They are always attempted to be focused in infinity. That means that the beam will stay parralel and thin all the way to infinity. In reality it is not the case, because of imperfection of emitter section, where KTP crystal transforms 1064 nm light into 532 nm light.
The constant spreading of the beam is called divergence. Measured in mRad, it shows how much the beam spreads, example: if laser's diverggence is 2.0 mRad, it means it will spread 2 mm for every meter traveled.
Most green laser modules are about 1.2 or 1.5 mRad divergence, which is as-is, you cannot change that. You can with external collimator lens, but the beam won't be thin 1.5 mm or 2mm (the part of the beam closest to you) but it will be bigger.
That, my friend, is the law of optics: the thinner the beam @aperture is, the greater the divergence.
You could get the 0.0 divergence, if the emmiter of the light you are trying to focus is infinitely small. We all know that is not the case. Chip's die in laser diodes are still 10x 30 micrometers (correct?) in surface. You cannot get the 0mRad in red lasers either. You can try that if you can have somebody on the other end of the street looking at the dot, while your laser is fixed in position. You will see that you cannot focus the dot in smallest spot for burning any more. As you turn the ring, dot will get smaller and smaller, until the point it will stop and spread out again. The smallest dot won't be small enough for burning. Not even close.
If you want a green laser that will burn matches (without focusing OFC), 150mW is the lowest power capable of doing that. But since they are a bit pricey, even 50 mW will make you a great green laser that will catch attention of everybody around!
