I've been running variants of this design around in my head for something like two years, but I haven't drawn schematics on the computer until now. The general idea is that you have a rotating laser (yaw) and a rotating mirror (pitch), then maybe some lenses and something to draw on. If you build it small enough, you can probably drive it with stepper motors and make a (VERY LOUD) monochrome vector graphics display on the inside of a pair of sunglasses. Build it bigger and you can project it on a wall or something.
(http://namcub.accela-labs.com/pics/vector-schem.png)
I have too little blood in my caffeine system currently, and don't trust my math, so I won't post my code for determining the position of points based on the rotation of the two pivots and the distance between the pivot and the {mirror,laser source}. If you have mirrors, steppers, laser diodes, and too much free time, feel free to play around with the design. I don't have the cash to build a prototype (though it is sure cheaper than a MyVu). If you happen to have a red laser, a green laser, and a blue laser (plus, again, lots of free time and some more lenses) you can probably build a colour version capable of displaying both primary and secondary colours of light (plus white), but that requires more optics experience and probably more maths than I am aware of.
Probably. Not sure. The main problem with just using plain DC motors is that they don't stop when you tell them to. I know for sure that floppy drives have usable steppers for this project. Scavenged servos from remote control cars/planes/whatever are also good. But, if you can drive it accurately, it's probably okay. (I'm not sure if a cdrom motor is just a DC motor for spinning the disk and a couple solenoids for moving the laser around on one axis -- in which case, it would not be suitable).
While a stepper has very smooth motion, getting one with a fast enough response to actually properly generate what you need you might need something expensive.
The original design (as of N years ago) had DC motors and bumpers. As long as the timing is reliable, you can probably turn the laser on and off while the thing is going. I just didn't think the timing would be reliable. A variant was kind of a pseudo-raster (two motors set to go back and forth out of phase, with the speed of one being N times the speed of the other, N being the number of rows, and the only thing being controlled is the laser) -- but that should probably be far less efficient, and one of the goals was to be able to automate particular types of things by memorizing the path (it was initially for a cheap augmented reality rig, so having a dedicated micro capable of remembering a series of paths and going through them on the command of another micro would be VERY useful since all the chips involved were planned to be AT90S8515s or something).
Gonna be big. And pricey.
This stunner thing is very simple by comparison, and is still costing an arm and a leg, and more time than I thought.
Quote from: Horrendous Foreign Love Stoat on February 15, 2010, 07:36:26 PM
I'm kinda lost as to how this would work out cheaper, as surely you'd need an arduino or knock off, at least to drive em? right?
Initially, I was planning to use an AVR -- specifically, an AT90S8515 (which I have two of, and which sold for something like $5, and more like $2 if you order ten at once). Currently, that line is being phased out, and they suggest the ATMEGA8515 (same pinout, twice the flash memory, twice the speed, slightly cheaper last I checked) instead. The arduino would probably be overkill for this project. If you managed to scavenge all the other parts (including gearboxes, lenses, whatever) a stereoscopic HUD of this type would cost you about $20 (three ATMEGA8515 chips plus shipping -- one for each eye, and one to keep they eyes synced up. Might actually be less, since you might trigger free shipping).
That said, mechanics isn't my thing, nor is optics, so appropriate motors and lenses may be super-expensive or super-rare. I was told that the steppers out of a floppy drive would be suitable for the project by an otherwise reliable source, but that was several years ago and I didn't have any schematics to show.
Edit: The programmer for an ATMEGA8515 (if you use a parallel port) is a couple capacitors, a resistor, a breadboard, and a 555 timer, iirc. There's a gnu toolchain for compiling code for that arch. I never got around to actually using either of my AVR chips, but in theory they shouldn't be particularly expensive.
Stepper motors might be overkill-you're just moving a mirror. You could probably use a speaker and a light weight mirror to point the light ray wherever.
You could probably use a pivot and a solenoid as well, and just keep the solenoid moving in and out at the appropriate rate. However, I don't know where to scavenge solenoids.
How would you do it with a speaker, by the way?
Note: This project will never get done.
It's an interesting thought experiment.
You are probably right. It took me three years to draw the damned thing up.
Quote from: Enki v. 2.0 on February 15, 2010, 09:23:55 PM
You are probably right. It took me three years to draw the damned thing up.
Now all you have to do is buy the hardware, and build it.
And like I said, that simple-ass stunner I'm working on is taking more time and money than I expected, by a long shot.
