Choosing the Right Infrared Emitter for DIY photosensor

[Chriss]

Hi!
After a while of digging trough the internet for a photo transmitter and receiver, but I'l give up and will make my photocell
from IR diodes and IR receiver.

The main reason for what I have to make my own sensor is because I can't find one for my purpose.

I will use this type of sensor in a seeder machine of a friend of mine, to count the seed and to alarm
when no seed is traveling through the system...

Based on the picture I attached. In an original build they use 3 or more LED's and 3 or more receivers.
I think this is the resolution of the reader.
The distance between are around 3cm.

Qusetion:
Do somebody can give me some advice what type and/or characteristic should I follow
by part choosing?

What should be the angle of emitting of the LED etc..

Or maybe better, if somebody have some experience in this and can tell me the
IR LED and IR receiver part number?

Any advice. help I will appreciate.
Thank's.

My best regards.

[Chriss]

Oh I forgot!
Maybe a circuit would be also nice.

Thank you.

[Kleinstein]

I would consider using multiple detectors and a single emitter at a slightly larger distance, so that the sensitive area is wide enough even at the emitter side. Likely a more small ample emitter (e.g. 10 -15 degree) should work best, just to well light the detector row. The minimum spacing on the detectors depends on the seed size - needs to be smaller than this.  2 Emitters might work too, but could reduce the signal strength as light from the other emitter might still hit the detector.

The suitable circuit depends on the speed the seeds drop through, background light and the way one want's to look at it. It could be just a phototransistor, a resistor and ADC input from an µC for each detector, but it depends.

[SeanB]

An old but very reliable detector is simply a light source, in this case incandescent lamp severely underrun at around 60% of rated voltage, or an IR LED or red LED run at around 30mA of current, and a solar cell that is driving a resistor load, and then capacitively coupling the resultant voltage change into an amplifier and then using that to detect the seed passing. Simple, rugged and reliable, and works on only the seed falling through the beam and attenuating the beam slightly, not actually obscuring fully. Solar cell is simply a wide section of cell, which develops around 300mV when illuminated, and placed behind a glass protective window as wide as the desired detector area. As you only are using a single cell section , a slightly larger cell with the mask obscuring the edges has little effect, just using a monocrystalline cell sliced to size.

[Chriss]

Thanks for the suggestions.

I thinking in the turn of Kleinstein.
That's a great idea I think, to put one light source and put many receivers.
That "many" receivers is actually the resolution of the detection, actually
if the seed is a smaller seed culture then I need more sensors, if the seed is
bigger in diameter I mean than I need lover number of sensors.

The data I would collect to a uC and then represent it to an LCD maybe to a laptop...

Actually, the sensor part should be interchangeable, because of the seed type...
That is also a good info, I didn't think in that direction...

Can somebody share with me a link about a ir diode and ir transistor pair what would be good to use?

Thank you!

[Kleinstein]

The IR sender could be about any IR LED (e.g. 850-960 nm) with a suitable angle to match the setup. So likely a 15 degree or similar type and maybe a 3 mm one if small seed are needed.
For the detector it depends on the minimal size to detect. So the BPW17 (2.4 mm wide) could be an option. The smaller the detectors the more you need. It would still work with larger seeds, its just a larger circuit.   So for a 30 mm wide slot this would be about 12 detectors and input.  With sufficient IR light intensity the signal should be strong enough to work with just resistor and ADC input or even resistor and schmidt trigger type digital input.

It could be a bit difficult to keep everything clean in a dusty environment. So a flat window might be an advantage over just the lenses from the LED / detectors.

Depending on the space available one could alternatively look at the scattered light instead of the seed blocking the path. This can be simpler with very small targets, but might need a more light tight case around it. This would be like a well defined light and only one sensitive detector, a little like the smoke detectors or dust particle counters.

[Chriss]

Depending on the space available one could alternatively look at the scattered light instead of the seed blocking the path. This can be simpler with very small targets, but might need a more light tight case around it. This would be like a well defined light and only one sensitive detector, a little like the smoke detectors or dust particle counters.

Hmmm...
This option vent also through my head.
When I saw on original seeder machine how that is made I moved away from that idea.

My question was, why the original seeder machine is not equipped with the idea based of the "dust particle counter" ?
What do you think, what could cos that ?

I think the dusty area is the reason and maybe the available space.

I checked on very expensive machinery too and they are also made with the same photo sensing technique like on the pic I uploaded.

[Kleinstein]

The particle counter way might be more expensive as it needs more optics to make a well limited excitation beam. It could also be more sensitive to dust.  The trade off also depends on the size to detect, the smaller the size the more attractive is it to look at the back scattered light instead of shadow. So if 6 detectors are good enough it likely just the simpler way.

[Marco]

I doubt the sensor from MC Electronics in your pic uses simple on/off gating. I think they differentiate+threshold detection.

If I were to experiment I'd parallel some BP 104-FS/FAS photodiodes, connect them to a transimpedance amplifier, RC differentiator on the output and a comparator for thresholding.

[Chriss]

Hi Marko!
Thanks for the infos, your theory makes also sense.
I will play with several setups and check what is the best option...

thanks for any help and advice.
My best regards.