A Review of the “K40”-type Laser Cutter/Engraver.

July 2015

When I was a kid, lasers had just barely transformed from science fiction to science. The laser printer was a miracle of modern technology. Now, well, I paid $25 (used) for the laser printer in my office, and it retailed for $6,500 new. Not long after landing my first real full-time job after college, I treated myself to my very own laser. Yea, this was way before laser diodes were available. It was (well, it still is) a lab-grade HeNe 10mW laser. Ten milliwatts. Cost me about $150, I think.

Now, Epilog will sell you a laser cutter that costs 40 times as much, but with a laser 4,000 times more powerful. Desktop laser cutters? Holy cow.

Nevertheless, that 35 watt Epilog laser cutter costs more than my laser printer did when it was new. For most people, buying their own personal laser cutter is still not exactly practical. Or is it? There are a bunch of companies offering 40 watt laser cutters on eBay for under $1,000; sometimes way under. Can you actually get a brand-new, real, working laser cutter/engraver for $500?

As a matter of fact, yes, you can.


Oh, let me be completely up-front with you. I am utterly certain that an $8,000 Epilog Zing is a far better piece of equipment than the laser cutter I now have in my workshop. It’s probably only about half as good, and I expect it will only last half as long, if that. The hardware is quirky, the software is annoyingly limited, and I have no doubt it’s going to require a lot of tender loving care to keep it running. Still, it could have been a lot worse and still have been a good deal.

There are a number of different companies selling what seem to be two or three slight variations on the same device. I think I ended up with a slightly older model; it has an analog power meter and a knob for adjusting the beam strength. The other style has up-down buttons for adjusting the strength, and a 3-digit digital power readout. As far as I can tell, though, the mechanical bits are pretty much identical.


It arrived in a big wooden crate. I had no issues with the shipping. The manual is...amusing. The main thing it tells you is how to align the mirrors. After shipping, you’d be a fool not to assume the mirrors are misaligned, so walk through the process and align them.

The hardware, for the most part, seemed very economical, rather than cheap. The case is solid. The wires are tied down and in wire looms. Some of the connections to the power board are hot-glued down for strain-relief. That’s a nice touch.

There’s a ventilation fan that drops into a set of guides on the back, and a pump to circulate water around the laser tube. The fan is too tall; it sticks up slightly past the top of the enclosure, and it doesn’t fit against the duct output hole very tightly. Also, when it’s in place it blocks the door to the laser tube. A classy laser cutter would have put the fan inside the case. I fabricated a little cover that screwed on to my fan to give it a better seal. You don’t normally need to get into the laser tube compartment. When you do, just lift the fan out of its slots.

It’s your job to figure out where to put a bucket of water for the water pump. The pump itself is just an off-the-shelf aquarium pump. I filled a five-gallon bucket about 2/3rds full of water, threaded the hoses through a hole in the lid, threaded the power cord for the pump out of the lid, and I was good to go.

The fan and the pump have their own independent power cords. You know, the standard two-prong US-style 120V wall outlet cords. I could have plugged those things into a power strip along with the laser cutter itself, and used the power strip switch to control everything, but instead, I cut out holes in my laser cutter’s case and installed a standard receptacle in the case, controlled by the laser cutter’s main power switch.

The manual warns that there are dire consequences if you let the laser tube get too hot. In a nutshell, if it overheats it could crack, which would destroy the near-vacuum gas mixture in it that spawns the laser beam. I mounted a remote-read thermometer in the water return to keep an eye on it, but as I expected, 3-4 gallons of water is such a large heat sink than even cutting on full power for ten to fifteen minutes at a time doesn’t heat the water up more than about fifteen degrees.

Unfortunately, there’s nothing to detect if the pump fails or the water stops circulating. You just have to pay attention. I have my return hose above the surface of the water, so I can hear it trickling and burbling when the pump is running. A fancier laser would presumably have a flowmeter or maybe a thermal overload breaker, and would shut off the laser if it were getting too hot.


