User Tools

Site Tools


Laser Cutter Technical Details

Technical details about the laser cutter itself.

Project Champion: AJ Fisher (contact via the mailing list or find him in person)

This link is to the documentation for the current upgrade of the laser cutter: project:cchs:laser_cutter:laser_cutter_upgrade. This will be integrated into the wiki when the upgrade is complete.

Current Known Issues

Circle/Arc Segmentation

Currently arc segments shorter than 1mm are cut as straight lines. This can be a problem for small features, for example small circular holes are cut as polygons where each side is 1mm long.

Short time workaround is to convert any small arcs to individual linear segments in CAMBam (ie do the curve interpolation in CAMBam instead of in the Marlin firmware). FIXME: Can anyone document the exact CAMBam process to do this? I think it's the opposite of “Arc Fit”, possibly just “Convert to Polyline”…

The long term/permanent fix is to change a compiled-in setting in the Marlin firmware, MM_PER_ARC_SEGMENT. The setting is currently 1 for 1mm minimum arc segments, but it should be 0.1 or so.

Just Want to use the laser cutter?

There are two other laser cutter pages:

  • laser_cuttergo here if you're a first time laser cutter user.
  • usage_details ← go here if you want more detail on using the laser cutter.

This page is for details on hacking the internals of the laser cutter.

CAMBam Processing Tricks

Multiple Passes

The laser can cut in multiple passes over the material, to keep a higher feed rate.

To do this, when working on the “Machining→Engraving” section in CAMBam, set the “Target Depth” to a multiple of the “Depth Increment” setting. For example, if the “Depth Increment” is set to 0.4 then set the “Target Depth” to -0.8 for 2 passes.


Current firmware is maintained here:

It is based on the 3d Printer firmware “Marlin”.

Source at - README includes a Guide to configuring and compiling.


Marlin firmware GCodes (non-standard)

Linuxcnc also has a good page on gcode at

  • G0 X0 Y0: Go to origin (X:0, Y:0)
  • G28: Go to home position (X: -12, Y: -14)
  • M114: Output current position
  • M3, M4, M106 - Turn on the laser. All identical in usage. S-code modifier for laser intensity.
  • M(3,4,106) S(0-255) Turn on and set laser intensity. 0 lowest, 255 full on. Example: M3 S128 = Laser on, half intensity.
  • Laser probably won’t turn on under ~S25, but this is not a substitute for M5/M107.
  • M(3,4,106) by itself is equivalent to M3 S255 = Laser full power on.
  • M5, M107 - Turn laser off. This should be in your header and footer of every design. It should be very near the start and end of your G-code file. Check it is there. Now check it again.
  • ! M3/4/106 S0 DOES NOT EQUATE TO M5/M107. Use M5/M107 always !
  • G92 - reset the laser’s 0 point. Move the Laser to your new X,Y position and enter the command “G92 X0 Y0” to set that as the current “home”.



FIXME: These updates were current a couple of years ago, lots of changes since then. Can anyone update with a summary of how the Red Sail has been modified?

Laser has been basically aligned.

During the process of trying to update the firmware, the controller board was bricked.

A group decision was reached to re-engineer the electronics, basing the new electronics on the LaserSaur design (LaserSaur compatible from an electronics layout, and therefore capable of using the LaserSaur code and toolchain).

New electronics have been designed by Luke Weston. These are currently being produced for CCHS by Freetronics. The cost of producing the electronics was covered for by funds raised from members specifically for this project.

Tech Specs of Red Sail M500

NB: Most of the below information dates from when the laser cutter was originally acquired. The machine has been heavily modified since then.

We have acquired a decommissioned, second-hand Red Sail M500 laser engraver. The M500 uses a 50W (optical power) carbon dioxide laser tube, and has a usable working area of 500 mm x 300 mm.

