K40 Chinese Laser Upgrade Plans


I did something bad....very very bad....I bought I Cheapo brand 40W Chinese laser off of Ebay. <gasp!> These machines are crap and leave a lot to be wanted. This post is somewhat of a brain-storming post as well as an upgrade path for me. I included A LOT of information in this post, and I probably SHOULD separate it into multiple posts, but hell, it will nice for someone out there to have all this in one place.

Why does anyone need a laser cutter?

While I love my 3D printer and CNC mill dearly, they just can't do it all. The 3D printer is slow and cannot do large, flat parts easily. The CNC mill can do large flat parts, but struggles with hard materials like acrylic. Milling is also limited to the radius of the mill bit which means I can't get tight inside corners or do very fine cutting detail. They have their uses, but if all I want is a project box or a sign, I have to go through a bit of struggling to get things exactly what I want.

Enter laser cutters. Laser cutters can cut just laser_cutting_snowflakesabout any material you throw at them. Acrylic, wood, leather, paper, you name it. About the only limitations you find occur when the material reflects the laser (such as metals), are non reactive (like glass), or give off truly nasty chemicals (like chlorine from PVC). They are extremely precise and stupidly easy to use. They can also cost *less* than a typical 3D printer or CNC mill, believe it or not.

If you are fortunate enough to have the money to buy a brand name laser, good for you. By all means buy a Boss or an Epilog laser and go to town. They are fantastic machines, but they simply cost an insane amount of money for the general hobbyist. Instead, you can find cheap laser cutters on Ebay for anywhere between $350 and $500 depending on shipping times and how much you want to get screwed over. They will work right out of the box, but a little tender love and care is needed to get the full potential out of them.

Stock K40 Overview

k40-ebay-laserGenerally speaking, all of the cheap ebay laser cutters are the same. There are two main variants you might find. The first and most common is the one shown at the right. It has a basic control panel with an on/off button, a "test" button, a "laser test" button, a potentiometer to change the laser-tube current, and an analog current meter. The second variant is pretty much the same but has a digital seven segment display that shows the current and a couple buttons instead of a potentiometer to adjust the current up or down.

Inside is a POS control board that controls the stepper motors, laser, and a connection to the host computer. A power supply creates the voltage and current needed to run the stepper motors as well as drive the laser tube itself.

Included with most systems is a water pump, duct tube, air blower, and software. The water pump is needed to circulate water through the laser tube to keep it cool. By all reports, this is total junk and should be replaced with something better. The air blower is purportedly ALSO junk (go figure) and doesn't attach well to most machines. The software included can work well for some folks and can do basic things like cutting and engraving without TOO much problem.

Inside the unit itself is the bed you place your lasering materials on. It is almost always a flat piece of metal that can (sometimes) be adjusted up and down with bolts.

Room for Upgrades

Unlike a brand name laser cutter, these machines are stupidly simple which means they are relatively easy to hack, upgrade, and modify. This gives us the ability to make the machine work EXACTLY as we want it to without anything more than a bit of wiring and some splurges on Digikey, Amazon, or Ebay.

These laser cutters consist of just a handful of parts:

     # A metal cabinet
     # Two stepper motors to move the laser in the X/Y directions
     # A power supply for the laser and the control board
     # A cheapo control board
     # A couple end stops to home the laser
     # A laser tube and generic optics
     # A couple switches and buttons on the control panel
     # Linear rails

That's it. There is next to nothing to them. Most of what you are buying with the machine is a metal box with a pile of generic junk inside. The fact that most of this stuff is dirt cheap isn't necessarily an issue though. Things like stepper motors and buttons will operate exactly the same regardless of the brand stamped on them.

Why upgrade then? Simple. While the machine will work out of the box, it will not work well. This is where I get my jollies when it comes to home projects. I love to take existing things and improve them wherever I can. The laser cutter rebuild is no different. And, at the end, I get to use the phrase "Yeah, I rebuilt a crap laser cutter into a thing of hacker glory."

