Scannermax 506s

[planters]

Hehe. I am the designer If you meant our mechanical engineer, he executes my vision.

That's great. But, aside from who does what, I take it my presumption is correct.

We've been doing it for more than 28 years. In the planetarium where I first worked, I implemented a full 45 degree set-back. For straight projection, 15 degrees is more practical. In any event, this technique has been available to everyone at all points in time. All someone would have had to do was think about the problem...

I did. I have only been wandering around in this hobby for three years and about two years ago I made a video describing this technique.

http://www.youtube.com/watch?v=a-PzKHnMS0s

My approach is a little crude, but at the time I was unaware that this had ever been done before or that manufacturers had been producing blocks that supported it. So, either I am brilliant ( my preferred theory) or this is pretty obvious and it's a shame that it has not been more widely adopted.

We'd get into a discussion to pick the best scanner and mirror set for the particular job

Agreed. When these are available I will be one of the first to take your recommendations and commission a system. I have a scanner-less projector built for just this purpose... waiting.

[The_Doctor]

...at the time I was unaware that this had ever been done before or that manufacturers had been producing blocks that supported it. So, either I am brilliant ( my preferred theory) or this is pretty obvious and it's a shame that it has not been more widely adopted.

Actually it's come up before. I modelled it extensively on an early version of SketchUp before Google got hold of it. I posted details here. Maybe the thread survives, but I nuked some a few years ago when I vanished that time. My version was to use either a 45°, 50° or 60° angle. 50° got the best compromise for clear aperture and scan angle but 60° was a close second and far easier to achieve with easily machined fixed mountings. I won't go into the details here, I talked at length openly with Bill (Pangolin) right here at the time, so there should be no doubt that this occurred. At the time I was surprised that Bill used only 15° off the original 90° when I saw that better compromises can be had with at least double that reduction. To be fair, I was looking purely at the optical situation, I didn't have to think of heat and mechanical construction problems. I was looking only at a best-possible case for the optical path.

In the end what matters is that galvos are efficient enough to dissipate heat easily without damage so that less demand exists for strict rigid moutnings. This may not be possible when aming for the fastest speeds, but with WideMoves, or the new Scannermax 506's this seems to be the aim, and it allows much more flexibility. The reason I went to great lengths to model it was partly the lack of hardware to try, it was to gain insights into what to buy. Also, it was based on the idea that a lack of opportunity to change anything meant I had to exhaust as many possible outcomes to prevent expensive error. But if you can buy cheap good quality scanners that dissipate little heat and allow 'cant angles' or 'setbacks' or whatever you want to call them, then it's better to figure it out empirically. After all, the mirror used to steer a laser onto a fixed height above a plane for combination in dichros sets an exact precedent for the means to direct the combined beam into scanners at whatever angle you want. In practise the aim is to let your fattest beam define required clear aperture, then pick your scan mirror angle and proximity to maximise the output scan angle. Work from source to output the way you currently align beam axes, so that errors don't creep into the process. The only time these scan system angles are a major issue is if you can't change them. If you can, it puts a lot less demand in specific mirror geometry. But like I said, it probably can't be this free (as in liberty, not beer) if you have to get rid of a lot of heat.

[planters]

[QUOTEIn the end what matters is that galvos are efficient enough to dissipate heat easily without damage so that less demand exists for strict rigid moutnings.][/QUOTE]

You want strict and rigid mountings for a number of reasons, but the most direct is that the position feed back is not inertial it is referenced to the base. If you watched the video I demonstrated the EMS 4000 scanners and the atypical mounting blocks that permit the angle to be adjusted in the field, but which are clearly less rigid than the typical billet blocks that are more familiar.

But if you can buy cheap good quality scanners that dissipate little heat and allow 'cant angles' or 'setbacks' or whatever you want to call them, then it's better to figure it out empirically.

