Results 1 to 2 of 2

Thread: Closed Loop Galvo driving question

  1. #1
    Join Date
    Dec 2006
    Location
    Pflugerville, TX, USA
    Posts
    1,977

    Default Closed Loop Galvo driving question

    Do laser scanners work better by feeding them a series of points at a rate less than than maximum speed or by feeding them a continuous signal?

    For example, lets say I have a sinewave. Do the galvos have better response if discrete points are fed to it at 15kpps or would the galvos have better response if a the same sinewave of infinite points is fed to it (ie from a signal generator)?

    Also, what happens when there are big jumps? Let's say there is a case when the galvo needs to go from minX to maxX in one step. Let's say the max step speed for that deflection is .001 second (I don't know if that is realistic or not). Would it be best to feed it two points (the first and last). Or would it be better to feed it a continuous signal that increases from minX to maxX in .1 sec.

    I know that with closed loop scanners when there is a large step the driver will shoot a lot of current into the galvos to get them to the spot as fast as they can. But if the signal feed is continous there would never be a large jump in input... On the other hand, the driver is usually comparing position vs input so maybe it doesn't matter.

    To simplify things, I am just trying to understand what happens when you feed a scanner rated at 15kpps a number of points much great than 15kpps but still within it's step limit... In other words, if the physical limit for the galvo for a step from minX to maxX is 1ms, does it make a difference if you feed it 2 points in 1 ms or 1000000 points in 1ms?

  2. #2
    Join Date
    Jun 2005
    Location
    SoCal
    Posts
    490

    Default

    Provided you are not overdriving the scanner, a signal generator will receive a kinder gentler response from the scanner...

    That "provided" comes with a significant caveat... The ILDA tuning test pattern is designed to calibrate to an acceptable level of "squash"... So software or hardware which does the "right thing" may (and probably will) not look right on scanners that have not been tuned for it.

    It is always better to feed it subpoints if your pps rate will allow it. Slowing the scanner before commanding an abrupt stop or turn will also make the scanners happy and live longer. In an optimistic universe, incorrectly overdriving a scanner will
    cause the scanamp to either blow a fuse or limit the servo gain.

    If you go from min to max in 0.001 seconds, there is enough time in that period using a 30k scanner to put in 30 tween
    points (assuming you aren't using any anchor points, or more conventionally, 25 tween points and say 5 anchor points at the end)

    To answer what I think you're asking. There is no scientific limitation against feeding say 500kpps to a set of scanners and just making the point distance smaller. The general reason this isn't common on conventional laser boards twofold. One is because of the difficulty of acheiving a consistant and constant datarate between the computer and the laser board. At 1Mpps using 16bit X & Y, and 8 - 8bit color channels requires 10MB/s (that's bytes, not bits) of seamless glitchless throughput. Not a "difficult" throughput as it can be solved using sufficiently sized buffers. The second problem which is the big kicker is that one quickly runs into diminishing returns due to the resonant frequency being around 10 to 15kHz. This means that the fastest signal that can actually be effectively quantized is at 2.2x the resonant frequency or around 33Khz... The current marketting blurb you see about 60k scaners etc is just that, marketting. It allows faster "tuned ILDA" rates by limiting the scanangle to a very small angle and (to a lesser degree) improving the range and quality of the servo gain and feedback portions of the scanamp.

    So the quick answer is. Things are the way they are because they had to standardize on something, and 30k was considered good enough to work with, given the physical limitation of traditional galvos.

    Hope that didn't muddle things up too much.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •