I have a need for a way to move a light around a geometry(such as on a circle(not a disk)).
One way would be to have a single LED and "cover" above the led. The cover has a small hole in it letting the LED light shine through. If hte cover can be easily positioned then it would produce the effect I want. Because the cover requires mechanical translation it is more difficult than what I'm interested in (unless there is some trick to moving such things without motors/pullys/etc... (I've seen some cool translation devices based on piezo's somewhere).
In any case what I was thinking about was if there was a way to electrically position light. Somehow using frequency resonance in 2D on an "led" plane(I'm just making up stuff here).
Of course I could just use a bunch of LED's but this doesn't give me the resolution I want, costs a ton, and requires a lot of circuitry.
Basically I just want to move a dot around a small circuit where the angular position is controlled by an electrical single. Nothing fancy but just something simple. I'm looking more about if it is theoretically possible rather than some actual device since I doubt there is anything out there that does this.
Any ideas how one can achieve this?
Just incase your confused on what I'm asking. You can also think of the knighrider effect from the car. In my case I want to be able to be able to have a continuous position(or close) and to be able to use a single signal.
How precisely does it need to be positioned? What are the consequences if the position of the light is slightly off? "Hunh, that light is slightly off" or "Doctor, the patient is dead."
What angular tracking rate do you need? How large of a circle? Dinner plate or football stadium?
How much light? Visible in room illumination or midsummer Mojave desert?
Making some reasonable assumptions, using hobby-grade servo motors might do the trick. Use a pair oriented X-Y to directly aim the light source or (slightly harder) use a fixed light source and the servos to orient a mirror.
Highly collimated LED laser focussed onto the plane of the display wall and a mirror galvanometer movement. Two of them at right angles and you can draw simple pictures or patterns as well.
Obviously it won't be used in critical applications but will be viewed so if it's off more than a few percent it will be obvious.
About an inch square(probably half if I were to actually implement it).
This is precisely what I'm trying to avoid. Motor's is the "standard" solution but expensive and bulky. If there were some way to "position" light then it be a better solution. There are many possibly solutions but I'm not sure there are any cheap ones.
The only practical solution I can think of is to use a fixed LED with the "cover" and hole and position the cover using some means. magnetics could possibly be used for both the linear and angular 1D case.
I'm looking more for a possible solid state monolithic method or something that approximates this well. What I'm thinking of is possibly using a standing wave that has a high resonant peak through a material that response to it(sorta like an LCD). You basically position the "voltage" along the material and the material responds to it. Not sure if this would be easily doable though since it might be difficult or require a lot of circuitry to create such standing waves.
The main point is that I want something simple. In this case it means few parts and low cost. Because I'm looking for more of a theoretical solution I don't mind pretending it is already in production. Obviously with enough money and techology one can shrink just about anything down to scale and say it will work. I'm looking for it to be practical too.
Think of a consumer appliance. Generally these have led's. In fact, a simple digital dB meter would be a good application(although I want mine slightly smaller than these). These are generally made up of a stack of LED's. I'd like a way to do this continuously with a simple way to control it. If one could easily create a standing wave it would be pretty cool to use.
Alternatively maybe one can use a voltage to control the angle of a solid state device such as an LED by changing some width in it's geometry. This would be shined on a mirror to increase the length.
Suppose someone showed you a small disk(say about 1cm radius and about 3mm thick) that had 5 or so pins on the bottom and they were able to use them to create a small light on the disk and position it anywhere. How would you think it could be done? Your best guess would be starting with something solid state. Possibly a very dense array of led's with a controller in it that translated the voltages into coordinates.
Lets try to get away from a "discrete arary" and "controllers" and come up with another idea.
That would work but might be too costly or not small enough(I'd like to fit all this stuff in the size of a dim(or slightly smaller). One could probably use piezo's to position the mirrors(using a "wedge" shape possibly).
Would be nice if there were less of a "mechanical" solution though. LED based technology should work if one could design an LED to be able to modify itself. e.g., having a "channel" in the led that could be modified with a signal which modifies the position of the emitted beam which may use a mirror to magnify the effective change.
While it would be nice to display full pictures I would need something about half the size(slightly smaller than a dime). I don't really need anything but to display a movable dot so feel that there is some simple way to do so although the technology may not be out there.
