| ??? 05/03/09 10:17 Read: times |
#165018 - Finally noticing the lack of information from the OP? Responding to: ???'s previous message |
Richard said:
This is 8052.com and Chico has indicated that's what he's using. And your point? An 8052 chip would not be limited to playing a static image from ROM if using an XY projector. Or was your claim that it can't? My note about bigger processors was about requirements to get moving output if switching from vector to raster. With vector output, an 8051 would manage find emitting changing/moving data. Richard said:
I've not said the "design issues" you mention are "gone" but I do maintain that they are, at this stage, totally irrelevant. Yes, I see. It is totally irrelevant that your laser pointer smeared over a 50,000 times larger area will make each point having 50,000 times less energy (but not look 50,000 times weaker). Haven't you noticed that the electron beam of a 21" CRT is many times stronger than the electron beam of a 9" monitor? The OP may happen to have a quite strong HeNe laser laying around. They are quite easy to come by. How do you modulate one if using it for a raster display? People using them for XY displays has to use a galvo or similar to block the ray. Where would Chico get a device that than turn on/off a light ray for a display running at 1.5 million pixels/second. Ah - I know. At this state, totally irrelevant. As long as Chico gets a rotating mirror from an old laser printer, everything will come together automagically... Richard said:
There are lots of things that can be done with other equipment, but Chico wants to do something with what he's already got. He wants to project an image onto the wall. He hasn't said he wants to project full-color real-time video. What he already got? Can you please repeat that list for me. Read the following posts again: http://www.8052.com/forum/read/164826 http://www.8052.com/forum/read/164919 The only thing I think I can deduce that he has is a HDD, an ADC and a L293. So return to me with what you mean with "what he has"? He has a scrapped laser printer with a rotating mirror? He has a laser with enough intensity to be visible when sweeping 50k pixels? Said laser is possible to modulate? At MHz speed? You are bettar than me at picking up facts from other peoples posts. Richard said:
He wants to project an image onto the wall. The correct quote was: Chico said:
I want to fix a mirror in the HDD head to deflect a laser, just direct driven, no gearboxes, as I need some speed, I think pwm should not fit. Where do you see something about a wall? One of the strengths of a XY projector is that it draws very nicely in air - how do you know he wants to project an image, and no just show patterns in the air? Richard said:
He hasn't said he wants to project full-color real-time video. We are finally getting somewhere. Correct! He hasn't! And I have several times claimed that video do require a raster display, but he might just want to draw a vector graph, in which case an XY projector will get way more of the laser light on the wall - you remember yrou suggesteed, but not confirmed, wall? Not only that. If he want to draw in the air, then he can't even draw photos or video. The concept of drawing in air works because of vector output, unless you happen to have a 100 times stronger laser and can afford to only let out the ray 1:100 of the time. Richard said:
Further, you're mixing technologies, for some reason. On one hand you say that you can't do fills, which, from where I sit, you certainly can, and easily at that, and then you say you can draw a line at 20 kHz. I have my doubts about that, but maybe ... You have better be more specific about what you mean by "mixing technologies". From my view, you are transfering old knowledge about CRT into the domain of lasers and expecting it to be true. I call that "mixing of technologies". The correct quote about 20k was: Per said:
I have seen home-built galvos reaching past 20k points/second, but most people have a hard time to get past 5k. I would say 9k is above average. It was from this post: http://www.8052.com/forum/read/164993 and regarding this post from Steve M. Taylor: http://www.8052.com/forum/read/164992 Photo of output from a home-built galvo at 18k: http://elm-chan.org/works/vlp/ilda18k8d.jpeg Full text, also containing description how to build the galvo: http://elm-chan.org/works/vlp/report_e.html Photos of output from a home-built galvo at 30k: http://74.58.33.200:81/images/RG...G_2305.JPG Some more photos and some other projects: http://74.58.33.200:81/RGBLaserProjector.htm My claim was that most people probably only manages around 5k vectors. Note again that the mirror has travel time, so the number of vectors7second will depend on the number of degrees of deflection. Richard said:
However, if you can draw a line, no matter the rate, what difference does it make whether the beam is on or off? Should be obvious. The time the ray is off represents time not producing an image. The quota between light on and light off represents how much of the original laser light (ignoring losses in mirrors and optics) that will reach your claimed wall. A laser fed a constant current will give out twice as much light energy when lit 20% than when lit 10% of the time. Don't you find that obvious? Next thing is "the rate". If I draw a line at 1m/s or at 100m/s with a fixed-intensity laser, the 100m/s trace will will have an energy density that is 100 times lower. I can compensate for this if I can draw it 100 times more often. Your raster display will for any frame rate require the laser beam to move much faster, since it always has to cover all parts of the surface. The XY display will only let the laser beam pass lines to draw, and the distance to move the ray in blanked state when moving to the next polygon to draw. This allows it to run with much lower speed, so it will project more intensity at any given frame rate. This is one of the reasons why it paints so well in air: If you draw a rotating star, you may have 10 line segments in the star allowing for either a very slow movement rate of the ray or a very high frame rate. But eacy of the line segments may get 10% of the laser light. If you draw the same star using a raster display, then an inscribed five-edged star of maximum size in a 256x192 display will light about 1.2% of the pixels (Yes, I drew the star and counted the pixels). So at best, the raster display will draw black for 98.8% of the time - this represents light energy wasted since you can't store this energy in a capacitor somewhere and reuse at a later time). Each line segment will get 0.12% of the laser energy, instead of 10% for the vector display. 