Building a laser cutter around a Coherent GEM-50L laser
To make the experience fit your profile, pick a username and tell us what interests you.
Finished up the cutting table, have not attached the fan to it yet, just need to make a couple holes to do that.
Now that's out of the way it's time to do some cut tests. For this I made up a little program that cuts and inch over and a quarter inch down and then steps up and goes an inch back and another quarter down at another speed increment. Keeps doing this for a few iterations so I cans where it cuts the cleanest. For this test I used some scrap .35mm latex sheeting, one of the main things I will be cutting with this.
I found that 100ipm at 12.5% duty cycle, 20khz, in Enhanced Pulse Mode made a nice clean cut.
I figured that if a dropped the PWM frequency down a bit I could use it to mark the latex without going though it. All I succeeded in doing was drilling a whole bunch of 100um holes in the latex. Whoops.
What I might try next is to lower the rep rate and feed rate down and use the higher power pulses to cut instead of a QCW beam, it might cut cleaner that way.
The thing is getting pretty close to what I consider done, still need to make an enclosure over the laser itself to keep crap out. I probably will just have something cut and bent though OSHCut.
Working on the exhaust now, I have some aluminum honeycomb coming a and ordered some extruded aluminum from Misumi to build the frame to hold it. For the exhaust fan I am using a fan setup that was pulled out of an old Spectra Physics laser diode power supply. A little over 200CFM. I took a piece of 1/2 acrylic and cut it to fit in the garage window frame and used the laser to cut the holes to mount a dryer vent exhaust to it.
Need to mess around with finding good feed rates for this, I was cutting slower than I needed to here.
Focus testing the laser. Made a little program that moves the z axis up .005" for every vertical line. From this I can find the narrowest line and find the focus offset from the nozzle tip.
Top test is with he 1.5" focal length lens. The bottom two are with the 3" lens. You can see how much deeper the depth of focus is with the 3" lens.
I got the laser connected to the mill and created a second profile in Mach to run the laser. Configured the Kflop output board to run the solenoids for air and to turn the laser on. Just using M3/M5 for laser on/off and M8 for nozzle air. Screwed the CDA (Clean Dry Air filter) to the wall and plumbed that into the air supply. Beam expander is installed and adjusted.
Ran into one issue. My mill is incredibly electrically noisy from the servo drives. I had the exact same thing happen with my telescope mount. Its causing the microcontroller to trigger randomly and turning on the laser. Needless to say not good. I threw in a couple .1 uf caps across the inputs and that seems to eliminated that problem.
Playing around I did find something interesting, I can actually cut thin metal. I managed to cut some .005" stainless and made holes in .015 steel.
I got the main head adapter mounted and installed. Machined it from some PPS plastic I had lying around. I did end up having to machine the nose of the 1.5" lens assy back a but to get the focal point outside the nozzle.
Also got the other head cleaned up and installed. The head was intended for 1064nm (YAG) and the optics were fried so I machined an adapter to hold a ZnSe lens in the old lens cell and bought a window to replace the glass one. I couldn't find a window in the diameter it was designed for so I took a larger window and ground it down. I made a wax chuck and used some mitee-grip wax fixture paper to attach the window to the chuck. Then used my Tsugami carbide tool grinder to ground down the outside diameter. Worked pretty well, some minor chipping but nothing even close to the beam path.
Last night I did some more work on the optics, got the basic alignment to the front plate done and also found it does have the Enhanced Pulse Mode that the larger version of the laser has. It sends twice the power to the laser which can increase the pulse output by 3x, so about 180w pulse out of the laser for this one. The PWM duty cycle is limited in this mode though to 50% so your average power is probably about the same.
I made up the block to hold the rest of the mirrors to the front plate. I just decided to use some Hysol 1C to glue the mirror blocks together. 1/2 hr later at 230f the blocks were nice and solid. I did have to make a spacer to clear the ATFR mirror block.
I also got some label tape for my label maker and made labels for all the controls on the control box.
