Soil optical sensor for measuring NPK nutrients (SOLVED)

Hey guys, did anyone here try to build optical sensor based on this research?Detection of NPK nutirents in the soil using optical transducer

I want to construct one, but have doubts after looking at schematics. The light detection part (LDR sensor) used for reading & detecting values from RGB LED seems to be enough to attain measured nutrient level from the soil sample.

However, after watching this video I can't decide which one to try and construct.
I need to have this sensor implemented as crucial component of my system. Please give me some advice or guidelines, I'm currently in a dead end :fearful:

Screenshot_1.jpg

Screenshot_2.jpg

have doubts after looking at schematics

Good eye! Theory and experimental verification is absent as well.

https://forum.arduino.cc/index.php?topic=494209.0

Seems only way to do it is by using color sensor that I linked in the video on the first post.

Actually, the only way for a hobbyist to measure NPK content is to pay for proper soil analysis.

Good summary of the difficulties in reply #3 here: Nitrate , Phosphate, BOD , TOC and ECOLI sensors compatible with UNO - Project Guidance - Arduino Forum

hey yall, please check this paper

it has the same method described in the video using TCS3200 color sensor

It is indeed amazing what one can get published, these days.

Yeah, sadly more interested on the money they get to publish the papers.

If it's feasible to identify the nutrients from the colour of a solution made with some stuff that comes with a kit, strikes me that an Arduino and IoT is a bit of overkill.

Seems like it's hardly more than a homeowner pool test kit.

Even with 3 things to measure, and I get that it's the combination of the 3 that informs the decision to add or to the soil, it should be feasible to compare the colours to the charts, and add or according to the instructions on the back.

Testing the color of the solution does not quantify the N, P, K values and automation is not possible since its a manual procedure.

But I do agree its an overkill.

I guess this is the best option we have as of now.

I'd rather opt for soil analysis to get better results and save my time.

lindsayBoxer:
If it's feasible to identify the nutrients from the colour of a solution made with some stuff that comes with a kit, strikes me that an Arduino and IoT is a bit of overkill.

Seems like it's hardly more than a homeowner pool test kit.

Even with 3 things to measure, and I get that it's the combination of the 3 that informs the decision to add or to the soil, it should be feasible to compare the colours to the charts, and add or according to the instructions on the back.

I've sent bradleybradms this already, since you mentioned overkill doing IoT analysis take a look at this

I am skeptical on the quantification aspect of that sensor since it is color based and on the color comparator chamber it shows 5 gradients of colors.

As you can see in the image of the color chamber there are 5 levels: Surplus, Sufficient, Adequate, Deficient and Depleted.

Did you get a chance to try it out?

Waiting for your results

The only way to know if the kit works at all is to compare kit results with those obtained by sending a sample of the same soil to a reliable testing agency.

Keep in mind that the kit tests for only certain ions, like NO3-, whereas agencies can test for total oxidized nitrogen.

If you want only category information like Low / Medium / High for NPK, then there is this sensor but it cannot be interfaced with arduino

Also, it does not quantify in units of milligram per liter (mg/l) or milligrams per kilogram (mg/kg).

For quantification,

I found this sensor which can be interfaced with arduino using a USB cable and quantifies all N & P & K contents in mg/kg (mg/l)

OP requested deletion of the topic as the question has been answered

Leaving it in place for future reference

OP - please edit the title of your original post and add [SOLVED] to it

The eBay ad sure looks like a scam to me. This sentence is a dead giveaway: "NPK is an important element in the soil".

There is absolutely no way that three conductive rods can be used to individually quantify nitrogen, phosphorous and potassium content in soil. US$250 completely wasted.

ok thanks