It sounds fairly complex to me, but as far as soil density is concerned, I would think that boring samples would have to be extracted and tested, or that some known echo technique be employed. Have you studied what the experts in the field are doing?

http://ieeexplore.ieee.org/xpl/logi...ore.ieee.org/xpls/abs_all.jsp?arnumber=406677

Just looking at this site makes my head spin; but doing your homework almost always pays off.

WBahn and MikeML and MrChips are guys whose input always fascinates me.
Experience and logic are guides that cannot be overlooked.
I expect that they are right....

 

All of the following are stated in my thesis topic proposal which was approved:

Idea: Soil moisture content and soil density measuring instrument that is portable.
Concept: The soil moisture content will be measured using a sensor that relies on the principle of soil conductivity. The soil density will just be predicted out of this soil moisture content since a lot of factors affect this soil parameter.
Theory: The moisture content of the soil decides on the resistance of the soil and hence the attenuation on signals traveling through soil or the attenuation the signals may suffer upon their impact on soil.

Design: Employing a bridge amplifier, an integrator, and a comparator, for detecting the moisture content of soil is suggested. This circuit has the advantage of detecting soil resistance with better accuracy and with wide-range linearity when compared with other similar works.

 

All of the following are stated in my thesis topic proposal which was approved:

Idea: Soil moisture content and soil density measuring instrument that is portable.
Concept: The soil moisture content will be measured using a sensor that relies on the principle of soil conductivity. The soil density will just be predicted out of this soil moisture content since a lot of factors affect this soil parameter.
Theory: The moisture content of the soil decides on the resistance of the soil and hence the attenuation on signals traveling through soil or the attenuation the signals may suffer upon their impact on soil.

Design: Employing a bridge amplifier, an integrator, and a comparator, for detecting the moisture content of soil is suggested. This circuit has the advantage of detecting soil resistance with better accuracy and with wide-range linearity when compared with other similar works.

In order to determine soil density you need determine its volume to weight ratio, and I fail to see how that can be accomplished with any degree of practical accuracy (unless you're just looking for a ballpark figure) if you don't extract a sample and measure it's weight and determine its volume. It's conductivity will not tell you this information all by itself, since it depends heavily also on its salts to moisture ratio. Unless there is yet another technique that I haven't heard about that can accomplish this (maybe pumping a sinusoidal wave through it, or maybe its capacitance? maybe even its permeability to sound waves?) I simply don't know how you will be able to extract that information.

 

All of the following are stated in my thesis topic proposal which was approved:

Idea: Soil moisture content and soil density measuring instrument that is portable.

Fine so far.

Concept: The soil moisture content will be measured using a sensor that relies on the principle of soil conductivity.

1) Where is this sensor coming from?
1.1) Is it one that is being designed/built as part of this project?
1.2) Is it off the shelf?

2) What are its specs?
2.1) What is the mapping from soil conductivity to soil moisture content that is being used?
2.2) What is the basis for this mapping?
2.3) What are the assumptions and limitations of this mapping?

Just consider the following thought experiment. Take some very dry, sandy soil and split it into two equal parts. Now take some deionized water and split that into two equal parts. Add some table salt to one of the parts of water. Now combine the soil and water to give you two soil samples that have nearly identical moisture content but one has salty water. Do you really expect these two soil samples to have anywhere near the same conductivity?

The soil density will just be predicted out of this soil moisture content since a lot of factors affect this soil parameter.

3) What is the mapping between soil moisture and soil density that will be used?
4) What possible basis is there for believing that this mapping has any validity whatsoever?
5) What is the basis for believing that all of those other factors can be ignored?

Just consider the following thought experiment. Take some very dry, sandy soil and split it into two equal parts. Now add a small amount of water to one and twice that amount of water to the other. Knowing that the moisture content in one is twice the moister content in the other, can you determine what the density of the soil is?

Theory: The moisture content of the soil decides on the resistance of the soil and hence the attenuation on signals traveling through soil or the attenuation the signals may suffer upon their impact on soil.

Very dubious theory. Prepare a soil sample that has a large amount of brass BBs added to it. Do you expect the resistivity of this soil sample to go down (since adding something to it that has no moisture will reduce the overall moisture content)?

Design: Employing a bridge amplifier, an integrator, and a comparator, for detecting the moisture content of soil is suggested. This circuit has the advantage of detecting soil resistance with better accuracy and with wide-range linearity when compared with other similar works.

6) What is the basis for claiming that a circuit (any circuit?) that uses these components will have better accuracy with a wide-range linearity?
7) Compared to what similar works?

Deal with these issues NOW, not when someone that is attending your defense brings them up!

Both of these seem to be highly dubious claims.
The claim that the soil density is predictable based solely on the mois

 

All of the following are stated in my thesis topic proposal which was approved:

Where did the circuit that was posted as part of this come from?



Can you even identify what the functional role of each of these three opamps is?

 

Where did the circuit that was posted as part of this come from?

View attachment 83075

Can you even identify what the functional role of each of these three opamps is?

