I'm looking for a way to build up some kind of laser seeker.
I mark a spot with a handheld laserpointer, an autonomeous rover detects the light reflection and moves towards the light... (more or less the same principle as used for missile homing - just a little less powerfull for indoor aplications and without nasty payload;-)
my basic problem of course is finding a sensor which is able to detect the low-power reflection of my handheld laserpointer (actualy I don't need an Array: one simple I/O sensor would already do it in my case; an analog sensor also measuring the light intensity would even be perfect)
Although I knew that regular systems like the ones used in missiles use very strong IR-Laserpointers in combination with an Array of photo diodes, i've startet with testing some photoresistor (as expected: won't work at all) and phototransistor (as expected: work's up to a few inches), because these things are a lot easyer to get in an local electronic store than high-end Photodiodes and my detection range will in either way be limited to 10 meters (so no need for texas instruments high-end solution;-)
Now I'm going to test out some photodiodes, but i've had a hard time to find this stuff...
does any of you have experience with a similar project using a photo diodes? where to get? Wavelenghts that work better than others? tutorials?
i've already considered a similar solution using a 200x200 CMOS and image processing, but it's not my favorite at all. I prefere the mechanical scanning solution if possible.
i've already got the "mechanical infrastructure" for scanning (i'm using rotating mirrors), i just could not find a sensor capable of detecting the small amounth of incoming light (i used a 550 nm green solid state laser so far because green light doesn't show up too often in an indoor environment)
Strange choice, I always thought, that less mechanical parts, better reliability. Other things to considered, as camera basically consist of 640x480 sensors, single ptotodetector would get only 1 / 307200 of light energy, and consequently must have very low noise level. Last time I've seen mirroring scanning device, it has to be filled up with liquid nitrogen in order to get desired noise floor. But there was not other alternative, 10.6 um long IR thermal vision different story.
Did you try using the photo-transistor with a lens to focus/capture more light? What about some kind of amplifier between the photo-transistor and Arduino?
Strange choice, I always thought, that less mechanical parts, better reliability
i neither agree nor disagree on that point in general...but considering the fact that even a camera uses mechanical parts for the gimbaled platform makes that question more or less obsolent at all;-)
Other things to considered, as camera basically consist of 640x480 sensors, single ptotodetector would get only 1 / 307200 of light energy
i don't think so ]
intensity = absorbed energy / Energy absorbing area
=> one large 1x1mm Photodiode gets exactly the same amount of intensity AND energy as 1x1 mm camera chip which provides 10000 small diodes
but you might be right about the noise...
Did you try using the photo-transistor with a lens to focus/capture more light? What about some kind of amplifier between the photo-transistor and Arduino?
well, focusing on a spot would be an option but it's limited for a seeking device (you don't know the spot to focus on exactly, otherwise the seeking would be obsolent => you'll have to focus on every possible spot one after another => slows down the whole seeking process;-)
Amplifier... that doesn't work at all: amplifing changes the amplitude of a signal, not the frequency
=> for an amplifier you always need a weak - but clear - signal in the beginning (which i don't have as there is no signal at all yet;-)
Quote
Other things to considered, as camera basically consist of 640x480 sensors, single ptotodetector would get only 1 / 307200 of light energy
i don't think so
intensity = absorbed energy / Energy absorbing area
=> one large 1x1mm Photodiode gets exactly the same amount of intensity AND energy as 1x1 mm camera chip which provides 10000 small diodes
You've missed a time in your formula. Sensor is receiving reflected energy. Let's make a laser pointer = 1 mm for a moment. Reflected energy = Laser Power * refl. coefficient . If I spaced camera 1 m from the white wall, narrowing vision field down to 640 x 480 mm, only one of 307000 diodes would received full amount of power, w/o interruption - energy would accumulate, or integrate in a single led all time till I've moved the pointer (CCD) .
With a photodiode there is no integration process, as time when flow of energy reach a diode strictly limited to scanning speed. It's the same , like using a long exposure photography against making a movie 30 fps. Lightning conditions of two process enormously different. The only case when it's true, "gets exactly the same amount of intensity AND energy" is focusing a laser spot to a diode, making resolution 1 pixel.
right... i've never consideret the exposure time... and in my case the marked spot won't even move in a significant way which would be perfect for a signal-integrating device. I think i'll give this CCD-Stuff at least a try thanks for that input!
Radio Shack sells an optical sensor that can sense over a Very large range, has 3 pins for gain control. The issue is not I think to detect the light but to detect where it is coming from. is your device to work in a 360 degree theater or are you just using it to guide it, once you've got it's attention?. whether by photodiode (Btw did you know that LED's make fair to good photodiodes? lost takes an op-amp) and what of reflections, again do you scan looking for the strongest one and if so over what range of received input power... different materials have different albedo's? (can't remember and my spell checker's an idiot, I know I programmed it) but the reflected energy might vary widely If you modulated your laser beam (IMO a necessity) and made sure It couldn't respond to frequencies much below 500 Hz (60 Hz line noise and modulation of light from CCFL lamps, Tv, Etc). It might be embarrassing to have the device home in on a night light (love at first blink? ... There are some interesting engineering challenges here... as to modulating the light you could use it as a steering command... wide PWM on the light to the right and narrow for left? for 'fine' tuning?
One of the problems that you might hit in this project is the inability to distinguish between the scattered light from your laser pointer and other stray light sources.
In a real-world application you might add a monochromator so that your robot would only look for the specific wavelength of your laser light pointer. It is possible to build a monochromator using a diffraction grating (such as the surface of a CD or DVD) angled correctly towards your detector, however you will lose a lot of light and it is not the easiest of projects.
I have heard it is possible to use LEDs backwards as photodiodes, however their wavelength-sensitivity backwards is not the same as forwards: they will respond to all light of a lower wavelength than their emission. Green light would also respond to blue, violet, etc. Perhaps if you had a blue laser and several blue LEDs (with accompanying amplifier circuits) on different parts of your rover head.
On the other hand, it is easier to filter out UV-blue-green then it is to filter out red, IR. Unfortunately, most things in our hot world give off some amount of infrared.
thanks for that input!