Detecting a pulsed laser in daylight

Jimmy W.

I have been asked to look at a project that will have many laser guns(<255) with many targets (~128) I just got clarification that this will be set up in daylight conditions, each gun will transmit its ID on the laser in Sony IR format to the receiver on the target, it was originally going to be indoors in dimly lit conditions, outdoor with high light conditions i now cant use a simple photoresistor, does anyone have any ideas on how to reliably pickup the laser in high light conditions?

Jimmy

Leon

Laser wavelength?

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

4Alex

@jimmy:

Is it a visible or infrared laser? Which wavelength? There's Fairchild detectors that are very discriminating. If infrared, these detectors will do the job, if visible (presumably red), you will need light filtering of wavelengths other than of your laser.

Cheers,

Alex

CounterRotatingProps

Hi Jimmy,

Sheild the detector with a black painted tube - a piece of cheap PVC pipe works great.

What color/frequency are the lasers to be?

Does rain matter? (Heavy rain will cut the beams and potentially cause false signal pulses, or disrupt things entirely.)

- Howard

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

StefanL38

every IR-remote-control for TV, stereo, DVD-player etc. uses a 36-40 kHz carrier frequency that the receiver can
filter the IR-signal.

Photoresitsors are reacting very slow. You should use phototransistors as a minimum

take a look into the datasheet of this

Infrared Receiver

which wavelength sensivity it has at your lasers wavelength

best regards

Stefan

Jimmy W.

CounterRotatingProps said...
Hi Jimmy,

Sheild the detector with a black painted tube - a piece of cheap PVC pipe works great.

What color/frequency are the lasers to be?

Does rain matter? (Heavy rain will cut the beams and potentially cause false signal pulses, or disrupt things entirely.)

- Howard

650nm for the red lasers
532nm for the GREEN lasers

I dont think tubes will work, that will make them have to shoot the targets dead on, they want a decent shooting angle.

They will not be used in rain conditions
@Stefan: Yes, I am looking at phototransistors now


EDIT: Whoops I meant to say GREEN, not red twice

Post Edited (Jimmy W.) : 8/1/2009 5:13:33 PM GMT

kwinn

Try an optical filter over the sensor and see if the receiver module for tv remotes will work with the red laser. If it does you are in luck as it has 40KHz filtering built in. If the remote control module does not work then you have a bit more work to do.

cessnapilot

Hi Jimmy,

You would like to detect from a distance, yes? A 100 m or, so. Start to check our good friends, the sensitive laser detectors in cars, then.

In your application, I think, ·lasers are (should be) in the infrared region to be eye-safe. Look for TAOS new, higly sensitive light detector chip the TSL237. Its spectral range from 320 to 1050 nm is maybe not optimal for those IR lasers, but it is perhaps good enough on the spectral edges. A 2D 'bull's eye' array· of those small sensors, where their signals can be correlated with the Propeller (yes, in DSP mode with high speed fixed-poin math) should be very responsive.

When your lasers are really in the red (630 nm), those TSL237 patches wil be the most sensitive (top at 700 nm), but for the light of the environment, too. Apply narroband(!) flexible red filter foils on the detector's fabric to improve S/N for the (hopefully) slightly diverging beams.

Or, choose a photodiode-amplifier hybrid with filter to be more cute. At the extremes you can use Peltier cooling of the device. The whole thing packs within a 1-2 cubic cm, and can be constructed durable for on-field application. Again, DSP will be necessary to pick out the signal from the cluttered radiation of the environment. I guess, to sense from 1-2 km, or from even farther, is realistic with that.

You can even build a laser distance·sensor for several km·with such a sensor + optics. The Prop·can handle those timings easily. ·A laser signal, reflecting back from·1.5 km, travels for about 10 usec. Counting a 120 MHz pulse wave from the shot until the echo, gives 1200 counts.· So, resolution is about 1-2 m with that counter in a COG. Or just take the difference of CNT readings, to make life easier.

If you accept help and assistance·in programming, let me know. A few free games will be enough for me.


Cheers,


Istvan


Post Edited (cessnapilot) : 8/1/2009 1:52:39 PM GMT

Jimmy W.

cessnapilot said...
Hi Jimmy,


When your lasers are really in the red (630 nm), those TSL237 patches wil be the most sensitive (top at 700 nm), but for the light of the environment, too. Apply narroband(!) flexible red filter foils on the detector's fabric to improve S/N for the (hopefully) slightly diverging beams.

If you accept help and assistance in programming, let me know. A few free games will be enough for me.

Istvan

Any idea where to get these filters? We are going with red laser @ 650nm and green @ 532nm, the teams will only be able to shoot the correct targets so filtering will be a bonus. Looking @ TSL237 it looks like these are light to frequency devices, I am going to be modulating the laser to send a byte in sony format to the target to identify which gun shot it. I guess there are way to still decode the signal if that were the case.

