galvo mirror mounting theory question

[tribble]

So, I'm curious about something counter-intuitive (to me) about galvo mirror mounting. Every galvo I've ever seen has a rectangular mirror (maybe elongated octagon if it has its corners clipped) mounted to the galvo shaft with the long axis of the mirror parallel to the galvo shaft. That means the short axis of the mirror is available as 'input aperture' and gets smaller as the galvo turns. The beam typically does not strike the extremes of the mirror closest to and farthest away from the galvo.

Seems to me that if you mounted the mirror rotated 90 degrees from the typical fashion, you would have a lot more mirror travel available with less 'wasted' space for the same mass of mirror.

Obviously there must be a good reason this isn't done... but I'm just curious what it is. The only thing I can come up with is that mounting the mirror sideways increases the inertial moment of the mirror mass and may slow the scanners down. Even so, it seems a square mirror would be lighter with less unused area left and right of the input side and allow for faster scanning.

Even on the second mirror, the deflected beam does not 'appear' to take up so much room...

Who can enlighten me on this basic part of setting up a good projector?

Thanks!

[james]

The bounce from the first mirror onto the second mirror is where you need that length. Think about it. The first mirror is scanning in one dimension onto the second one. In most cases the galvo mirrors are exactly the same so they match mass and inertia. But that is not always true, sometimes the first mirror is just a little square.

James.

[-bart-]

The x and y mirror are identical in form so they have identical mass/inertia properties when the galvos are tuned. The angle of incedence for a galvo xy pair is best kept at 45 degrees for minimal geometric distortion.

[norty303]

Some higher end scanners do use different shaped mirrors exactly as you describe. I suspect the reason you don't see it on the cheaper scanners is due to the reasons stated above regarding difficulties tuning different weight mirrors. But they do exist.

[mixedgas]

Also the inertia is worse for a side mounted mirror. Any mass hanging farther from the axis of the shaft greatly increases the inertia. (Squared function if I remember right) You want the mirror inertia to be between 1/10th and 10 times the shaft inertia for best performance.

In the old days we trimmed the first mirror to be as short as possible, with 12K scanners, you needed every last bit of performance you can get. Now its easier to just match the inertias for tuning.


Steve

[tribble]

Interesting, thanks guys!

What are they making galvo shafts out of these days? I assume it's generally steel... has anyone tried using carbon fiber rod? Should be less mass but greater rigidity. Might be an interesting project on a cheaper pair of galvos in case something goes wrong...

[-bart-]

While we're at it, can anybody shed some light on asymmetric mirror designs ?
such as these
or these and these ?

[mixedgas]

[QUOTE=-bart-;248807]While we're at it, can anybody shed some light on asymmetric mirror designs ?

Roughly:

Reduced inertia and enhanced stiffness for larger diameter beams.
Those go hand in hand with the tilted scan heads that reduce tangent distortion.

See the galvanometer scanner chapter in "Laser Beam Scanning" Edited by G. F. Marshall, Published by Marcel Dekker, Inc., Ny, Ny, 1991 ISBN 0-8247-7418-3
The galvo chapter was written by Dr Montagu, one of the founders of General Scanning.
It will give you the equations for mirror geometry.

While I would not pay for a new edition, a used book place might get it for you cheap.

Steve

[mixedgas]

Interesting, thanks guys!

What are they making galvo shafts out of these days? I assume it's generally steel... has anyone tried using carbon fiber rod? Should be less mass but greater rigidity. Might be an interesting project on a cheaper pair of galvos in case something goes wrong...

The shaft is part of the magnetic circuit, on smaller, faster galvos, so switching to CF is a no-go. On a larger, industrial, galvo, its a maybe... I'm not aware of CF being that strong in torsion, either.

Steve

[planters]

While we're at it, can anybody shed some light on asymmetric mirror designs ?

That is exactly what norty was talking about. The low end scanners use identical mirrors for convenience. The mass/inertia arguments above are correct, but the overriding principal is that you do not want to be swinging any glass that is not reflecting light. This of course has to allow for foreshortening as the mirror rotates through its range. The OP's description of the typical CN scanner is in fact a pretty poor design. I believe it is a cheap attempt to duplicate an elliptical mirror assuming that at 45degrees AOI that we are scanning round beams (rarely these days). The X mirror will always require a smaller mirror and the galvo/amp could be down sized, but the X/Y speed equality means the weak link is the Y axis and it's this amp/motor that will limit the system performance.

has anyone tried using carbon fiber rod?

The rod has such a small rotational inertia, having a small radius that I wouldn't look there. Also, this is way to integral to the galvo. I think the place to look is the mirror. We want them big and fast (can't be too thin or too rich). Mechanically, they would be better if metal; aluminum or beryllium or even sapphire ( this last is more practical). They should taper in thickness from the central axis to the high velocity edge. And they should be shaped to better match the beam (probably a simple square most of the time). Such "super mirrors" could easily be after market for any scanner. This would help a lot.

Of course, better amps and motors are important and there are competing suppliers, but the all important "propeller" has been under engineered by all but the top end manufacturers.