Scanning a high power beam
Unfortunately, scanning a high power beam isnt going to work very well, as the total integrated light per pixel will be much less than continuously illuminated pixels.
Example:
Suppose your light source on a "traditional" setup is 1000 watts. Also suppose you have that light distributed evenly across your 1 million pixel display. This would be an average of 0.001 watts, or 1 milliwatt, per pixel.
Now suppose you had one powerful laser/device/thingy. You're going to have to continuously scan that thing around fast enough to cover all 1 million pixels each frame, effectively illuminating each individual pixel for only 1/1,000,000 of a second for every total second of real time. Therefore to reach the same integrated power per second on each pixel as in the continuous case you would have to have a 1,000 W laser/device/thingy. No way an LCD pixel is going to survive a 1,000 W laser hitting it, even for nanoseconds at a pop. So you would have to live with whatever intensity the pixels could survive for short bursts without degrading the panel, which is bound to be MUCH lower.
You could try and "meet in the middle" by using a bigger spot size and therefore illuminating several clusters of pixels at once, but you're still going to quickly hit the upper allowable intensity limit on the LCD panel rather quickly. Plus the hassle of setting up a galvo/scanning device is one more source of expense/bulkiness/failure.
Another possibility is a magic phosphor like material (on a panel in between the strobing laser thingy and the lcd) that will continue to glow white and emit close to the integrated energy of what you hit it with, and also can be hit by lots of power for short time periods and still respond. This is actually how most white LEDs basically work (blue light hitting a coating that glows white). Still going to have to live with the intensity threshold/minimum exposure time/glow time constant/etc of such material, plus the ever present added expense of the galvo/modulator/whatever.
So in short, I think that we are already near the upper limit of what intensity an LCD panel can sustain without damage, and switching to a strobing style format would demand more intensity (albeit in "pulses"). So, cost/hassle of galvo mechanism notwithstanding, the real limit is that we just can't jack up power enough on the panels without damaging them.
Unfortunately, scanning a high power beam isnt going to work very well, as the total integrated light per pixel will be much less than continuously illuminated pixels.
Example:
Suppose your light source on a "traditional" setup is 1000 watts. Also suppose you have that light distributed evenly across your 1 million pixel display. This would be an average of 0.001 watts, or 1 milliwatt, per pixel.
Now suppose you had one powerful laser/device/thingy. You're going to have to continuously scan that thing around fast enough to cover all 1 million pixels each frame, effectively illuminating each individual pixel for only 1/1,000,000 of a second for every total second of real time. Therefore to reach the same integrated power per second on each pixel as in the continuous case you would have to have a 1,000 W laser/device/thingy. No way an LCD pixel is going to survive a 1,000 W laser hitting it, even for nanoseconds at a pop. So you would have to live with whatever intensity the pixels could survive for short bursts without degrading the panel, which is bound to be MUCH lower.
You could try and "meet in the middle" by using a bigger spot size and therefore illuminating several clusters of pixels at once, but you're still going to quickly hit the upper allowable intensity limit on the LCD panel rather quickly. Plus the hassle of setting up a galvo/scanning device is one more source of expense/bulkiness/failure.
Another possibility is a magic phosphor like material (on a panel in between the strobing laser thingy and the lcd) that will continue to glow white and emit close to the integrated energy of what you hit it with, and also can be hit by lots of power for short time periods and still respond. This is actually how most white LEDs basically work (blue light hitting a coating that glows white). Still going to have to live with the intensity threshold/minimum exposure time/glow time constant/etc of such material, plus the ever present added expense of the galvo/modulator/whatever.
So in short, I think that we are already near the upper limit of what intensity an LCD panel can sustain without damage, and switching to a strobing style format would demand more intensity (albeit in "pulses"). So, cost/hassle of galvo mechanism notwithstanding, the real limit is that we just can't jack up power enough on the panels without damaging them.
scanning lasers
If you have fast computer-controlled laser scanning, then the LCD would serve no purpose! Red, green, and blue laser diodes can all be modulated with very high speed data. So you could just modulate each beam as needed to generate the pixels, combine them, and use the spinning mirrors to scan them across the screen. If you do it fast enough with enough lumens, the human eye's light pigment persistence will work fine, instead of messing with another phosphor. You wouldn't even need a projection lens.
