Being quite new to the game, I did some research and tests, so I am documenting here what I understood as it might be useful for others.
Laser diodes are normally quite easy to use and, for low power and relatively high frequencies, they offer a good alternative to CO2 laser tubes.
For the application I have in mind I will need to use a 60W CO2 tube, but since I am not sure the concept is sound, before investing in that kind of material I wanted to proof the concept with cheaper and easier to use lasers.
There are two things that actually affect my test :
Obviously the power, for a cheap laser (about 300$ shipped, with driver, power supply and good protection glasses) there is no way you can compete witha 60W tube.
And then the frequency.
Non pigmented plastic is quite transparent to visible light, while it absorbs pretty well in the deep infrared range.
The diode I got is at 808nm which is a wavelength sitting in the "near infrared", it is absorbed by plastic materials only if they are colored, possibly black.
The deep infrared CO2 lasers work at about 10.000nm, which means they have a wavelength which is more than 10x the 808nm ones.
That makes a lot of difference.
Still, if I use black targets -which is kind of ok for my test- I should be able to achieve something.
Operating the laser itself is quite easy, just plug in the power supply connected to the driver (this one works at 12V) and it basically fires out right away.
WARNING : even few milliWatts can easily damage your eyes permanently, wearing proper protection goggles is mandatory.
No, it is not enough to make sure you are not firing the beam directly to your eyes, as the beam can bounce against any target and being scattered to your eyes, it just takes a small portion of the beam sometimes to be a big issue.
Also, make sure the goggles you use are specific for the frequency you are using : 404nm protection goggles would not help much if your laser is at 808nm.
Particularly, if you are using near infrared diodes, like mine, an additional danger comes from the fact that the beam is not easy to spot, since the frequency is almost invisible for the human eye (cameras can detect it way better).
That said, laser diodes can come with some different features :
Optics : depending on the power and frequency the lenses can be plastic or glass. For high power (more than 0.5 -1W) you normally have glass lenses, like mine.
Normally you should have some kind of regulation on the lenses, to adjust the focus, which is really important to properly use your laser.
Driver : Technically you apply the correct voltage to the diode and it emits the beam, pretty simple.
However a " driver" circuit is normally used to deliver a constant current to the diode (you can build yourself one of these circuits, there are plenty of schematics on the web normally based on the LM317 - just google the topic-) else your diode might have a short life span.
Power Control : The laser diode is an LED, like any LED it is possible to dim it using a PWM signal. For this reason many lasers come with two additional wires connected to the driver, they are labeled "TTL" and are supposed to be the PWM signal input.
I believe there is a pull-up resistor at the end of those wires (at least in MY driver) because if you do not provide any signal, then the laser is on 100%, if you shortcut them (grounding the signal, I take) the laser goes at 0% (off).
The ttl signal in my case is 5V, 3.3V compatible, so it can be driven by any MCU directly, the suggested frequency says >20KHz.
I completed my PWM setup (my stellaris pwm library here), but did not test it on the laser yet.
Cooling : Laser diodes tend to get hot and heat generates a lot of issues.
For starters it might actually damage your diode, but another effect is that the emitted frequency starts drifting off, presumably reducing the emitted power as well.
Typically you really want to run your laser at about ambient temperature, not exceeding too much 30C and if so not for long time.
Cooling becomes a real issue when you are planning to use the laser for long periods.
My laser is supposed to be able to operate at 100% for 30 minutes and then rest to cool down for 5 minutes, it just has a small fan.
I am staying on the safe side and operate it continuously for not more than 5 mins at the moment, but I ordered a few TECs (peltiers) which I plan to use to cool down the assembly.
Now, that said, the first thing I had to do, which proved to be a bit more complex than I expected, is to set the proper focus.
As you cannot see the beam when you are wearing protection goggles (and you DO NOT want to remove them), focusing the beam by looking at the target is pretty much impossible, maybe you can do something using a camera.
The other thing is that if you can chose the distance it is easier to set the lens focus and then move the target towards or away the laser until you get the desired focus.
But how do you detect the "desired focus" ?
I placed my laser on a table, taped down a reference with a mark every 5 cm, then moved some target objects on the reference.
The closer you are to the focus, the fastest the target object starts burning, I used a piece of wood for this rough measurement.
Once you get roughly the range in which you believe your focal distance is (between 40 and 55 cm for me), you can assume the correct value is somewhat in the middle.
I then took a black object and moving it I found that the best results were at 47cm, to prove that I let the beam punch a hole in my target, the smallest hole will be the one where the focus distance is optimal.
You can see the difference between the holes on the left and on the right, clearly the left one was way closer to the optimal distance.
At this point I knew roughly the range I would need, so I started building a stand that would allow me to run my tests in a more comfortable and repeatable way.
I came out with this :
It still allows me to move the diode up&down, is uses a glass work area where I can easily place my target objects and holds in place driver, power supply (and in the future the stellaris launchpad for the PWM)... granted, nothing fancy, but quite functional in the end.
One issue you may want to consider is how to attach the diode part as you may want to avoid something that prevents a good thermal dispersion (i.e. avoid to attach it to a wooden support that blocks the heat dispersion on one side).
Now I need to finish the interface program for the stellaris, it is written in Java and communicates via UART with the stellaris, the host computer will be an Olimex A20 running Debian Linux, that will give me a graphical interface to control the power, temperature drive eventual additional fans and peltier cells.
I am planning to record a video to demonstrate the rig, hopefully in few days.
Update : here's the video