The manufacturer's specifications are here:

  • Control system: 32 BIT DSP
  • Working size: 11″x 19″(300mm x 500mm) with double side open door
  • Up and down work plate: 8″ (200mm)
  • Laser tube type: CO2 Glass sealed laser tube
  • Laser power: 50W
  • Driver type: Micro Stepping Motor
  • Max speed: 800mm/s
  • Resolution: 1000 DPI
  • Location precision: <0.01mm
  • PC Interface: USB2.0 and USB Disk
  • Software: Easycut 5.3
  • Image form: HPGL, BMP, GIF, JPG, JPEG, DXF, DST, AI
  • Warranty: 12 Months (3 months for Laser tube and mirrors)
  • Machine Dimension: 1150*670*890 (mm)
  • Package Dimension: 1270*790*1150 (mm)
  • Package: Standard Wooden case for export
  • N.W. : 85KG
  • G.W. : 125KG


Unit set up, with the cooling pump etc. attached.

Cooling water is supplied from a large tank of water containing a submersible pump (I guess it doesn't strictly have to be a submersible pump, it could be a pump outside the reservoir), and the cooling water output is returned to the reservoir, which should be of a sufficiently large volume that it stays at an appropriately low temperature.

This heatsinking reservoir and pump is part of the system which we will have to set up.

Note also the extraction fan on the back of the unit. Extracting the air from the cutting area during operation is very important, otherwise smoke and gases from the burning of the work material can easily cover the sensitive laser optics and stuff them up. This is especially important when cutting halogenated plastics such as PVC, since the hydrogen chloride in the resulting off-gas is especially corrosive to the laser. (Some people say that halogenated plastics should not be cut at all.)

Usually, air is drawn down through the “honeycomb” structure that the work piece sits on and out through the extraction fan, which both removes smoke and fumes and helps to hold the working material in position. Sometimes, an “air knife”, consisting of a jet of compressed air escaping through a small nozzle, is in front of the lenses and optics, keeping any dust or smoke away from the optics and preventing any contaminants from settling on the optics. This is what the air pump input depicted in the above photo is for. A small air compressor is also needed, along with the water pump, to supply the air.

Note that Andy's photos show that there is a blue air hose that enters the laser cutting head area, where the output lens is, but this hose is broken off the little hose barb. That's what this air jet is for - to protect the output lens. This needs to be fixed up. On the back of the unit there are three mains outlet sockets - one Type I Australian (also used in China, I think) one, and the two international-combination Type A/C sockets. These three sockets are supposed to power the extraction fan on the back, the water pump, and the air compressor. We could just use plug-in international plug adapters on the two A/C sockets, but I think that's a little bit dodgy, especially since they're unearthed. Perhaps these should be rewired to a pair of new earthed Type I sockets on the chassis, so that the air pump and the water pump can be plugged in reliably and with their earths connected.

Here are some references that deal with beam alignment:

This one seems generally useful, too:

The mirrors in the beam delivery system are made of a material that reflects the 10.6 micron (far infrared) light from the CO2 laser very well with no absorption… a wafer of silicon with a sputtered reflective coating of silver on the surface is a common example of a mirror material. There is usually a thin dielectric coating over the surface of the silver to help protect it (zinc selenide is the usual material, one of few materials that is really transparent at this wavelength. ZnSe is also the material that is used, usually, for the lenses and other transmissive optical elements at this wavelength.)

The mirrors should be handled with great care so as to not get any fingerprints or contaminants on their reflective surface, as these kinds of things generally need to be cleaned in special ways so as not to damage them.

At this wavelength, the beam is totally invisible, both to the eye and to most laser power meters and similar instruments. The common trick of using an ordinary visible-spectrum CCD to monitor the presence of the beam works for near-visible IR from laser diodes and IR LEDs operating at about 800 nm, but it will not work at all at 10 um. The simplest way to find out where the beam is is to carefully put a piece of paper in the beam (taking care to avoid the potential for invisible but dangerous scattering of the beam), and after an appropriate length of time, the burn pattern tells you where the beam is.

This document describes (with pictures), the alignment of a CO2 laser using a bit of thermal-printing paper:

A Metcard, which is made of thermal-printing paper, might work well for this purpose.

