Miniature Desk-Top Laser-Cutter

[wefalck]

A few month ago I acquired a KKMoon K4 3W miniature desk-top laser-cutter and it has proven to be a useful investment. Therefore, I would like to share a few operational insights, though you can find a variety of ‘test’ videos and the like on the Internet.

As with many Chinese products of this kind, it comes in various guises and configurations that may be mechanically identical or not. The traders’ descriptions are often somewhat haphazard and also suffer from translation issues. I am not sure, whether KKMoon is a trader or a manufacturer, their Web-site does not actually list these laser-cutters.

Prices between the different offers on the Internet marketing platforms can vary as much as 30%. However, I paid just over 100€, shipping included.

Image of the laser-cutter as advertised

 

 

 

The stated main specification of the machine I bought are

 

 

- Size: about 155 mm x 166 mm x 143mm

- Weight: ca. 600 g

- Laser Power: 3 W (3000 mW) – blue = xxx nm wavelength

- Engraving Area: about 80 mm x 80 mm (3.1" x 3.1")

- Engraving Depth: about 1 mm /0.04" (Adjustable in the range of 0-1 mm)

- Mechanical resolution: 0.05 mm = 512 dpi

- Supporting System: for Windows XP / 7/8/10 and MacOS 10.10 or later

- Supporting Image Format: JPEG / JPG / PNG / BMP

- Connectivity: Micro USB B to USB A (cable included)

- Frame Material: ABS

 

 

The laser-cutting system consists of three main components that determine its capabilities: the mechanics, the control board, and the software.

 

 

Mechanics

 

The mechanical resolution of 512 dpi is not that brilliant, if you compare this with modern scanners or printers, but then mechanics have their price.

The 3 W diode laser has an adjustable focal point.

 

 

Control Board

 

I know next to nothing about electronics and commercial products, such as the control board that is being used in this machine. It would be particularly interesting to know, whether the board could be driven by other types of software. Perhaps someone from the Forum community has insights into this.

 

 

Driver

 

The software consists of two components, the driver and the cutting software itself. The driver is a standard piece of software under MS Windows and either comes with your MS Windows configuration or can be downloaded from the software producer’s Web-site. The driver runs under MS Windows XP/7/8/10. I am using an oldish mini-laptop with MS Windows XP on it. The driver unfortunately does not run under MS Windows emulation Parallels under MacOS 10.7.1, nor under the iOS for the iPad pro. The cutting software, however, seems to run in Parallels under MacOS 10.7.1. It should also run under MacOS 10.10 and higher, but I could not test this.

 

 

Cutting software

 

The cutting software is a very simple piece and is based on bit-image processing. In other words, the image is processed line by line from the top down and whenever a black pixel is encountered, the laser flashes. As noted above, the software can handle JPEG-, JPG-, PNG-, and BMP-files, but not TIFF. Images of up 1600 x 1600 pixels can be processed.

There are three variables that can be adjusted to control the cutting process: the laser power in %, the cutting depth in 0.01 mm increments, and contrast (0 to 256). It is obvious, what the power adjustment does and I assume the cutting depth is determined by the length of the laser pulse. The cutting speed cannot be adjusted explicitly. What influence the contrast setting has is not completely clear to me, as the screen appearance of the image changes, even when I use a 0/1 b/w bit image. In practice, however, it does change the width of the cutting traces.

The image to be cut can be freely moved around the cutting area of 80 mm x 80 mm on the screen.

 

Screenshot of the cutting software user interface

 

Set-up

 

The machine is mobile and in principle does not require any special set-up apart from a flat surface. However, any energy penetrating the material cut will be taken up by the surface on which the machine stands. This means that the material has to be fire-proof. I happened to have a piece of roof-slate at hand, which turned out to be very useful for the purpose. Pieces of marble or tiles would do as well.

The laser beam needs to be focused onto the material to be cut. The machine comes with a piece of black cardboard for the purpose, but this is thicker than many of the materials to be cut. It is better to focus the beam on the material in question. The laser spot is very bright, making it difficult to see, whether its size is minimal. I found it useful to illuminate the cutting area with a strong table lamp so that the contrast is reduced during focus setting.