Well, you said something about "theoretically possible, I'll shrink it later". There you are -- use an OLED, or a high-resolution LCD, and switch pixels on and off. Pretend that it's only active in the ring that you want, then when later comes have your parts made up with _only_ the ring active.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html
Yes it was. You're getting answers that are well within the parameters that you've stated, but you're rejecting everything you've gotten so far. Quit complaining and start cutting metal, or you'll never get anywhere.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html
I've explained what I want to do and the only person that seems to get it is Martin Brown. I said I would like a solid state version. It seems to me you need to stop wasting peoples time and go "cut metal" yourself.
A dime is what about 1cm diameter x 1mm thick. I can't see you being able to do it without enormous expense miniaturising things.
The buds for walkman should provide you with small cheap voice coils and mechanics and a bit of mylar film on one of them would be a start.
Other options would be a pair of narrow slits at a shallow angle so that a small linear shift perpendicular to the slits moves the crossing point where light gets through a long distance.
If you described why you wanted to do this it might be possible for others to suggest something more appropriate.
I know some hand-stuff PCB stuffing operations use a LED to point to where the part goes on the circuit board. Then, the workers grab the part and insert it. Later, it's off to the waveflow... Is this what you're talking about?
I don't understand your comment: "Basically I just want to move a dot around a small circuit..." What small circuit?
Anyway, another idea (maybe?) is I've seen those "LED clocks" for lack of a better description. They have this rotating or oscillating "stick" and embedded in the stick is a bunch of LED's. As the stick rotates (rapidly), the LED's turn on or off based on their angular position. The human eye "translates" these on/off LED's vs. time & position into legible clock readout (or text, etc...)
Otherwise, I'm afraid I don't know what you're looking to do. Some other posters have mentioned OLED, and that could be reasonable (in terms of $$ and headaches), but again, not clear on your desired outcome.
Either the light source must move, or the target must move (or both). Either option may include mirrors, etc... The only other option I see is to use multiple light sources.
Not necessarily. Obviously using todays technology that is the only way have.
I'm interested in a new way using current theory to develop something more practical. It is really more of a thought experiment about something I would use in practice if it were out.
The idea is to use current technology to create exactly what I'm looking for rather than using current products to create something that isn't.
As I pointed out in some post, suppose you want to create a dot on a small screen and be able to move it around continuously(or almost). How would you tackle this? What if no moving parts and only a few signal lines can be used? What if no complex controller and/or driver can be used? What does this mean? For starts it means you have to probably come up with some base material to use that will you can manipulate to do what you want. (This is probably why some are having such trouble with this question).
The best thing I can do is offer my line of thinking(theoretically valid or not):
Take a rectangular piece of light emitting semiconductor. Send a pulse down it along the anode channel(think of it as just a stretched out LED). As the pulse is traveling down the anode channel the voltage across the diode is larger at that location.
< pulse
-------------------- < anode
-------------------- < cathode connected to ground
from * to ground the voltage is high. Therefore the local region of this led should emit light rather than the whole channel. It will depend on the pulse size in some way and unfortunately travel at the speed of light.
In this case we get a "kightrider" effect as the light moves from left to right with the pulse. Probably on far to small of a scale to see. This should work in theory I believe without much problem besides scale. One end of the anode would probably be connected to ground through a resistor. You can think of it being like a transmission line.
If we were able to setup a standing wave with a single resonant peak and control it's position then we could control the position along the LED strip. I'm not sure if this is even theoretically possible to do but maybe some adaptation could work.
This uses current theory and technology for the device but may require a complex control circuit for setting up standing waves if it is possible to do so.
Another possible way would be to use the similar concept above but have a movable jumper between the vcc and the anode. The moveable jumper's position could be controlled by a voltage which would accellerate it and it's position could be computed and should be rather predicable(or some feedback could be used). This would be more of a mems application though and I'm not sure of hte lifespan of the jumper or sliding contacts.
-------------------- < Vcc * < jumper
-------------------- < anode
-------------------- < cathode connected to ground
Anyways, I'm sure there are other ideas that would "work" too. Theoretically this stuff should "work" as there is nothing advanced here. Just basic physics and electronics. If it did work well then it wouldn't be hard to create(not much harder than a normal LED, which I guess is pretty easy).
These also work on the small scale and can be used a wide range of 1D geometries with low power consumption and size.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.