If you have fast computer-controlled laser scanning, then the LCD would serve no purpose! Red, green, and blue laser diodes can all be modulated with very high speed data. So you could just modulate each beam as needed to generate the pixels, combine them, and use the spinning mirrors to scan them across the screen. If you do it fast enough with enough lumens, the human eye's light pigment persistence will work fine, instead of messing with another phosphor. You wouldn't even need a projection lens.
Direct Laser Projector
It has been tried, albeit on a diy mono level;
http://heim.ifi.uio.no/~haakoh/avr/
And the research is there for full blown video;
http://www.photonicsonline.com/content/news/article.asp?DocID={EA2E56F4-057C-11D4-8C30-009027DE0829}&Bucket=&Featured=&VNETCOOKIE=NO
Attempted it myself with the removed laser scanner gubbins of an old HP laser printer on a VERY basic scale.
Conclusions (DIY);
1. Dangerous! Any direct viewing of the device could be hazardous, appreciate that this hazard exists now with MH projectors, but retina damage is more of an issue with larers.
2. Image would need to be scanned at a higher resolution to avoid extreme pixelation/screendoor, alternatively you could difuse the beam slightly, but would defy the point of having that high res capability in the first place.
3. Would require a dual motorised scan system. Something akin to the old VHS helical scan system would be best. The issue here would be that we would be moving away from LCD/optics and into high speed motion control.
4. It can be done and most probably will happen on a commercial scale. No high heat ouptut or high end optics, just three laser units and two spinning mirrored motors - just a matter of time. the biggest problem I think from a DIY point of view is the control gear, i.e. getting control info from a video signal.
When I first starting playing with this idea, I was convinced I had found the Utopia of DIY projection, alas brute force Lumens is the only way at this stage in the game if you want to produce a high quality image.
Despite my conclusion that MH is still the best way forward for cost effective image projection, I am still committed to playing with laser projection. I would encourage people to do the same, strip the imaging laser assembly from an old laser printer, replace the laser unit for a standard laser pointer (laser printer lasers do not operate in the visable spectrum and by the time you manage to drive the circuitry you will wish you hadn't bothered) and have a play.
I'm waffling now. Basically, there is some mileage in this topic, even at a DIY level.
It has been tried, albeit on a diy mono level;
http://heim.ifi.uio.no/~haakoh/avr/
And the research is there for full blown video;
http://www.photonicsonline.com/content/news/article.asp?DocID={EA2E56F4-057C-11D4-8C30-009027DE0829}&Bucket=&Featured=&VNETCOOKIE=NO
Attempted it myself with the removed laser scanner gubbins of an old HP laser printer on a VERY basic scale.
Conclusions (DIY);
1. Dangerous! Any direct viewing of the device could be hazardous, appreciate that this hazard exists now with MH projectors, but retina damage is more of an issue with larers.
2. Image would need to be scanned at a higher resolution to avoid extreme pixelation/screendoor, alternatively you could difuse the beam slightly, but would defy the point of having that high res capability in the first place.
3. Would require a dual motorised scan system. Something akin to the old VHS helical scan system would be best. The issue here would be that we would be moving away from LCD/optics and into high speed motion control.
4. It can be done and most probably will happen on a commercial scale. No high heat ouptut or high end optics, just three laser units and two spinning mirrored motors - just a matter of time. the biggest problem I think from a DIY point of view is the control gear, i.e. getting control info from a video signal.
When I first starting playing with this idea, I was convinced I had found the Utopia of DIY projection, alas brute force Lumens is the only way at this stage in the game if you want to produce a high quality image.
Despite my conclusion that MH is still the best way forward for cost effective image projection, I am still committed to playing with laser projection. I would encourage people to do the same, strip the imaging laser assembly from an old laser printer, replace the laser unit for a standard laser pointer (laser printer lasers do not operate in the visable spectrum and by the time you manage to drive the circuitry you will wish you hadn't bothered) and have a play.
