Here are some guys who could build you a compact, and i have one in my plans too (with active cooling and drivers integrated) but not ready in the next months. but yes with 405

Good news. I hope one day we will combine our efforts to make a full colour scanner.

15625 Kpps (15.625 Mpps) Scanner

3D Visual Effect by Kuntman's Scanner

http://www.youtube.com/watch?v=yRNsjvzmKRI

Sweet, keep up the great work.

15625 Kpps (15.625 Mpps) Color Projector/ TV:
http://www.youtube.com/watch?v=NTpwygS8baY

3D Visual Effect by Kuntman's Scanner:
http://www.youtube.com/watch?v=yRNsjvzmKRI

Videos:
http://www.youtube.com/watch?v=DmCzjv7Q20s
http://www.youtube.com/watch?v=HleDvzLk7SI

Photos:
http://www.photonlexicon.com/forums/...hp?albumid=104

This is a nice PDF i found over the Microvision Pico Projector:

http://www.opnmagazine-digital.com/o...n200905-dl.pdf

they have some text about the sinusoidal duplex scanning method sltvm uses, too!

Do you have any pictures that are actually worth looking at?? I cant see anything in your posts...utterly worthless pics!

you could use the prism mirror combiner thing in a 3 lcd projector to combine the lasers into one white one

We've had this guy over on PL, he will not release ANY technical information. He was banned from PL and I think he should be banned from here too, I think his intentions are just to tease/troll.

I make use of an ECG Galvo salvaged from an old ECG machine.
It has a nice capacitive feedback.
I intended to use it wth my color laser projector.

More pictures are available in the album

Hi, I'm still new at lasers, so I hope I'm not doing something wrong.

I thought about the refresh/galvo speed problem. Maybe I have a simple solution for fast refresh: two polygon mirrors, one for the X and the other for the Y axis. They only have to continuously rotate in order to give a full two axis refresh rate (obviously one axis has to turn slower than the other, so that you plot the next line only when the upper one is completely drawn).
I think that the major problem is that mirrors have to be custom made (and one of the two has to be taller).
If we ask some optic manufacturer I think we will have a good price on prototypes and final products.

Or we can make our own DIY first surface mirrors mounted on a polygonal shape turning at an high speed.

I hope you could understand my Engrish and, more importantly, my idea

I have considered that a few times, and it would work fine. Although, to get a good resolution, you'd need a polygon mirror with more than 8 mirrors, all at a slightly larger angle to each other. If you got a scanner with say 40 mirrors, with each one angled slightly more than the last you could scan a square, and just modulate the laser.

Even if you just got 1 scanner with like 40 mirrors, you could do it, as you don't need a separate Y scanner to generate a picture.

The reason for so many mirrors is if you only have 8 mirrors, you only get 8 horizontal lines, which means shitty resolution. The more mirrors the better the resolution.

Hi, if I understood you wanted to use only ONE polygon mirror, which gives you both the axis. My idea is much simpler: you have to use TWO polygon mirrors, one per axis. Your problems of inclination will only become problems of finding the right rotating speed for the two motors. No need for a polygon with many sides, even a triangular one (e.g. A silvered prism) can do this job, it's only a matter of speed as I said.

The part the most difficult is IMO the sync between mirrors and lasers, as the thread says laser diodes have no problems to reach switching speeds around the MHz

maybe a stepped motor, where you can control the position, can solve many problems.

...only need two constant rotating speeds, one slower than the other

Yes, you are absolutely right. ECG galvos are not fast. That's why I intended to use it as a slow axis.

Polygonal mirrors may also good for the slow axis. But, for an angle of 30 degrees of total optical deflection, you will need a polygonal mirror with 24 faces. As far as I know, such a polygonal mirror is not available in the market and it will not be easy to construct such a mirror with great precision for a DIY experimenter.

Maybe I came with a conclusion: if we take a 100 sides custom made polygon mirror (we have to ask some optic store for prices) we have something similar to a cylinder (which reduces friction with air) that needs to rotate at "only" 9375 RPM to give a clean 15625 Hz horizontal refresh.
For the vertical refresh the solution is a simple, economic galvo.

It could be expensive...

What do you think about it?

Maybe MEMS or AOE are still the better choices...

EDIT: I saw on ebay that you can easily find 40k+ RPM dc motors. At a STABLE frequency of 37500 RPM we will only need a 25 sides polygon. But I think that at such an high frequency there is a loss of stability (it's not so simple to track that speed, maybe with another cheap red laser and a photodiode) and a lot of noise.

It could be expensive...

end quote


I hope you like air bearing, air driven, rotary polygon scanners running in vacuum or helium..

These guys make the state of the art...

Manufacturer of Polygon motors, precision laser beam delivery systems and laser scanning components - lincoln laser

I do not think you'll find inexpensive 35K rpm 24 faucet scanners on their website though.