The material to be cut needs to lie absolutely flat. I have been thinking of making some clamping rails or similar. It turned out that short tabs of cellotape are quite sufficient for the purpose. The small pieces of material are just taped down at each corner onto the slate.

 

 

Cutting times

 

I did not make systematic tests, but the examples shown here took about 10 minutes to cut. I would estimate that covering the full 80 mm x 80 mm cutting area would take in the order of about one hour.

 

 

Steering wheels cut from 0.15 mm thick Canson paper (120 g/m2) (cutting area about 40 mm x 40 mm)

 

 

 

Capabilities

 

Whether a material can be cut by laser depends on a number of properties of the material in question. First of all the material must be either combustible or it must be able to be evaporated. The material must be capable to absorb enough energy to reach its combustion point or its evaporation temperature. Whether a material can absorb enough energy depends in turn on a number of factors.

A key factor is its albedo, in other words, how well the material reflects or absorbs light. Bright and shiny materials reflect most of the light, as do white and light coloured materials. Hence they are not absorbing enough energy. Conversely, dark and in particular black materials absorb most of the light that is shot at them.

Another factor that determines how much energy is needed to combust or evaporate it is its volumetric density. Compact materials with no pores contain more mass per volume than porous materials and hence need more energy per volume to combust or evaporate. The volumetric heat conductivity is also important. If the material conducts heat well, the energy transmitted may become dissipated before it reaches the flash-point or the boiling-point.

While in theory virtually all materials could be cut with a laser, in practice the available laser may just not be powerful enough.

In practical terms this means that it is not possible to cut metal and transparent or translucent materials with this small laser. The 3 W laser just does not impart sufficient energy to melt and evaporate metals. Not surprising though. Plexiglas or tracing paper let all or too much of the light pass and therefore cannot be cut.

Bakelite paper has a high evaporation temperature and is translucent. It can be cut through in thicknesses of up to 0.1 mm, but edges become charred. A strategy can be to only cut part through and then brake off the part along the cutting. This works only for simple shapes with straight edges and not too small parts.

 

 

 

As set of doors (ca. 11 mm high) cut from 0.1 mm bakelite paper

 

White polystyrene is too reflective and is only lightly engraved, if at all. I did not have black polystyrene at hand to try this out.

 

I would abstain from cutting PVC due to the generation of toxic and corrosive combustion products.

 

I have not tried ABS or Lexan, but would expect similar issues as for polystyrene.

 

Celluloid might cut well, if you have a coloured variety. Transparent celluloid, including drafting films such as Ultraphane, will not work. The high flammability of celluloid may be an issue.

 

White paper works moderately well due to its high reflectivity. An important factor is also its weighing and seizing. Weighing with barite or titanium oxide makes it more difficult to cut, as both materials are refractory. A seizing with glue or plastic polymers increases the volumetric density and therefore make the paper more difficult to cut.

Coloured papers and cardboard work best, but thicknesses above 0.5 mm become more difficult to cut. The deeper the cut the more charring of the edges will occur, loosing precision in size and reducing the minimum size of features that can be cut.

I have not had the opportunity to cut wood, but I would expect that low-density woods cut better and then hardwoods. The size limitations are likely to be similar to those of cardboard.

 

Cork should cut reasonably well, but I have not tried it myself.

 

 

Drafting for cutting

 

As for any other ‘machining’ operation, the ‘tool’ diameter is an important consideration. The effective diameter of the well-focused laser-beam is in the order of 0.1 mm. These leads to the rounding of internal corners in this order of magnitude, but the actual rounding depends also on the size of the opening to be cut. Smaller openings may have a more perceptible rounding than larger ones.

In practice, the charring of the edges leads to slightly larger openings than those drawn. Thus the diameter of e.g. holes needs to be drawn 0.1 mm less than required. Similarly, slots should be chosen 0.1 mm narrower than the nominal width.