I think the first opamp is the brdige amp, the second is an integrator, and the last one is a comparator. I have built and simulate those circuit using Multi-SIM and I plotted the results. The graph is almost linear. I am now in the process of prototyping the circuit in a breadboard to test it. I will be using your suggestions stated above for soil test samples. Thank you so much.

I found the diagram in the web as a PDF. It was a proposal in India, I guess, which was not completed. So I tried it since according to the author, it is okay. The data acquired matched with the circuit operation description.

 

I think the first opamp is the brdige amp, the second is an integrator, and the last one is a comparator. I have built and simulate those circuit using Multi-SIM and I plotted the results. The graph is almost linear. I am now in the process of prototyping the circuit in a breadboard to test it. I will be using your suggestions stated above for soil test samples. Thank you so much.

You plotted the results of what versus what?

 

You plotted the results of what versus what?

You mean what is the 'reference'?

There is an argument that, say dinosaur bones are said to be 4-million-years-old, but what is the reference used to calculate this age?

 

You mean what is the 'reference'?

There is an argument that, say dinosaur bones are said to be 4-million-years-old, but what is the reference used to calculate this age?

No, that is not what I mean.

I mean what is being plotted on the x-axis (the independent variable) and what is being plotted on the y-axis (the dependent variable). The final stage of that circuit is a comparator, yet the claim is that the results are linear. That implies that the final stage is NOT behaving like a comparator (which is possible given that it's output is fed back into the earlier parts of the circuit).

 

the attached figure was the result in my simulation through Multi-SIM. I varied Rx using different values of resistors and I also tested it using a potentiometer. But all of these were in simulation only not in actual.

You mean what is the 'reference'?

There is an argument that, say dinosaur bones are said to be 4-million-years-old, but what is the reference used to calculate this age?

ahh.. yes, I forgot to include the x-y axis parameters. It an the oscilloscope snapshot. The x-axis is the time domain and the y-axis is the amplitude. Th graph shows that the resistances of Rx are converted into frequency. There is almost a constant amplitude.

 

No, that is not what I mean.

I mean what is being plotted on the x-axis (the independent variable) and what is being plotted on the y-axis (the dependent variable). The final stage of that circuit is a comparator, yet the claim is that the results are linear. That implies that the final stage is NOT behaving like a comparator (which is possible given that it's output is fed back into the earlier parts of the circuit).

ahh.. yes, I forgot to include the x-y axis parameters. It an the oscilloscope snapshot. The x-axis is the time domain and the y-axis is the amplitude. Th graph shows that the resistances of Rx are converted into frequency. There is almost a constant amplitude.

 

ahh.. yes, I forgot to include the x-y axis parameters. It an the oscilloscope snapshot. The x-axis is the time domain and the y-axis is the amplitude. Th graph shows that the resistances of Rx are converted into frequency. There is almost a constant amplitude.

According to my source, the output can be fed to the attached circuit to become an input to microcontrollers and the like.

But I cannot simulate both circuits at the same time because an error occurred in the Multi-SIM.

 

So what is it that is it that is "almost linear"?

Linear means: y = f(x) = mx + b

(the whether nonzero b is allowed depends on how what type of linearity you are talking about).

For these results you plotted, what is 'y' and what is 'x'?

For the scope traces you plotted, the period at R=1kΩ is about 27 ms. For R=1.4kΩ it appears to be about 52ms, and for R=2kΩ it appears to be about 77 ms.

So what is it that is proportional to what?

 

So here is an hardware block diagram for an possible digital system.

 

Going back to the original question, it looks like the circuit really is a current or resistance to frequency converter. At the R5-D1 node there should be a square wave that swings from +4.7V to -4.7V. You can eliminate D2 so the negative swing is down to only -0.7V, which might be safe enough for your uC input. If not, then drive an NPN inverter to keep all of the output above ground, feed it to the uC, use one of the internal timers to measure the pulse width, and convert that to whatever kind of output you need.

ak

 

So what is it that is it that is "almost linear"?

Linear means: y = f(x) = mx + b

(the whether nonzero b is allowed depends on how what type of linearity you are talking about).

For these results you plotted, what is 'y' and what is 'x'?

For the scope traces you plotted, the period at R=1kΩ is about 27 ms. For R=1.4kΩ it appears to be about 52ms, and for R=2kΩ it appears to be about 77 ms.

So what is it that is proportional to what?

I will send again the oscilloscope results sir with two square wave connected from Vin of the comparator and the output.

 

I will send again the oscilloscope results sir with two square wave connected from Vin of the comparator and the output.

 

Sending the oscilloscope traces again will not answer my questions.

You said that you "plotted the results" and that they were "almost linear".

Show us THAT plot, along with clearly labeled axes and the best fit linear relationship that is being claimed.

 

Going back to the original question, it looks like the circuit really is a current or resistance to frequency converter. At the R5-D1 node there should be a square wave that swings from +4.7V to -4.7V. You can eliminate D2 so the negative swing is down to only -0.7V, which might be safe enough for your uC input. If not, then drive an NPN inverter to keep all of the output above ground, feed it to the uC, use one of the internal timers to measure the pulse width, and convert that to whatever kind of output you need.

ak

The output of the sensor will be connected to the attached circuit sir.

 

Hello guys! This is an update.
I found this article/study in the web.