1000 mW Laser Engraver

A while back I bought an Neje DK-8-KZ 1W laser engraver. It was on sale, and I was curious about it. At only 1W (1000 mW) I knew I wasn’t going to be cutting anything except paper with it, but I was looking for another way to create engraved labeling on plastic enclosures. Well, it can do that, with some limitations, and it works pretty well on wood, but my biggest aggravation has been just getting it to actually put the engraving where I want it to go, and also deal with the automatic scaling it employs.

The unit itself is very compact, as you can see in the photo below. It consists of an acrylic frame and two stepper motor mechanisms for the X and Y axis (there is no Z axis, it doesn’t need one). The stepper and lead screw assemblies look like they were based on old floppy disk drives, but I can’t tell for certain. Both axis use the same mechanism.


These are the specifications for the DK-8-KZ. I’ve pulled them from the supplied documentation, fixed the grammar and spelling mistakes, and reformatted them for readability.

Supported OS: Windows XP, 7, 8, and 10, Mac OS X
DPI: 350dpi
Laser Wavelength: 405nm
Laser Output Power: 1000 mW
Temperature: 400 C
Image Format: JPG\BMP\PNG
Image Size: 512x512 pixels
Motion control: Raster scan
Input: Dual USB (5V 1A)
Supply Voltage: 4.2-5.5V
Maximum work area: 38 x 38mm
Maximum ON time: < 1 hour

I’m not sure why Linux isn’t supported, since some of the software is built using Qt and MinGW. Oh well, I keep an old Dell alive just so it can run Windows XP for the stuff I get that hasn’t yet moved into the 21st Century.

OK, now let’s take a look at the electronics. The controller PCB, shown in the photo below of the engraver with the back panel removed, is a tiny thing. The microcontroller is an STC/IAP 15F2K61S2, produced by STC in Beijing, China. This is an 8051 variant, similar to devices produced by several other chip manufacturers (Silicon Labs, Cypress, Atmel, Maxim, TI, and others). The USB interface is handled by a CH340G chip, which is common in a lot of stuff these days.


The stepper motors are controlled by four TC117HS motor controller ICs, two per stepper motor. These parts are typically listed as toy motor controllers, and the official datasheet provides no diagrams as to what is really inside these things. I couldn’t find an English version of the datasheet but I suspect that each one is an H-bridge, so two would be able to control a 4-lead bipolar stepper motor, which happens to be what the engraver is using.

Lastly, there are two USB connectors; a Type A and and a mini. The type A is used for power only, whereas the mini is the communications channel between the host PC and engraver. Actually, the use of the type A USB connector is not in accordance with the USB specifications. It should be a type B. That doesn’t hurt anything,  but it does mean that if you ever lose the supplied A-to-A cable, you’ll need to start hunting for a replacement.

So, in summary, the controller PCB has only one easily recognizable part on it (the CH340G), and even that can be a problem with older operating systems. Hence the need for drivers for Windows. Nonetheless, it does work, and while I might not be happy with the closed-source nature of it all, I can live with it for now.

The physical design is interesting in that it uses the minimum number of components possible, and re-uses the mechanisms for the X and Y axis. This is a fixed-position gantry design, wherein the work table is the Y axis, and the laser head is the X axis. The stepper mechanism really looks like a design from an old-style 8″ floppy disk drive, but it may be something made specifically for this application. In any case, it works, and it’s simple. Just a piece of stamped sheet metal, and lead screw, a couple of slide rods, and a mounting plate driven along by the lead screw. The design does not incorporate limit switches–setting each axis to its home position is just a matter of over-stepping it into the end stop and letting it chatter for a bit.

The frame is made from CNC-cut black acrylic pieces, and the T-slot scheme is used with screws and nuts to hold it all together. It’s very sturdy, but the acrylic won’t take a lot of abuse before it breaks (I already managed to break off one of the screw hole alignment tabs). Acrylic can also be a real pain to drill, so adding something to the unit, like a power switch, will be an exercise in graduated drilling with a non-solvent lubricant to make a clean hole without chipping or fractures.

The specs for the engraver claims that it has a working area of 38 x 38 mm, but in practice it has a working area of 37 mm by 37 mm. Fortunately there are large openings in the frame at the front and rear. This means you can mount a work piece larger that the 37 mm Y axis size, but narrower than the width of the openings in the frame. So long as it doesn’t interfere with the frame, and doesn’t weigh too much, you can engrave something on one section of it, reposition it in the Y axis direction, and then do some more engraving. For a lot of the small-scale microcontroller projects I do this is fine. The diagram below shows the work table, the X and Y axis movements, and the work area.


The front and rear openings in the frame are 118 mm wide, and since the Y axis table doesn’t move sideways you could theoretically put a 110 mm wide work piece on the table. Provided, of course, that you can devise a way to hold it down. The method provided with the engraver is a couple of rubber bands (yes, really). The Y axis platform is about 77 mm wide, but the X axis carriage (with the laser head) only moves to within 20 mm of each side. Hence the 37 mm range on the X axis. The same is true of the Y axis, which also has only 37 mm of travel. This means that if you want to do work close to the edge you will need to physically move the work piece and then reset the starting point for the laser. It looks like there’s room inside to extend the travel of the X axis by at least 10 mm on each side. That would help make this engraver a lot more useful. I haven’t taken any measurements, but it may also be possible to use a larger transport mechanism for Y axis.