The laser sends a pulse for each black pixel encountered. When converting vector drawings into bit images, the question arises of the actual size of the parts that appear white in the final image to be used in the laser-cutter. This may depend on the line thickness chosen and the kind of drafting program. I found that I needed to experiment with the cutting parameters (power setting and contrast) and in some cases needed to redraft (parts of) the drawings in order to arrive at the correct size. Several iterations may be needed to arrive at the correct size. This also depends on the material, thicker material requiring more adjustments.

Every part that is black in the drawing will be burned. In order to reduce the laser time and the fumes generated, it is good practice to fill in any empty space. While this would be good practice in photo-etching too in order to safe etching fluid, often this is not done. However, when converting drawings for laser-cutting it is a good idea to fill in the empty spaces.

I use a 2D CAD system for drafting (EazyDraw™). This program allows the drawing to be exported into picture formats such as JPG. The resolution for this step has to be chosen so that the final part has the correct size for a resolution of 512 dpi or 202 pixels per centimetre. This means that a part that is 1 cm long should be 202 pixels wide in the JPG etc. file. In order to reduce the area to be burned, I usually import the image into Adobe Photoshop Elements™ and whiten all the respective areas. Sometimes is also convenient to draw the parts in solid black, which then necessitates their inversion in Photoshop. I typically export the drawings at 1024 dpi and then reduce the image in Photoshop to the desired width in the number of pixels as calculated for 512 dpi after the post-processing has been done. This allows me to ascertain that the drawing has the desired size. In this way it is also easy to produce cutting designs in various scales from the original drawing.

As the cutting happens on a flat surface and there is no mechanical interaction with the material, the cut pieces do not move from their place during the cutting process. Therefore, retaining tabs, as you would need in photo-etching, are not needed and the parts can be completely cut out. This avoids the problem of distortion during separation from the fret, particularly of very small parts.

 

 

A typical JPG-image as used for the cutting process (size around 35 mm x 30 mm)

 

 

 

Safety

 

Lasers are dangerous for the eyes and you are advised to consult the respective guidance on laser safety.

The laser-cutter comes with a green protective glass on one side. I also bought a pair of green safety-glasses for adjusting the laser focus, as viewing the focal point through the shielding glass is inconvenient.

The combustion fumes of certain materials can be a nuisance, noxious, or carcinogenic. In any case they are smelly. As noted above, it is wise to reduce the areas to be burned in order to minimise the amount of combustion products. For certain materials some kind of forced aeration may be needed, or you need to set up the laser-cutter outside.

Some materials may also be a fire hazard. However, none of the materials I worked with seem to have been problematic in this sense. There would not be enough mass to sustain a serious fire, but a fire-proof base is important.

In any case: never leave the machine running unobserved !

On the Internet you can see people, who have encased their cutters and added forced ventilation to it. Whether such arrangement is warranted, depends really on how intensively you use it. In my case it just runs occasionally for a few minutes at a time.

 

 

Conclusions

 

This technique cannot fully replace photo-etching to produce small, complex and delicate parts, but is is a versatile ad hoc option requiring little preparation in comparison. The cost of materials is minimal and therefore that of trial and error. There are no chemicals to manage safely, but fumes can be an issue.

There is no equivalent to the ‘surface etching’ process, parts are strictly two-dimensional. As in photo-etching, there is, however, the possibility to build up parts from several layers.

Metal surfaces and its edges can be made very smooth. Achieving the same effect with paper or cardboard is difficult, even when treated with wood-filler to produce some sort of compound material that can be sanded. In some applications that surface roughness does not matter or may be even desirable.

The mechanical resolution of 512 dpi and the diameter of 0.1 mm of the laser-beam impose limitations to the minimum size of parts that can be produced. Laser-cutting with such small desk-top machine cannot compete with commercial etching processes using high-resolution masks.

In scratch-building, when parts need to be developed as the building goes on this kind of laser-cutting certainly is a useful ad hoc and flexible process.

