB in the bin. Somehow in this mess, I need a new front panel, a controller for the motors and laser modulation as well as lots of microswitches on the doors. Perhaps I’ll use hall-effect switches and magnets since I suspect the flimsy steel doors won’t have enough weight to them to reliably close against a micro switch.

Laser Tube Bay

The business end – the CO2 laser tube and primary mirror assembly

So, the 20kV supply connection is just wrapped around the tube and hot-glued onto the power connection? Awesome.

The water cooling loop is literally a couple of pieces of flimsy silicone tube pushed through some holes in the case. No clamping, no strain relief, no check valves, or flow indicators. Nothing.

There is no tube temperature sensing. Nothing to stop you from putting your face in the beam path. No interlock on the door.

The water cooling system is simply a pond pump in a bucket of water. No filtering, no water chiller, no temperature sensing, and no safety cut-out should the tube overheat and crack, shorting 20kV into the cooling water. Scary.

The Plan

  • Add an open-source controller, likely just an Arduino Mega running GRBL controlled with LaserWeb on the host side
  • Enclose the laser supply connectors
  • Re-design the front panel
  • Add interlocks on all doors
  • Add decent earth bonding everywhere
  • Do a better job of the water connection to the tube
  • Build a return-flow water chiller with closed-loop temperature control
  • Add tube thermal sensing
  • Create a level, adjustable-height bed
  • Add an air-assist jet to the focussing lens assembly
  • Ditch the extraction duct and fit a more powerful fume extraction

Not much then 🙂

From the parts bins

A few parts I had lying around

I had a good rummage in my parts bins and found a few bits I could use to help keep the cost down.

On the left is a 120x240mm radiator, and a pair of old 120mm fans. The latter are powerful and noisy, so I will need speed control there. I also found a fused IEC inlet, an illuminated switch, and a few couplers.

On the right, I found an Arduino-compatible RAMPs-type shield, which can host some stepper motor drivers. This will be enough for controlling the machine. I also found a small air pump (a rotary type) and an air filter, which may or may not be good enough. I even found a couple of air/water valves and some micro switches.

The next job is to hit eBay and find the rest of the parts I need.