10/0.12 means that the vector display projects 83 times more energy on your claimed wall. But wait a minute. I think your take on this is "totally irrelevant". And you back it with so strong arguments. If I could just find those arguments... I perceive that you believe that you can draw only a few lines frequently enough to accommodate the persistence of the human retina Don't believe. Know. And it is my guess that you have never visited a disco, or you would know too. The eyes have their own "phosphor" in them. They report average energies. Have you tried photographing a new CRT monitor yet, verifying my claim that the phosphor in todays CRT monitors is very fast? If I take a photo of my monitor, I can get pictures showing only a few scan lines lit. The other scan lines have already dimmed away, in wait for the next pass. This works because we do not need any persistence in the CRT - the eye manages on its own. Same thing with a multiplexed LED display. No persistence in the diodes, but with high enough frame rate, we can't see any flicker. Simple shapes would be represented by a small and simple bitmap, for a monochrome raster display. Such a display could easily be updated from time to time if it were necessary or desirable to do so. And what would be the reason to build a raster projector showing extremely dim static images? The cheap optics of a $100 dia projector should manage 2k horisontal resolution, and at orders of magnitude higher intensity. The only reason I'm taking the stance that the bitmapped/rasterized display is easier is because it requires only one degree of freedom for a voice-coil-driven mirror Correct. That part may represent a simplification. But you claim "with what he has". He has a rotating mirror. Or rotating prisma? Or a laser printer to extract one from? The rest is just electronics. Mixing of technology. A lot of it is not electronics, but optics and simple physics. Remeber the little 1:50000 reduction of energy/area? And are you sure you can modulate the laser at that speed? Remember when you in a different thread wrote: Richard said:
If this has "typical" sorts of LED's in it, it takes somewhere on the order of a microsecond for one of them to come to full brightness, once the current is flowing. But your suggested raster display would run with 1.5MHz pixel frequency. Are you changing your arguments depending on what thread you are in, or do you have reasons to believe that a high-intensity laser diode will be significantly faster than a LED? The quote was from: http://www.8052.com/forum/thread/160811 I'm not saying that the lasers can't be modulated at that speed. I'm just interested why you never once feel interested in backing your claim, given the fact that you, yourself, have had doubt about modulating normal LEDs at these frequencies. And your switch to a raster display do limit that 805x processor to show a static image. How do you know that is ok with the OP, given the fact that most people interested in laser projectors want them for their strengths, and not for something they are not suitable for? As always, there are many ways to "skin a cat," but in this case, it would be a good thing to start with something that Chico can actually build. I doubt he has access to high-precision machinery. Moving two mirrors requires not only that he build and precisely control a two-axis system, but that his construction be precise, say, to 0.01mm. $10 working low-end implementation that does work: http://spt06.chez-alice.fr/00/scan1.htm This obviously did not required any high-precision machinery (just two speakers), but seems quite nice anyway: http://video.google.com/videoplay...1774962691 If we ignore the size constraints, this one is quite buildable with simpler tools: http://hacknmod.com/hack/how-...e-lighter/ The green laser here is a bit expensive, but red (DVD burner) or purple (HD or BD drive) can be had cheaper. Claimed build time: 20 min: http://zedomax.com/blog/2008...projector/ This is another simple one, where a piece of a CD is used instead: http://www.wonderhowto.com/how-to/vi...ls-209693/ A bit more high-end (seems to have bought the dichormatic mirrors): http://www.laserblog.de/2008/05/m...dvd-diode/ Another one showing what happens when the deigner do have the skills and the tools: http://hackedgadgets.com/2008/07/2...projector/ Richard said:
Above all, let's not get off into bigger processors and more complex systems. We are not. Maybe you are. AFAIK, all he wants is to project an image onto his room wall. All you think he wants... Richard said:
Yes, coarse resolution makes the lines "jaggy" as you put it, but any other display projector that uses mechanical means is often quite non-linear as well. Non-linear and jaggy are two different issues. Richard said:
Do you know what sort of laser Chico has? We don't know if he has one at all. Only that he have tried to use L293 to control a motor. Richard said:
Do you know what sort of voice-coil and driver he has available? Correct. Nothing known, but indications that he may have tried a HDD. Doesn't that mean that we don't know if he has a rotating mirror? Richard said:
Do you know anything about his projection target? Correct - that "wall" has never been mentioned. It may just as well be smoky air or something else. And if the target surface is not flat, then linearity or pictures may not make sence, while vector output would still have the same psychedelic effect. Richard said:
Do you know what images he wants to project? Exactly - we don't even know if he wants to projects images, in the meaning you seem to have locked down on. |
