Maybe @JohnJohn can split this thread… I don’t think, as users, we can do anything about it.
Nobody with a machine similar to yours is going to read this or at least there will be less, everybody will think it’s for an AtomStack…
You have never mentioned what speed you are running.
Most of these machines have an lps (Laser Power Supply) that creates a high voltage to cause your tube to lase. This is for a DC excited tube. It has to go from a low voltage to up around 30kV to cause your tube to lase.
This response time of the lps will determine how fast your laser can fire…
All of these I’ve seen for these machines, if stated, is 90% of the placard voltage <= 1mS. If it’s 1mS then the best you can expect is to be able to toggle the tubes output is, in 1S, 1S/1000 (= 1mS) or 1000 times a second for 1000dpi.
You can see that speed and response times are important… I don’t know of any dc excited, co2 that can respond any faster, they might exist somewhere…
Agreed, if you look closely at your ramp test, I think you’ll see it’s acting like it’s pulsing instead of a continuous wave (cw) output. If you can interpret the dot pattern, I blew it up…
I think your ramp test indicates you have a problem from the lens back towards the tube.
I have a compound lens, that’s not as short as yours and it’s a much more fine line. Yours seems like there is an issue with the beam, alignment or it’s cleanliness.
I’d start at the beginning and when I get to the problem I’d stop and fix it. 
- How long have you had this?
- full alignment, including the beam pattern before m1 (mirror 1) TEM mode. The beam must be in TEM0 resonance for it to operate properly.
Do you know how to align it or confirm it’s alignment?
Lets hope this gets moved, so you get feedback from others…
Good luck