I disagree. The value of the CAD work that Bill demonstrated is that the blocks have to be fabricated and the mirrors have to be cut: not too short and thin to miss beam or too fat and long to load down the rotor or interfere with each other.

I think you make the point. This is not a great discovery, it has been discussed before and it is a good idea. Cambridge does it. EMS now does it and Scanner Max will do it as well. How much you want to make a bet the clone'rs will soon do it too?

[The_Doctor]

Oh... by 'not rigid' I don't mean lack of stiffness, but flexible configuration. (Besides, I always assumed that position sensing was done entirely with respect to internal mechanics and that stiff mounting was a requirement). I just meant that if you can rotate the scanner body on one mount, and adjust the off-axis tilt of the other scanner, you can adjust the angles as desired. The problem is that the fixing, once set, would usually be ona mount that while stiff, lacks a big thermal path to the base plate, limiting the kinds of galvos this will work for. Eyemagics, WideMoves, Scannermax 506, great, but for a CT6215 you'd need a big hefty block deliberately made for something specific. That might be partly why Bill chose 15°, because if you go for more you rapidly start running into users who want less! The cloners will no doubt want to try this as you say, but they'll be beset by the same restrictions. More, if what Steve said about their coil-winding methods is anything to go by. I reckon they'll mostly settle for the simple life once they get a few burned galvos sent back by anxious punters.

EDIT:
About CAD, I did it on the same basis as your disagreement with my statement. But I'm not talking about galvo design. Or mirror design. Assume that's already done. Now, most PL members would rather buy some galvos and try them than spend the kind of time I did in SketchUp. When people choose a mirror size they find that they still have to gauge the scan angle they can get without spill, usually by empirical testing. That's what I meant, that for the end user, that kind of testing will usually be fastest. There are lots of small details on mirror geometry and mounting that affected my models. Stuff I took weeks to explore, stuff I could have seen in an hour or two IF there were things I could afford to try first hand. CAD is great for designers because they're making stuff that doesn't exist. For this reason an end user would only do it if it were easier than finding money for scanners. Or perhaps if they were very very interested in doing it anyway.. Ok, there is a third reason: it's a great way to start finding out how to minimise projector housing and aperture window sizes for best output angle... But even that can be faster done empirically if you have scanners.

[Pangolin]

My approach is a little crude, but at the time I was unaware that this had ever been done before or that manufacturers had been producing blocks that supported it.

Yes, although I did it some time around 1986 (and not so crude either-- I wish camera/phones had been around then) I too was unaware that anyone had done it in the past. Later, in a General Scanning brochure I saw that they had done the same thing, in a way really super similar to how I had done it. Since Jean Montagu really and truly invented the whole galvanometer scanner industry, I assume it was he who was the very first to do this. He also had some brilliant engineers working for him who might have suggested it to him. Jean is retired now (and I heard one person assuming he had actually passed), and some other engineers who worked for him are now 70 years old. So it's hard to get data on exactly who invented it.

So, either I am brilliant ( my preferred theory) or this is pretty obvious and it's a shame that it has not been more widely adopted.

Watching the video it becomes obvious that you are brilliant. There is no doubt about that! Really! No doubt! What you did, you did in a vacuum and that's commendable.

Regarding "not more widely adopted", I think the technique is really not so obvious. Plus I think people are simply taking the simplest approach. It's a whole lot easier to make an X-Y block completely orthogonal (only 2 setups, or perhaps even 1 if you have a fancy milling machine) as opposed to trying to incorporate some kind of "set back". Moreover, even mirror shaping isn't terribly terribly non-obvious -- at least not to folks who are brilliant

When these are available I will be one of the first to take your recommendations and commission a system. I have a scanner-less projector built for just this purpose... waiting.

I hope it's obvious that we've made special arrangements here and there when needed After all, Steve has a pair, as does Norty, as do several projector manufacturers and individuals. Get in touch with me directly and we'll figure out what you need and how to get you there.

