Rounded Rectangle: The New Laser Project (3)


CNC is only limited by our imagination.

Rounded Rectangle: CNC Home.
Rounded Rectangle: Construction 1.
Rounded Rectangle: Construction 2.
Rounded Rectangle: Construction 3.
Rounded Rectangle: Construction 4.
Rounded Rectangle: Construction 5.
Rounded Rectangle: Routing & Cutting.
Rounded Rectangle: Vinyl Cutting.
Rounded Rectangle: Lithophanes.
Rounded Rectangle: Engraving.
Rounded Rectangle: Making a Vinyl Cutter.
Rounded Rectangle: Soft & Hard Issues.
Rounded Rectangle: Micro Stepping.
Rounded Rectangle: Bob the Mill.
Rounded Rectangle: Making Printed Circuits.
Rounded Rectangle: Making a 4th Axis.
Rounded Rectangle: Tool Position Setting.
Rounded Rectangle: Low Power Lasers.
Rounded Rectangle: Gallery.
Rounded Rectangle: Related Links.
Rounded Rectangle: Site Map.
Rounded Rectangle: Sam.
Rounded Rectangle: Optic.
Rounded Rectangle: The X Files.
Rounded Rectangle: Bertie.
Rounded Rectangle: The Rover Jet Engine.
Rounded Rectangle: The RF Laser.
Rounded Rectangle: Various Projects.



copyright (c), all rights reserved.



There are dangers associated with this project of which you should be aware.


1) Extremely high voltages are involved which, given the right conditions, could cause electrocution.

2) The laser beam will burn and could cause serious personal injury, especially to eyes.

3) Many materials when heated or burned produce fumes, vapours or particles which are extremely toxic.

4) Combustible materials may catch fire when being cut with a laser.


Do not attempt this project unless you fully understand the risks involved and are satisfied that you can implement all necessary safety precautions to prevent injury to yourself and others.


Switching a laser under Mach control.


With a laser the depth of cut is related to the time of exposure so that, for the same power setting, slow federates will cut deeper than fast federates. If there is any delay between the laser being switched on and the subsequent axis movement a deep spot is produced at the beginning of the kerf and likewise if there is any delay between the axis movement stopping, at the end of the kerf, and the laser being switched off another deep spot is also produced. This effect seems to be no problem when cutting right through or when profile cutting but it does cause problems and creates poor looking work when engraving (for example, a fine line engraved into a soft wood appears to have a dot at each end of the line). So to produce neat work, a method of rapidly switching the laser coincident with axis motion starting and stopping is necessary.

The existing spindle commands M3/M4 and M5 are really not suitable, mainly because of the delay they exhibit (after all they are designed to switch a spindle, which already has inherent spin up / down delay). Likewise the M7/M8 and M9 commands have similar (albeit not so much) delays. A command set which shows much less of a delay is the undocumented M11P1/M10P1 combination which will switch the Output #1 on/off in time with the next axis movement. The M11P1 turns the Output #1 on and the M10P1 turns it off - a different Output # can be switched just by changing the P#.

Because Mach is a constantly evolving product - some of it’s routines / subroutines may be rewritten with each new revision in order to create improvements in both speed and function, however, in recent versions changes have been noted in the way Visual Basic routines, in particular, behave (what used to work perfectly now needs some re-coding to work properly again – the Auto Tool Zero macro is a typical example of this).

Recently I noticed some unpredictable behaviour with the M11P1/M10P1 command set and found consistent results could no longer be achieved with my Mach R3.042.040 version. Whilst these inconsistencies were being investigated, in an attempt to establish the cause of the problem, the complete solution had not yet been found so it’s onto ‘plan B’ for a more reliable method of control. I now seem to have gone full circle here and ended up looking back at using the Z axis direction pin to control the laser (as I did with the low power diode laser) and this was probably the best and most reliable overall option. The technique here is to edit the existing GCode so that the +Z moves are just one step (in my case +0.002) and the –Z moves are also just one step (-0.002) then set the Z axis direction pin to the laser trigger pin allocation. The time delay between the Z axis completing the two virtual steps and the laser switching is so small that it can almost be considered instantaneous and when I return to engraving the fine line in a soft wood there is no longer a dot at each end of the line.