For refs on whats possible, Find a copy of Laser Beam Scanning, published by Marcel Dekkar, edited by Gerald Marshall

Steve

Thank you. I'll try to ask for a prototype cost... sigh...

I was googling "mems mirror" and I came across this image


It's the classic magnetic dipole that I studied in physics. What do you think about making a larger one axis model? We theoretically need six things (more to add):
1) a lightweight mirror (mylar film?)
2) thin magnet wire
3) a rigid support
4) a constant magnetic field (neodymium?)
5) maths XD
6) building skills

We need also to find a feedback system (maybe another laser) in order to calibrate motion vs current.
Bye, Riccardo

Good idea. But you should sacrifice the linearity.
Linearity is the main advantage of rotating polygons

I know, it has to be digitally controlled to apply some corrections. Rotating polygons are way too expensive for this project.

Maybe I can take advantage from the opposite principle: use the magnet as a mirror with an external electromagnetic field induced by a solenoid (shaped as a U) placed under it to control it, in a context similar to the one in the upper image (but reversed).
I hope you understood my idea...
If I have some time I can do a simple 3D model.

I think there will be no problem to make it oscillate at high frequencies, as it doesn't need to make a full angle of rotation but only few degrees.

Someone knows where I can find a neodymium magnet with a mirror-shaped surface? Can I use silver nitrate to turn it into a mirror?

I saw you are a university researcher, can you help me with some maths?

Thanks, Riccardo

EDIT: here the render for my idea. The U is the solenoid, which induces a EMF which actually moves the magnet (the round mirror in the center)


ANOTHER EDIT: In this other example I took two ferrite cores and a rectangular shaped neodymium mirror. Normally the mirror can rotate freely, but when an electro magnetic field is inducted by the semi-toroids the mirror has to rotate around its hub. Regulating the current flux inside the solenoids you can control mirror movements. It's important that the magnetic force is enough strong that it can overcome the moment of inertia and the friction. The two semi-toroids are slightly separated so that the laser beam can pass between them.
Can you help me mathematically find the right current function in order to have a constant angular velocity?
The realization is very inexpensive: two toroids, a rigid support, a hub, some magnet wire and the neodymium magnet (cheap on ebay, it can be bought silver or nickel coated, so it's even a good mirror). Well, it can be easily done with less than 10$.
In the rear of the machine a cheap red laser can be used with some photodiodes in order to have a feedback system


I also found a very lightweight neodymium magnet, easily turnable into a rotating mirror
https://www.hkcm.de/HKCM_datasheet.php?l=en&id=4286

Yes. If you let me know mass and dimensions of your magnet mirror and the torque constant, I can calculate the optimal drive frequency for you.

Thank you for your help, but I have no clue how to find the torque constant, as it is the opposite of a motor. In addition it is not supposed to make a full rotation but only few degrees, so it's impossible to have a true constant angular velocity

For the laser feedback system I think I will use a photodiode array (maybe this one) with an Atmel or a PIC microcontroller.

Maybe I can try to build it and experimentally find with this feedback system a way to have something like a constant angular velocity (if the coil can produce a strong enough magnetic field it can overcome some non-linear factors and make the magnet have a constant angular velocity for a while). For example, if I take the angle vs time plot and I interpolate its function, I can find how the electromagnetic flux (and then the current into the electromagnet) has to behave in order to achieve a constant angular velocity.
It's a little tricky...
What do you think?

You don't need to know the angular velocity. Just try to measure the torque required to tilt the mirror one degree.

I realized that in the examples I made earlier even forces are nonlinear, so I think I'll return my path of the magnetic dipole. Yesterday I realized even that I can use directly a neodymium rigid support, in which B is supposed to be very linear and strong. I also thought about using the current carrying wire as the mirror hub. As I have to use a lightweight mirror (I think an acrylic one), I don't know if I can make more than one turn around the mirror, so I will probably need a lot of current, which I can obtain with a low voltage and a relatively large (14 AWG?) wire.
At this point, with a constant magnetic field, it becomes much more simple. Like a motor
Once I'll know the relationship angle vs time I will be able to digitally correct the laser modulation. Laser feedback concept can still be applied to find the mirror inclination.

I made another model, which uses the current carrying wire in a constant magnetic field principle. It uses 28 cubic N52 neodymium 10mm cubes to have a constant 0,5T magnetic field around the wires. In the model the two wires are formed by 50 turns of 24 AWG wire each, which will carry a maximum current of about 3A. Each side is capable of generating a 7,5N force under these conditions. A similar layout can work even with 20 magnets only, but with a force per wire of 5,2N. I think it's strong enough to let the mirror vibrate at the wanted frequency, maybe even an higher one. I hope there won't be problem of resonance...


Sketchup:
Download mirror.skp, upload your files and earn money.

It can work?