I have a little bit of past experience using a similar commercial laser engraver/cutter of about the same size, (not made by Red Sail) and we used CorelDraw to control it. (This was a university-owned machine, and it wasn't really my business to stuff around with the existing software system control PC that was set up and working with the machine, so I didn't really play with it in depth.) We simply used vector artwork drawn in CorelDraw, with one of three colors - red, blue or black. The three colors specified whether the laser was to either cut, engrave, or burn imagery onto the surface of the material along that line. When using the laser in the “surface-burning” mode, it could be used to “print” effectively on materials such as plywood, without engraving into the surface to any noticeable depth. Filled-in regions were made by raster-scanning the laser across the region.

Imagery from CorelDraw was sent to the laser by some sort of export or output… it may actually have been a printer driver, and you printed out to that device to talk to it. I can't remember now, to be honest.

I hope we're not locked in to using Windows to operate the machine's software… although I have a suspicion Windows will almost certainly be required (running on a VM should be fine).

The commercial laser engraver I've worked with before had a little 5 mW visible red laser diode with the beam coming out through the output optics, co-axial with the beam from the CO2 laser. This was used to see where the beam was going, and to get the focusing and so forth correct, so the beam was focused appropriately on the work piece, before turning on the CO2 laser. The CO2 laser was interlocked to the enclosure cabinet (i.e. no laser power with the enclosure open), but the low-power visible diode was not interlocked, so the work piece could be put in and the laser focussed to the right height with the enclosure open and the alignment laser on. But I'm not sure if the Red Sail unit has one.

The relatively long-wavelength 10.6 micron (far IR) output from a CO2 laser will not usually penetrate beyond the first few layers of skin. Thus, it will create a minor, shallow surface burn that will heal well. Still hurts, though. To give you an idea of what a laser of about this size does to skin, refer here:

If a scattered or reflected beam is observed, the 10 micron light can't reach the retina at the back of the eye, since it is absorbed by the vitreous humor. Exposure to the beam will still potentially severely damage the eye, though, by burning the cornea at the front of the eye - especially at the high power levels usually associated with CO2 lasers.

For 50W of optical power, the electronics must drive 5x-20x that much power into the laser tube, and the 50 W laser tube might be typically running at a voltage of about 10-20 kV at a current of roughly 30-80 mA. (The starting ignition pulse for the laser tube would typically be about 30-40 kV.) The overall efficiency of the laser tube will typically be about 5-20% for a CO2 laser.

The high-voltage high-energy electronics used to power a gas laser is generally more dangerous than the laser's radiation.

The efficiency of a CO2 laser is actually very high compared to other gas lasers, which is why they are the most common laser type for high-power cutting and materials processing. Efficiencies of 0.1%-0.01% are common for He-Ne lasers, Ar ion and Ar/Kr ion lasers. (My Ar/Kr ion laser draws about 1-2 kW out of the power point, for about 250 mW optical output, and the rest of the power keeps the room warm.)

Commercial laser engravers are generally not considered dangerous in terms of laser radiation… because the laser is within a completely enclosed, interlocked enclosure, so the beam can't get out into the room, unlike (for example) research lasers, which are relatively dangerous because the beam freely comes out into the room. But if you're taking the unit apart and operating the laser with the enclosure dismantled, then you're potentially exposed to the laser beam, which you aren't during normal operation. Although almost all common materials will absorb and block 10 micron light, pretty much any surface the laser beam hits, even specular surfaces, will potentially reflect it quite well. As with all Class IV lasers, as a general rule, indirect, reflected or scattered light from the laser is potentially dangerous at such power levels. Scattered or reflected beams can still have potentially dangerous power levels.

Therefore, running the laser with the enclosure taken off and the beam exposed should be avoided if at all possible. However, for alignment of the optics, it will probably inevitably be required, just a little bit.

It's worth reading all the following posts (mostly from CNCZone and similar sites) regarding the M500. A couple of people have not-very-positive things to say about it:

Also, for a wealth of further background reading on laser technology, Sam's Laser FAQ is always worthwhile reading, if you're bored:

project/cchs/laser_cutter/technical_details.txt · Last modified: 2015/10/07 15:25 by