An alternative option, which is also worth investigating, is the B axis direction pin. This is probably much easier to implement where the B axis ‘steps per’ are set to 1 and the B axis direction pin is mapped to the laser trigger (pin 16 in my case) then any GCode is edited to replace the Z axis moves with B1 to turn the laser on and B0 to turn it off. Again this command can almost be considered as instantaneous and tests have shown it to have equal results to when the Z axis direction pin was used.

After some discussions (Jan.2012) Artsoft and the software team have re-looked at the Mach3 code and have successfully restored the M10P1/M11P1 command set to it’s previous glory. I have now had the chance to carry out a number of tests with the latest revision R3.043.056 and can confirm that there are now no noticeable delays with the laser switching times and that performance is equal to the results I obtained by using an axis direction pin. The first advantage I have found from returning to using an Output to control the laser is that, in an emergency, the software polled EStop will always turn the laser off.


Finding M codes that are not commonly used for other functions was quite a concern during the early development of Mach4 and after some discussions between the various interested parties it was decided that the commands M11P1/M10P1 should be replaced with M62P1/M63P1. So for Mach4 use the command M62P1 will turn on Output #1 and the command M63P1 will turn off Output #1 both (as with Mach3) taking effect at the instant of axis movement.

As mentioned elsewhere Tweakie is not really suited to raster scanning work but it can be done if you are prepared to wait forever and here is the proof. This image was lightly burned into the surface of a piece of obechi, it is approx 150mm x 150mm and took around 40 minutes to complete (not quite forever but it seemed like it).

To produce this work I used the Impact/Laser Engraver plug-in within Mach3. It is probably better suited for impact engraving than it is for laser use but it does work and the results are not all bad. The plug-in converts the image to 8 bit greyscale and allocates a number of dots or impacts to each of the shades - unfortunately I have only been able to make use of any 5 shades of gray within the scale at any one time. This is probably due to the response time of my psu or possibly even the laser tube itself which is not reacting at the slower prf end of the scale.

The variable pulse repetition frequency system used here has to be for impact engraving but for lasers a pulse width modulated output with a reasonably high repetition rate would work a whole lot better.

However, I must not complain, this plug-in has been provided, free of charge (together with it’s source code for those who wish to make changes and are good with C++) and thanks to some recent changes many of the original bugs have now been removed.



61) My first attempts at cutting acrylic resulted in relatively poor edge quality but after I had seen videos of professionals cutting at a much slower federate, than I had been using and achieving near perfect results I reduced my federates and power settings - the result, well it’s 100% better. It would seem that although I can vary the speed and power settings and although the power density may well be the same between (high power / high speed) and (low power / low speed) the thermal transfer rate of the material being cut is a constant so all materials will have a preferred cutting / engraving speed, acrylic being no exception. These little bikes were cut just for fun, like most of my stuff, they have no practical purpose.

62) Not sure of the type of wood I used but these little ‘interlocking construction’ boxes were easily made using a nifty little program, found on the net, which is available for download here . Excellent program Rahul, it was fun to use.

63) Now this little bike is entirely different, it is edible! and was made with cane sugar, in two layers, using a technique known as ‘selective laser sintering’. (more to follow on this process once I have constructed the necessary parts and jig to contain the fusible material).


Laser Cutting Acrylic


Cutting acrylic is probably a subject which could fill a whole website so I will try to stick to the important things. Not all acrylic is the same and the cast stuff is considerably better to use than the rolled / extruded stuff. Even then, there is still quite large differences in the properties between the products from different manufacturers resulting in different laser settings having to be used for each.

I have found that to obtain the best results, acrylic has to be cut slowly using the minimum amount of power necessary and the work has to be supported in a manner that prevents reflections from the table or supports which would otherwise spoil the finish on the underside. The air-assist needs to be kept to the minimum flow rate that is necessary to protect the lens or an ‘air-knife’ which does not direct the airflow at the work surface would probably be even better (if air is directed into the kerf it can cool the edge too quickly resulting in a milky colour as opposed to crystal clear).

