Seed Counter for a Pumpkin Planter

I have a corn planter that I'm going to use to plant pumpkins from seeds. The planter is capable of doing so but pumpkins are spaced out so far that it is not effective to get out and check the row to be sure it is dropping seed. In corn, it is only 6" between seeds. Pumpkins can be 48".

So, I want to use seed flow monitors to visually be able to see when a seed drops. It has an electric eye that detects the seed passing through the tube. The sensor has 3 wires. Red and Black for Positive and Negative power as well as green for "Signal". I'm not sure if it gets passed a current or if that current gets cut when a seed is detected. I'll test to find out.

My question is: How would you envision using an Arduino to accomplish this. There are only two rows that need to be monitored so I'll have two sensors. They can be powered by 12v on the tractor.

So, essentially, the Arduino could just have the signal wire go to it. From there I could read the signal and turn on an LED when the signal indicates a seed. At least, this is my thought.

Any direction that anybody can give? Thoughts?

Bob

1 Like

Perhaps a better system would be a warning if a seed did NOT drop when expected. Since your planter knows 48", include that in the not-dropped code.

Hello TheRobertRalph
My answer is:
Post your current sketch, well formated, with comments and in so called code tags "</>" and a schematic with seed sensors, not a Fritzy diagram, to see how we can help.

Have a nice day and enjoy coding in C++.
Дайте миру шанс

My first question would be: How do you know you have moved 48"?

Can you measure the speed of the tractor, do you have other ways of measuring this?

1 Like

That measurement is all based on a ground driven set of gears. I'm not worried about the spacing in terms of this project. I am pretty confident the sprockets and chains work as intended. If the seed pickup in the meter misses one every so often, not a big deal. I mostly care if I don't see a see planted for awhile... Either the meter isn't picking up the seeds or the tube is plugged. THAT is what I care about as it relates to what I'm trying to build.

Assuming you can detect a falling seed, the Arduino still needs to know what "a while" is, i.e., the speed of the tractor and the spacing. If the speed is constant, then it should not be too hard to implement.

[edit]
In that case you could measure the time between two falling seeds and when this exceeds a certain threshold, turn on the LED.

Hi @TheRobertRalph
Welcome to the forum.

Have you Googled:

Seed Counter Arduino

Tom... :smiley: :+1: :coffee: :australia:

Hmm, I could restate my question this way: I have a signal wire that is coming out of the an seed monitor that is 12v. I have 12v + and - fed to it and when a seed passes, it send a signal on the signal wire.

I can I read that through A0? What resistors do I need in between that signal wire and the A0 pin to get it to 5v so it can be read and shown as a 0 to 1023 scale?

IF the "signal" is indeed 12 volts, (you have not told us), then you need a VOLTAGE DIVIDER circuit to reduce the voltage of the signal. But first, measure the "signal".

When I measure that signal by putting the Negative from the meter on the Negative and the Positive to Signal, I get 5.50 when the meter is set to 200m in the volt section of the meter (sorry, not sure what that is).

Are you sure that is not "mv" for millivolts Since the signal is most likely a short pulse, you need an oscilloscope the see and measure it.

Here is a quick video of what I'm doing:

Hi,

Can you please post a link to specs/data of the sensor?
Or at least a schematic of it?

Is its output a 0 LOW and 12V HIGH signal or a variable level as the seed goes by?

In the video, place the LED emitters across from the sensor window, as it would be in the application, then pass an opaque object between them.
In the video your are using ambient light and that the result is not very instructive.

Thanks.. Tom... :smiley: :+1: :coffee: :australia:

fellow farmer here, interesting project! Is it a vacuum plate seeder?
I have some thoughts, but I'll directly answer your question first:

as others have said gonna need the specific sensor information.

why is your mm set to 200mVolt range? It looks to me like it peeks @ 6mV (0.006 volts)...thats nothing, something seems wrong there. I'd also expect that voltage to be stable with your finger over it...a little fluctuation, sure, but I dont see any reason that voltage should be fleeting(really need more info on sensor to truly make that call, but its a good guess) You gotta figure out that sensor first. if its outputting properly (i dont think it is), the only way your going pick that up is by tying it to an external interrupt, allowing the signal itself to trigger the arduino to read. not to hard to do, but also probably not an appropriate course of action.

