How to build a laser light show system?

[randall_ducky]

The hardest part of building a laser show system is finding the scanners at a reasonable price. It turns out I have them starting at $10 each. <p>If you want to build a laser light show system I have some extra galvanometers you may use. I've experimented with generating laser show patterns quite a bit and can help those interested in building their own systems.<p>The basics require..<p>1. a laser (diode with power supply will do 3-5 mw minimum) available for 5-25$ seach a laser surplus store like mwk or meredith instruments.<p>2. 2 front surface mirrors 1" x 1/2" will do<p>3. 2 scanning galvanometers (for x and y axis).
These are typically the most difficult parts to find but you can find some on ebay starting at 10$ each. search under "precision scanner"<p>4. A 50 watt stereo that drives left (x axis) and right (y axis).<p>5. Some music with pretty good bass.<p>I've even done some computer control images but does anyone else have a good method for creating digital images?

[Chris Smith]

Randall, that makes a nice analog show for the beginner. <p>For "Sign Writing" where you project words or pictures onto a wall you need to get into writing code or emulating it. <p>Its simalar to printer software, where the picture or text is moved exactly like a printer head, only it calibrated slightly different. <p>I would immagine the Proliferation of the PIC controllers onto the market would be ideal for this next step in evolution.

[jollyrgr]

If you just want dancing lights and don't care about what type of picture it draws, you can use cheap mirror squares or strips (like those found at makeup counters or in shaving kits) and rubber cement them to a speaker. Set speakers at right angles to one another and drive them with tone generators or even music. True this is not real fancy, but it is cheap and readily available.<p>For parallel scanning, you can snag the motor/scanning unit out of a "dead" laser printer. You could probably get a clunker from a computer store for free.<p>Chris, you bring up an interesting point but I'm not sure how to implement it but know it can be done. For laser shows with pictures, I've thought of using the "yoke coils" from a CRT and driving speakers instead. (No, I'd not connect a speaker to the yoke coil outputs of a monitor.) But you hit on something that I KNOW can be done but I don't know programming well enough to carry this out. And with my speaker idea, here is what would happen.<p>Soundcards are used as modems, digital capture devices, and so on. How about a program that drives the Left (X) and Right (Y) axis of the soundcard outputs to move speaker coils. The speakers would move mirrors. A laser pointer would then be shined through this mirror combination. The drive/modulator for the laser pointer on and off would come from either a serial or parallel port data line. I know this can be done and it could be done under DOS for that matter. But I don't know how to program such a program. I'm thinking of drawing pictures in something like Microsoft PAINT. Simple line drawings. The scan of the X Y coordinates for the soundcard would come from the information in the bitmap image. Drive the speakers at 15Khz for the X axis, and something like 60Hz for the Y axis (sync'd like a TV picture). The "beam on" information would come from the location of the lines in the drawing. This would pulse the laser on and off, just like an electron beam.<p>Any ideas on how to get such a program?<p>[ September 30, 2005: Message edited by: Jolly Roger ]</p>

[Chris Smith]

Jolly<p>Probably the best and possibly the easiest method is to combine a bit of mechanical leverage for a wide swing in the mirror deflection area, combined with a Peizo element as the driver or linear actuator.<p> All the high tek laser movers and positioners use the Peizo Crystal to advance and retrieve their actuators, and basically because the Peizo grows and shrinks with voltage in a almost in perfect linear manner, the rate of the mirror deflection shoud be easy to track and keep porportional. <p>A few thou with a leverage method can be ten thou of deflection or even hundreds of thou deflection depending on the leverage ratio. <p>With small light weight mirror materials you should be able to move the mirrors pretty damn fast. <p>The soft ware program is like the printer, X&Y in nature but Im not a programmer [yet] but I do have a retired silicon valley programmer working with me on another project involving the Stamp / PIC with the hopes of later making it into a mass production project. <p>By the time the project is done, I should be up to speed on how to program these movements,.... with any hope? <p>Chris<p>[ September 30, 2005: Message edited by: Chris Smith ]</p>

[philba]

Raster vs Vector<p>The way the printer (and crt monitors) works is called raster imaging and it takes fairly precise control and timing of the beam positioning (deflection) and the beam on/off timing but the technique is well understood. Just remember that with the mechanical leverage advantage comes a proportional increase in positioning error and dampening is a very important factor. Programming the raster scanning is also very well understood in the literature and fairly easy. Authoring of the content is pretty easy - you can use any digital image out there. all you need to do is convert to your particular format.<p>It does have some disadvantages in this situation: aliasing and image brightness. On the brightness front, your overall brightness is the inverse square of the image size for a given beam intensity. Also, you 'waste' brightness on all the un-illuminated pixels. For an NxM pixel array, each pixel can only be 1/(M*N) th the brightness of a stationary beam. For example, drawing a line will only cause a small percentage of the pixels to turn on, say 10%. This means that you are only using 10% of potential image brightness. Aliasing is a harder issue to deal with but there are lots of antialiasing techniques in the literature. <p>The few laser shows I've seen have used vector (or stroke) graphics to produce images. Think of the beam as a pen and how you would draw a picture. This preserves brightness by not wasting time on scanning dark areas. The same line from the above example would be 10 times brighter. The x-y mechanism takes more mechanical sophistication and precision, though. This was the dominant graphics technology in the 50s and 60s - there should be lots of info in the literature and many expired patents in this area. Authoring is problematic - I'm not aware of authoring tools for vector graphics. It might be possible to take direct pen input from, say, a palm device and convert into stroke format for an x-y device. It would be way cool to have a palm pilot, write something on the pad and have it displayed immediately.<p>[ October 01, 2005: Message edited by: philba ]</p>

[Newz2000]

HP (I think) used to make plotters - like printers, except they used a number of pens to draw images. It was fascinating to watch them work because they drew a picture much the same way an artist would, just much faster.<p>Some of the early create-a-card machines used these plotters. You could watch them choose a color from about 12 pens and then the machine would scribble the design of the card onto the page.<p>I've seen other devices work this way as well. Instead of shonwing a concentration of dots they show strokes.<p>The four most popular vector formats DX? used by autocad, EPS used by Adobe Illustrator and others, WMF used in MS Office clipart and the new guy, SVG, a web-friendly replacement for EPS and WMF. I guess Macromedia Flash format is also vector based, so call it 5 instead of 4.<p>I suspect it would be possible to read the format of these and make it into a picture with some mechanical device like a laser show. Dot matrix patters would be easier, but the vector format type would look much nicer.