I am getting to the point where I need this thing mounted to the mill to finish the alignment and figure out the rest of it. I knew I wanted to mount it to the dovetail for the ram, but not sure how. the dovetail is way bigger than any cutter I had and on top of that it's a non standard dovetail angle. First thought was come up with a plate and a couple round bars that would go in the dovetails, some bolts would push them into place, I had some 30mm round linear rail that would work nice. That is until a hack saw just bounced off it. Looking around I saw a chunk of some sort of continuous cast hollow bronze I have had lying around for years, I have about 6' of it. Perfect size to fit in the dovetail. Took a couple 6" pieces of that and faced and chamfered the ends. Had some 20mm thick aluminum plate to attach them to. I figured if I made one side fixed and the other with holes that fit the bolts tight at the bottom and counterbored so they are looser at the top and then have the hole centers slightly smaller than the center to center distance of the rounds in the dovetail that when I tightened the bolts on the loose side it would draw it in and tighten it all up. And boy does it. It's rock solid. I have another chunk of aluminum plate that will attach to this to make a shelf that the laser assembly will sit on.
The beam expander came in and with that in mind I laid out the beam path and machined the base plate tp hold everything. The plate has a hole for the water fittings and then mounting points for the second fold mirror, beam expander rail, and for the final beam expander optic. I did mess up on the location for the second fold mirror but easy fix, though I did snap off the drill on the last hole, three screws are more than enough to hold down the mirror mount so no biggie. Got the head mounted using spacers and some spherical washers to make sure there is as little stress on the head as possible. First two mirrors are installed as well, next step will be aligning these two with the hole in the end plate and then installing and adjusting the beam expander.
Machining the base plate
Mirrors mounted ont he kinematic mounts.
Head mounted to the rail as well as the end plate and the beam expander rail
First and second fold mirror.
Did the final debugging of the control box. Electrically everything was pretty good. The only issue I saw was that when I set a 50% dutycycle it was around 7% off and this offset was across the range. So I fixed that and software and it all works now. Hooked it to the laser and it works. Turned it down to 10% and was getting about 6 watts so that's good.
Next I need to test the ATFR mirror position, I took a couple beam tube ends and put a small piece of tubing in between and stuck the ATFR on the mirror and then held the power meter head on the exit. Held it horizontal where it absorbed almost everything and then vertical were it passed almost everything. That verifies that the mirror works like I thought and will be the second to last optic in the beam path.
Now I need to start laying all that out, I have a 7" wide x 1/2 thick piece of aluminum I am going to mount it all too. I have a beam expander set from a friend who has a store on ebay coming as well, once I have that I can figure out how high off the plate I need to mount the laser to clear the hose fittings and get everything else lined up.
I got the optics mounted. They are attached to an aluminum stud which mounts inside the mirror block. Machined a fixture out of scrap plastic to get everything aligned. After that I coated the optics with a special polymer optic cleaner that is also good for storing optics.
There are two normal gold plated silicon mirrors, one phase retarder and one ATFR. The ATFR and the phase retarder are what make up the cut cut enhancer. The ATFR reflects in s-polarization and absorbs the 90degree off p-polarization. The beam is reflected off this and hits the phase retarder which converts linear polarized light into circularly polarized light. When this hits something reflective when cutting the beam bounces back and is converted from circular to p-pol which gets absorbed by the ATFR.
Since the ATFR can absorb several watts I made a custom holder that uses some silver heat sink paste and a couple springs to hold the mirror in place against it.
A friend gave me a nice coherent Gem-50L co2 laser and so I could build a laser cutter. It’s a pretty nice laser, rated 50W and doing 62. Datasheet here: http://www.istcgroup.com/pdf/GEM_50L.pdf
I have decided to make it as an attachment for the mill which will attach to the dovetail on the ram.
First thing I need to do is get the laser under control. I need to be able to run it at low power to check the orientation of the ATFR mirror that stops back reflections from the work. It needs to go in a certain position relative to the orientation of the polarization. Of the output beam. If it’s 90 degrees off it will absorb 97% of the beam instead of reflecting it.
To control the laser I am building a small box that handles the enable, mode, and PWM signals that the laser needs to work. For this I reused a little Teensy 3.1 board and pcb I made to attach to the back of a vacuum fluorescent display. I made another daughter board to attach to it that handle the IO. I really should have just designed a new board and send it out to JLCPCB with amount of time I reworked and debugged finding and fixing he mistakes I made when I laid out the Pcb years ago. I still might to clean it up. It’s a mess.
Become a member to follow this project and never miss any updates