Bill's 30 minute video is here: http://www.youtube.com/watch?v=BiaM0adidIo
At the start it describes very clearly the KPPS relation to scan speed. Scanners are physical things, so the physical quantities are specified. Once you know those you can deduce what is possible. Tuning the amps for gain and damping to get correct display at 8° is based on these physical details, not the other way round (If you used a heavier mirror you'd have to tune them again).
In other words, stating a KPPS value first is putting the cart before the horse. Worse, the comparison is meaningless if you can't verify that the scanner's physical specs are even capable of what is claimed for KPPS. So Bill starts with the raw physics, and works from there.
At the start of that big video is a very nice description of the link between KPPS and raw scan speed capability. (Best I have seen yet, as it happens). I'll paraphrase it, but in my own way so you can compare this post with what he says. Think of a square. The distance along one side is exactly the reciprocal of the square root of 2 of the distance between diagonally opposite corners. This is useful because the ILDA pattern draws the circle OUTSIDE the square, touching the corners. When you speed up the scan to the point where it touches the INSIDE of the square at each edge, the circle's diameter has shrunk by exactly the reciprocal of the square root of two, which also happens to be the -3dB reduction in original amplitude (scale) which defines the cutoff point of a filter. Scanners have mass, they act as low-pass filters when their inertia resists motion, so this is entirely appropriate. There are TWELVE points drawing that circle, so the frequency at the -3dB point for the 8° scan angle is the KPPS value divided by 12. It's linear, and exact, so 1KHz for 12K scanners, 2.5KHz for 30K scanners, etc.
Now, the resonant frequency of the scanners will depend on the physical abilities of the scanner, torque etc, first, and on amp tuning (gain, damping), second. I imagine that any scanner tunable to 30K at a wide angle may be tuned to 40K at a smaller angle, but how much smaller, and how good it looks, depends on its torque and heat dissipation and a lot of other physics.
It may be hard to compare scanners in detail unless their torque, and any resonances, etc, are well measured and known. So the only easy decision to make is: do not trust scanners for which such important physical data is either unknowable, or is widely varying between two supposedly identical scanners. You can maybe trust them up to a point, but Bill's already said several times recently that it can be a lottery regarding lifetime, large signal performance (ability to make large jumps in a well-controlled manner), etc.
That's a rough guide from me. This is new to me too, a lot of it anyway. But that's partly why I did it. It IS fairly easy to grasp so long as you put the peices together in the right order. Don't start with a declaration of KPPS because it's the wrong end of the stick. The only reason that 30K is often quoted first is because people chose it and stuck with it as a standard, And the only reason they did that was to make sure a show made for one 30K scanner looks right on another 30K scanner. And given that doubling to 60K is equivalent to doubling frequency, that is just one octave, not as dramatic is it seems because to go one more you'd have to double again, four times the original frequency. Not linear! In other words, 40K is not one third of the way between 30K and 60K! It's less, in all practical considerations. Trying to make a 30K scanner design do 40K costs more than it gains, because a 30K scanner is built with an entire set of physical qualities aimed at exactly that. To make a real 40K scanner still isn't worth the hassle of going back to the drawing board, and a derivative of an old CT6800 30K scanner won't do it right even if the copy was exact.
This probably isn't the neat short answer you wanted, but you'll gain more by looking at this stuff till your expectations change. Make sure you get that bit about the circle, the 0.707 (70%, -3dB) in the ILDA pattern because it's neat, easy, and if you can base the change in your thinking on that bit the rest will probably change more easily.
If anyone who knows this stuff well spots a flaw in what I said, feel free to spell it out because I'm trying not to misguide either myself or anyone else.
Reply With Quote
Originally Posted by dnar
