Coherent 40 Watt CO2 Laser

Not the laser we were working on, but a very similar one.

Dennis & Glenn have a desktop size, CNC laser cutting machine with a busted laser. Epilog, the maker of the machine wants $1500 for a rebuilt laser. That's a lot of scratch for a couple of guys fooling around in the garage. Let's take a look and see what we can find. What we find is a Coherent 40 Watt CO2 laser that is put together with about a hundred screws. I think the precision mind set that is needed to make a laser work got carried over into everything they did, so don't use one screw when 10 will do. The laser itself is a chunk of aluminum about 18 inches long and four inches square. Most of it is heat sink fins. Lasers are not very efficient, maybe 10% of the input gets converted to output. This laser produces 40 Watts of infrared radiation but consumes something over 500 Watts. All that other energy gets turned into heat.
    We were thinking that if we took it apart enough we would eventually get to the glass cylinder in the center, but 75 screws later all we had removed was the RF circuit board and it's solid aluminum case. We were down to an aluminum block with one electrical contact point in the top center, one sealed tube where the air was pumped out and CO2 pumped in, and one hole where the beam comes out. We stopped at the end caps.
    The end caps (see picture) are kind of odd. It looks like there are two small cylinders about a half an inch in diameter in two holes. There is an 1/8 inch gap between the cylinders and the holes all the way around. There are three screws arranged perpendicular to the center axis of the cylinders. This looks like the mirror adjustment. Looking at the end caps it occurs to me that the cylinders are not separate pieces but are integral parts of the end caps. The gap between the walls of the hole and the cylinder appears to have been machined out leaving just a thin piece connecting the plug in the center to rest of the cap. Adjusting those three screws will essentially bend the inner surface of end cap. Since the amount of adjustment needed is minuscule this could work. I'm thinking if we opened the end caps all we would have found would have been the RF antennae inside an empty box.
    To get it working, I think all that would need to be done would be to pump in some new gas.