Edited March 13, 2020 by wefalck

[bruce d]

Thank you very much. That is the most useful description I have seen of how these tools can help a modeller. 

 

Bruce

[modeller_masa]

I've never seen such detailed and comprehensive review. Thank you.

[Justin P.]

Fantastic information and well structured for future updates!   Thanks for taking the time.

[VTHokiEE]

Awesome review! I’ve had a desktop laser cutter on my mind for a little while. I was looking at something like this:

 

https://www.amazon.com/Orion-Motor-Tech-Upgraded-Engraving/dp/B07SQQMN3F

[Gregory]

I have had one of those for a while.  They do some serious cutting.  ( not metal )

You need to be ready to do a little modification to make it versatile, but you can google " K40 " laser modifications,

and see a lot of useful information.

 

I have heard some people having a bad experience with out of the box alignment problems, or not working at all.

I was fortunate not to have any problems other than the learning curve.

[wefalck]

Actually, it is not a K40. The 40 stand for 40W and this is a completely different anyímal.

 

Once I had it hooked up to my wife's old MS Windows XP mini-laptop and downloaded the driver, I was seriously working on it within minutes. The learning curve was not with the machine itself, but hitting the right cutting parameters for the material and drawing in question.

[Gregory]

Sorry for the confusion..  I was replying to VTHokiEE regarding the laser he linked to on Amazon..

[VTHokiEE]

49 minutes ago, Gregory said:

I have had one of those for a while.  They do some serious cutting. 

Thanks, that’s great to know! I don’t mind tinkering (I have a 3D printer that needed a decent amount of modifications). As soon as I can think up how to arrange the garage I think I’ll pick one up.

[wefalck]

I have just added to the original post a screenshot of the cutting software user interface

[mtaylor]

13 hours ago, VTHokiEE said:

Awesome review! I’ve had a desktop laser cutter on my mind for a little while. I was looking at something like this:

 

https://www.amazon.com/Orion-Motor-Tech-Upgraded-Engraving/dp/B07SQQMN3F

 

Looks similar to what I have.  If you get one, and decide it needs "upgrades" <cough, cough> Go here for parts (https://www.lightobject.com/) as they have excellent after market products like tubes, lenses, power supplies, water chillers, etc.    (Note: You will need a way to keep the beast's tube cool.)

[BANYAN]

Great review and guide Eberhard, that is very useful.  I am seriously thinking of ditching the PE process for this wherever I can.

 

cheers

Pat

[wefalck]

I didn't get around to give this a try, but I have the idea of covering some brass with a thin layer of black paint and then burn it away for etching. This would do away with the step of having to create masks. It should work for single-sided etching, but getting the brass aligned properly for burning on both sides would be a bit of challenge.

[bruce d]

Why not? One of the many different processes used commercially to create masks was to apply a vinyl sheet and cut away pieces with a blade in a plotter: exactly like modern vinyl sign-writers.

As for registration, if you have the ability to make a jig and produce holes with dead-accurate spacing, you can do one side of the workpiece and then flip it and the other side will be in perfect register. This is also based on an old commercial etching practice for small workshops like jewellers. 

Of course the laser cutter must have good repeatablity but I assume this is not a problem. 

[wefalck]

Indee, I could make a jig to locate both, the cutter, as well as the sheet metal, but the problem I expect is with the alingment of the images on the screen. One has to load the left and the right sides consecutively. However, with a carefully designed procedure it should be possible.

[bruce d]

9 minutes ago, wefalck said:

the problem I expect is with the alingment of the images on the screen.

In my time as an etcher, when dinosaurs roamed the earth, one of the first registration systems I used was very much like we discussed above. The key to success was that all pieces in the system were consistent: any workpiece would fit the jig and anything registered on the jig would align with any other piece even when flipped over. This worked despite having the added complication of using photographic film that had emulsion on only one side and the final images had to end up emulsion-to-emulsion.

