galvo mirror mounting theory question

[tribble]

http://optics.org/article/34451

Interesting article. I notice figure 4 has some of that ribbing I was talking about in the OP, but it looks more molded in than cemented on. Sadly, figure 4 is not described in the text. I like your description of the tapered mirror substrate, but it's beyond my capabilities to fabricate such a thing. Perhaps this is an area beyond the reach of most laser hobbyists (although I am sure there are some out there with the equipment and expertise to investigate).

But, practically speaking... it seems that if mirror mass doesn't come in to play in a noticable way in at least some systems, we wouldn't see different speed specs for the larger mirror sets on, say, the DT-40s.

http://shop.stanwaxlaser.co.uk/dt-40...-dt40-16-p.asp

Of course, with high-end units already offering 60Kpps, maybe mirror optimization is not of interest to scanner manufacturers?

[norty303]

I wonder whether air resistance of a bigger mirror ('paddle') has a greater effect than the inertia, or that in combination they make the difference we see with bigger mirrors?

I do know that I hear a very different sound from my DT30's with 7mm mirrors compared to the same scanners with a 5mm mirror.

[-bart-]

Yeah galvomirrors in a vacuum chamber !
(not the entire galvo, that'll instantly toast them)

Would be interesting to see how fast one could tune a galvopair with the mirrors removed. (by looking at the feedback signals with a CRO.)

[mixedgas]

The possible solution is available in micro-machined Beryllium and Silicon Carbide mirrors. They can have the stiffening ribs on the back, or a tapered thickness etc
There are companies that make them on short order production. However it would be a group buy with a steep entry price.
They show up in FLIR systems and other very critical applications for larger galvos, mainly for the military.

The pure Silicon ones break if you sneeze. I've had a few surplus ones come in. I would not want to take them on the road. I realize the G-forces during scanning are probably sharper then handling impacts, but past experience says projectors take major hits in the field.

This is a downside of any very thin mirror. I have surplus bio-medical scanners with much thinner then normal mirrors. Again, not something roadworthy without machined ribs.

Cambridge does offer some specialty mirrors for this sort of thing. However I have not seen them on smaller scanners in a long time. Probably because the cost-benefit analysis showed the "D" shaped mirror of the right custom glassy material was a better deal.

The "Windage" as its known in the polygon scanner business WOULD show up on a oscilloscope measurement if you mounted the scanner in the vacuum and did a large step. Is it significant, but not very.

"Laser Beam Scanning" does have a few blurbs on math for calculating the error induced in the mirror surface figure during scanning. For short throw install, its not that relevent at all. On a long throw install, on the order of 400-700 feet, then you start seeing bearing noise and position sensor noise in the graphics, followed by mirror issues. I have seen this in one of my own shows (Long Ago) on a 150 foot tall screen at a 700 foot throw. But this requires a large frame ion laser or a Verdi class DPSS and a client who really wants big graphics.

Most people here would benefit simply by having better power supply stiffness. This comes from massive low ESR electrolytic caps within inches of the amps and larger diameter wiring with high grade connections. Wiring that avoids sharing ground returns between amps.

Remember, depending on the galvo, you can see surges of 4-6 amps peak at sharp corners. Any high resistance in the wiring fights what the amp intends to do during these surges. High means more then a few 10s of milli Ohms.

This is why I worry about people who are so proud of building micro sized projectors. They dont have room for the caps.

There is a nice pic of a machined metal mirror for a large scanner in the upper left of this picture.

http://www.camtech.com/products/mirrors.html

Steve

[planters]

I agree with Steve on each of his points. Better power supplies are SO MUCH easier than ribbed SiC mirrors. The actual edge velocities are low in M/sec and so windage is going to be negligible. But, the mirrors being used now a long way from optimal. Simply changing to sapphire and thinning them a little (same effective stiffness) is a no brainier aside from a source (and there are a fair number of them). Chamfering or better beveling the outer edge on the back surface is a relatively simple pre-coating step that would have a lot of benefit. Finally, I would get rid of the 45 degree angle cut on the two corners furthest from the drive shaft. This is where my nicely centered and fully filling beams spill off when scanning. The other two inboard corners are OK, in that they reduce weight and can't be used due to the shaft coverage of the mirror on this end.

[mixedgas]

Only thing with having mirrors made is the coating needs to be good. What to watch is how the reflectance spectrum changes with angle.
Make sure you have Rugate or similar style gradient coating stack to handle the angle change.