Jimmy

4Alex

@Jimmy: I get my filters from Edmunds Optics at www.edmundoptics.com. They have flexible and solid plastic sheets. The solid sheets are harder to cut in small pieces as it breaks like glass. If flexible works for you, that's what I would use.

You can use standard photosensors from Fairchilds, Optex, Panasonic, etc. Then you have to deal with the electronics and mounting. It may sound a bit odd at first but have a look at IF-D93 from Industrial Fiberoptics. On Digikey, p/n is FB122-ND with datasheet. I've used these components both for inexpensive on-site plastic fiberoptics links and also for other goodies. Completely plastic, easy to cut, modify, adapt, and install in custom enclosures. There's even a mounting screw hole. They are sensitive photodarlington designed for transmission through plastic fiber optics in visible and IR. The inner male mating is easy to break and that would leave you with a flush surface in front of the photosensor where you could glue your optical filter. Plastic-on-plastic is very easy to glue. Anyway, that's how I would do it. As for your modulation question, these parts are intrinsically designed for modulated light signals!

Cheers,

Alex

Post Edited (4Alex) : 8/1/2009 7:42:20 PM GMT

cessnapilot

Oops, I thought it's a target ON/target OFF detection, even that would not be easy. To identify shooter from the faint signal in the noise is more formidable task. Would you like to detect 'Blue on Blue' shots in this color coding?

Phil Pilgrim (PhiPi)

You don't want a light to frequency detector. It won't be fast enough to decode a signal with. A light to voltage sensor, along with a fast ADC would be much better. That way you could demodulate the signal using simulated I and Q mixers.

Without shielding the sensors from direct sunlight, you will have some real problems with sensor saturation. Once the sensor reaches its maximum output due to too much light, you will not be able to detect any modulation riding atop the ambient light.

One thing to keep in mind is that laser beams spread over long distances, whereas bullets do not. Therefore, you will likely record many more "bulleyes" with a laser than with real ammo.

-Phil

Jimmy W.

I have decided on an array of 13 sensor with a hard filter(to help prevent any beam distortion that may come from flexible filters) , the sensors will have a setback of 1/4 inch from the filter and the filter its self is 1/2 inch set back from the face of the unit, I am planning to implement a 1 inch hood above the sensor in an attempt to not interfere with shooting angle. I am not convinced that the right pickup is IF-D93, as its maximum response rate as 1khz, I am looking for a much faster data rate to ensure reading of the transmitted id, I would like to transmit it several times to ensure pickup. I know from spin I can do 120khz, as I have done a little irda work with it before. I want to transmit an attention preamble, 2 bytes( the id, and the inverted id) and a lead out. I would like to transmit that package at least twice, even with noise I can do error correction once I have captured 2 sets.
At the start the the day all targets and guns are on a common charging/data bus, they get all their RTC clocks synced up, the worst case scenario is the target cannot decode the ID, and has to ask the main processor via zigbee who shot within a few ms of it getting shot and then doing catchup and logging the data.
I am thinking about using some cheap programmables to do ADC and noise filtering and then shooting back data to the prop, I am going to see how to lay that out.

@Cessna, "blue on blue" targets would get stopped by the filter now, counting a total miss.

Jimmy

cessnapilot

Well Jimmy, if a Green team player hits with his laser a team-mate by accident, or because of a·lady, that should be taken seriously. Otherwise they will be crossing friendly lines of fire all the time, making the whole game unrealistic. But, it will be a game, so it may be acceptable.···

Jimmy W.

@Cessna: I would love to detect cross target hits, but I dont see it realistic to do so with the notch filters in place

Carl Hayes

One solution would be to use a prism (or a diffraction grating) to select only the wavelength you're interested in.· You can make it extremely selective that way, and get rid of nearly all noise.· For a target that will accept light from various directions you can use a curved first-surface mirror or similar, and aim the spectrograph at it.· That will lose you some light, but your s/n ratio might be very good anyway if your frequency selection (wavelength discrimination) is tight enough.· If you lose 60 db of noise and 20 db of signal, you're better off by 40 db, bottom line.· Many receiving systems work that way; Beverage receiving antennas are an example, being very insensitive but very quiet.

This can be very cheap to build; you'll need a curved mirror that will reflect IR or whatever colors you use; a·prism or grating that can pass those colors; a piece of pipe to separate the prism or grating from the detector; and the detector itself, which will be by a considerable margin the most costly part.

The longer the pipe (or other tube) separating the prism/grating·from the detector, the greater your wavelength/frequency selectivity will be.

You can even put the curved reflector out somewhere by itself and aim the spectrograph at it from a distance.

All this is brainstorming, of course; actually I have no expertise at all in building such stuff.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net