[fl-consult] published this interesting RGB laser diode projector. The build uses three lasers, 532nm green, 660nm red and a 405nm blue diode from an XBox 360 HD-DVD drive. Aside from the salvaged diodes, it uses some off the shelf hardware to power and scan the lasers to make the display. Details are a bit lacking, but google translate helps a bit. If you’re not quite sure what’s going on: the three lasers bounce off of a set of mirrors that scan from side to side as well as up and down to create images.
19 thoughts on “RGB Laser Diode Projector”
Add a smoke machine between the projector and the projection screen :)
I can’t wait to see this attached to a shark’s head.
Its a shame this sort of rig is only useful for vector graphics. Would be nice if you could make an actual laser tv projector. Once you get past a certain point the scanlines are too wide for the image to make sense.
(Scanlines = the black gaps between horizontal lines in the image)
If there were more ReGIS animations out there it would be worth programming a mcu to parse ReGIS commands and render the “frame” out to the mirrors.
@grizzlyadams: I couldn’t agree more. I suppose all you’d need is some faster mirrors, and some equally fast processor to tie them all together.
The lasers can be focused, just like any other type of light. In this manner, we can get a large or small “dot”.
So then, it should simply be a matter of pulsing on and off with the rotation of the mirror to produce a rasterized image.
Someone needs to write a MAME driver for this so he can play Asteroids on his wall. ;)
GrizzlyAdams: I guess the problem is total power output. If you wanted a fully filled in image, you might want say 10W of visible light output – there aren’t many lasers that can output that for a reasonable cost.
@#2 – ROFLMAO
I love the Austin Powers reference!
Well, for those who thought it could never be done to make a real projector:
It can, but it’s extremely expensive for cinema-style ones. The upshot, though, is that images are so much more colorful and vibrant.
This project looks a lot like the RGB laser I have been working on, except that it has a bit more power and is much larger. Kudos for getting the alignment on the first try!
As to doing image projection, it is entirely possible to project images using a rgb laser, either by using a set of DLP chips, a mems based scanning mirror, or even a rotating mirror and galvo assembly, but you run into problems that you need very fast bitrates into the lasers (30fps at 1024×768 is 23 millions pixels/second, DPSS lasers can’t change their output that quickly, nor can many diodes), and an overall lack of power output. Most of these lasers only get well under 1W of overall power ouput, which is about as much as a single luxeon LED, not that impressive for a big display…
Give it time, there are already talks of getting some high power 570nm laser diodes to get multawatt lasers within the price range of a decent TV.
Someone did do a vectormame port for a laser controller, LazerMAME, but the project has been dead for a long time. (The project made me a little sad – it couldn’t handle the two Atari vector games I _wanted_ to play, Star Wars and the Empire Strikes Back.)
also look up Metatroy you tubes to see video of laser projector.
Also look up on Youtubes Metatroy to see the
Z*Tron Vision laser projector.
why not get yourself some old glass He-Ne laser tubes and figure out what gases would make a good blue, fill the tubes with a high quality He-Ne gases an play around with the high voltage supply’s try the DLP chip set with a MIX of all three what could happen you could make a “frickin laser” that could blow up a 3rd world or something lol
the reason why DPSS can’t be switched quickly is due to the way they are driven- i.e. a >0.5W IR laser with active temperature feedback to keep the cavity at the optimal temperature for 532nm production.
one way to get around this would be to add a small LCD panel in the beam path, and increase the drive voltage to the panel as well as running it at around 50C to make it switch more quickly.
I’m interested to see if this is the same technology going in to the current laser projector units. I’ve been looking at the aaxa l1 laser projector and it seems that the technology is very similar. Was wondering if you would agree…
Raster scan laser video projectors have been made, but the power required is immense. The one I saw details on used large frame argon ion lasers for green and blue with one pumping a dye laser for red. The thing required tens of kilowatts, water cooling, and was the size of a small car. It could be done today using solid state lasers, but a better approach is to use the lasers as a light source and create the image with a DLP chip.
You can’t change the color of a HeNe tube by simply changing the gas. The color is determined by the mirrors, with HeNe tubes available in IR, orange, yellow, green, as well as the common red. These “other color” HeNe tubes are much less efficient so they are a much larger more carefully constructed tube for a given output. It is not possible to get blue from one.
Until relatively recently, the only practical way to get a decently powerful blue beam was an argon-ion laser. Most big laser shows out there still use a mixture of argon and krypton ion lasers, occasionally a single laser with a mixture of argon and krypton using multiline optics. The argon provides the blue and green with the krypton producing the red. These are separated, modulated, recombined, and then scanned with mechanical galvos to produce the display. Some use a PCAOM to modulate the beam while older setups used mechanical galvos to do that too.
What if you combined the scanned output of a relatively small area, split the beams out into grid arrays, and pumped the output through fiber optics into the back of an ‘opaque’ display board. Basically the setup would be like LED displays, but instead of LED’s you would have clusters of end-glow fiber optic cables, three fibers for each pixel, one blue, one red, and one green. It would require two scanner sets; one to project the base image into a smaller fiber optic receiver panel, and the other to split each ‘pixel’ of the grid up into a 3×3 or 5×5 copy of itself in order to scale up the size of the display. Its true there would be a slight loss of resolution with the scaling, but if you start out with small enough pixels (ie. focus the beams to a small enough point, you could compensate enough to make a clear display at massive scale.) The fiber optic scaling rig would help to offset the limitations of the laser switching speed, and the scanners would also lend a hand in the switching if you have a ‘dead spot’ target that would effectively block one point of light, making the beam appear to “switch off” when aimed at that spot then “switch on” when aimed back into the grid.
Just a thought…
Please be kind and respectful to help make the comments section excellent. (Comment Policy)