Scanner Mirror Angles

Working on a new idea here..
Ok, everyone knows that we lose a lot of "Big Red" because of the 45° scanner angles, right? Ok, what about a near zero degree angle design?
Thoughts?

Attachment 40214

Old idea. :)

From notes I made 6 years ago (And I'm not the first to try)...

Orthogonal (90°) Standard Geometry:
-------------------------------------

The mirrors are 6.35mm wide (1/4"). The layout is optimised for an optical scan angle of 45°, a mechanical angle of 22.5°, or +/- 11.25°. The minimum length of the second mirror is limited by the minimum separation between the mirror centres (and wanted scan angle), in this case 7mm so that they can rest in any orientation without collision. This layout with mirrors 6.35mm wide will allow a maximum beam width of 3.5mm to scan through 45°. The ratio of 3.5 to 6.35 is 0.551.

Non-orthogonal 60° Scan Geometry:
-------------------------------------

This uses the same mirror width of 6.35mm, with the layout optimised for an optical scan angle of 45° as before, but the beam path is based on 60°, not 90°. As the first mirror is placed forward and raised, it risks intruding on the clear aperture of the lowest horizontal scan. This complicates the choice of mirror separation. The mirrors rotate around their centres of mass, so the front surface reflects from a point that shifts left when deflecting right, or up when deflecting down. This helps the clearance, which is good when the mirror axes are 8mm apart and when the first mirror is correctly aligned front edge to front beam edge.

The incident angle to the mirrors is closer to the normal, so the beam need not spread so much on the surface, and the clear aperture increases to 4.5mm. The ratio is now 4.5 to 6.35, or 0.709. This is nearly a 33% increase.


Aiming for less degrees than 60 is pointless because the first mirror will intrude so much on the other's output space that no business results. Also, with 60 degrees it's fairly easy to do. The first scanner needs to be tilted so a new mounting block is needed, but the second would just sit as usual but rotated 30 degrees from normal orientation. Any round clamp makes that easy. Less easy is getting the beam to the first mirror because it must come in at two angles, one slightly forward, the other slightly downward, in combination. Nothing that can't be done though, with one mirror in a good mount capable of that steering.

Also, you might get by with 50 degrees, I worked that out too, but it doesn't take much reduction before it really impacts on your total scan angle.

Hmm, interesting. Thx for the information Doc . I'll have to keep working on it :)

There's an old thread somewhere on PL. Bill Benner was talking to me about this thing on it. They use a 15 degree reduction in first mirror angle but for issues related to mass, inertia and speed. Which I never entirely figured out because in many posts on PL there is much talk of the mirror inertia being the least of it, more concern given to rotor weight.. (My guess is they are trying to match inertias so that scan speeds draw a line the same way in either direction to avoid inaccuracy, something not so easy to do, when your mirrors are not identical..) Anyway, when scanners (as many do) come equal with thin mirrors on thin shafts, all that goes out the window. :) So people might as well do as you want to do, just to get some more clear aperture at the expense of a bit of scan angle. And 60 degrees is about as easy to model and make as it gets, all other choices after 90 degrees considered. I'll guess that for long range beam shows with narrow scan angles, 45 degrees might be fun. I don't think I ever modelled that though, so I have no clue of actual figures involved in the compromise. Whatever you do, aim to use mirrors with corners cut off such that you capture the beam well on the first, while also preventing its corner from fouling the reflection off the second. That's the most critical bit.

EDIT:
I did find some more info... Two bits. (Might be all its worth, so caveat emptor...)

First:

Code:

---

Aperture    Scan      Beam
Fraction    Angle    Diameter

Full RA    53.4°    4.5 mm
0.75 RA    61.0°    3.375 mm
Half RA    68.4°    2.25 mm
0.25 RA    76.0°    1.125 mm

(RA = Rated Aperture)

---

I can't remember the anchoring context for that, but if you take the ratios for scan angle and beam diameter, and adapt them to some known scan mirrors (both being equal), you might be able to predict some outcomes. I think this might be useless, as likely only applies to 90 degree paths.

Second:

Quote:

---

The 90° array allows wide angles, but poor light gathering. The clear aperture is extremely limited with large scan angles.
The 45° array is harder to build, limited to 45° optical scanning, but the clear aperture is over 5 mm on a 6.35 mm mirror.
The 60° array is even more awkward, but is a good compromise, allowing 60° scanning of a 4.5 mm beam on a 6.35 mm mirror.

It is best to make the 45° array and give it the capability for scanning 60° or more, because a narrower beam will allow
much wider scans to be made. The rated scan width is only a limit at maximum aperture.

---

My opinion about those angles changed later, I'd now favour 60, not 45, because at the time I was assuming that DPSS with thin beams would take off, but we now know that diodes with fat beams are more often used. So the two lines in the first paragraph give the big clue: They start with a (weird?) assertion of the scan angle being chosen to be the same as the beam path angles, but it does show that narrow beam path angles can help with fat beams (which you already know) but that the gains quickly fall off into 'diminshing returns' territory if you push too hard. But 45 degrees for 45 degree scan angle, using a 5mm beam on a quarter-inch mirror can't be all bad. I'd say it was worth machining a mount just to try it. But unlike the earlier notes, I have no word about mirror separation, but I think I always assumed that avoiding any chance of collision was vital, because my scanners could settle at any position if power failed. So allow a 0.5 mm gap and this scheme might actually work. BUT, the notes in my first post were based on some detailed modelling that took into account the cut corners on mirrors like those on WideMoves and many Chinese scanners. These one in this post are older, and weren't.