Timing is almost always trivial. Computer hardware is quite good with timing. The problem is to get any data to display. That data would be either video (in which case MPEG is a good encoding) or the processor must render information in the video memory. Which requires a fast microcontroller or a graphics accelerator. A PC has ample power to do it, but currently no good output interface for the data, since the current technology isn't suitable for building a laser-based raster display connectable directly to the graphics card. Do not forget the problems with finding lasers suitable for a raster display, given the different requirements of a raster display. Right now, a laser projector can't light a larger area with any reasonably priced lasers. No need to discuss designs for the lasers 5 or 10 years into the future. So it is better to discuss the "now" instead of how trivial it is to work with a raster display requiring components that you can't buy. To build a laser-based raster display right now capable of lighting a suitable area (many home-built XY projectors can draw many square-meter big images without intensity problems, since they don't try to light an area), we would have to use multiple lasers taking care of parts of the image. Or we would have to get extremely (!!!) expensive, and hard-to-use lasers. How much does it matter how simple it is with a raster display, if you need to tripple the mains voltage to power the laser, and then cool away 100W+ to keep the laser cool enough that it still resonates? With luck, you may get away with Peltier elements (and remember how much power they require for each W of heat moved away) but with bad luck you may need really heavy-duty cooling.


And how do you think the MCU gets the data to refresh with?


You are mixing technologies in your debate. Don't make too big analogies between a projection TV and a laser projector. A projection TV is basically a normal video projector (LCD, DLP, CRT, ...) projecting an image. It just a question of having a back-projection instead of a front-projection. I can take my projector and reconfigure it for back-projection (it will mirror the image) and replace the normal screen with one intended for back-projection. But that is irrelevant to a laser projector, since the laser projector does not light a surface, but render with a ray. It is similar to a CRT, but there are a number of differences, and - more importantly - we don't have suitable lasers yet for the task. A $1000 blue laser is fun to play with. And it is very dangerous to look at. But it still has a very low intensity if you need to light up a square meter of a wall. The efficiency of a 200W projector lamp in a LCD projector is quite low, but it does put out a significant number of W past the LCD and colors flters, ... Yes, the electron beam of a CRT also has very significant energy. But it is a bit easier to control, and you do have 10mm of glass between you and the beam in case the monitor fails to bend the beam. And the important thing: It is - with current technology - easy to produce an electron beam of suitable intensity. That is the reason why CRT displays have been in existence for so many years.

With the current limitations of laser projectors, a XY-projector still has large advantages over a raster projector.


But I think you did not read what I wrote. It is irrelevant how easy it is to spin that mirror. To be meaningful, you still need suitable lasers. With the current technology, it is way more practical to do XY projection. Just because you have worked with raster displays doesn't make them the best since sliced bread. You concentrate on the trivial parts, while ignoring the big problem: What laser to use. Most people with a bit of mechanical and electronics skills reading this thread and wanting to build a laser projector can build an XY projector quite easily. And it will be fun and useable. A raster projector can also be built, but it would probably not be fun or useable, because of the problems with finding suitable lasers. How fun do you think it would be to build a laser projector displaying a 100x100mm picture on the wall? How fun do you think it would be with several square meters of real-time-generated output?


Once again - you are ignoring the big thing: What laser to use, to supply the light for a raster display? 50 years ago, it was trivial to produce a strong enough electron beam. Then it was the electronics that was the weak link. For a laser projector, it isn't. It's the laser. So it really doesn't matter what people did with a CRT 50 years ago. We just do not have lasers available that are suitable for a raster display, unless you limit your resolution in one or both axes very significantly, or limit the output to a quite small area. Unless you can afford building a system with many small lasers driving a limited number of scan lines. But before you say: That's ok - do check the price of the lasers. And once more - a small micro can easily play with the galvos for a laser projector - it is easy to scroll pre-generated data.


No, it is not the only way to go. But it is currently not even a good way to go, for most applications. If I can project a 100x100mm raster image on the wall with a laser projector, it would be way easier to use a 19" cheap LCD. If the projection part was the required magic, then the cheapest LCD projectors would win several times over for emitting raster data.

But a cheap home-made XY projector can do what the 19" LCD or the cheap LCD projector can't. Anyone interested in building a laser projector is most likely interested in letting it do what no other technology can match. So they are most likely interested in an XY-modulated projector.