Oh, I can say a lot about this.
The parameters you need are the 1/e2 beam diameter (in mm) and divergence (in mrads, which is just another unit of angle like degrees). It would need to be very clear whether the mrad figure is for the half angle or full angle. Half angle is the angle from the centerline, full is twice that (side to side)
The high quality RF CO2 lasers (usually metal, though there is a ceramic type too) always give detailed specs on dia and divergence. Unfortunately,I have never seen any mfg of a glass HVDC-excited laser provide a beam spec on their tubes.
It is actually pretty low, though. I can say divergence on a “130W” Reci (glass tube mfg ratings are fictional units) isn’t very significant on a 1400mmx1000mm bed (2.4m travel)
What would the beam being “too big” mean?
Well, first off, a beam doesn’t inherently lose cutting power as it gets larger. And attenuation (loss of energy due to the air absorbing some energy) is not significant for clean air, but smoky air is another matter.
What happens is the beam assumes a larger dia when it hits the output lens. This does something opposite of what you’d expect, due to “diffraction rules” and weird “light is a wave” math-y b.s. When you double the beam diameter on a 2" focal length lens, the beam STILL focuses at the same point 2" from the lens. But, picture a 5mm dia beam hitting the lens, refracting, and forms a light cone 5mm dia 2" long, then if it continues past the focal point it forms the same cone inverted. Versus being 10mm at the top. Doing the math, a 5mm dia cone 2" long means the focal cone has is tan-1((beam dia/2)/2inch)= 5.62 deg half angle. 10mm is 11.24 deg half angle.
There’s 3 important effects here:
- The beam doesn’t actually make a focal “point”. It has a beam waist with its smallest dia at the focal point. And oddly, right near the focal distance, the cone stops coming to a point as instead has a nearly constant dia for a short distance, which mostly dictates how much your focus can be off and still wok, the “depth of field” or “depth of focus”. DoF scales with the focal length of the lens divided by input dia of the beam . So the 10mm beam has half the DoF. It’s very picky about the focus being exact. Warped wood naturally creates focal problems.
- Here’s the hard part to believe- the beam waist dia is dictated by well-defined diffraction rules. The smaller, the better, in theory. The beam waist dia scales up with the lens focal length, but also goes DOWN when the beam dia entering the lens goes UP.
So, the wider 10mm beam entering the lens produces half the focal spot diameter. The area of that focal spot decreases by a factor of 4x and the energy density increases by a factor of 4x.
Incidentally, this is exactly the same effect you get from switching a 2" focal length lens with a 1" focal length lens, except the 2" lens with a 10mm beam dia will still focus 2" from the lens regardless of beam dia.
But 1" lenses are only good for surface engraving on aluminum with high resolution, and not much else. But the loss of DoF is not the only problem.
The thing is, CO2 lasers cut depth by a mechanism few people understand- it must be able to form a channel with near-vertical walls, then the laser energy is hitting the walls at very shallow angles and actually reflects off the walls losslessly down the channel, making a straight channel of about 0.2mm dia. But, to get captured, the critical angle is about max 9 deg from vertical. This comes down to a 2" lens with the beam dia of most glass CO2 lasers. 1.5" and 1" lenses- or a smaller beam dia on a 2"- can do fine engraving but cannot form a cut channel so they can’t cut. A 4" focal lens sounds better for capture angle, but it won’t cut as well- the focal spot dia is 2x and energy density dropping to 1/4 makes it bad for cutting most materials (thick foam is an exception where the 4" lens works)
Anyhow, the limiting factor is actually optics. You must get all the beam energy within the 20mm dia of the lens, minus the area where the mount holds it- more like 18mm. Why do we use 25mm mirrors with a 20mm lens? Well, the mirrors are used at 45 deg angles, so sin(45)*25=17.7mm before the beam is off one edge of the mirror. You can get 24mm-25mm lenses and larger mirrors, but they’re more expensive and the mirror mounts and cutting heads that hold them are notably more expensive.
And FYI the 1/e2 beam dia is not that ALL of the beam energy stops past that dia. It’s a cutoff point chosen for “most of it”. There’s still usable energy a bit further out.
So, is the limit 18mm? Or, to be able to capture all the energy, does the beam need to have a 1/e2 dia smaller than 14mm? In practice, that’s not what limits it. It’s how well you can align it. It can be VERY touchy to stay aligned at 3000mm, but with a good, stable frame, and good, stable mirror mounts and high accuracy linear rails, yes it’s very possible to build even larger. Also note that the timing between the two parallel belts (usually that’s Y, but would be X in your #2 option) has to be very low tolerance. If one’s looser than the other, during accel the length will stretch a bit and it won’t be at the perfect 90 deg so cutting corners at high speed at the far side of the gantry becomes problematic.
Bottom line, a 20mm lens has a max 1/e2 beam dia of about 10mm if you can keep alignment near perfect.
If you hit the lens off-center, it will still focus at the same point. So it vectors off at an angle to get there. It will form a slant, noticeable on thick materials. Won’t matter for cloth.
A “beam expander” does a neat trick- it will trade off to larger beam dia, but with less divergence. A 2x expander with a 1/e2 beam dia of 6mm at 2mrad divergence becomes 12mm with 5mrad divergence.
That’s generally worse for a glass CO2 laser. Metal RF-excited CO2 tubes typically have a smaller beam dia but worse for divergence, it will widen and cut very differently at 1000mm vs 0mm.
But, does it have to be that way? My technical experience with a focusable flashlight says otherwise. If you increase the distance between two lenses, the beam converges.
Your guess would actually be correct! If you can change the distance between lenses, the beam expander is now a “collimator”, and you can take out the divergence, even make it converge, though, as you may have noticed with the flashlight, it will not be able to converge to a small point.
ZnSe CO2 laser collimators are rare. Only one company actually makes them for general use. There’s even a 1:1 collimator- basically the same beam dia out, but you can trim the divergence. You generally don’t need one for glass HVDC tubes, you need them for RF-CO2. A >3000mm path length actually could be a case for using one with a glass HVDC tube.
) all of it.
