I needed a laser cutter/engraver to engrave on a product that I'm designing. There are vendors that could do this, but there are several reasons that we wanted a laser in-house. We started looking around at available laser cutters in the 40W range with at least a 12x18 table. The cheapest start around $3,000-$4,000 for units from China, which have a lot of issues, and often need to be partially rebuilt. The cheapest US made lasers start at about $8,000, and generally around $15,000 for the specs we wanted. We knew that we could also design one from scratch, but the design would be time consuming, and only valuable if we wanted to get in the business of selling laser cutters. I'm also intimately familiar with the amount of iterations generally required to make something function better than a prototype, and I quickly changed my mind when I saw that the people at Buildlog.net had created an open sources design that had already gone through a few revisions. It's still an awesome experience to build something like this from the ground up, and know how everything works.
BuildLog.net is an open source laser cutter design community that has already gone through several prototype revisions of a laser design to arrive at the 2.x design, which is relatively well refined. There was still a lot of engineering and refinement that went into the construction of the laser. A lot of the small details were not well thought out. The base construction price for all the parts that we didn't already have came out to about $2,000, much less than anything that can be purchased.
Note: If I were to build this again, I would redesign it in a modular format, such that the electronics are completely separated in their own enclosure, and the mechanical system and laser is in another enclosure. This would make it easier to work on the electronics, keep them away from fumes, and allow the electronics to be used with other mechanical/laser systems, like a larger laser cutter gantry, or a plasma cutter.
Jelly fish flames wood box. Cigar box which was sanded, engraved, and then coated with wax/orange oil. I only learned later that the discoloration around the engravings can easily be removed with a wet rag. The box was made in 3 separate steps. Engraved the top, rotate 90 degrees, engrave the right side, rotate 180 degrees, and engrave the left side. Unfortunatly, pictures don't do it justice. It looks amazing in person.
Power: 50% (considering that 60% appears to be max power, ~20ma, while cutting)
Other side of the box.
I had a stepper motor and matching belt pulled from a surplus printer that fit the Z-axis bracket, but did not have the matching pulley. The belt is a 1.5mm pitch, which has tiny teeth. 1.5mm pitch pulleys are also not readily available, so I modeled one in Solidworks and printed it out on my 3D printer. I was unsure of how the teeth would resolve due to their tiny size, and the resolution of the printer. The printed part turned out to be functional. The teeth are rounded quite a bit, but the belt still engaged properly.
The stock air assist nozzle left a little to be desired. Being made from a chunk of aluminum, it was heavy, and the barb fitting was straight instead of 90 degrees (which made the air assist tubing stick out too far).
The DSP laser controller has an input for the autofocus sensor. I thought it would be handy to have the autofocus sensor integrated into the nozzle. The nozzle shown below is the result of these design requirements. It has a spring loaded tip. When the table is raised, and the part touches the tip, the electrical contacts inside the tip are lost, and the DSP autofocus input is triggered. The table then lowers a specified distance away from the tip to achieve proper focal distance.
The laser mounts supplied with the plastic parts kit from Buildlog.net were functional, but left a little to be desired.
Alignment systems that are not subject to high forces, and use 3 point alignment method, generally have 2 points that are adjustable, and the third is compliant. It prevents over constraints, and is easier to adjust, because only 2 points need to be adjusted, and the third automatically moves to compensate.
If the laser mount was designed like this, then it would be easy to align the laser without the worry of breaking it. Also, it eliminates the need to loosen one screw as the other is tightened.
I thought about wrapping the tube with a strip of rubber at the contact points, to allow for some compliance and prevent slip. However, this makes all the points compliant, and adjustments would not be as exact. Also, over time, the rubber can creep, and the laser could loose alignment.
Without knowing how strong the laser tube is, and not wanting to test that, I was concerned about using a spring loaded point. I also did not know what the required strength of the spring would be to keep the laser in place.
A third of the mount with one of the screws was cut off. Using an extension spring that wrapped part way around the tube would help distribute the point force. It also allows for compliance if the frame is twisted, and will prevent the laser from breaking.
After trying to use Mach3, I decided that the 2012 Laser DSP controller from LightObject.com would be a much better solution. However, I wanted to keep the parallel port input of the laser control board used for Mach3, incase I needed to use Mach3 for a custom purpose in the future.
The solution was to wire everything from the Laser DSP into a parallel connector, and run a parallel cable to the laser control board. This also reduced the amount of re-wiring to practically nothing.
Below is the DSP with the parallel connector mounted to a bracket.
In order to run the USB connector from the DSP computer, there would be a cable dangling out the back of the laser. This is not a clean setup, so I created a panel mount USB A to B connector. That way I can run a USB cable from the DSP to the inside panel mount connector, and then connect a seperate USB cable between the computer and the laser.
The DSP laser controller keypad arrives without a housing. The edges of the pcb are exposed, so they will be visible unless flush mounted. Using Solidworks and the 3D printer, I quickly threw together and made an angled housing for the keypad. The housing needs to be printed in 2 halves. The halves capture the keypad pcb and hold it in place. The 2 halves have 3 pins that align the 2 parts to each other, and several holes in the bottom for threaded inserts as a way of attaching it to the laser panels.
