Homemade Scanning Laser Color Projector/TV

[kuntman]

In this project, by means of two mirrors (which can be tilted about two perpendicular axis) we deflect the laser beam so as to scan a rectangular screen and modulate the laser source to obtain a colored image on the screen. We don’t make use of sophisticated devices such as MEMS micromirror and acousto-optical-modulators.

Advantages of our technique are:
- Total cost will not exceed 10 USD in mass production. (Mechanical part costs only few dollars)
- Low power consumption (Approx. 4 W.)
- Pictures can be performed on any screen
- No lenses are needed for focusing
- Scanning had can be made as small as a cigarette box.

Max. speed of the fast axis mirror is 25 KHz. That corresponds to 50000 back and fro lines per second.
In order to make use of the conventional video signal, the resonant frequency of the present scanner is tuned to 15625Hz.
Therefore number scanned lines in each frame is 625. Refresh rate is 25 frames (50 fields) per second.
Vibrations of both axis are resonant, therefore sinusoidal.
Slow axis is driven by piezo.
Fast axis is elecromagnetically driven.
Laser modulation is analog.







http://www.photonlexicon.com/forums/album.php?albumid=104

And related videos on Youtube posted by sltvm2007 and sltvm2008

 

[mgb_phys]

The fast scan mirror is tricky
You have to scan a line quickly ( ie in 1/30*625 s) but then get back the start in one flyback time - can't remember what that is for pal but it's only about 1/50 of the line time
So you can't use a tuned resonant scanner, if you use a spinning mirror you need a huge f-theta lens

You will also need 3 colour lasers and some optics to superimpose them.
Red is easy, green is getting easier but solid state Ti-Sapphire (blue) replacements are still expensive, especially if you want watts of power which you will need to get a bright spot.

 

[kuntman]

The fast scan mirror is tricky
You have to scan a line quickly ( ie in 1/30*625 s) but then get back the start in one flyback time - can't remember what that is for pal but it's only about 1/50 of the line time

There is no trick
Scanning scheme is as follows:
First line of the first feld is scanned from left to the right.
Second line of the first field is scanned from left to the right with the video content inverted in time.
Same technique is applied to the successive fields which are scanned up and down and up...

 

[mgb_phys]

Thats a good idea

 

[berkeman]

There is no trick
Scanning scheme is as follows:
First line of the first feld is scanned from left to the right.
Second line of the first field is scanned from left to the right with the video content inverted in time.
Same technique is applied to the successive fields which are scanned up and down and up...

I believe there is a fundamental problem with resonant scanning. How do you intend to get your scan lines parallel? You need linear, single-direction scanning to get that.

 

[kuntman]

I believe there is a fundamental problem with resonant scanning. How do you intend to get your scan lines parallel? You need linear, single-direction scanning to get that.

When the beam scans sinusoidally this does not mean that successive lines will not be parallel.
However the content of the video will be somehow distorted.

You may view the picture of scanned lines from the last link.

 

[mgb_phys]

You can linearise the f-theta out in software to some extent.
This is patented (at least for scanners that measure a surface) - I had to design a multi resonant scanner system once to cancel out the effect without changing the timing.

For the vertical (field) axis you don't need to use a resonant scanner - an normal coil driven one is fast enough and the field flyback time is long enough,

 

[kuntman]

I tried to make a linear slow (vertical) axis which uses coils etc. It is easy to actuate the mirror to move linearly, but it is almost impossible to stop it and get back to the starting point during the flyback time because of the considerable moment of inertia.

 

[armlaser]

Thanks for sharing. It's really a low cost