I can see your idea working in practice. The starting point, from which you proceed with all possible accuracy, is to see if your software generates a consistent image. In other words, if you make a template to use for etching masks, will it work every time? If it will, then put some 'X' marks at the same point in each corner. Each piece of brass must have holes of a consistent size to correspond with the 'X's, again only possible with a jig.

Your jig to hold the workpiece in the printer will have pegs in these same locations and must have some way of being located repeatedly in the printer.

The last step in repeatability is not an obvious one but must be considered. The material to be etched must be flat or distortions will creep in. Even a small curl at the edge can prevent a piece from sitting properly in the jig(s) and make a mess of all the work put into getting good registration.

For the mask material, I believe you will be able to simply use a can of spray paint.

 

By the way, you have got me thinking and I will be interested in how you get on if you try this. Maybe I will follow your path ... ? You go first, I will make a cup of tea and watch.

[bruce d]

On reflection, I see this process can be even more simple than I originally thought. In the past for conventional etching processes I had to bring together images and metallic components in perfect registration. To do this I had to juggle photographic material and photosensative coated metal sheets. You do not have these same materials so you do not have the same issues. If your software can produce mirror images that can then be repositioned to allign the 'X' registration marks, it will be possible to make a more simple fixture to position the workpiece for plotting.

2 hours ago, wefalck said:

I could make a jig to locate both, the cutter, as well as the sheet metal, but the problem I expect is with the alingment of the images on the screen.

If you are going to proceed let me know. If the software can't do that trick then it is back to 'eyeball Mk 1' registration techniques.

[wefalck]

The problem is that the cutting software does not allow to superimpose images, it does not have 'layers', as CAD software would have. Therefore, I am afraid it would be 'eyeball MK1'-registration ...

 

If you do surface etching, it is not just mirroring the drawing. You will have to produce two different drawings that have to register perfectly. A very laborious procedure with a lot of scope for error. I usually make one drawing, mirror it and then change it.

[Gregory]

Do you have a lower end 2D photo editing program with layers, such as Photoshop Elements or Corel Paint shop?

Edited March 15, 2020 by Gregory

[wefalck]

I actually do have quite a bit of experience in doing the drawings for etching and have done some reasonably successful etching at home. The drawings as such are not the problem. My main obstacle was always to produce masks in which the black was sufficiently dense and uniform. The wet process part can be mastered, but is a nuisance (though I am actually a sort of chemist by training).

 

I try to begin with that side of the drawing that has most details on it. For instance, the 'outside' of parts with rivets and such surface detail. I complete all the drawings that are going onto one fret. I arrange the drawings and then draw the fret together with register marks (a white dot with a cross-hair) in each corner. All this is 'grouped', as my CAD software calls it. The group then is mirrored. After ungrouping I can edit out all the detail, or add more detail, for the opposite side of the parts. By proceeding in this way, I ensure a (theoretically) perfect register of the drawings.

 

The problem could be in practice, that the cutting software does not align two drawings perfectly to the zero point. Or the somewhat flimsy construction of the cutter does not ensure that the carriage with the laser returns to the same origin within a 0.05 mm. With the small parts I am producing, I need such close tolerances.

 

I have no immediate intentions to try this out, as my needs are more or less satisfied with the pieces cut from paper.

 

The problem with building from original drawings is that you have to develop all the model drawings and in order to keep track of things, this is done in an iterative fashion, iterating between drawing and building. Hence, I have been trying to turn etching into an ad-hoc process, as you would use say a lathe or mill. Otherwise, etching becomes very expensive, as you might not get it right in the first try.

[Ras Ambrioso]

Wefalck: You are not only a great craftsman but also a great teacher. thanks

[Gregory]

On 3/15/2020 at 10:02 AM, wefalck said:

The problem is that the cutting software does not allow to superimpose images,

Have you seen LightBurn

 

It is a very robust cutting program with layer support.  it supports many different controller boards.

 

It has limited editing features, but You can import multiple images and align them very precisely on the cutting grid.

 

It also has a very good tracing tool for defining cutting outlines of objects that are not vectorized.

 

Edited May 24, 2021 by Gregory