Hello, so I've been working on this project for a precision planter in which I need to know whether there is seed passing through a certain point on a tube or not, and the program should keep count of the seeds passing through.
I've been thinking of capactive sensors but that would be very inacurate since seeds will be passing at a high rate (affected only by gravity), and they will be very close together. So a capacitive would be always detecting.
Laser beam detection is the other option, but since it is only one beam it wont be able to count doubles or smaller seeds like canola, or the seeds passing asside the beam.
I've also ruled out piezoelectric sensors since I cannot disturbe the motion of the seeds by making them bounce, because that would result on bad seed spacing down the row.
Hope someone could help me out here..
Thanks
Lasers are used in the movies because they look cool. The real world uses IR LEDs which aren't lasers and aren't visible to the heroic thieves.
You might get somewhere with a QRE113 sensor. That has the LED and sensor in the same package. I suspect that you are going to end up buying lots of sensors and testing them. One source of sensors is to tear apart an old laser printer or copier. The sensors which detect the paper are forked so the paper travels between the emitter and receiver. They are available in thousands of different sizes and configurations.
If the seeds are too close together that two might appear as one long one then you need to make them move faster to expand the gap between them. Or maybe use a number of sensors to gather more information on the tumbling seeds.
Please post photos of your setup, showing the channel the seeds pass through.
How controlled is this passage? Small seeds are indeed hard to detect optically as even if passing right through the beam they don't block it completely.
Piezo bounce is the most reliable counting in this situation - any way you can do the count first, and the lining up later? Or bounce them in a very controlled manner? A bounce doesn't necessarily make for big changes in drop time.
MorganS:
Lasers are used in the movies because they look cool. The real world uses IR LEDs which aren't lasers and aren't visible to the heroic thieves.
You might get somewhere with a QRE113 sensor. That has the LED and sensor in the same package. I suspect that you are going to end up buying lots of sensors and testing them. One source of sensors is to tear apart an old laser printer or copier. The sensors which detect the paper are forked so the paper travels between the emitter and receiver. They are available in thousands of different sizes and configurations.
If the seeds are too close together that two might appear as one long one then you need to make them move faster to expand the gap between them. Or maybe use a number of sensors to gather more information on the tumbling seeds.
Thanks for your response.
What do you think about using a camera?
is there any way to program arduino so that it identifies the seeds falling in the image and also its speed and direction?
Im new in this arduino world, so I have no idea on how I would be able to accomplish that.
With cameras, which may or may not work depending on how fast the camera itself is, you need much (MUCH) faster processing hardware than an Arduino. Maybe, just maybe, something like the 1.4 GHz RPi can count seeds with a high speed camera. It'll indeed have to track seeds (so measuring speed and direction becomes trivial) as they fall through the view of the camera to prevent double counting.