Control Board Replacement

moshieThe control board in these machines is junk for my purposes.  My goal in owning a laser cutter is to be able to cut out precise pieces for other projects. I want to be able to make things like project boxes, robot parts, and more, all of which require exact geometries. The included control board (either a "Moshie" board pictured left, or another POS) is designed for people who want to engrave iPhone cases and luggage tags for their Etsy store, and as a result, they do not do well in the precision department. Above all, they are not modifiable. I like to add things. I like to code. I like to have things I can modify and reprogram 😉


Alternative Boards

Thanks to the 3D printer revolution, there are a large number of options when it comes to CNC control nowadays. Better yet, there are a large number of CHEAP options available that have a lot of users behind them (read "there are a lot of support forums online when I screw things up").


One of the options I am looking at is the ever ramps1-3_finfamiliar RAMPS platform. The RAMPS board was designed for 3D printing in mind but should work fantastically for my purposes. It snaps onto an Arduino Mega or Due and has quite a few different breakouts for motors, motor drivers, and sensors. The beauty of this platform and the reason I love it is that you can place your own stepper drivers onto the board. This means that as more advanced stepper drivers, such as the SilentStepStick, come onto the market (or I blow one up in a poof of magic blue smoke), I can swap them out easily.  They are also VERY heavily used in the 3D printer market which means they cost very little, are easy to get, and are well supported by the programming community.


42-tm_thickbox_default-jpgThe Smoothieboard was one of the first widely accepted 3D printer boards that came onto the market after RAMPS. It was created in response to the surge of Delta-style 3D printers that needed just a little more processing power.  They also featured a few nifty features like easy configuration in a text file instead of needing to search for settings through the Marlin firmware most commonly used with RAMPS. I have to say, it is a beautiful little board and one I am very tempted to get. My two main issues with this board, however, is that the stepper drivers are soldered on and it is rather pricey. On the first point, there is an option to place your own stepper drivers in slots, but I'd almost rather not deal with that.

Custom DSP Boards

If you do a web search for "K40 control boards" you will come up with a few upgrade kits that have "DSP" stepper drivers and fancy control panels. They promise buttery smooth movements and low sound. If you are not electronically gifted and have another $400-$600 to blow, these might be a good option. Given I want this to be a budget project and that I won't be able to modify these much later, I'm going to have to say "no".


I'm going to have to go with the classic RAMPS board. While the Smoothieboard is certainly nifty, I don't need super crazy processing power to do a two-axis CNC machine with PWM for the laser.  The price point of a RAMPS system is around $30 for the board and Arduino clone and $15 for two SilentStepStick stepper drivers.

I will definitely say that the SilentStepStick is a bit of a huge factor in this decision as well (and no, I'm not being endorsed for saying that). They drive steppers extremely quietly. Low noise = low vibrations = cleaner and more precise cut. For me, this is a no brainer.

The RAMPS board I will likely use: http://amzn.to/2rnP3LC

Generic SilentStepStick's (same driver chip): http://amzn.to/2qoVZKc



Control Panel Upgrades

The control panel (the part you actually see and interact with) is sad. Most units come with just a few basic components:

     # An ON/OFF switch
     # A laser ON/OFF toggle button
     # A laser test fire button
     # A potentiometer knob to adjust the current going through the tube
     # An analog current meter (ammeter)

Yeah...not exactly up to my standards. A few upgrades are in order.

Safety Upgrades

First and foremost is safety. We are dealing with a high powered and invisible laser. It cuts, it burns, it slices and dices skin, eyes, and anything else in it's path. Should I have the machine open to replace materials or debug and accidentally place my hand on the control panel, I could accidentally turn on the unit and fire the laser. A software glitch or a pet might also accidentally start the laser unexpectedly. We need to fix this.

To solve the issue of safety, a few additions/replacements need to be made:

System ESTOP Button

An ESTOP button is a big red mushroom shaped button that latches when you push it in. estopThey are used in industrial machines as a way to stop and de-energize a system. For my purpose, I will use it in one of two ways.

1) If the mains power consumption is low enough and I can find a contact that can support the current needed, I will wire the mains power through the button. That is, if I press the button, the 120VAC being fed to the power supply will be cut off. I will also wire the laser ON/OFF circuit through a second contact so that the laser cuts out before the power supply has a chance to drain any stored power.

2) If the mains power consumption is significantly high, I will wire just the laser ON/OFF circuit through the button so that it acts as a safer laser ON/OFF switch. A circuit through the second contact would also be wired to the new control board to stop motor movement.

If I can, I will shoot for option 1. Given the tube fires at a max of 40W, that will give me around 0.5A at 120VAC with the power losses in the power supply (I'll measure and report this later). I don't trust the engineers who designed the laser power supply to have a fail safe system that won't continue firing the laser after the signal was removed or that won't fire on its own. There is too much at stake if they screwed up either the engineering or the wiring. It also allows me to stop the machine quickly and easily should something go massively wrong like a fire.

Laser Enable Key Switch

In addition to the ESTOP button, I will be replacing the laser ON/OFF toggle switch with a key switch. In a nutshell, you need a key to turn the laser on. Once you turn it off and remove the key, the laser will not should not fire. This is great for a couple reasons. 1) Since there isn't a button you can accidentally push, you can't accidentally push it. 2) If I have company over that gets a little drunk and curious, I can remove the key and hide it so they can't burn a hole in their fingernails or set fire to my apartment.

Covered Test Fire Button

Alright, alright, part of this is me wanting a goofy movie-style "FIRE!!!" button, but there is a practical use. I don't want to accidentally push it. If the button has a cover that keeps you from pushing it, again, you can't accidentally push it. So, double whammy - I get to live my action movie dream every time I use it and it increases the safety of the machine.

Current Settings and Measurement

Potentiometer Replacement

Looking through schematics others have posted online, the current control knob on the machine is just a basic voltage divider circuit. Turning it one way or the other will sweep the voltage read by the power supply between 0V and 5V. Potentiometers are great for inexact things like volume control, but not for laser power control. They have two issues for this application. 1) They can wear out over time or oxidize. This means that there is no guarantee of a smooth transition of power as you rotate it. A bare spot internally might cause an intermittent short and therefore a massive jump in the laser power or else a short that takes out the power supply.  I don't like either.  2) They are not precise. When it comes to engraving, I want to have the ability to dial in the power for any given material. Once I find a setting that works, I want to be able to return exactly back to that setting.

To fix the potentiometer issue, I will be adding a small Arduino board to control a DAC (digital to analog converter) alongside either an LCD or 7 segment display. The DAC will allow me to send a precise and repeatable value from 0-5V and therefore a precise and repeatable power level. The power level would be displayed on the display. The actual input device will be a rotary encoder read by the Arduino.

Current Sensor Replacement

Analog meters are cool, but they kinda suck. They aren't usually calibrated well (if at all), and can fail quickly due to their mechanical nature. I've used them in novelty clock projects and they just don't last that long. To replace it, I have two main options.

The first is to place a 1Ohm resistor where the analog ammeter is and measure the voltage across it (it would have to be amplified to read well). I'm not a huge fan of this option. Should the laser tube fail or short for some reason, I will be left with around 40,000V being sent to a control board which will result in a fun explosion and an unsafe machine.

The second option is to utilize a hall-effect current measurement board. These come in two flavors. The first flavor is a non-contact sensor that wraps around the current-carrying wire and sense the magnetic fields there. The second flavor is placed in series with the circuit. It has a thick, half-circle shaped loop with the hall effect sensor underneath. As current passes through the half circle loop, it creates a magnetic field that is measured. Both are safer options, but they are susceptible to electromagnetic noise which the laser and its power supply will make a lot of. They can also lack in accuracy for lower currents which we will be using for the laser.

At the moment, I am leaning towards the second option - a hall effect sensor based unit. Should I have a failure somewhere, I'm much less likely to fry everything. I need to do more research to figure out what sensor units are available. Regardless of which I pick, they will be read by the same Arduino that is running the DAC and the value displayed on an LCD or 7-segment display.

Other Goodies

Water Flow, Control, and Temperature

I've seen a couple examples of people adding temperature and water flow sensors to their machines.  I like this quite a bit. It is extremely important to keep the laser tube cool. Should it overheat, it will go down the tubes. Knowing the temperature of the water reservoir and knowing that the water is in fact flowing is very helpful.  I will be adding temperature and flow indication to the unit. The temp will be found with a thermocouple and the flow with a simple flow switch. Both can be found online for cheap.

I will be adding to these as well. A liquid level sensor in the reservoir will tell me if the water is getting low. A DPDT (double pole, double throw) relay will be placed in the laser enable circuit as well as the control board ESTOP circuit to stop the laser and relay if the water isn't flowing, the temp gets too high, or the water level too low. A button added to the panel will be added to the panel to turn the water on/off via a dedicated relay.

I want to use a relay instead of just wiring a bunch of sensor switches in series is for a couple reasons. First, wiring a bunch of sensor switches is begging for some horrible debug issues. Should any one go down, you have do dismantle the entire system to find the problem. Having everything go back to a master node that monitors things and either lights up indication LEDs or outputs on a screen makes things a lot easier. Second, it allows me to use analog sensors and tweak them as time goes by. If I don't like an enable setpoint, I simply change the code. It also allows me to add warnings and alarms in the future should I want them.

I may also add a pressure sensor that would be used shut off the flow in the event a hose gets disconnected.

Fire Detection and Ventilation Control

Should the laser cutter catch fire, I would like the system to stop running and cut off the air ventilation system so it doesn't make the fire burn hotter. Normal methods of fire detection become difficult in laser cutting applications. Smoke detectors will not work as the unit creates smoke during normal operation. IR (infrared) detectors don't work because we are cutting with IR and burning the material we are lazing. UV (ultraviolet) detectors don't work for reasons similar to IR. What is left is internal ambient temperature. A thermocouple placed in the machine will watch for temperatures to get too high which would indicate a fire.

If a fire does occur and the temp increases, a relay would remove power from the blower and a second relay would put the system into a soft ESTOP just like with the water monitoring system.



Mechanical Upgrades

I don't have the machine yet, so I don't know the mechanical limitations of the machine. It may very well be that everything is pretty decent and doesn't need a replacement. But, since I bought cheap Chinese crap, it could go completely opposite.

Air Assist Nozzle

This is a must. When you cut or engrave a material, you are doing one of two things: you are precisely burning the material or vaporizing it into a gas. An air assist nozzle blows air onto the cutting/engraving spot to help remove the fumes from the area. This allows the laser to focus more power directly onto the active surface as well as keep the smoke and fumes away from the expensive mirrors. In short, this means faster cutting/engraving, at lower powers which prolongs the life of the tube and gives you a better result.

I found this air assist and will be adding it to my build. http://amzn.to/2rnv8vu

Stepper Motors

Generally speaking, a stepper motor is a stepper motor. They are stupidly simple things and pretty hard to screw up. Unless there is a flat tooth or something in a motor, I expect to leave these pretty much alone.

A possible upgrade route would be to get stepper motors with either higher steps-per-revolution or higher torque. Both would allow me to cut more smoothly and the second would allow me to (possibly) run at higher speeds. Neither are actually necessary as I can simply slow down the machine, adjust the driver current, or adjust the acceleration in the firmware.

Another possible route would be to add a gear box to the steppers to add 2:1 or 3:1 gearing. Again, this might not be needed, but would increase both the motor resolution and torque of the system. The only downside to this comes from slack that will be found in the gearbox. Slack, in a nutshell, happens when a gear rotates the opposite direction and the small gaps in the gears allow the motor to spin a small amount without turning the drive gear. What this means is that I could get really smooth curves as long as the motors don't have to reverse direction.

Belts and Timing Gears/Pulleys

Cheap timing belts have stretch and cheap timing pulleys don't have good bite. The stretch will cause slack issues like I talked about with the gear box at higher speeds as well as possible oscillations as the carriages bounce like a ball. Good news is that both belts and pulleys are cheap and easy to find due to 3D printing, so this wouldn't be hard.

Second Y Pulley

The machines I've seen typically have one Y (think "North" and "South) drive system on the left side. If the carriage assembly is well done and tight, there may not be any issue with this. If it is sloppy, I need to add a second Y pulley on the right hand side. In this way, both sides get pulled at the same time which can stop skewing as the cut/engraving moves to the right side of the machine.

Z Axis Bed

A flaw of this machine is that there is no easy way to adjust the focus height of the laser. Adjusting the focus means raising/lowering the material bed (if the machine supports it) or screwing with the optics. Neither of which are simple. So, if you want to engrave an 1/8 inch thick piece of plastic and then a 1 inch thick piece of wood, you have to change the settings. A Z axis bed would move up or down as needed to make sure the engraving/cutting surface is where it needs to be with ease.

Honeycomb Bed

Nearly all machines I've seen have a solid, flat, sheet metal bed that you place the cutting/engraving materials on. This works for engraving, but terrifies me for cutting. If the laser is cutting all the way through the material, it has a chance of bouncing off of the metal behind it and becoming a danger to nearby people should the reflected beam find its way out of the machine. It will also trap smoke and fumes underneath the material causing burn and scorch marks. Replacing the sheet metal bed with a honeycomb allows the laser and fumes a place to go.

Placing ceramic tiles upside down underneath the honeycomb bed (with the backs facing up) would absorb any stray lasers and safely scatter anything it can't absorb evenly in all directions. The idea here would be to stop reflections entirely.  Another option is to get graphite plates specifically made to absorb IR lasers. These can be a tad pricey though...

The aluminum bed I will likely use: http://amzn.to/2pR65Ri



Lasers and Optics

I don't expect the laser and optics to be up to par. Given this is a laser cutter, these are all fairly important.

Laser Tube

Nearly all of these machines ship with poorly made, poor quality, reject tubes. Actually, this is pretty common for a lot of electronics - factories will make components for a customer, test them, give "quality" parts to the customer, and then sell the rejects to budget Chinese companies. Chances are, the tube in the machine will either not be marked with a manufacturer, have the manufacturer mark scrubbed off, or have the sticker hastily removed with all the residue left behind. What I'm getting at is that the tube should be replaced at some point.  I will run this tube into the ground, but sooner than later, I need to replace it with a professionally made tube.

A new tube ranges from $150 to $250 for a decent one, so this will wait a bit.


s-l1600Mirrors and lenses are a pretty huge deal for a laser system and it is important to have good ones. Why? Because every time a laser bounces off a mirror, it looses a small amount of energy. For example, a  cheap mirror might have a reflectance of 95%. That sounds okay, but when you consider that three of them are used in the system, it adds up. The laser after the first mirror will be at 95%, after the second will be 90.25%, and the end laser output will be 85.7%. In edition to losing power, this also means that the mirrors will heat up and degrade faster. Replacing the mirrors with better ones helps to fix this.

Lens upgrade if needed: http://amzn.to/2pRy1V5

Laser Combiner

IR is invisible to the human eye. This means that the only way to line things up without changes to the system is to move the laser, do a test fire, and see where you get a burn mark. This means that to line up your laser, you have to burn something. With a laser combiner, you can add a low powered, red laser into the same path as the IR laser beam. The end result is that you can align mirrors and center your materials without risk of injury or damage to the material or machine.

Other Upgrades

There are a few other bits and bobs that I want to update or improve. These are things that don't really impact the functionality of the machine, but may help either its aesthetics or general usability.

Replacement Window

I'm not a huge fan of the orange viewing window. Given that I paid less than $400 for the machine, I very highly doubt there is anything special about this window except the color. From the sources I've found, acrylic in and of itself is opaque to the wavelengths put out by the laser. What this means is that I should be able to drop in a 3-5mm clear acrylic window without fear of eye damage. Of course, I need to do a bit more research/testing to verify this, but I'd really like a clear window.

Internal LED Lighting

Lighting is crucial and a pretty cheap upgrade. I will buy some LED light strips and mount them inside the cutting chamber so I can align and calibrate the laser easier.

Mirror Alignment Covers

To align the laser mirrors, you typically place a piece of tape or paper in front of a mirror and fire the laser on low power. You can see where the laser is by the burn mark on the tape/paper. Unfortunately, the fumes and smoke from this process can contaminate mirrors are lenses. An easy fix for this is to 3D print covers that fit over the mirrors to keep smoke off.

Auto Laser Focusing

There is a great little sensor board from Sparkfun that has a time-of-flight (TOF) distance sensor on it. Simply place it on the laser head carriage, and viola! With some simple math, you have the distance between the material and the laser lens.

By | 2017-05-16T13:11:28+00:00 November 20th, 2016|laser cutting, Project Logs|8 Comments

About the Author:

I’m an electrical and controls engineer with experience in C++, C#, Sequencer Programming (ie “Game Loop), a little bit of graphic design, 3D modeling, and bits and bobs of a lot of game design.


  1. Joel June 10, 2017 at 3:51 pm - Reply

    Nice summary! Thanks for all the tips. I am about to be embarking on the cheap-o Chinese laser cutter journey and want to order all of the upgrades at the same time that I order the actual cutter. How did RAMPS work for you… ? I was considering getting a smoothieware board myself, but am not fully convinced.

    • Austin Steingrube June 15, 2017 at 1:31 pm - Reply

      It actually works quite well. I am considering, however, to switch to a GRBL board which should supposedly work better with LaserWEB. While it is WAY late, I’m working on an article and a video of the full upgrade. Got wrapped up in a few other projects, heheh.

      Also, at the moment, I would NOT get a smoothieboard (for any projects). Beyond any etching/cutting issues it has, the firmware configuration isn’t very flexible. Namely, you can’t have two drivers move the same axis (which most 3D printers and K40’s use). And if you DO decide to get a Smoothieboard, get one from RobotSeed and not Uberclock. The guy at Uberclock is not exactly great at responding to customer support or timely shipping.

  2. Herman Arango September 2, 2017 at 12:58 pm - Reply

    Need to upgrade my recently delivered Chinese 40W laser. Not a techie, need something “plug and play.” Any recommendations?.

    • Austin Steingrube September 2, 2017 at 5:10 pm - Reply

      If you want plug and play, I’d stay FAR away from the cheapo laser cutters. Go with a company like Full Spectrum, Boss, or Epilog. The issue with these, beyond what I write up in this post, is that the quality control is horrible. My unit shipped with bad wiring, ruined optics, and several other issues. I talk about it a little bit in another posting.

      One of these days, I’ll do a full write up with pictures, schematics, and videos. It may be a while though – I’m currently in the process of changing jobs and moving. I wanted to get this laser completely done earlier this year and I put a lot of work into it. Unfortunately, it got put on the back shelf while I tended to a few high priority and difficult jobs at work.

      • Herman Arango September 2, 2017 at 6:22 pm - Reply

        Thanks for your response. Already have the machine. Dealing with all issues

        • Austin Steingrube September 2, 2017 at 10:13 pm - Reply

          Yeah, unfortunately there isn’t a quick and easy thing for these. At a minimum, you should replace the control board and install an estop button that cuts power.

          • Herman Arango September 3, 2017 at 3:27 am

            Which control board do you recommend?
            Stop button?
            How difficult?


          • Herman Arango September 3, 2017 at 8:37 am

            Hopefully one (board) not requiring much tinkering other than removing old one, installing new one and sfotware but no modification of either hardware or software.

Leave A Comment