Seed monitoring

[hogar]

Can someone help me to make a seed monitoring system for my seeding machine, something like this:
http://www.electrolee.co.za/PM8E.htm or http://www.roninpfs.com/images/documents/Planter_Monitors_PM100_Info_Brochure.pdf


with the optical sensors resistant to daylight which register the seeds fall from the machine in the soil
it is important to me next:
 LED indicator lights shows every kernel as it falls into the ground.
 LED's indicates an alarm condition at that specific row


I hope that somebody help me

Best Regards

[DarkClover]

I don't know what exactly a seed monitoring system does but maybe you can explain
it a little bit particularly ?
The devices found under your links don't seem to be very difficult to build. The most
work will be the recognition of the seeds and therefore the selection of proper sensor system.
How does your seeding machine look like ?

This seem to be a very interesting project. I hope many other people will join this topic.


Greets DarkClover

[hogar]

I put pictures of my corn pneumatic drills, the role of monitors is to observe the falling seeds into the soil and count the seed of corn
I hope that now is a little easier to understand the role of monitor

[hogar]

The sensor would be placed in the circled area but will be exposed to daylight and dust please help to to solve this problem the main purpose of the monitor is to oversee the proper functioning of the whole machine because if the seeds fall correctly means that the machine working correctly I tried the IR transistor and diode as sensor but without any result .. I will put circuit soon, so you correct my mistakes

[DarkClover]

Okay so far...
For a position like this (exposed to daylight, dust, dirt and so on) I have two ideas which should be cheap and easy to implement.
First is a IR light barrier with daylight blocker and/or polarized light recognition. This makes sure that no other light will
interfere with the sensor. The only problem could be the dusty environment. My second idea would be a ultrasonic solution.
Such sensors are insensitive to any light and sorts of dust but they don't like humidity, if there is any humidity at that place.

I searched on the internet for sensors and found out that 90% of them are based on the light barrier principle. The more
expensive versions (10%) have multiple light barriers or ultrasonic.

So it's up to you to chose a system. The next step is to pick a proper sensor.

[hogar]

I guess that would be the best solution was to IR with multiple light barriers do not know how to resolve the the influence of daylight, physically or electronically.
Maybe they can come to a complete device and copy the circuit but is based on PIC16F628A microcontroller, but unfortunately no code if it would help to complete the project

[DarkClover]

Well the daylight blocking maybe can be done with a foil out of the same material
the plastic masking of IR-remotes are. It's kind of red-violet color and I know that
it is to block interference with other light sources.

Also I've found a cheap method to make a test daylight blocking filter.
Just use the magnetic part of a 3.5" floppy disc to filter the daylight parts.
Take a look at this: http://labs.sypherus.com/cms/Coffee%20Table/Day%203

For coding I can't really help because my platforms are AVR and since two weeks
the ARM. I only can help with some pseudo code.

[Ichan]

Hi Hogar,

How big is the diameter of the tube? And what is the smallest and largest size of the seed?

This should be interesting.

-ichan

[LabVIEWguru]

Do they fall fast enough and enough of them to use a small plastic paddlewheel (like a water wheel) attached to an encoder? (very cheap, very efficient, breaks easily if not careful)

How about a webcam in a cylinder with an LED strobe light? I remember seeing some software recently that detected movement from a webcam. (can not break, but maybe dust and contamination is a problem)

If you want to be really complicated, (maybe I should work and try to sell these things!) Analog Devices sells a "Capacity - to - digital" d/a. You could have two screen wires separated in a cylinder. As the seeds fall through, it changes the capacity. I know there is water in the seeds because they measure the moisture content of the seeds when they bring them to market. (I grew up in Ohio) Dust may be a problem for the capacity system.

If the inside of a cylinder is painted black, does enough light reflect from the seeds to give an indication on a photosensor?
(probably how the commercial version works) Like Dark Clover suggested, I saw a science television show that flowers and seeds look very different to birds and insects because of IR or UV light. Perhaps some member knows about this.

Just thinking about your problem.

[hogar]

I have hardver of unit based on
  microcontroller can set the scheme but do not know
  to program microcontroller, the device is in operation with IR transistor and
  diode

[Magnox]

The AD capacitance-ADC's are interesting devices - I was looking at those the other day and thinking up a multitude of potential uses. They are remarkably sensitive to minute capacitance changes - femtofarads! This might be a good application of them.

The catch might be the slow-ish sampling rate - the fastest are 200 samples per second so a long 'capacitance tube' might be needed if the seeds are travelling quickly.

Some commercial units use an electromagnetic sensor, again taking advantage of the seeds water content compared to dry dust particles. That might depend on how much iron oxide is in the dust though.

[Ichan]

Problems with the use of IR light on this case will be the variance of ambient light and environment condition like dust etc.

The idea is to use the IR sensor in analog way, feed to the ADC of the micros - while the output level of the IR source can be adjusted by DAC or PWM. Other IR sensor for environmental reference point might be needed.

-ichan