Short distance Light beam control (10cm) - what is best to use in On/Off mode?

[flotulopex]

Hello,

I have to make a kind of keyboard where the buttons are replaced by seven small light beam control (barrières lumineuses - Lichtschranke) having to check a distance of 10 centimeters (is the beam interrupted or not = On/Off).

It's going to look like a light beam grid - four rows and three colons.

I've been looking around in the web for hours and can't make-up my mind (to much information kills the information...) on which system is best to use.

In my opinion, the simplest, cheapest & smallest way to do would be to use common leds with a narrow beam as emitters and, on the other side, as receivers, photodiodes.

Another way would be to use IR leds. The advantage here seems to be that I could "code" the light beam to avoid interferences and make it more reliable. But I never used them up to now so I don't know if I can use the same IR leds as emitters and as receivers and if this solution is better or not.

Does anyone of you had some experience with this he could share?

Thanks a lot for any advice.

[rhino]

Hello,
Another way would be to use IR leds. The advantage here seems to be that I could "code" the light beam to avoid interferences and make it more reliable. But I never used them up to now so I don't know if I can use the same IR leds as emitters and as receivers and if this solution is better or not.

We use a product like this in our manufacturing processes for error proofing parts picking. There is an array of IR LEDS and receivers... I think upto 150 in each axis. Basically creates an IR grid system. An external controller will send the expected pick string to the unit. It will light up the appropriate bins with light bars for the operator to pick from. Once the operator sticks his/her hand in the lighted location, it turns off the light for that bin and the operator moves to the next lighted bin. If he/she were to stick there hand in the wrong bin, the system recognizes this and beeps and them and begins flashing the light bar for the bin they are supposed to pick from. We do have interference problems from time to time due to receiver sensitivity. For this, there are jumpers to adjust the emitter power.
If I were designing a small scale version, I'd think about pulsing the emitters at a different frequency and detecting that frequency for the appropriate receiver. Don't know if that makes it any more simple or robust, but maybe something to think about.

[skimask]

Hello,
I have to make a kind of keyboard where the buttons are replaced by seven small light beam control (barrières lumineuses - Lichtschranke) having to check a distance of 10 centimeters (is the beam interrupted or not = On/Off).
It's going to look like a light beam grid - four rows and three colons.
I've been looking around in the web for hours and can't make-up my mind (to much information kills the information...) on which system is best to use.
In my opinion, the simplest, cheapest & smallest way to do would be to use common leds with a narrow beam as emitters and, on the other side, as receivers, photodiodes.
Another way would be to use IR leds. The advantage here seems to be that I could "code" the light beam to avoid interferences and make it more reliable. But I never used them up to now so I don't know if I can use the same IR leds as emitters and as receivers and if this solution is better or not.
Does anyone of you had some experience with this he could share?
Thanks a lot for any advice.

I don't know if this is what you're looking for...but I thought what I built was cool...
www.srt.com/~jdgrotte click the LED Touch Sensor link and watch the video (which is really bad quality, someday I'll update it).

[flotulopex]

Cool stuff, skimask.

I've seen your video some times ago and your explanation about how to do/use it is really great

In my case, distance between the "finger" and the LED is greater and needs more precision. Take a look at the drawing I made to see what the distances will be (more or less).

Each beam is 1 centimeter distant from another and the distance from one LED to it's corresponding photodiode is about 10 centimeters.

I'll have to "cut" two beams (row/colon) to get a correct coordinate.

So I need to have a very narrow beam to locate precisely where my finger (or an object) will be positionned.
<img src ="http://www.picbasic.co.uk/forum/attachment.php?attachmentid=1569&stc=1&d=117747937 7">
Again, the question is: what light emitting and detecting system is the best?

[Ioannis]

Why don't you modulate each pair with different carrier? Also I think you have to play with the power that Tx LEDs will emit.

Ioannis

[rhino]

Why don't you modulate each pair with different carrier? Also I think you have to play with the power that Tx LEDs will emit.

Ioannis

That's a better description of what I was trying to say.

[skimask]

That's a better description of what I was trying to say.

I think I smell what you're cooking. I saw it on one of those shows talking about Ford Motor Company's assemblies processes.
How big are the boxes themselves?
How many are there going to be?

[flotulopex]

Sorry, I don't understand your comments...

What's a carrier?

Why working on the LED's power?

What are the bigs?

[mister_e]

Carrier = Porteuse (ou fréquence porteuse)

Roughly, it's the frequency you modulate your signal. like an FM/AM radio station.

Why playing with the Power of the LED... more power (current) = usually better range, lower power, lower range.

[skimask]

Sorry, I don't understand your comments...
What's a carrier?
Why working on the LED's power?
What are the bigs?

Carrier - the underlying information mode of transport. For instance...air is the carrier for sound waves, we modulate the air when we speak. Like AM radio, 900khz is the carrier, which we modulate the amplitude of to get the audio information from. Make any sense?
With an IR LED, the way it's most commonly done is to modulate the IR LED with a ~38Khz signal, then turn that whole signal on/off and a certain rate.
If you've got one IR LED working at 38khz and another one next to it working at say 50khz, the two probably won't interfere with each other as opposed to having two of them modulated at the same frequency next to each other.
A good example would be you and wife going out with another husband/wife couple. Both him and his wife are talking at the same time, both are transmitting different modulated carrier waves (sound), but you're only tuned into him, and that's easy (maybe 'cause it's easier for the guys to tune out the girls sometimes! ...oh that's gonna start something I can feel it now! ). If you were in that same situation talking to 3 other guys (with similar voices), it would be a lot harder to figure out who's talking to who.

Power - maybe because some of the light might spill over to other IR detectors...

Bigs - fixed it...should have read 'How big are the boxes themselves?'

[rhino]

I think I smell what you're cooking. I saw it on one of those shows talking about Ford Motor Company's assemblies processes.

Yep... probably the same system. The approach that Skimask suggests would do the trick. If you stagger the frequency to say every other emitter/reciever pair that are next to each other, you effectively reduce the possibility of picking up a stray signal from an adjacent pair. The other thing is to look into some kind of filter lens or plastic that will filter out stray IR similar to what you see on you VCR/DVD IR receiver.

[mister_e]

The other thing is to look into some kind of filter lens or plastic that will filter out stray IR similar to what you see on you VCR/DVD IR receiver.

i'll second that one. i'll read back the whole thing but i think some sequential scanning & Beaming may help too...

[flotulopex]

Okay for all your infos.

I think I'm going to start with IR and make some tests.

Just found an old remote where I can recover an IR-LED...

[Ioannis]

Also you may switch off the pairs and have on only one at a time. Of course this has to be done fast enough. Like scanning a keyboard. I suppose that is what Steve implied?

Ioannis

[mister_e]

Yup, that's what i think... never tried it so far.