Well, buy the hardware, built it, and write the code to make it into something other than a doorstop ;-)
Quote from: Enki v. 2.0 on February 15, 2010, 09:29:51 PM
Well, buy the hardware, built it, and write the code to make it into something other than a doorstop ;-)
Wait. The code isn't even written?
Do you at least have the logic worked out?
It's trig. Other than accounting for the radius of the axles, it's one line of C code in a loop.
Quote from: Enki v. 2.0 on February 15, 2010, 09:18:30 PM
You could probably use a pivot and a solenoid as well, and just keep the solenoid moving in and out at the appropriate rate. However, I don't know where to scavenge solenoids.
How would you do it with a speaker, by the way?
You can scavenge solenoids from car starters in auto junk yards, from battery operated clocks and, I believe, from old disk drives. And also from my brother's garage if you're willing to travel to Utah for one.
Hmm. I'd love to see something like this in car windshields.
Hint: a speaker is a very sensitive solenoid. They don't particularly like DC through them, but that's easy enough to account for. You'll need a DAC and a uC and that's pretty much it. Maybe I should get in on this...
Also--the geometry approximates linear for incident angles close to the normal.
Quote from: Template on February 16, 2010, 12:23:56 AM
Hint: a speaker is a very sensitive solenoid. They don't particularly like DC through them, but that's easy enough to account for. You'll need a DAC and a uC and that's pretty much it. Maybe I should get in on this...
Also--the geometry approximates linear for incident angles close to the normal.
Oh, okay. I dig. I wasn't sure if you were talking about using it as a solenoid, or if you were planning to do something funny with sound frequencies (in which case I'd suggest a peizo disc instead).
Quote from: Enki v. 2.0 on February 16, 2010, 12:59:27 AM
Quote from: Template on February 16, 2010, 12:23:56 AM
Hint: a speaker is a very sensitive solenoid. They don't particularly like DC through them, but that's easy enough to account for. You'll need a DAC and a uC and that's pretty much it. Maybe I should get in on this...
Also--the geometry approximates linear for incident angles close to the normal.
Oh, okay. I dig. I wasn't sure if you were talking about using it as a solenoid, or if you were planning to do something funny with sound frequencies (in which case I'd suggest a peizo disc instead).
They're both linear actuators known to work into the kHz range...
I don't know that I have the equipment or fine motor skills to build a rotating mirror that can be driven by a peizo disc.
Quote from: Enki v. 2.0 on February 16, 2010, 01:38:29 AM
I don't know that I have the equipment or fine motor skills to build a rotating mirror that can be driven by a peizo disc.
:piss:
Use two solenoids to actuate two single-axis mirrors (or one double-axis mirror--trickier).
[*LASER*]:---\(Pivot Mirror 1)
|
(P M 2)\-----|SCREEN|
Quote from: Doktor Howl on February 15, 2010, 07:40:51 PM
Gonna be big. And pricey.
This stunner thing is very simple by comparison, and is still costing an arm and a leg, and more time than I thought.
the article about the home-made stunner said the hardware cost somewhere around $250 total, IIRC?
Quote from: Horrendous Foreign Love Stoat on February 16, 2010, 08:59:00 AM
QuoteInitially, I was planning to use an AVR -- specifically, an AT90S8515 (which I have two of, and which sold for something like $5, and more like $2 if you order ten at once). Currently, that line is being phased out, and they suggest the ATMEGA8515 (same pinout, twice the flash memory, twice the speed, slightly cheaper last I checked) instead.
shit. those are disposable cheap. :D cheers.
http://www.atmel.com/dyn/products/product_card.asp?PN=ATmega8515
http://www.avrfreaks.net/index.php?module=FreaksTools&func=viewItem&item_id=145
These would be the AVR resources. The $5 pricetag was from jameco, iirc.
Okay, so I had a little too much caffeine, and half-built the thing (giving no heed to the fact that I had none of the required parts save a DC motor with a cam), unbuilt it, rebuilt the same half differently, rinse and repeat about six times. Having broken it enough times, I think I can safely say that if I get around to trying to build it some time when I'm not too manic to take the time to fabricate a stiff rod of the appropriate length and count the teeth on the gears, I could probably do it in an evening. The source code for the driver on the AVR should be easier, though I would probably need to know which method of driving the mirror and laser are being used and how many rows and columns should be used.
Protip: really ancient CDROM drives (the kind that look about twice as big as they should) often have a lot of gears and shit inside (a lot more than necessary). A smart man (read: not me) would just rearrange the gears and remove the stuff that would get in the way on the innards of such a drive and implement the display that way, then reassemble the drive and stick a lens where the door for the drive bay would be, rather than breaking the damned thing and scattering the gear assemblies all over the floor.
