Scannermax 506s

Quote:

---

Originally Posted by The_Doctor

It's 7 years since we discussed it, but I found that 4.5mm diameter beam would work to 60° off a 60° angle arrangement (i.e. 'setback' of 30°) if the mirrors were like those on WideMoves, a quarter inch wide.

---

Quarter inch wide is 6.35mm. A 4.5mm beam reflected off of a mirror whose angle of incidence is 45 degrees will make a spot that is 1.414 times (secant of 45 degrees) the 4.5mm on the mirror itself. 1.414 * 4.5mm = 6.36mm. In other words, with the scanner mirrors sitting in a neutral position, you actually have a little bit of overfill from the very start. When you rotate the mirror to an angle required to reflect it to a peak-to-peak angle of 60 degrees, it means the mirror must be twice as wide as the incoming beam size. So in other words, the MINIMUM SIZE of the X mirror would have to be 2x4.5mm = 9mm, and this does not account for any kind of "chip margin" on the mirror since the coating can tend to flake off the very edges of the mirror...

If you've got your Sketchup files or some pictures that you can show, these would be very useful in the discussion, and you'd have a shot at proving me wrong. Otherwise the screen shots from Solid Works, along with the fact that the two largest companies that make laser scanners use 15 degrees should give some degree of confidence that 15 degrees is about as good as you can get for a 60 degree scan angle.

Bill

Quote:

---

Originally Posted by Pangolin

Quarter inch wide is 6.35mm. A 4.5mm beam reflected off of a mirror whose angle of incidence is 45 degrees will make a spot that is 1.414 times (secant of 45 degrees) the 4.5mm on the mirror itself. 1.414 * 4.5mm = 6.36mm. In other words, with the scanner mirrors sitting in a neutral position, you actually have a little bit of overfill from the very start. When you rotate the mirror to an angle required to reflect it to a peak-to-peak angle of 60 degrees, it means the mirror must be twice as wide as the incoming beam size. So in other words, the MINIMUM SIZE of the X mirror would have to be 2x4.5mm = 9mm, and this does not account for any kind of "chip margin" on the mirror since the coating can tend to flake off the very edges of the mirror...

If you've got your Sketchup files or some pictures that you can show, these would be very useful in the discussion, and you'd have a shot at proving me wrong. Otherwise the screen shots from Solid Works, along with the fact that the two largest companies that make laser scanners use 15 degrees should give some degree of confidence that 15 degrees is about as good as you can get for a 60 degree scan angle.

Bill

---

Well, you're right, I dug out some old files to answer and found an error in my memory, the scan angle was 45° not 60°, but the other figures are correct. Does that look more plausible? :) I've attached the SketchUp files too, and added the original text notes I made below. The file "60° Scanner Layout.skp" in the RAR file shows an attempt to model different shaped mirrors for each axis but I can't remember the outcome, but you can look at it yourself now. The other (older) SketchUp files show very basic shapes based on the WideMove thickness and width, with empty edges showing position of mirror planes for the appropriate mechanical angle. The models are crude but good enough to show the the extent of light spill and incursion by the X mirror when using the angles I describe. I'll edit the earlier post to fix the scan angle error.

Original text notes: (Summary: indicates that 60° (30° setback) for both mirrors allows 4.5mm beams to do what a max of 3.5mm is allowed for at 90°.)

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 separation 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.

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, and it means that for a given optical scan angle and beam width, the mirror width can decrease by one quarter. That's a lot of mass reduction, and a lot of potential speed increase. The short first mirror will lose a quarter of its previous length, so its mass will drop to little more than half.

TEST: Try to optimise the 60° layout to shorten the second mirror.

EDIT:
Maybe ignore the last bit in the notes, about mirror mass... I understand now that the rotor mass is far more significant, based on things I have read since. These are old notes, but what they say about beam width, scan angle and mirror width still works. With scanners like WideMoves with very small rotors this mass issue might still be relevant too.

Quote:

---

Originally Posted by The_Doctor

Does that look more plausible? :)

---

I don't have Sketch-up. Never have... But moving a 3.5mm beam through a 45 degree angle seems plausible with those mirrors. BUT REMEMBER -- at nominal (no deflection) you're actually slightly over-filling those mirrors with a 4.5mm incoming beam diameter. You therefore can't project a 4.5mm beam through substantially ANY angle with those mirrors.

Now let me show you something that bothers me. You may recall a while ago, one particular company claiming to be able to project a 3.8mm beam through a 60 degree angle at a scan rate of 60K. I didn't believe it... So I acquired a pair of those scanners, and mocked up the configuration in Solid Works. You can see the results here. Then, when I wrote to them privately and showed them what I found, they sent me another set of mirrors -- this time guaranteed to "project big beams through wide angles". Once I got that set I once again mocked them up in Solid Works to see what they would do. You can see the results here.

So this company publicly claimed that they could project "big beams through wide angles", but the reality is that they could only project a THREE millimeter beam through a FOURTY degree angle with their original set, and then, after they revised the mirror set, they could still only project a THREE millimeter beam through around 46 degrees optical, and frankly NEVER at 60K! Still and yet, this company has never revised their literature online to let people know that they made a mistake.