The cutter connects to your computer via USB, although for no reason I can fathom, they put the wrong kind of USB socket on it. The laser cutter should have a USB “B” socket, which is square-ish. Instead, it has a USB “A” connector, the rectangular kind that’s on your computer. Thus, in order to plug it in, you need a USB cable that’s rectangular on both ends, which, fortunately, they supply.

In a message to the seller, I mentioned “I wish you had not used the wrong kind of USB port.” They replied, “could you please let us know which kind you use?” In response, I sent them a link to the USB specification, which not only has a section about what not to do, it even explicitly states that a USB “A” USB “A” cable should never exist. So don’t lose that cable!

The unit is primarily designed for engraving, not cutting. The work surface is set up with a rather small spring clamp, about 80 x 100 millimeters. Fortunately, it can be removed by unscrewing it. This leaves you with a nice aluminum work surface with a great big hole in it.

Speaking of holes, there’s a round hole about 2.5” wide in the bottom of the enclosure. With the clamp removed, it’s possible for the cutting laser to shoot right out the bottom of the box and start burning whatever’s next. I taped a piece of aluminum foil over mine to prevent that. Even at the focal point of the beam, this laser can’t cut aluminum foil, and by the time it reaches the hole it’s spread out quite a bit.

The focal point, by the way, is about 3mm above the surface. That’s a good thing. It means if you cut 3mm acrylic, the smallest part of the beam is at the surface. You definitely shouldn’t plan on cutting anything thicker than 6mm with a 40 watt laser, so that’s a good place for it to be. If you’re planning on cutting paper, vinyl, or something else thin, you might want to put something underneath your cutting medium and raise it up.

There is a weird bug in the controller. At any cutting speed from 5.25mm/sec to 45mm/sec, it’s fine. At 5mm or less, the cutting head actually speeds up a little, and it also ignores any origin offset I’d configured in the software, causing the cut lines to be out of place.

That’s a problem because the air intake for the ventilation fan actually sticks out into the cutting zone slightly. The top of my intake has quite a few lines burned into the paint where the laser’s sliced across it. Why would they put the air vent where the laser can hit it? Beats me.

That’s not the only place where there are no guard rails, so to speak. When the machine starts up, the cutter head moves back and to the left until it taps two limit switches. “Hey, I’m in the upper left corner!” But there are only two limit switches. The controller board will happily attempt to move the cutter head down or right further than the mechanism can physically move, if you tell it to, resulting in a nasty-sounding grinding noise as the belt jumps against the drive gear, or some such thing. Once that happens, the cutter head isn’t in the position that the controller thinks it is, so all further cuts will be incorrectly positioned.

I believe that the Epilog unit has servomotors, so its controller knows where the cutting head actually is, and can tell right away if anything has interfered with the head’s motion. With the K40, you just need to be smart enough to not tell it to cut past where it can actually reach. On mine, that means it has a functional cutting area of 185mm by 310mm-ish. I haven’t taken the time to figure out exactly how far I can go in the X direction yet, but it’s at least 310mm, and probably not much more than that. I could probably get the Y direction up to 200mm if I cut the air intake vent back a bit.

The other way you can throw the cutter head out of alignment is by trying to move it too quickly. The software defaults to a maximum head speed of 500mm/sec. The software’s a bit optimistic. Now, in engraving mode, the cutter head is sweeping back and forth in the X direction, raster-scanning an image onto the surface. I have successfully engraved with a setting of 450mm/sec, but at 500, the momentum of the cutter head is too much for the drive belt, or the pulley, or some part or t’other, and it tends to jump a tooth every now and then.

When cutting, 50mm/sec is faster than it can handle. Cutting involves whipping the head around in both the X and Y directions, and it can’t slow down for corners because that would change the cut depth. So if it hits a right-angle turn too fast, once again, the gear slips, and all further cutting is misaligned. If you want really clean lines, even 40mm/sec might be too fast. I can see little ‘teeth’ on my lines when I’m cutting something like paper that fast. The teeth are caused by some part vibrating after being whacked by the sharp turn. If I slow down to 35mm/sec, they tend to disappear.