Bill

PS: Planters, in watching your video and looking at the reflected image of one of the scanners, it looks like the "seal" has been broken. Did you open it up just out of curiosity?

[Pangolin]

At the time I was surprised that Bill used only 15° off the original 90° when I saw that better compromises can be had with at least double that reduction.

Ahh, but there aren't.... 15 degrees is really a kind of "magic number". If you go any more than that, you're going to re-reflect off of the X mirror. In fact, with the 3mm mirrors that Cambridge ships, they do indeed re-reflect off X at one angle. If you take a close look at all of the screen shots in that Excel spreadsheet, you'll see that we've made a special cut in the X mirror to "get out of the way" of the downward reflecting beam from Y. (Cambridge didn't do that on their simple 3mm mirror set.)

The best numbers are 15 to 18 degrees, and for 60-degree projected angle in all directions, 15 is the number. It's what General Scanning used, what Cambridge uses, and what we use. We've studied this extensively. Back in 1986 I used 45 degrees, but only because the entire projection was upward toward a planetarium dome.

In the end what matters is that galvos are efficient enough to dissipate heat easily without damage so that less demand exists for strict rigid mournings.

If you look at the totality of the Compact 506 specifications, you'd come to the initial conclusion that the heat results should be pretty similar to a 6800-style scanner. But it just doesn't work out that way... While Steve was at our office, I tried something that I had never tried before. We ran our "creation" demo at full size and full speed on the scanners. Most scanners would be rounding all to hell and power limiting, but we projected the whole thing at full size. At the end, I got up from my seat and put my hand on the X-Y mount. The scanners were ice cold! They were also sitting on a stack of circuit boards and then taped to my projector base (just as a convenient way of getting them into the beam path). Steve and I were both pretty pleased by that...

Bill

[Pangolin]

The value of the CAD work that Bill demonstrated is that the blocks have to be fabricated and the mirrors have to be cut: not too short and thin to miss beam or too fat and long to load down the rotor or interfere with each other.

When designing mirrors we start with only the mirrors in freespace. The scanners are added later...

I think you make the point. This is not a great discovery, it has been discussed before and it is a good idea. Cambridge does it. EMS now does it and Scanner Max will do it as well. How much you want to make a bet the clone'rs will soon do it too?

ScannerMax HAS BEEN DOING IT for a long time. Remember we showed this at Photonics West tradeshow last year. In fact, you might be surprised at the number of private emails it too me to convince Tom from EMS that this was a good idea...

Yes, I think people eventually come to the conclusion that this is a good idea. But some people take longer than others For cheap/easy manufacturing, I'm sure some companies will never do it, even though it brings performance benefits. Same thing with scanner mirrors. We have NEVER used the same mirror for both scanners, but nearly everyone else in the world does (at least all of the cheap guys)...

Bill

[Pangolin]

More, if what Steve said about their coil-winding methods is anything to go by. I reckon they'll mostly settle for the simple life once they get a few burned galvos sent back by anxious punters.

We're patenting our coil winding technique, and the fixture to make them. Within the patent application (which will publish soon, so everyone will be able to see it), we actually have photographs (it will be a rare case where there are not only illustrations, but also photographs) in there of other people's coils and our coils. Even the coils made by the best (hint -- longest in the industry) guys have turns that "compete for space" with other turns. The result looks like rush-hour traffic, with cars (wires) changing lanes and such. Coils made in your average Chinese scanner look worst, because they're wound by hand, and then after they're removed from the manual winding machine, they're further manipulated by hand to get into a final form. If you watch the video on the ScannerMAX YouTube channel, I talk about the heat required inside the scanner (and thus to be dissipated by the coil). Because of the fourth-power-rule, there is very nearly zero heat build up at 10K and 20K. Heat grows as a fourth-power of scanning speed. This shows why it's so easy to make a 10K or 20K scanner.