The laser does have a finned heatsink, but there is no fan to help keep it cool. There is also no fan to blow away noxious smoke when doing a long-step burn. A quick remedy for both situations would be to install a couple of small 5V fans, or put a small desk fan off to the side of the unit and have it blow through the frame, which is what I do. Excessive smoke can interfere with the beam, so keeping it blown away is a good idea.

The specs claim that the unit requires 1A at 5V DC. Neje recommends a 5V USB supply (a wall transformer with a type A socket) rather than attempt to run the unit from a PC, and I would have to agree with this. Some PCs have high current capability on the primary USB ports on the front of the box, but some don’t. A wall-wart power supply is a better option.

I should also point out that there is no power switch on the engraver. There is a button on the top, but that is used to set up for a run and then start the burn. It allows the engraver to be loaded with an image and then used without a PC attached. This might be useful if you want to make a lot of one particular thing, such as at a county fair, or in a knick-knack stall on a dock where a cruise ship ties up, but I personally don’t have much use for it. A power switch would be more useful to me, so I will most likely add one in the near future.

Transferring a simple image to something like a piece of wood or leather is a straightforward process. Load the image into the printer, center whatever you want to engrave on in the middle of the Y-axis table, and light it up. Problems arise when attempting to accurately position the work and get the laser where you want it to go, and then doing that with repeatable positioning.

This engraver doesn’t use absolute positioning nor does it employ vector motion control. It employs raster scanning. Ugh. Not only that, but the software supplied with the engraver will scale the original image down to put into the 37 x 37 mm work area, as shown below:


The rocket ship image is supplied with the engraver on a microSD, which is where the rest of the software resides. And, speaking of software, this unit only works with Windows, so there’s a device driver that needs to be installed before you connect the engraver. The documentation that comes with the unit on the microSD is sufficient to get it up and running. It’s obvious that someone made an effort to create useful instructions. It isn’t perfect, but it’s definitely better than nothing at all. Be sure to look at the pictures, since those are actually a key part of the documentation. There are also some videos on YouTube you might want to look at it (of course, some of these are better than others).

I ran some tests on the scaling to see how it worked. Starting with a 512 x 512 pixel image did indeed result in a 37 x 37 mm engraving space. Actually, it was closer to 36.5, but I think that is due to the raster converter rounding the resulting scaled image down.

In Visio, if the screen resolution is set to 96 DPI (dots per inch) then a 512 x 512 image will be shown as 5.3333 x 5.3333 inches, or 135.466 x 135.466 mm. This is important to keep in mind when attempting to print text using the engraver. If we divide the claimed resolution of the engraver (350 DPI) by the native image display resolution (96 DPI for a Visio image) we get a scaling factor of 3.65. That means that whatever you see in your graphics tool will be 3.65 times smaller when engraved. In an early experiment I created a test image with a 12 point typeface, and the text came out so small it was almost unreadable. The scaling takes a 12-point font to about 3 point size. So, if I want to have a 10 point font on an engraved panel, I need to start with a 36 or 38 point typeface. Unfortunately, the software provided with the engraver offers no help with scaling or dimensioning.

As I stated above, engraving uses a raster scan approach. To set up a run you first drag and drop an image into the main window of the control application’s GUI, as shown below (there is no “File” menu drop-down). Then you transfer the image to the engraver’s microcontroller, set your burn time (the default is 60, but it goes to 240), make sure your work piece is positioned where you want it to be, and click the “Start” button to light it up.


Despite missing some things I would consider essential (like a “File” drop-down) and the ability to do absolute positioning, the control tool is actually fairly intuitive. Note that the buttons shown with grey text in the screenshot above will change to black and be available for use after the image data is transferred to the printer’s microcontroller.

During further testing I discovered that the image created during the scan does not hold its orthogonality. In other words, what should be a perfect square with right-angle corners comes out as a parallelogram leaning slightly to the left. The offset is subtle, but it is a consistent and repeatable error. It’s not readily apparent unless you are attempting to create an image with long vertical lines and right-angle corners; a grid pattern, for example. If I had to guess, I would say that the error is due to an alignment issue with the Y axis mechanism. I’m still working on identifying precisely where the error is occurring. In any case, it doesn’t really effect an image as the error is very subtle and difficult to notice.

If you look at the control application GUI you will notice three buttons in the box labeled “Location Mode”. These are “To Origin”, “To Middle of Image”, and “Carving Preview” (I guess “carving” works, but it strikes me as rather odd–not the term I would have picked). If you use the buttons in the box immediately below, titled “Movement”, you can shift the beam point across the work table in the X and Y axis. Just bear in mind that the default origin is with Y and X at their maximum (0, 0) extents. So from the default origin you can only move right, not left, and down, not up. If you position the locating beam point at the middle of the image (the “To Middle of Image” button), then you can move it in either X or Y.