This is what we're battling in the marketplace... You can take our specifications to the bank, and if we say it projects a fricken 4.5mm beam through a 60 degree range of angles, it fricken projects a 4.5mm beam through a 60 degree range of angles OR MORE! (since we tend to under-promise and over-deliver)

We do real engineering on this stuff. We don't just throw it together, hope for the best, and then have our customers tell us what the products will do...

There have been a number of good questions brought up here, and some debates set forth by skeptical people. QUESTIONS ARE WELCOME! And I'll do my best to communicate the science behind what's happening, and why we do what we do.

I only ask that when others are hyping this or that or the other thing, you examine their products with equal skepticism.

Bill

Quote:

---

Originally Posted by Pangolin

I don't have Sketch-up. Never have... But moving a 3.5mm beam through a 45 degree angle seems plausible with those mirrors. BUT REMEMBER -- at nominal (no deflection) you're actually slightly over-filling those mirrors with a 4.5mm incoming beam diameter. You therefore can't project a 4.5mm beam through substantially ANY angle with those mirrors.

---

At a standard 90° that's true, but that wasn't what I said. At 90°, 3.5mm diameter and 45° full scan angle IS the upper limit. The question is what is possible at 60° beam path angles, not 90°...

If the beam is coming at the mirror from 30° instead of 45° off normal incidence it won't fill the mirror width until that mirror rotates the extra 15° to cause the total width to be filled. As the mechanical half angle is 11.25° for the optical full scan angle of 45° there is clearly adequate angle available either side of centre to allow this, IF you use the 60° beam path ('setback' or 'cant angle' of 30°) as I described it. I provided the files at your request. Those, or even just the figures on the notes, will support what I said. I accept that with the square beams people use today there may be a small bit more spill than with the round beams used when I modelled this stuff, but not much, as exploring the first SketchUp file in the RAR file shows. Find a SketchUp viewer, or a friend who has one. I've done all I can, all I was asked to do. Play around with the 'Orbit' tool and the X-ray display mode to help with examining the model. The main thing to look at is the outer excursion limit frames shown, based on where the edge of the beam hits the relevant edge of a mirror.

Quote:

---

Originally Posted by The_Doctor

At a standard 90° that's true, but that wasn't what I said.

---

See, and that can be a problem. The details of what you "said" could be easily missed... That's why I prefer pictures, and, when necessary with individual clients, prefer to communicate through Skype, where I can hold up a scanner, a mirror, etc. and you could see it in real time. (really see it!)

Quote:

---

Originally Posted by The_Doctor

At 90°, 3.5mm diameter and 45° full scan angle IS the upper limit. The question is what is possible at 60° beam path angles, not 90°...

---

Doc, it is clear that you understand what's going on here. In fact, and ironically, it appears as though you've done more engineering on mirrors and X-Y mounting than most scanner companies in the whole world! You should hire yourself out as a consultant to scanner manufacturers. God knows they need it!

The WIDTH of the mirror is the problem, and it's a problem for both X and Y.

Merely "setting back" the X scanner helps the reduce the requirement for Y mirror width but does nothing for X. What is possible is to use an additional mirror to bring the beam into the X-Y stage at a downward angle. This would help the X mirror (not require as much width from the X) but you'd also then have to avoid reflecting off of the Y mirror at entrance. What's also possible is to rotate the entire X-Y mount so that no additional mirror is needed, and then perform geometric correction to counter-rotate all of your graphics going into the X-Y system. But the problem you said you were having in a separate thread (lines at non-horizontal and non-45 degree angles appearing curved) would become worst.

I'd say that what you have done was a worthwhile exercise to see what you could do with your old scanners and old mirrors, given no constraints on additional system complexity (additional external mirrors, additional machining, additional graphic corrections, etc.). The fact that you had to go thorough all of this shows that you're doing the engineering instead of the people who sold you the scanners...

Wouldn't it be better to buy something that was enginered to work in the first place? :D

Bill

Airline located the scanner amplifier and my neighbor retrieved it from them.

I just need to solve one heck of a SHG crystal problem at a customer's site then I can drive home from South Bend and run the tests.
Fixing a 200,000++ dollar install overrules the hobby, sorry for the delay. I'll have to drive home in the results of the Blizzard.

Montagu has a chapter on the design of Low Inertia Scanners in Laser Beam Scanning, G. Marshall, Editor, Marcel Dekkar, New York, New York, publisher that covers the geometry of the scan head in detail. The book was published right before Cambridge Scanners took off in the market. So it mainly covers torsion bar based scanners. However the mirror math is described in exquisite detail as is the image distortion that can result.

Steve

Quote:

---

Originally Posted by Pangolin

Doc, it is clear that you understand what's going on here. In fact, and ironically, it appears as though you've done more engineering on mirrors and X-Y mounting than most scanner companies in the whole world! You should hire yourself out as a consultant to scanner manufacturers. God knows they need it!