1. Aluminum Egg Crate Return: ~$25 http://surpluscityliquidators.com/view_product/152513/ for $9 + ~$15 shipping.
On the Buildlog.net forum, they report good results using this aluminum egg crate return. Supposedly, it's commonly used for fluorescent lighting fixtures, and from what I can tell it's the same thing that http://lazergrids.com/, but they charge a premium for cutting to size.
You may want to check a local electrical/hvac supply store. I found one in Atlanta, called:
1333 Logan Circle
Atlanta, GA 30318
However, I called them to inquire, and at $75, I think I'll risk shipping damage from Surplus City Liquidators. UPDATE: I purchased and received the egg crate return from SurplusCityLiquidators. It arrived quickly and in good shape. It's actually painted white, although the paint is not very thick. Total with shipping came to $23. Hopefully the white paint does not have detrimental effects.
Here's an alternate source if you need a lot (share with friends). It's 2 pieces of 2'x4' grid for $57.95+shipping = ~$75.
2. http://lazergrids.com/ - $80 Considering that it's the same as the Aluminum Egg Crate return, this is expensive stuff for the dimensions that I need.
3. Laser-links.net - $225 durable stainless links, but very expensive
4. Expanded Aluminum - $62.99 Available from McMaster.com. This is the standard "vector cutting grid" used in laser cutters. Fragile, and needs to be fully supported. Laser cutter companies often charge hundreds of dollars for this stuff.
The aluminum egg crate from SurplusCityLiquidators was cheap, and should work great.
Mach3 uses the Parallel Port to communicate directly with a CNC motor controller board. This requires that the software be able to take full speed control of the parallel port. If there are any delays, steps or other info will be missed.
Laptops have a lot of power management controls that can slow down communication over the parallel port. Here are the steps that worked for me to be able to run Mach3 on a laptop. In the end, the main problem appeared to be the processor wanting to sleep all the time, so the key was a program which kept the processor active and running at full speed all the time.
RightMark CPU Clock Utility was the critical component that got it working perfectly. It's possible that it's the only thing that you need to do, but I had already performed the previous steps.
The 40W CO2 laser tube being used for this laser is water cooled. Using surplus parts we already had, we went a little overboard with the water cooling system. It uses a radiator from an industrial dehumidifier, and a water pump from a surplus contact lens assembly line machine. A reservoir is custom built from a piece of 4 inch PVC pipe, and uses a flow switch as feedback.
Most inline red laser pointers require an extra mirror that is transparent to the CO2 laser wavelength, but reflects the red laser beam. The mirror is setup at 45 degrees to the path of the C02 laser beam. I don't really have much space to easily mount that, nor do I want to buy more mirrors and mounts. This device allows for an inline red laser pointer, without adding an extra mirror to the setup. The red laser drops down into the laser path when the door is opened (and the laser is off). It is designed for the Buildlog.net 2.x laser, but can be adapted to other designs.
It was made on my UP! 3D printer. If you don't have a 3D printer, let me know and you can buy a set from me. Here are the files if you would like to print it out on your printer:DOWNLOAD: 111023 inline red laser pointer STL files
CAD showing the door closed and laser in the up position. The door closed (laser interrupt) switch is by the lever part, not by the door. So the red laser has to be up and out of the way of the main laser beam in order to close the switch.
CAD showing the door open and the laser in the down position. Now the laser is inline with the normal beam, and will create a red alignment dot on the part to be lasered.
The laser tube mounts and mirrors must be aligned before the laser can be used. It's kind of difficult when the laser you're using is invisble and can burn something within a millisecond. I designed and made a mount for a red laser pointer to be used to help in aligning everything before inserting the laser tube. It is also useful for test running the whole machine prior to laser install.
The files can be downloaded here:Laser Tube & Mirrors Alignment fixture Solidworks and STL Files
The way this works is that the piece to the left goes closest to the exit of the laser tube, and the piece to the right holds the red laser at the opposite laser mount, close to the rear of the tube. The left piece has a pinhole in the center. The laser must be aligned to shine through the pinhole. This assures that the beam is properly aligned down the center of the laser mounts. Then the mirrors can be aligned. Once everything is set, one screw on each laser mount is unscrewed, the alignment tools removed, the laser tube inserted, and the screw retightened. Everything should be properly aligned at that point.
NOTE: The left print looks thin in the center, because it missed a few layers due to my error while swapping filament mid print. The machine print's beautiful... I'm the one that messes up.
The side and bottom panels are cut from 1/8inch HDPE. This material is inexpensive, easy to cut, and durable. The pieces I ordered from McMaster already came close to the correct size, so there was minimal cutting required. The top is Grey Tinted Polycarbonate, because I wanted the coolness going on inside to be visible. It's not as dark as I was hoping for.
An air filter is installed on the right side. Clean air is pulled in over the electronics. Hopefully this will reduce the amount of dust in the unit. The air filter is a standard 10"x20"x1" home air filter purchased from Home Depot.
I'm considering mounting the exhaust to the underside of the Z table. This way it can be used as a suction table also to hold flexible stuff flat.