also how you're powering it..maybe its 12v because your giving it 12v, but it will run on say 5v..in that case, if you change the sensor supply to 5v, you -could- run it directly into the arduino..though I wouldn't. The 5v rail is notoriously noisy, better to drop it 3.3 and use the AREF pin to reference 3.3v instead. So you need a voltage divider feeding A0, think of that like a T on a water line with valves on either end. The value of the resistors is like opening/closing those valves to control how much goes each way. You want to ensure you drain enough off, so to not flood your arduino (5v max).

here is a simple calculator to assist you

use the calculator to find a good ratio, give yourself a bit of head space & breadboard the voltage divider by itself. supply it with 12v (or whatever your sensor is) & use your multi meter in the places you'll connect to the arduino(not yet connected) to make sure you've properly dropped the voltage. those pins will handle 1-2v over 5, but 12 will fry them for sure.

some general concerns about your project :

where exactly will the sensor be located? in a tube of some sort? what do the walls of that tube look like? can a seed even be detected by that sensor? inside of the tube might just reflect light around the seed l, the change @ the sensor may be difficult to distinguish from noise. I also worry about the dusty environment this tool needs to function in & how your thresholds for detection will change as dust collects on the sensor.

seems you're using a visible light pv cell, may I ask why? Those are great for turning a light on at dusk, but that spectrum is gonna be flooded with interference out in the field. IR would be a better option IMO, but still not immune to what I outlined above.

am I missing something? why would you expect a pulsed output from this sensor? While there are ways that could be happening, it looks like a simple PV resistor to me...and if he is giving it 12v, he should be getting much more back in at least 1 state. Details are vague on the sensor so far, so maybe I'll have to eat crow...but thats how I see it.

This is the sensor I'm using. There aren't any specs or schematics available on the website. However, the company that makes them is close by and they are friendly. I'm sure I could get something from them.

I'd imagine the signal varies as the object goes by. I placed the LEDs across from the sensor and turned off the lights. As I wave something by it, the reading quickly go up and then come back down. I'll ask the manufacturer for sure.

This is a finger pickup planter. John Deere 7000 and this sensor is made for that seed tube. The mounting holes are there and everything. So, this will work in that application. The problem is that it is a six-row planter and the very old seed monitor will emit a high pitched squeal if a row "isn't planting" for some reason. Well, planting pumpkins, due to spacing of the plants, I am using row two and five and won't be planting the other rows.

So, I'm trying to come up with another method. This is the approach that I'm taking. Instead of the typical ways of monitoring seed drop (or no drop), I just want an LED that blinks when a seed passes the sensor. This way I can have some assurance that it is dropping.

Part of my anxiety beyond typical desires to have seeds plant, is the finger pickup seed meter in the box is new to me. It is designed for pumpkins so it should work but I don't want to leave anything to chance.

This should be fairly simple to code once you can establish what the signal is. If you can just find out what voltage it is, that would be enough to set up some testing with the Arduino and a voltage divider.

Hi,
A picture of the seed counter;


It has an LDR sensor.
Three wire output.

Possibly red = positive supply, black = negative supply, for the LED emitters.
Possibly green= LDR, black = LDR.

If you connect the DMM in OHMs mode to green and black, do you see the resistance change with the amount of light you shield from the LDR window.

Tom... :smiley: :+1: :coffee: :australia:

where in the seed tube would those normally mount? Near the hopper? I ask because , Im not entirely sure those sensors were made to sense individual seeds, but rather let you know when your hopper is empty. I think that because in sever places they mentions the sensor is designed to tell you if a row is being planted to not.

try what TomGeorge said in the previous post. +12v to red, black looks like a common ground with your signal coming through green.

what is the range of those readings (resistance with no power & voltage when given power)? from there we can backtrack what voltage you'd expect to come through the singal line. I still think that should be must more than what your seeing & not just a blip...UNLESS thats not just an LDR & they have utilized a latching ciruit or used shmidt triggers to send a pulse. I couldn't find a datasheet for the sensor, maybe you have one that came with the product? You're gonna have to tinker with that thing until you get a proper output. I would also be doing this testing in the seed tube(on the bench) such that the environmental conditions most closely replicate your final usage.

assuming the max signal voltage is 12v, you could run a divider with r1=1000, r2=350 ohms (380 gives you exactly 3.3v max). pretty much any two resistors of that ratio will do the trick, I would stay out of the megaOhm range though.