Here’s the good news. As long as I’m not trying to cut too fast or too slow, the cutter head is actually remarkably precise. I’ve double-cut some pretty elaborate patterns, and the head will come back and retrace its path close enough that I can’t see any error with my naked eye. I keep meaning to find out what the specified minimum step size is, but whatever it is, it’s dang small, and the repeatability is excellent.

One last missing safety feature: there’s no cover interlock. Fancier cutters won’t energize the laser unless the lid is down. Now, it’s actually rather tempting to leave the lid up for little jobs. For one thing, you can get a closer look at the cut point. And one might think there’s no particularly compelling reason to put the lid down, since it’s got a great big clear window in it anyway. So I thought, until one day I noticed a smudge on the window lid pane. When I tried to clean it off, I found that it was actually a rough spot on the inside of the polycarbonate window. Apparently, the laser beam hit something fairly reflective, and ricocheted upward. It was absorbed by the polycarbonate, but the polycarb got hot enough to melt a bit. If that had happened while I had the lid up and my face down there, well, I’m betting it would have stung more than a bit. Keep the lid closed while cutting!


Alas, the supplied software is less impressive. First of all, there’s no instruction manual of any kind. There is simply a CD-ROM and a special USB dongle. Oh, and a fairly useless video that only showed me what I found perfectly obvious anyway.

The first challenge was figuring out how to activate it. You have to enter the model number and ID number of the laser cutter. This info is only found on the circuit board inside the machine. And I had to blow the dust off an old Intel carcass in order to have something to run Windows on. At some point I’m probably going to try moving everything over to a virtual Windows machine, but that may or may not work with the USB dongle and such.

As mentioned, the system’s original intention was to engrave signature seals. The software will import a wide range of raster formats: jpeg, gif, tiff, png, bmp, and more. However, if you want to cut, you need to be able to load a vector format, and it didn’t seem like it could read any vector formats of any kind. Eventually I discovered that it can read one : Windows Metafile (and its younger cousin, Extended Metafile). So now I save my work as an Illustrator file, move it to the Windows machine, load it into CorelDraw, save it out as a wmf file, then load that into LaserDRW. There’s also a CorelDraw plug-in for burning from within Corel, but it seemed a hair less functional. There’s also another stand-alone program on the included CD-ROM which looks more or less like a less-capable version of LaserDRW.

I’ve seen various people online complain about how the software seems to cut everything twice. I had the same problem, but I eventually figured out what was going on. If you want to cut out, say, a circle, then you need to set it up in your drawing program with a solid fill, and no stroke. Once converted to a .wmf or .emf file, LaserDRW will trace once around the edge, cutting a circle.

If, on the other hand, you want to cut an X, then you’ve got a problem. When I’m preparing a file for a vinyl cutter, I just put two lines in my file, and make them the “Cut” spot color. What kind of line (stroke) or fill the shape might have is ignored; the vinyl cutter just follows the lines. The laser engraver, alas, isn’t dumb enough. It’s trying to be “helpful.” If I put those two lines in a file and ask it to cut them, it’s going to cut both of them twice. More precisely (literally!), it’s going to cut the left and right sides of each line once. Since I’d mostly been making my lines “1pt” or “hairline,” I couldn’t tell at first that it was trying to accurately replicate the *thickness* of my lines. If you send it a 28pt line, it will, in fact, cut a rectangle.

And don’t think you can get extra-clever and make super-duper thin lines. I have found if the line is too thin, the engraver starts jumping and skipping around erratically. Probably a result of mathematical rounding errors causing the line’s sides to cross over each other.


So, you can only cut closed curves. This is generally not too severe of a limitation. I’ve been doing some paper sculpture work where the cutting was not suitable to be described by closed curves, but I just sped up the cut speed, dialed the power down a bit, and let the machine cut all the lines twice. A 0.1 mm wide line is effectively the same as zero width, since the laser beam itself is around 0.2mm wide at its tightest focus.