Quote from: Enki v. 2.0 on February 20, 2010, 02:37:30 AM
Okay, so I had a little too much caffeine, and half-built the thing (giving no heed to the fact that I had none of the required parts save a DC motor with a cam), unbuilt it, rebuilt the same half differently, rinse and repeat about six times. Having broken it enough times, I think I can safely say that if I get around to trying to build it some time when I'm not too manic to take the time to fabricate a stiff rod of the appropriate length and count the teeth on the gears, I could probably do it in an evening. The source code for the driver on the AVR should be easier, though I would probably need to know which method of driving the mirror and laser are being used and how many rows and columns should be used.
Protip: really ancient CDROM drives (the kind that look about twice as big as they should) often have a lot of gears and shit inside (a lot more than necessary). A smart man (read: not me) would just rearrange the gears and remove the stuff that would get in the way on the innards of such a drive and implement the display that way, then reassemble the drive and stick a lens where the door for the drive bay would be, rather than breaking the damned thing and scattering the gear assemblies all over the floor.
Rows and columns are raster concepts. We're talking about a vector display. The instruction set is roughly {turn on, turn off, move(xy vector), return to center}, right? Repeated/arpeggiated to make an actual image...
Something like that, right?
Quote from: Template on February 20, 2010, 03:11:33 AM
Quote from: Enki v. 2.0 on February 20, 2010, 02:37:30 AM
Okay, so I had a little too much caffeine, and half-built the thing (giving no heed to the fact that I had none of the required parts save a DC motor with a cam), unbuilt it, rebuilt the same half differently, rinse and repeat about six times. Having broken it enough times, I think I can safely say that if I get around to trying to build it some time when I'm not too manic to take the time to fabricate a stiff rod of the appropriate length and count the teeth on the gears, I could probably do it in an evening. The source code for the driver on the AVR should be easier, though I would probably need to know which method of driving the mirror and laser are being used and how many rows and columns should be used.
Protip: really ancient CDROM drives (the kind that look about twice as big as they should) often have a lot of gears and shit inside (a lot more than necessary). A smart man (read: not me) would just rearrange the gears and remove the stuff that would get in the way on the innards of such a drive and implement the display that way, then reassemble the drive and stick a lens where the door for the drive bay would be, rather than breaking the damned thing and scattering the gear assemblies all over the floor.
Rows and columns are raster concepts. We're talking about a vector display. The instruction set is roughly {turn on, turn off, move(xy vector), return to center}, right? Repeated/arpeggiated to make an actual image...
Something like that, right?
Right now, the mirror and the laser are driven by a single DC motor with cams and gears and shit. So, it's a pseudo-raster system, since all I control is when the light goes on.
Quote from: Enki v. 2.0 on February 20, 2010, 01:50:51 PM
Quote from: Template on February 20, 2010, 03:11:33 AM
Quote from: Enki v. 2.0 on February 20, 2010, 02:37:30 AM
Okay, so I had a little too much caffeine, and half-built the thing (giving no heed to the fact that I had none of the required parts save a DC motor with a cam), unbuilt it, rebuilt the same half differently, rinse and repeat about six times. Having broken it enough times, I think I can safely say that if I get around to trying to build it some time when I'm not too manic to take the time to fabricate a stiff rod of the appropriate length and count the teeth on the gears, I could probably do it in an evening. The source code for the driver on the AVR should be easier, though I would probably need to know which method of driving the mirror and laser are being used and how many rows and columns should be used.
Protip: really ancient CDROM drives (the kind that look about twice as big as they should) often have a lot of gears and shit inside (a lot more than necessary). A smart man (read: not me) would just rearrange the gears and remove the stuff that would get in the way on the innards of such a drive and implement the display that way, then reassemble the drive and stick a lens where the door for the drive bay would be, rather than breaking the damned thing and scattering the gear assemblies all over the floor.
Rows and columns are raster concepts. We're talking about a vector display. The instruction set is roughly {turn on, turn off, move(xy vector), return to center}, right? Repeated/arpeggiated to make an actual image...
Something like that, right?
Right now, the mirror and the laser are driven by a single DC motor with cams and gears and shit. So, it's a pseudo-raster system, since all I control is when the light goes on.
OK. So, like the laser barcode reader at a grocery store? You should be able to minimize parts if you do it right. A barrel-shaped faceted mirror might make things lots easier...