To the laser show software companies

[Pangolin]

Hey Bill..

..I'll put this in the que for an answer in ~2-3 mo's, since I know yer 'not on PL often', but that'll be 'bout when I'm ready for the info, anyway, so.. no worries..

AHA, yes, this is my first time back in I think 3 weeks. Anyone needing faster and direct service can contact me at Pangolin by email, Skype, Telephone, etc. I am sure there are many folks who can tell you that they called on the phone and got me directly. In fact, just tonight we had a client call needing a LivePRO license. He called Saturday at around 7PM and yep, I was here. Got his license right out...

Gotcha.. And, this is presuming the 'standard' mirs you'd ship-with.. However, see-below..

..Reasonable conclusion: This does-include the 506, yes? Please verify..

Well, we have a variety of mirror sets in stock and we can put virtually any mirror set on any scanner. We can even put a solid shaft on the Compact 506 too if we wanted to.

The '506 is a unique scanner. Because of its rotor dynamics it can handle unusually large mirrors. The picture in the datasheet with a 1 inch round mirror is not an exaggeration. We can get 1 millisecond small steps with that mirror, or around 2.2 millisecond full-field moves.

We originally made it to simply be a very light scanner suitable for use in medical handpieces. For those applications weight is the absolute most important factor, and speed is really not important at all. So we made this very light (13 grams) scanner, assuming that it would only find a home for the slow, lightweight applications. But when we cranked it up with normal Cambridge-style amps, we were surprised to get exactly Cambridge-style performance.

..if the "Saturn 5 is designed for up to 8mm mirrors", and we wanna go put a honkin' clam-shell mir-mount on it, a-la the 6220-style.. is this gonna create 'tuning issues'? ie: When you say 'UP-TO' (..as-if addressing a 'design limit'..) is 'up-to' speaking more-to the 'stated performance-aspect limits'? (ie: 'The Saturn 5 will give you up-to XXkpps with our 8mm mir'..) Or is it more like 'using over 8mm mirs is not-recommended.. at-all?' (oscillation-issues, etc..) - My point-being... If we wanted to use, ie: a 6220 clam-shell-style mount, on a Saturn 5, with, say, the 10mm-aperture mirs I referenced.. Would there be 'risk' with that much mass? (..even-if 'scan-speed' was necessarily-lower than max-potential.. which is fine for our app.. I'm concerned with max-accuracy / lowest-drift..) ..That all-said..

Hehe. Sometimes it's hard for me to understand what exactly you're saying

In any event, if we put a solid shaft on a scanner, you can put any mirror you'd like on it, including 6220 mirrors since they are made for a 3mm solid shaft. The only questions are: 1) The amount of torque that the scanner can produce, and; 2) the total system inertia (including both the rotor and mirror). Yes, "rotor dynamics" issues also come into play, as they do with literally anything with a rotor. However, our rotors are stiffer and stronger than others, which makes our resonant frequencies higher.

I suspect that with a Saturn 5, you'd get at least similar performance to what you're getting from a 6220. However, if you "do it our way" and not use the clam-shell-style mirror mount, and rather glue the mirror into the slot of our extended-back shaft, then you'd likely wind up with superior performance. If you study Cambridge mirror/mount combinations, especially those for the 6220, you'll find that the mount inertia is really close to the mirror inertia itself. So by losing the mount, the torque-to-inertia ratio is increased, allowing you to scan faster or wider.

..OK, but what is the 'difference' between the 506 'Pro', (with a DSL-amp or whatever would-be best..) and a Saturn 5? At least in-terms of 'accuracy' / lowest-drift, *with* the honkin' 10mm mir-rig, ref'd above, in-play.. I'm just after what would be "the best scanner" for the job... Again, not worried about 'cheapest', etc..

Pointing accuracy is dictated by the position sensor. So pointing accuracy of all scanners in our Saturn series, and also the Compact 506 with our Pro sensor would be the same. The big difference is speed and angle at which you can project. (When I say "speed and angle" I mean both at the same time.)