An aggravating feature is that even if you create a small image (something like, say, 64 wide by 256 tall), the software will insist on centering it. One way around this is to start with a 512 x 512 blank square in your graphics tool and position your artwork within that. Another approach is to create a small image with a minimum margin and then use the movement controls to position it. Just bear in mind that the GUI will not tell you where the image will actually end up–you will need to visually place it by looking at the idle beam point while moving it around. That basically means that if you want to do any kind of specific placement you will need to jump through some hoops with a ruler and a calculator to get there. Be sure to keep notes, they might come in handy later on.

During operation whatever is in the image space in the middle of the GUI will be filled in with red as the engraving progresses. The restart button does not repeat the engraving, it simply resets the unit to the default origin position without erasing the stored image. You can over-trace an image by simply clicking on the “Start” button without physically moving the work piece, but the display won’t revert back to back without some goading. If you temporarily move another window over the control GUI before restarting then the red fill-in will be reset back to black when the overlying window is moved away (or minimized) and the GUI does a redraw of the display.

I’ve tried a variety of materials, and I’ve found that I can cut manila card stock, like the paper used for file folders, and if I’m willing to re-run an operation a few times I can work through thin balsa wood. Setting the burning time (the slider at the bottom of the GUI) to a higher value will cause the laser to dwell longer and take slower steps, which puts more accumulated power at the focal point during each step time. However, because of the low power the cut can be rather nasty since the laser is spending longer doing the work on any one spot and creating carbon. It can also put marks on the plastic Y-axis table surface. Finding the sweet spot for the step time for a given material takes some trial and error. Keep notes.

I could replace the controller PCB with something like a Smoothieboard, which is what will happen to the CNC router I bought, but I’m not going to bother with it for this gadget, at least not for a while. An Arduino Uno would also work, and the AVR MCU wouldn’t really be breaking a sweat. In any case, I’ve figured out how to roughly set the image position using the movement buttons on the control application’s GUI, and I can work with that for now.

Besides a power switch, I’m considering replacing the Y-axis table with something like a piece of 1/8″ birch plywood with pre-drilled holes for miniature clamps of some type. Thin aluminum is also a possibility, but I don’t think I want to deal with the specular reflections at UV wavelengths. The ultimate limitation on how thick I can make the work table is the focus adjustment range of the laser optics, so that’s another project when I can find the time.

The Neje web site is located at http://www.trusfer.com/, and Firefox thinks it contains unwanted or malicious software. Well, it might, but so far I’ve not encountered anything evil. The folks at Neje are apparently aware of this, and the site contains a rather vigorous denial of anything malicious. They claim the positive hits for malicious software are due to the programming language they are using. That may be, but as always, proceed with caution. I unpacked everything on a Linux image running in an isolated virtual machine, mainly so I could look at the documents and images. If anything was infected it vanished when I deleted the VM.

The documentation from the Neje web site was interesting. Turns out that there is a version of the engraver (the DK-8-3) that does support G-code, but it only has a 300 mW laser. They also have a 1.5W battery-powered model (yes, really!) that uses Bluetooth. I suppose that one is intended for stalls selling tourist trinkets so folks can get vacation selfies burned onto a wallet or cell phone case.

If you’re interested, I bought my Neje laser engraver from GearBest. A quick look shows that they currently have them on sale for $70. Eventually I will invest in a larger CO2 laser engraver/cutter, but for now I have other tools for cutting and shaping.


The Neje DK-8-KZ laser engraver is an interesting gadget, and you can do some useful things with it. But you won’t be doing anything large, or with a high degree of positional control or repeatability. Unfortunately, these last two criteria are what separates toys from tools, and by this measure the DK-8-KZ falls more into the toys and novelties category than the tools department. It’s great for putting a cool design on a cell phone case, or making Christmas ornaments from thin plywood or balsa pieces, but otherwise it’s rather limited.

For engraving my recommendation would be to spend a little more and buy something a bit larger, like this open-frame laser engraver. Tools in this class go for around $250 and have lasers with 2 to 2.5W output. Or, if you really want to cut stuff, then do some research and see what others have to say about some of the low-cost 40W CO2 laser cutters available. These go for around $600 to $800 or so, and if you’re willing to put in some time and effort to correct some minor design deficiencies and make some needed mods, then you can have your own usable laser cutter for under $1000. There are lots of videos and how-to presentations on Hackaday, Instructables, and YouTube that cover Chinese laser cutters. Makezine has a nice, short overview article on laser cutters that includes a mention of the low-cost Chinese units.

In any case, just bear in mind that you probably won’t get the same level of accuracy and performance with an imported $500 laser cutter that a full-up commercial tool is capable of providing. But, then again, the full-up tool will likely have a bigger CO2 laser, precision parts, a large (and I mean large) metal enclosure, an optional service contract, and a price tag of around $5000 (or more). Caveat emptor.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow Crankycode on WordPress.com

Little Buddy

An awesome little friend

Jordi the Sheltie passed away in 2008 at the ripe old age of 14. He was the most awesome dog I've ever known.

%d bloggers like this: