K40 Laser Engraver/Cutter Upgrade and Installation
I decided to get one of the ubiquitous cheap chinese K40 lasers and see what i could do with it.
I did not plan to reinvent the wheel on the conversion and setup of the machine as there is a lot of information out there pertaining to this. I plan to document my conversion as it differs from the others both mechanically, electrically and cosmetically.
There are two basic variations of the K40, Analog Controls and Digital Controls. I elected for Analog. I found mine on eBay and it was $313 shipped from a California warehouse. I will not bore you with more than one "here it is the day it arrived" pictures as it looks just like all the rest of them....
The working envelope of the laser left a lot to be desired, so it was ripped out along with the large metal duct that protruded into the work area. A better ventilation adapter with a remote fan was installed. The original fan was large and prone to failure based on everything I read on the internet forums.
I elected to ditch the original controller and the lousy Corel Draw 12 and plug in routine as sourcing the program and then still not being able to vary the laser power in software did not appeal to me. The universal opinion of the software/controller combination is "it sucks" and I never even bothered to try it. The machine was only fired up to verify it worked.
I elected to run the SBASE MKS v1.3 a clone of the Smoothie Controller as the desired Smoothie model I wanted was not available at the time I decided to purchase the laser There are a number of variations on the Smoothie Ware boards, this and the Cohesion being the most common. Read up on Smoothie
The Sbase boards are readily available and generally costs around $50. This is a clone of the Smoothie Ware board and is reasonably compatible. Expect some pushback on the Smoothie forums if you ask for help. The people making this board have not played entirely fair with the open source practices.
Setup of the controller was pretty straight forward and simple. I had to flip the direction of the Y Axis stepper and make an adapter Y cable to plug in the X and Y axis Min end stops.
Along the way I purchased a kit of plugs and sockets for both the JST-XHP and the Black Single Row headers found on PC motherboards. I also upgraded to a ratcheting type crimper for these as well.
To facilitate connections from the board to the outside world I got a Micro SD Card extension and a panel mount USB B extension. I designed the part with Fusion 360 and printed it on my 3D printer. I have posted the file on Thingiverse The SD Card Socket is a snug fit through the opening and a bead of super glue keeps it in place. This give a nice finished look to the outside of the machine and allows for easy removal of the memory card as well as access to connect the PC during use. I elected to use the LightBurn software, which I highly recommend!!
After a few months of use post conversion, there were issues concerning intermittent controller crashing. I decided to run with a Smoothie Board which I finally found on eBay.
The conversion documented here applies entirely to the use of the Smoothie controller. First step was to gut out the original K40 controller and control panel. After the ripping and tearing was completed, I commenced to update the control panel of the laser to support the Smoothie:
A piece of thin aluminum plate was machined on the CNC mill and then covered with a sheet of Weatherproof Silver Polyester Laser printed material (Online Labels # OL177SP) so the labeling would look nice and professional. Now that I have the laser and if I were to remake the control panel, I would use Cermark to burn the labeling onto the aluminum plate. A sample of the Cermark on Aluminum is shown here. Right side image shows the back of the display prior to the commencement of the wiring. The original panel meter shown here, was replaced with one that had a full scale of 20ma as the tube current would typically be 15-16 ma. This provided a more pleasing scaling.
.
.
The original Pot was re-used along with the master On/Off switch. The E-Stop switch has not been wired yet in this view. The E-Stop kills everything but the chiller and cabinet ventilator to prevent tube damage and to clear the air in the cabinet. Right side shows the functional panel in operation. The small digital display in the upper left is for a thermocouple that is installed in a whirly-gig flow meter behind the unit that provides a visual indication of water flow. As a rule the meter on the control panel, sensing the output temperature of the tube runs within a half to one degree of the chiller's temperature display. I could not resist the homage to Douglas Adams...
.
Then I installed the Smoothie Board. When I first tried to do this the 4 channel board was universally not available and the 5 channel quite a bit more expensive than I wanted to invest. Adventures with the controller chrashing with the Maker Base unit really taught me a lesson.
Here is a bench test of the Smoothie Board 5C. On the right is the board mounted to a metal plate that allowed me to use the original mounting holes in the cabinet, keeping the outside neat without any extra holes. For anyone interested, here is my final configuration file. Bear in mind, the configurations include the settings for a 4th axis, which I am using for a Z Table, described later.
Next was to get rid of the cheesy power input socket on the back of the unit and add some power outlets. Multiple outlet sockets needed to be added to control the Chiller, Vent Fan, Air Assist and a Vacuum Pump. The vacuum pump is for fixtures to hold flat plates. The input receptacle was replaced with a heavier duty model that does not have a light duty fuse holder incorporated. A fuse holder was installed in a more convenient location on the right side of the cabinet.
A drawer type platform was installed to allow easier access to the back of the machine and also to allow long pieces of stock to be cut in my plate shear which is mounted to the bench to the right of the laser. I 3D printed an air duct adapter as the supplied ventilation blower has been known to fail early in life. I mounted an inline 4" duct fan under the wall cabinets and used flexible foil duct to connect it to the outside vent fitting installed in the wall.
I elected to get a "real" chiller that has a compressor to help keep the tube cool. The smaller chillers that are more commonly available are thermo-electric Peltier units. The unit holds 8L of water. I use distilled water with a couple of ml fish tank algaecide. I ditched the original failure prone vent fan and installed a remote duct fan on the wall behind the unit. Here's the adapter and duct installation. Tube cover was removed to facilitate the installation. The unit is rolled out from the counter in this view. This shows how the moving platform facilitates access to the rear of the unit.
.
The whirly-gig flow meter. Decidedly low tech, but gives an immediate indication of water flow. The thermocouple for the temperature display in the control panel is mounted here. The flow meter is mounted in the return line from the laser tube. I use the temperature display on the chiller to indicate the input temperature of the cooling water to the tube The chiller has an alarm output for low water flow. This is connected to a Smoothie input to shut down the laser in the event of a low flow condition.
.
The completed laser engraver....
Power Supply Upgrade
8/10/2022
When i built the the laser I had intended to replace the power supply as the original "blue box" is generally considered to be a failure item and is usually replaced shortly after getting the unit if you use it very much. Well, the laser ground to a halt in the middle of a job three years later. I just ordered a new, upgraded power supply without doing much in the way of diagnostics. I mean, hey, what shoots craps first on one of these machines? ROFLMAO
However, I should have paid attention to the mA meter as it was still flowing current. As part of diagnostics I put a piece of cardboard between the primary and secondary mirrors and got no burning. The REAL problem? A broken primary mirror. I had this happen with the secondary mirror shortly after getting the machine operational.
I ordered the power supply from OMTech Laser They had a very nice price to begin with AND it was on sale! Connections needed to be changed a little on my installation to retain both the Test Fire and Enable switches.
The original power supply was hard wired to the the tube. The new power supply included a pigtail with connector to attach to the laser tube. As finicky as the original connection to the tube is, coupled with the fact it has been gooped up with silicone sealer, I elected to not use the pigtail. I purchased a high voltage connector I found on Amazon: MCW laser High Voltage Wire Connector (set of two) about $15.
The new P/S is installed and running.
Mirror Alignment Tool
8/14/2022
While cruising around on the OMTech site I found a laser pointer type mirror alignment device. American Photonics C02 Laser Alignment Tool I elected to buy a basic model as opposed to the one that was almost double the price. This required making an adapter.
I used my Elegoo Saturn Resin Printer to make the adapter. I use the Siraya Tech Simple Clear resin. The resin is easy to work with, clean up and it machines beautifully! I tapped a couple of 4-40 holes for the securing screws. I happen to have a second tertiary mirror / lens mount, and just stole the adapter ring from it to make a permanent tool. I discovered that I needed to move the screw holes closer to the blue adapter to better grab the tapered snout of the adapter. The corrected version is shown below.
Another view of the assembly. The tool comes with a sheet of 12 button cells, I think they are telling you something about the battery life! Here's the alignment tool installed after removing the Air Assist nozzle.
Z Table Construction
The original working platform that was installed in the laser was basically designed to hold small parts for engraving, with little flexibility. I believe the machine was designed for doing small badges and signs based on the spring loaded part of the opening. Most of the working area was lost.
I removed this and then designed and built a Z Table that could be controlled by the software. The table is loosely based on a table that is based on 1" aluminum tubing for a frame. By going to my flat plate design I gained about 1-1/2" of travel. Fully lowered the table is about 1/5" tall. The Thingiverse post for the parts is located here.
.
I 3D printed four corner brackets which us stacked pairs of "skate board" bearings. Pockets were sized so the bearings are a mild press fit. Short pieces of Acme lead screw and some timing belt pulleys complete the assemblies. One end of the four shafts were drilled and tapped for 10-32 machine screws to hold the shaft into the bearing packs. I used Helicoils to provide a tight grip of the screws that hold the corners to the bottom plate. Is is actually not needed, but as I was designing this as I built it, I figured I would be taking it apart multiple times. This proved to not be the case.
Top side shot before the upper deck has been installed. Top and bottom plates are made from 1/8" aluminum sheet. Simple belt tensioning setup eliminates the need for precise motor placement and/or belt calculation. Drive is a 2:1 reduction. The four neodymium magnets hold the table into position in the bottom of the cabinet and make it easy to remove if needed. Future plans for a rotary fixture made me think about the ease of which I could change the work area.
.
A dam made from Frog tape was used to contain the chips while drilling the hole to clear the stepper motor. I elected to have the motor protrude from the bottom of the table as opposed to hanging out to the side. This allowed the table to be larger in width. In the right shot, the magnetic feet have not been installed so I did not have to fight them while setting up the location and other issues. With the magnets, the table absolutely does not move.
.
Detail of upper limit switch mounting. Striker for the upper limit switch attaches to the frame of the laser. Right side shows the lower limit switch and adjustable striker. To maximize the travel the adjustable stop was implemented.
.
The strikers are make to be interchangeable depending on which work surface is being used. Neodymium magnets are used to hold them in place. The threaded rod allows fine tuning to the adjustment to achieve good focus on the part. When starting the laser, I run the "Home All Axis" command in the Smoothie Controller. The focus is set to the top of the work surface being used and during use the thickness of the material is entered into LightBurn. The Z table politely moves the required distance when the job is run.
.
The assortment of strikers is stuck to a metal bar behind the laser. The colors are arbitrary and were based on what was loaded in the 3D printer (or my mood) at the time. For different work surfaces or fixtures there will be a specified striker. In the case of a fixture that is placed on the bare bed or the honeycomb substrate, the Home All function is run with the specified striker. I also added a fence at the X0 and Y0 edges of the table.
.