Once our patent application publishes, you'll be able to see what Steve saw with his own eyes, while looking at our coils under a microscope. They are PERFECT with every turn nested in between two other turns and really and truly (not just words on a brochure) "no cross points". (...if you haven't noticed, it bothers me when people exaggerate on their brochures...)

Bill

[The_Doctor]

15 degrees is really a kind of "magic number". If you go any more than that, you're going to re-reflect off of the X mirror.

It's 7 years since we discussed it, but I found that 4.5mm diameter beam would work to 45° off a 60° angle arrangement (i.e. 'setback' of 30°) if the mirrors were like those on WideMoves, a quarter inch wide. This encroaching of the X mirror was the thing I examined at greatest length. One thing that made me call off the effort was the realisation that if mirrors are shaped differently per axis for optimal momentum and equal tuning per galvo, then it likely won't work as well. I did say that mirror geometry was a significant point to watch. I should have said specifically that this 60° scheme would work well only on the assumption that the mirrors were thin, equal, with clipped corners like those on WideMoves and most cheap Chinese scanners, which does allow scope for a lot of people to have a go at this. Even the amount of 'cut corner' on the mirror may make a difference, but maybe not that much, in this context. You were working with a starting basis of mirrors being very different, so I didn't expect the same results. I hoped there may be a bit more than 15° but I didn't model it with much detail to explore it. I think I got about as far as making models of the smaller wider mirrors then started doing other stuff.

[The_Doctor]

When designing mirrors we start with only the mirrors in freespace.

Interesting. That's what I did. SketchUp models of disembodied mirrors at various angles for scan and orientation... I just never took it as far. Not even to the point of looking at the matching or inertia in different shapes. After all, I was just looking at optimising the scanners I had for best effect, and considering the mount I'd have to make for one of them if I actually used my idea.

EDIT:
Understood, re scan speed and heat. Not lost on me.. Is similar thinking, if not exactly so, to the resistance of movement in air at changing speed. Not linear. Re coils, I'll be interested to see the patent once ready. I have wound some neat coils by hand, but there's always the change to the next layer to screw things up, so long coils with few layers are always easier. I never found a perfect method, just some that limit the number of places where it can go nasty. But the smallest thing I ever wound was a high-permeability ferrite ring.. Unless you count an open core RF coil or two which I don't. I found that on a straight long coil it can be fast by hand, by controlling the feed angle of the source wire. What you said about scanner makers modifying previously wound coils is odd. Odd that anyone would do that, I mean. I always wind a coil in place directly to where it will permanently remain. I'd always start with that basis. Winding on any kind of former with more than one layer gets interesting at times, but I usually only wound one-layer coils, sometimes for making my own springs as well as electrical coils.

What you said about heat really interests me. While I know that thermal coupling between coil and case is the most important thing, and that slower speeds just don't get that hot, it is still impressive, the combination of speed, scan angle, and heat, as described by Steve, and your reply to me earlier tonight. I'm specifically interested on the accuracy which Steve says is very high, very linear and repeatable. These are the kinds of things I wanted to see, and was so disappointed in not finding in the WideMoves. Repeatability was ok there, but they hated starting up from rest, at any speed! Made a pig's ear of linearity. Steve said that circles with 5-6 are always circles. If by this he means that they are regardless of scan speed and angle, even if drawn small, then I'm convinced. WideMove only drew clean circles if entirely ballistic, anything slower and they'd get flattened at top, bottom, sides... if 506 solves this, you're welcome to send me a price (or estimate within a few tens of bucks) to convince me to start saving. I know it's an unorthodox request, but it would make me happy. This is maybe the one thing that might make me rethink a decision to quit lasers. I won;t get into building to fund them because that will be like spending to save, an illogical pursuit if I can;t add lots of value without spending to get it, but if I can get past my WideMove scanner disappointment by using the 506 instead it might make a huge difference to my whole thinking about involvement with lasering.