Another annoying quirk of the software is that it doesn’t respect the dimensions specified in the datafile, either vector or raster. When you first start up LaserDRW, it defaults to a 40mm x 40mm imaging zone. You can change this to whatever you want in the “Layout” menu option; it then draws a new grid on the screen to represent your imaging area. When you load an artwork file, it will resize it to completely fill one of the directions. So if you use the default 40mm x 40mm, and load an oval that you’d set to be 10x20mm, it will appear in LaserDRW as a 20x40mm oval. You *can* resize it in LaserDRW and put it back to the correct size, but I generally just try to make my patterns so that they come out at nice round numbers which I can remember, and then set the imaging zone to match. Otherwise, I’ll forget to change it for one of the parts, and then it comes out the wrong size.

Engraving, as mentioned above, is an entirely different process. The laser head moves like an ink-jet printer, back and forth, firing the laser according to the bitmap image loaded. Now, the laser has a knob (for the older models) or some buttons (for the newer ones) to change the power of the laser beam. Mine has the knob and an analog needle meter, with a scale from 0-30mA. Turning the knob all the way up puts the needle at around 27mA, which is presumably “full power” for the laser. It’s supposed to be a 40 watt laser, but I’ve seen people online claim that that’s optimistic; it might be more like 30-35 watts. Nor do I have any particular reason to believe that 13mA is “half-power.” I have no way to confirm if the laser or meter have a linear response. Thus, I can tell you what settings have worked for me, but only in terms of “milliAmps,” which are pretty arbitrary.

I mention this because when I first tried engraving, I was using gray-scale images, assuming that the board would modulate the laser power based on the gray value in the picture. Eventually I got smart and fed it a test pattern, and discovered that everything was getting posterized; anything lighter than 50% gray was 'laser off,' and anything darker than that was 'laser on.' Alas.


So, what can you cut? Well, not metal. Rather to my surprise, even kitchen aluminum foil transfers heat so fast that it is impervious to the laser beam. Mostly what I’ve been cutting has been acrylic plastic and paper.

A brief aside; there’s a web page out there with a fair amount of pretty good advice, and one enormous error, wherein somebody will tell you how you should never never cut Plexiglas, because it’s not acrylic, it’s polycarbonate plastic. This is entirely and completely wrong. Plexiglas, Acrylite, Lucite, and Perspex are all brand names for acrylic sheet plastic.

Another aside: all acrylic sheet is metric. I get most of mine by buying scraps from local glass shops, and when I asked about 3mm acrylic, they look confused. “Eighth inch?” “Oh, yea, we got that.” But “1/8-inch” acrylic isn’t 0.125 inches thick, it’s 0.118 inches thick (according to the label on one sheet), which works out to 2.9972 millimeters. Knowing how thick it is, is really important if you’re trying to make pieces that interlock well. I’ve picked up 2mm, 3mm, 4mm, and 5.5mm acrylic. The 2 and 3 are easy to cut. The 4mm is tricky, and the 5.5 requires two passes.

Surprisingly, cranking up the power does not help. I cut all my 3mm acrylic at “10mA”, which is (probably) about 33% of full power. Cranking it up causes problems like having the acrylic re-weld itself together after being cut, and distorts the cut edge. I suspect it’s because when the laser cuts the acrylic, that creates a plume of vaporized plastic, and until the plume clears, more power just heats the vapor, transferring heat to the sides and generally being unhelpful. So my settings for cutting 3mm acrylic are a travel speed of 6mm/sec and a power level of 33%ish (10mA). For 4mm I’ll nudge it up to 12mA and slow it down to 5.5mm/sec. For engraving, power is between 5mA and 10mA, and head speed is 400mm/sec (as fast as it can go without errors).

I have learned to leave the protective paper on the bottom side of acrylic. When it cuts all the way through, some of the vapor is pulled along underneath the sheet before going out the exhaust vent, and it tends to re-deposit on cold acrylic and fog it up. I’ve tried leaving the paper on the top as well, but some acrylic has a plastic protective cover, which can fuse with the acrylic when being cut, and the extra layer of material (especially paper) seems to make it harder to cut cleanly through the sheet.