For example, as you'll see in the most recent video I made, the Compact 506 is running with our 3mm mirrors at 15 degrees. But when we use our mirrors for 5mm square beams that we'll be shipping to Norty (again "real values" here -- not just the size of the mirror) then the ILDA test pattern can only be projected at 8 degrees. If we increase aperture size, mirror inertia grows really dramatically. So the angle basically shrinks.

The Saturn 5 is a really long scanner, so it can produce a lot of torque. There are 83 laminations in a Saturn 5 but there are only 12 laminations in a Compact 506. So it produces much less torque.

..Oh, and "the job" is mechanical-blanking, with a 'v-block, vertical-retro' mount, 10mm-aperture mirs, that will 'blank' individual R, G, and B beams, *before* converging them into 'white'.. (thus, the need for 'max-accuracy / lowest-drift'..) Yes, the 'convergence', post-blanking will all be quickly-adjustable for the far-field RGB mash-up, but.. Obviously, the less 'pre-convergence galvo-drift' in the blanking-stage, the better, since that will = less convergence-adjustment, overall.. Am I making any sense?

Thanks for the answer, whenever..
j

The way you're talking about will require immense inertia, and also high RMS power (a lot of heat) in the scanner.

I think there's a much more clever way of accomplishing that. I'd be happy to discuss this with you either by Skype or email.

Bill

[Pangolin]

Pangolin; always late but sure as hell worth the wait!

Hehe. On that I will DEFINITELY agree!

Part of the reason is, after we start the development process, we keep rolling new ideas into it. QuickShow started as a "simple program", but now look at it? More than 400,000 lines of code and two full years of development (and the reason it wasn't more than 2 years is because it was built on a platform we already had slated for BEYOND)...

Once we saw how impressive the performance of the Compact 506 is, we wanted to develop a small, inexpensive "credit card size" amp. So that's where the idea started. But that's not where the idea ended... Even as the layout is being done right now, we're still adding things to it.

As Absolom said, it will DEFINITELY be worth the wait!

Bill

[Pangolin]

nice job bill, that looks great!

will you publish the turbo tracks mods so i can do the same with mine?

Generally here's what I am doing:

Make the feedback of the PD amps around 30K. Normally they are 50K to 100K but the signal coming from any of our position sensors is much greater, so the amp doesn't need to much feedback.

Do an additional change (single resistor) to increase the range of the "Position Scale" adjustment -- again, to compensate for how much "signal" our position sensor delivers.

Change the feedback components around the power amplifier -- normally configured to drive 4.5 ohms / 100 microhenry, and make it work with 3 ohms / 450 microhenry. Normally this is a single capacitor, or addition of resistor/capacitor in series.

Generally that's all that is needed. I did essentially the same thing with Norty's DT amps.

Bill

[absolom7691]

I know you're time on PL is limited but is there an eta yet for the new amps and how will the performance compare with the TT2 as displayed in the video?

[Pangolin]

Nice. If you can do that slowly rotating line thinger I mentioned, or any other unblanked slowly rotating polygon, I'm very keen to know how accurately they preserve the shape of straight lines, especially near corners.

I made another video. This time showing the scanners we'll be shipping to Norty, and driven by DT40 amps. Toward the end of this video I did the experiment you're mentioning here Doc.
http://www.youtube.com/watch?v=gImSsPJCxw4

If you're asking the scanning system to do what can be termed "ballistic moves", then straight lines may become curved. This is because of the X axis only needs to move a little, but Y axis needs to move a lot, then the X axis will finish it's move much sooner than the Y axis. Basically the X gets there sooner. But this only happens for "ballistic" moves (at times it can be said that the servo is operating in the large signal domain). When the servo is not executing ballistic moves, then the speed of both X and Y should be the same, even when the move distances are different.

Now -- the real question is -- WHEN does the servo exit the small signal domain and enter the large signal domain? This is THE key question. You see, you can take a crappy scanner, and get it to have really impressive seeming SMALL SIGNAL performance. But if it exits the small signal domain quickly and gets into the large signal domain, then basically you'll start to see the real limitation of THE SCANNER itself.