I'm waffling now. Basically, there is some mileage in this topic, even at a DIY level.
lasers?
Since the topic has slipped sideways into lasers I have some questions.
As Guy Grotke pointed out earlier an attempt to use cheap "white" leds wouldn't work since the color isn't true. What about lasers, How red, green or blue is a red green or blue laser?
Supermarket laser scanners cover about a 4" wide horizontal area but seem to flicker/strobe, do other people see this or just people like me with poor perception of the color red? How much work would it take to get one to scan in the vertical as well? How oblique of an angle can a laser strike at and still retain a dot shape? I was thinking that if this was doable it would need to be rear projection for safety. I have a feeling this is all way beyond diy.
Since the topic has slipped sideways into lasers I have some questions.
As Guy Grotke pointed out earlier an attempt to use cheap "white" leds wouldn't work since the color isn't true. What about lasers, How red, green or blue is a red green or blue laser?
Supermarket laser scanners cover about a 4" wide horizontal area but seem to flicker/strobe, do other people see this or just people like me with poor perception of the color red? How much work would it take to get one to scan in the vertical as well? How oblique of an angle can a laser strike at and still retain a dot shape? I was thinking that if this was doable it would need to be rear projection for safety. I have a feeling this is all way beyond diy.
Lasers
Lasers can be pretty darn narrow in bandwidth (i.e. most of the light is of one wavelength, a "pure" color). They (especially solid state lasers) also can be turned on and off very rapidly (nanoseconds or faster). But the intensity required per pixel will still be huge in order for persistence to work. As I mentioned above, a 1 million pixel 1000 W continuously lighted screen is effectively providing 1mW per pixel. If a laser(s) is scanning all 1 million pixels one at a time then duty cycle on each pixel is 1/1,000,000. Therefore will need a 1000 W laser(s) doing this to provide same integrated energy per pixel over time. 1000 W lasers (especially solid state visible) are insanely expensive and big, not gonna happen anytime soon. Plus, as mentioned, a 1000 W visible laser will fry you very easily (for comparison "high power) laser pointer are around 0.005 W).
Lasers can be pretty darn narrow in bandwidth (i.e. most of the light is of one wavelength, a "pure" color). They (especially solid state lasers) also can be turned on and off very rapidly (nanoseconds or faster). But the intensity required per pixel will still be huge in order for persistence to work. As I mentioned above, a 1 million pixel 1000 W continuously lighted screen is effectively providing 1mW per pixel. If a laser(s) is scanning all 1 million pixels one at a time then duty cycle on each pixel is 1/1,000,000. Therefore will need a 1000 W laser(s) doing this to provide same integrated energy per pixel over time. 1000 W lasers (especially solid state visible) are insanely expensive and big, not gonna happen anytime soon. Plus, as mentioned, a 1000 W visible laser will fry you very easily (for comparison "high power) laser pointer are around 0.005 W).
some inspiration
a bit of ispiration to keep the LED experiments going
fresh from the CES
LED powered mini DLP projectors
http://hometheater.about.com/b/a/137964.htm
a bit of ispiration to keep the LED experiments going
fresh from the CES
LED powered mini DLP projectors
http://hometheater.about.com/b/a/137964.htm
Umm, I'm new at speaker building but I figures I'd post here because I would like to know how you would integrate LED's with a speaker so that when the speaker has a peak siganl the lead flashes to the beat. I'm curious about this after I glanced one of these shelf stereos at Best Buy
http://www.sharp.ca/products/index.asp?cat=40&id=632
I thought it was a neat idea and I was curious as how you would wire in the LED with your speaker.
Later on I might even build a set of speakers with Led's like these just for the hell of it.
http://www.sharp.ca/products/index.asp?cat=40&id=632
I thought it was a neat idea and I was curious as how you would wire in the LED with your speaker.
Later on I might even build a set of speakers with Led's like these just for the hell of it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.