When cutting, the laser vaporises the material in the kerf and melts the material either side and it is this melting which, providing you get the federate right, creates the polished edge. There is a down side to the melting as it makes the cut slightly narrower and can sometimes make it difficult to remove the finished part from the stock - it doesn’t actually bond it just has very small clearance. Increasing the laser power setting will increase the width of the kerf and increase the final part clearance if it is found necessary.

Software is a problem area as they all seem to treat curves and circles without the proper respect and the final resolution of the edge is generally poor. I did a lot of experimentation by cutting out 50mm diameter discs from 6mm acrylic and then comparing the edge finish - the results, in a way, were rather surprising. A typical, drawing program DXF, processed into GCode with LazyCam produced very poor results yet a circle drawn from 128 straight lines saved as a DXF and given the LazyCam treatment was pretty good. Using more lines approx. 400 produced excellent results. And as I would expect, the Mach wizard ‘cut circle’ produced the best results as the GCode created here is a G2 circular interpolation which gives an extremely accurate approximation of a circle.


64) Another raster engraving, this time in clear acrylic. Incidentally, this is my first encounter with the elusive problem commonly referred to as ‘banding’ (the visible, horizontal, contrast variations or bands). There are many theory's as to exactly what causes this phenomenon and there may, in fact, be more than one cause (different laser manufacturers even disagree about the cause of this) but this is something I intend to get to the bottom of.

65) Mayan (Aztec) calendar laser cut into Traffolite engraving laminate.

66) “His and Hers”. Lasers and acrylic were just meant for each other.

67 - 69) Printed circuit boards, ready to etch (see “Making Printed Circuits” for further details of the process used here).

Rounded Rectangle: 1
Rounded Rectangle: 2
Rounded Rectangle:



These are some of the test pieces which show the near perfect edge finish that can be obtained, straight from the laser, without any polishing or re-working needed.





70) CO2 lasers are pretty good at ablating inks and dyes and especially good at removing the dye layer from anodized aluminium. This image is about 60mm x 90mm on a scrap piece of black anodized and for my first attempt at this it turned out surprisingly well.

71) Gearotic Motion is a serious software program for creating many types of gears in various formats including DXF and GCode. There is also a lighter side to the software which allows for the creation of non linear motion such as with these strange gears which have been laser cut from 6mm acrylic. Hard to believe they rotate and mesh perfectly but they do.

72) A few more, non-linear, laser cut gears. These have been mounted onto an acrylic base.



How fine a line can a laser draw or engrave ?


A focussed laser beam has a minimum diameter which is related to the focal distance of the lens that is used, the smaller the focal distance the smaller the spot size and visa versa. With a 70mm focal distance lens the minimum spot size is approx. 0.1mm and it would be reasonable to assume that this would be the width of the finest line that the laser is capable of producing. Surprisingly, perhaps, this is not the case and although I have not established the finest line that can be produced it can be considerably narrower than the beam diameter. This only occurs with a moving beam and it is a function of the relationship between power and speed – the longer a point, on the work surface, remains in the beam of the laser the greater it will be affected by the radiation and visa versa. Looking at the circular footprint of the focussed beam it will be evident that a line through the centre is longer than a chord, therefore, a travelling beam will always impart more energy to the work at the centre of the line than it will at the sides. If the laser power is carefully reduced or the federate is increased there comes a point where only the very centre of the beam has any major effect on the work. ie. The line drawn becomes narrower than the beam diameter.

I have always been fascinated by the drawing of M.C.Escher and the lizard is probably one of his most famous tessellations. These little chaps are about 50mm across and have been cut from 3mm frosted acrylic, just for fun.

For any that are interested, the official M.C.Escher website is here


Just a trial to get a feel for the problems encountered when inlaying wood into wood. The insert was cut from 2mm Obechi and the pocket into a leftover piece of Teak, both parts being covered with paper Transfer Tape (the stuff used for applying vinyl signs) to minimise surface discolouration from smoke etc. The insert was fitted using PVA and once dry the paper removed and the insert sanded flush. Not bad but not perfect either - the blackened edges caused by the laser are quite clearly still visible in the finished work.

Never leave an operating laser machine unattended.


Rounded Rectangle: 4
Rounded Rectangle: 5