---

I can't claim that level of understanding.. :) The moment I'd have to consider flexure in a mirror at high accelerations, or heat dissipation from a coupling between coil and stator body, I'd be no better off than they are, having to try stuff and see it. The nice thing about flat planes and straight beams is that SketchUp allowed me to model them easily enough to reality-check empirically with no loss other than time.

Quote:

---

The WIDTH of the mirror is the problem, and it's a problem for both X and Y.

Merely "setting back" the X scanner helps the reduce the requirement for Y mirror width but does nothing for X. What is possible is to use an additional mirror to bring the beam into the X-Y stage at a downward angle. This would help the X mirror (not require as much width from the X) but you'd also then have to avoid reflecting off of the Y mirror at entrance. What's also possible is to rotate the entire X-Y mount so that no additional mirror is needed, and then perform geometric correction to counter-rotate all of your graphics going into the X-Y system. But the problem you said you were having in a separate thread (lines at non-horizontal and non-45 degree angles appearing curved) would become worst.

I'd say that what you have done was a worthwhile exercise to see what you could do with your old scanners and old mirrors, given no constraints on additional system complexity (additional external mirrors, additional machining, additional graphic corrections, etc.). The fact that you had to go thorough all of this shows that you're doing the engineering instead of the people who sold you the scanners...

Wouldn't it be better to buy something that was enginered to work in the first place? :D

Bill

---

Well, yes, it would. >:) I need to be sure though. Looks like 506 will be worth a shot for me, but I have to say I like to know a price up front because it's the only way I can calculate how and when I can reach it.

Point taken about the seeing of actual stuff. One thing I showed in the SketchUp files but did not state directly in words in the notes or any recent posts is that the input beam doesn't enter the X mirror space parallel to the base plate, but is angled down 30° off it, steered that way by a mirror. So that helps the X mirror the way its own setback helps the Y mirror. The 60° path allows the input beam to avoid the Y mirror's end. It's tight, but doable. One nice thing I found with the 60° scheme was that the precision by which it neatly folds into place avoiding nasty clips and collisions seems to be inherently part of the geometry. It's not as elegant as snowflake geometry, but with care and accuracy it does 'just work' in a satisfying way. And of course any thinner beam will get more scan angle so all will benefit from it somehow. There may be a 'perfect angle' but it might be some nasty irrational fractional one, but 60° is a lot easier to machine. Another advantage is that dielectric coatings may lose less light on cheaper mirrors (especially in blue wavelengths), because of a closer-to-normal incidence.

Depending on actual mirror shapes and thicknesses, I predict that many scanners will gain from the 60° angles by about 33% increase in clear aperture for a set beam width and scan angle (specifically 45° full angle, may differ with other scan angles). Light spill or beam clipping will be very minimal, and alignment would be no harder than what people do now to get clean coaxiality on their last dichro.

EDIT:
That other thread, with the lines appearing curved when not horizontal or vertical, was a different thing.. I narrowed it down to WideMove scanners not starting up from a resting moment early enough. I haven't got them anymore, but I did look inside them briefly. There was no debris, delamination, that I was warned may exist, so I never figured out a cause. It was mechanical though, they'd track cleanly to rest, they just didn't start moving again until the force moving them reached a significant threshold. This would have shown up regardless of beam path angles. Given that Steve said the 506 has high linearity and repeatability, it looks like they will solve this.

Quote:

---

Originally Posted by Pangolin

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...


Bill

---

WHAT ? UNBELIEVABLE ! SIMPLY UNBELIEVABLE ..... and greatly disappointing, considering it comes from a man that wants to be a ''game changer''

Criticism in products is happily accepted, but ''inventing'' ways to look superior in personal level is just pathetic.

All is not lost though...At least there is some guilt visible : the ''k'' in ''took'' is missing....


Tom

Sounds like someone is being a bit hit on the fingers that their market is being overtaken.
The eye magics are great but for a company who hasn't got a product hands on to judge you've sure been eyeing the new range of pangolin scanners quite abit.

Quote:

---

Originally Posted by EyeMagic

Criticism in products is happily accepted, but ''inventing'' ways to look superior in personal level is just pathetic.

All is not lost though...At least there is some guilt visible : the ''k'' in ''took'' is missing....


Tom

---

Coming from the company who says on their site (and I quote) : "and they are not just DC motors with feedback" http://www.eyemagic.gr/laser-scanners.html
Isn't that a bit too much to say about your competitors?

Quote:

---

Originally Posted by masterpj

Sounds like someone is being a bit hit on the fingers that their market is being overtaken.
The eye magics are great but for a company who hasn't got a product hands on to judge you've sure been eyeing the new range of pangolin scanners quite abit.



Coming from the company who says on their site (and I quote) : "and they are not just DC motors with feedback" http://www.eyemagic.gr/laser-scanners.html
Isn't that a bit too much to say about your competitors?

---

Dear Masterpj,

I think you missed the point. All is not ''business'' and ''money'' dear Masterpj...At least for us.

By the way, the statement ''not just DC motors....'' is true and goes for some companies (in the past I have to admit) that were using step motors and such to make scanners.

Tom