See, none of this is found on any datasheets, and it is only through much testing and experimentation that you can see the performance, and how the performance differs from scanner to scanner (and brand to brand). Indeed I've seen a lot of scanners that claim to be able to operate faster than 30K -- AND WILL (but only for small signals). Large signal performance is sacrificed.

I hope the videos demonstrate how even our cheapest scanner has really pretty decent large-signal performance, and how the transition from small signal to large signal is practically not noticable for normal imagery.

I know you're time on PL is limited but is there an eta yet for the new amps and how will the performance compare with the TT2 as displayed in the video?

Performance will be SUPERIOR! mostly because the T2 design is more than 20 years old and we've learned a few new tricks in the last 20 years In fact I was thinking about this in terms of progress. 20 years later we have a design that works better, weighs probably 1/50 as much, and taking probably 1/50 of the volume of space...

Best regards,

William Benner

[drlava]

Hey, It's good to see these moving along nicely. Can't wait to pop some into a micro projector. Nice job.

[White-Light]

The video of Norty's scanners just blew me away!

The way the ILDA patterns hold together at even insane angles is amazing as is the fact that the scanners don't simply melt down. Most scanners would be screaming like hell at 15 degrees and their rated speed, not pull out to 50+ degrees and still be quiet / not shut down, never mind maintaining an almost perfect ILDA pattern whilst pulling those angles. The scanners seem to have very impressive protection / reserves and a very linear response to angle changes.

Looking great Bill.

BTW, just out of interest are those the Pro or non Pro versions?

[norty303]

Looks really fantastic, although my insistence on the large aperture mirrors may have been excessive! They look massive!

I'd be interested to hear what Bill thought about the DT40 amps for driving these, and how they compared with the TT2.

Like you said Al, I like the fact these appear to be able to be overdriven/misused for short (and maybe long!) periods, as these may be used by less experienced/conscientious hands.

[Pangolin]

Hi Norty,

The design of both amps is really very similar, only because of history and where all of these designs came from (hint, look at the silk screen printing on the earliest scanner amps and see who's name appears there). BUT, the T2 can operate in what's called H-Bridge mode. This effectively doubles the amount of voltage that is available to drive the scanner. The DT amp is a simple single-ended driver. The result is that the T2 could drive the scanner wider and if I wanted to change a few more components (to make both more voltage and more current available) then the result would have been even wider. Right now the current is artificially limited to 4.2 amps on both. On the T2 I could have easily doubled this if desired.

I wanted to see how wide I could drive the big mirrors with the T2, but unfortunately the crappy cables Steve sent me effectively prevents that, due to a phenomenon called "coupling" (where the drive current couples back into the position signal). Coupling exacerbates any problems, any resonances, etc. I did what I could with the cables Steve sent me, and even completely re-wired them. Still, there are long patches where wires are sitting in mid-air without any shielding, and even the patches that are shielded are not really done correctly. So...

Norty sent me DT30 with supposedly wide mirrors (God I hate the misconception that mirrors that *look* wide can actually project beams over wide angles...). Those mirrors could MAYBE project a 3mm beam through a 40 degree angle, and plus much light simply went right through the mirrors. I took pictures. Plus, even with those "wide" mirrors, the scanners only produced the ILDA test pattern at 5.5 degrees. The inertia of our mirrors is perhaps four times as much -- maybe more, since our mirrors are made of silicon and around 50% thicker. Still we can project faster and wider, even with this tiny scanner.

Al, these are the cheapy scanners, not the Pro ones. The position sensor on our Pro ones presents a phase which is completely the opposite of how everyone else does it. Using a cheapy sensor allows us to simply use the scanner on existing amplifiers. That's actually one of the keys we found, to producing a cheap scanner.

Norty, please look at the video and confirm (preferably by email) whether THIS is the beam entrance you need or if you need the opposite. Then I will mount the new mirrors I received and get these right off to you.

Bill

[planters]

Bill,

Thanks for the updates. The videos prove once again that a picture is worth a thousand words. It will be interesting to see the progress as the "credit card" amps are incorporated and then a pair of the Saturn V motors are driven by some powerful amps and later when the Saturn V specific amps are used. I noticed that Scanner Max will be exhibiting @ Photonics West. What will the company be showing/bringing?