Cutting paper is tricky. Too little power, and it doesn’t actually cut through. Too much, and it bursts into flame. Just the right amount sometimes doesn’t quite cut through AND it bursts into flame. Again, the big problem seems to be the vapor plume. The paper is vaporized, the vapor goes straight up into the beam, gets heated even further, and bursts into flame. Frequently there’s a little flame jet shooting up from where the beam’s hitting the paper. The flame jet doesn’t cause a problem unless the cut pattern I’m doing allows the paper to curl upwards. If the paper curls up, the cut edge lifts into the flame jet, and then I’ve got an actual fire burning in my machine. This is why you shouldn’t leave the room when you’re cutting. I’ve also had acrylic plastic burst into flame a few times, especially if the laser is cutting right at the edge of the piece.

I suspect it would be much much easier to cut paper if this machine had come with an air jet mounted on the cutting head, like the ones found on better laser cutters. Even a fairly gentle stream of air blowing onto the contact point would nudge vapors away from the beam, greatly reducing the odds of igniting them. The other trick I’ve found for cutting paper is to change the “Pixel” value (under Advanced Parameters on the cut/engrave requester) from “1” to “2”. You see, although the laser seems to be running continuously, it’s actually being pulsed at very high speed. I think (because, as mentioned, there isn’t any documentation whatsoever) that setting Pixel to 2 fires the laser less often. The shots still overlap enough that it cuts all the way through the paper, but it seems to reduce the odds of igniting the material. So my settings for cutting paper are Pixel: 2, power at 3 or 4mA, and speed at 30mm/sec.

Thin plywood is a very popular material for laser cutting, but I haven’t done much of that, mostly because I hate the charred edge that a laser puts on it. My paper shows a charred edge too, but it’s fairly unobtrusive.

I’ve successfully cut foam core, but it’s extremely tricky. I think, once again, it’s the vapor that messes things up. If the laser’s dialed up enough to cut all the way through, the core (which is, of course, the part that’s foam) pulls away from the edge for quite a ways. I know it shrinks when it gets hot, so I figure the hot vapor is the problem. In order to get a decent edge, I have to cut 1/8” (which is, actually, one eighth of an inch thick) foam core anywhere from 3 to 6 times on a very low power setting; maybe 3 or 4 mA. So: power level 3-4, speed 30-40mm/sec, repeat until cut is completed.

At one point, I thought I had a brilliant idea. I cut a piece of 1/8” glass (with a standard glass cutter) down into a sheet small enough to fit into my laser cutter. I figured if I lay that on top of a piece of paper, the glass should seal off oxygen from the paper well enough to prevent it from igniting. Unfortunately, it turns out that, although glass is quite transparent to visible light, it’s actually fairly opaque to infrared, which is what the laser uses. So instead of cutting the paper, the laser etched the glass. In fact, it was doing a really beautiful job of etching the glass, but because I’d used a really thin pane, the glass couldn’t handle the thermal shock, and shattered.

I haven’t tried cutting fabric yet, although I’ve heard it works very well. I have “branded” leather. Engraving mode, 4-5mA, 400mm/sec. It’s better to start w/not enough power, and just re-run it if it isn’t dark enough, because if you go too hot, it turns the top layer of the leather to carbon ash, which rubs off on everything. Even if you get it spot on and create the really dark brown that isn’t quite fully carbonized, you still need to seal the leather with polyurethane spray or something similar, to keep the burned part from rubbing off.

I’ve also had fun lasering food. Cheddar cheese didn’t go so well; the laser ended up cooking it as well as cutting it, and it tasted a bit burned. On the other hand, flour tortillas cut really well.

I keep cotton swabs and isopropyl alcohol by the machine, and clean the collimating lens (the part of the cutting head that’s pointed downward) fairly often, since the cutting vapor generally swirls around a lot before getting sucked into the vent. Airflow is probably the aspect of this machine that is most in need of improvement.

Finally, I’ll note that there are actually a number of upgrades available from U.S. companies to improve these machines. Better stepper motors, a new controller board, a cutting table that can be adjusted by the computer for different cutting heights, the missing air jet for the cutting head, and more. When I can afford to, I am definitely hoping to upgrade. It looks like replacing the controller board will mean I’ll be able to cut with a single pass without having to have a closed curve, for one thing. Yay!

Even with all the quirks and flaws, I am still very satisfied. If I’d spent $2,000, I’d be annoyed, but for $482 (including shipping), I do not regret this purchase at all.