I mention this because I saw it. Brand "C" went through fits improving their early 20 watt show coatings, getting rid of the varying red leakage as the angle changed. You could watch the red in the image go up and down with angle. But in the end they did, and got the industry far less loss.

Now 0-45 degree 100% mirrors are the norm, but a few years ago they were still expensive.

Oh well, I laser to build and only two days to do it.

Steve

[planters]

Steve,

Good point. Coatings for 45 degrees +/- 10 degrees AOI is probably a good target. Typically angle tolerant coatings spec 0-45 degrees and the problems start at 45 +.

As far as the laser is concerned... sit ups, push ups, vanadate...sit ups,push ups, vanadate... now repeat. Can't wait!

[buffo]

I am a little skeptical of Bill's claim that the mirror's rotational inertia is on the order of 1/10 that of the rotor. Having seen the new Pangolin scanner motor, I was impressed with its small diameter and this combined with their approach to accommodate larger diameter (diode array) beams..

I believe his comment about the mirror's inertia relative to the rotor was for the base Cambridge 6800 design. The new Pangolin scanner is a good bit different by comparison, and yes, while I'm sure the rotor is still heavier than the mirror, the ratio of the two masses is probably a lot smaller now than it used to be.

At least in my experience with the EMS 4000 series amps and motors, running with the large aperture 14mm minor diameter mirrors, I begin to see significant performance compromise above 18K. I believe, but I don't know this for a fact, that the 7mm and 5mm mirrors with these same motors and amps as an example, do not preform the same in terms of speed (by a large margin).

Large mirrors are more of an issue today than they used to be, because the rotor mass has been reduced a lot compared to the old 6800 design (which dates back to the 1990's, remember). But yeah, Bill admitted that even his new scanners will take a performance hit if you fit them with the larger mirrors. And the new Cambridge 6215H's also take a significant hit when you upsize the mirrors.

As far as the design of the mirror goes, the first approach should be to minimize inertia and maximize stiffness by optimizing it's shape and dimensions and second to choose the best material that is practical for fabrication.

Very well put. So many Chinese mirror manufacturers skipped the maximum stiffness part! I remember in 2007 (at the first FLEM), Bill actually suggested adding a thin layer of epoxy to the back of my original DT-40 scanner mirrors to stiffen them up. When I did, sure as hell, the slight "fuzzyness" in the image went away. It was caused by *vibration* of the mirrors themselves. (Never would have called that...)

I wonder whether air resistance of a bigger mirror ('paddle') has a greater effect than the inertia, or that in combination they make the difference we see with bigger mirrors?

I have a feeling the air resistance would be at least an order of magnitude below the inertia....

Hmmm... The maximum small-step bandwith of a 30K scanner is 2500 Hz . The distance moved is probably less than 1 cm, but let's use 1 cm anyway. That yields an absolute maximum mirror tip speed of 5000 cm/sec, or 50 meters/sec, which is around 112 MPH.

OK, that's the tips, and that's assuming a 1 cm travel, which is definitely worst case. So what is the air resistance of a tiny mirror at 112 MPH? I can't see that being a larger force than what would be required to rotate the entire rotor (with a mass that is 30x to 50x that of the mirror alone) back and fourth at 2500 Hz... Sound reasonable?

I do know that I hear a very different sound from my DT30's with 7mm mirrors compared to the same scanners with a 5mm mirror.

Agreed. But this is probably due to the different mirror mass changing the resonance of the rotor, which would also change how the bearings reacted, etc, etc, etc. And there's no doubt that the scanners are working harder to move the bigger mirrors in the first place. I just don't think that air resistance is the dominant factor in the sound change.

Adam

[planters]

Adam,

Thanks. But, I take issue with your back of the envelope. Even a 1cm wide mirror scanning a across a 30 degree optical angle (15 degree mechanical) will have a maximum tip speed = to 1cm x 3.14/24 = 0.13cm x the frequency. If the full angle scanned can actually be scanned at 2,500Hz full cycle then you get 6.25M/sec. peak and 3.13M/sec. average. Right? At the same time the acceleration forces I estimate will be on the order of 6,500Gs; 6.25M/sec. x 1/(1/10,000 sec.).

Are these scanners actually capable of 2,500 Hz full cycle/full range?

[mixedgas]

Are these scanners actually capable of 2,500 Hz full cycle/full range?[/QUOTE]

Exactly once. ...

If Spec ever fixes the image upload, I'll post a curve for older galvo.
I do have 3 watts of other stuff plus a broken spark plug to do tonight, first....

Steve