AC current source ? any idea for soil resistivity purpose

[Overspeed]

Hello

As I want to make some trials of soil resistivity measurement during summer time

I have a square wave inverter circuit with a frequency around 130 Hz ( can be adjusted by software )  with an output voltage of 70 Volt ( I can change the transformer ) but 70V look enough

I need a circuit able to generate a courant source of 10 mA ou 20 mA , I have no knowledge in AC current source , does someone can provide an idea of circuit ?
OR
Perhaps it s possible to simply limit the current at the input of the inverter ? as that a DC current ?

Thanks in advance
OS

[moffy]

The diodes have been chosen for convenience, but this circuit should work, it turns a DC current limit into an AC one, just make sure the current limit can drop the full voltage safely.

[Overspeed]

Hello

Thanks but it s seem that change AC to DC , I want to keep AC current to avoid soil electrical problem .

I link a schematic of the 2 solutions I plan and two possible circuit to control the current at the DC level

let me know if this circuit can work without transformer problem in the inverter

Regards
OS

[tszaboo]

That's a lot of electronics. You can just capacitive couple a PWM signal. It's not going to be sine wave, but who cares? The chemical process doesn't.

[moffy]

Hello

Thanks but it s seem that change AC to DC , I want to keep AC current to avoid soil electrical problem .

Regards
OS

If you look at the trace you have AC in, then current limited AC out, that is the green trace. The four diodes just switch the direction of the current source in sync with the AC input.

[Sensorcat]

If you want to measure the electric properties of soil, there's no reason to have 70V between the electrodes used. A few volts are absolutely sufficient. Perhaps you came to the conclusion that you need 70V because you have a large electrode distance in mind? Better have them only about 1 cm apart, because that way, you can keep the electrodes small, have a moderate, easy to measure impedance, and have full control over the soil sample you are measuring, plus some confidence that this sample is quite homogeneous.

I wrote impedance, not resistance, because it is preferred to measure with pure AC, no DC component, to avoid corrosion and a flow of ions that, on the long run, changes the soil sample. Doing so, you can conveniently model the soil as a parallel resistor and capacitor. By measuring amplitude and phase, both components can be identified. The electrodes can even be covered with an insulator if this is carried out properly.

[Overspeed]

Hello

Thanks but it s seem that change AC to DC , I want to keep AC current to avoid soil electrical problem .

Regards
OS

If you look at the trace you have AC in, then current limited AC out, that is the green trace. The four diodes just switch the direction of the current source in sync with the AC input.

Hello
Ok as that a simple circuit I will test it , no fancy IC

Regards
OS

[Overspeed]

That's a lot of electronics. You can just capacitive couple a PWM signal. It's not going to be sine wave, but who cares? The chemical process doesn't.

Hello

On all the soil resistivity tester I have checked they use AC , some with low voltage but most of them with quite high voltage , I use PWM with a non symmetrical duty for other resistivity tests but for soil I have always read that better and safer to use AC as AC cancel the soil charge

the circuit posted are not complicated and can easy allow to set a current value

Regards
OS

[Overspeed]

If you want to measure the electric properties of soil, there's no reason to have 70V between the electrodes used. A few volts are absolutely sufficient. Perhaps you came to the conclusion that you need 70V because you have a large electrode distance in mind? Better have them only about 1 cm apart, because that way, you can keep the electrodes small, have a moderate, easy to measure impedance, and have full control over the soil sample you are measuring, plus some confidence that this sample is quite homogeneous.

I wrote impedance, not resistance, because it is preferred to measure with pure AC, no DC component, to avoid corrosion and a flow of ions that, on the long run, changes the soil sample. Doing so, you can conveniently model the soil as a parallel resistor and capacitor. By measuring amplitude and phase, both components can be identified. The electrodes can even be covered with an insulator if this is carried out properly.

Hello
Yes that long distance ( 1 meter pitch ) purpose is to locate foundation of a old building in a field , my inverter can modified easily to reduce voltage in case of need .

My first goal is to built a working system '' quick and dirty '' without the need to run complex design and improve it after I have a batch of current sensor with various amp op I can use for free , so just need to control the current

I agree with impedance / resistance as that AC and not DC , i will measure variation in one batch of measurement and not really absolute values as the soil humidity is not a constant .

Last generation equipment allow to measure induced polarisation too .

Regards
OS

[Sensorcat]

... purpose is to locate foundation of a old building in a field , ...

In that case, ground-penetrating radar is what you need.

[Overspeed]

... purpose is to locate foundation of a old building in a field , ...

In that case, ground-penetrating radar is what you need.

Hello
Not really electrical resistivity is used by a lot of companies for various purposes ( archaeology , mineral search . water search ) that a powerful mean  , not spectacular as GPR but in most case interesting and powerful as that easier to go deeper or to cover very large surface

Technologies allow now tomography and induced polarization but also inversion software
https://resipy.org/gallery/nb_02dc-topo.html

That also far more cheaper than a GPR

Regards
OS

[Kleinstein]

For safety reasons I would keep the voltage / current a little lower. E.g. less than 40 V / 5 mA.
One may not need a true current source. A simple bridge ciruit with voltage source and resistor could be enough. The nonlinear response should be easy to correct.

For looking into the depth the soil resistivity makes absolute sense, especially with more electrodes to measure the portential.

One may consider to use more than 130 Hz to get a bit more separation from 50 Hz main to allow easier filtering, though a 50 Hz notch should aready work well the 130 Hz.

It looks like an interesting project.

[Overspeed]

For safety reasons I would keep the voltage / current a little lower. E.g. less than 40 V / 5 mA.
One may not need a true current source. A simple bridge ciruit with voltage source and resistor could be enough. The nonlinear response should be easy to correct.

For looking into the depth the soil resistivity makes absolute sense, especially with more electrodes to measure the portential.

One may consider to use more than 130 Hz to get a bit more separation from 50 Hz main to allow easier filtering, though a 50 Hz notch should aready work well the 130 Hz.

It looks like an interesting project.

Hello
2 pins resistivity use high current but close distance , 4 pins (wires ) soil resistivity use quite high voltage and low current , the rule is no body is in the measuring field and all equipment are battery powered so floating to the earth .

I have a 36V DC source I can link it to my pulsed mosfet generator ( not inverter ) and make some tests of resistance , main problem with soil is the fact that keep the electrical charge so pure DC is not the best to make measurement and more measurement in a chain / string of spike

I link a doc with most common resistivity value

Regards
OS

[moffy]

I designed a moisture meter several years ago and AC was preferred because of battery type EMFs being generated by the electrode in the soil, it also helps to reduce erosion of the electrodes. But I just used an AC coupled 5V square wave and synchronous demodulation for detection.

[Overspeed]

I designed a moisture meter several years ago and AC was preferred because of battery type EMFs being generated by the electrode in the soil, it also helps to reduce erosion of the electrodes. But I just used an AC coupled 5V square wave and synchronous demodulation for detection.

Hello
I agree on electrode erosion saving but do you have a circuit for the synchronous demodulation ? here I use a 4 wire configauration so that will fit after voltage amplifier by a op amp ? or you connect straight on the output voltage ?

Regards
OS

[moffy]

The following reference shows a single op amp synchronous demodulator: https://electronics.stackexchange.com/questions/437940/synchronous-demodulation-with-single-supply-op-amps

[Kleinstein]

If the interfering signal is not very large one can do the synchronous detection in the digital domain. So read the AC signal fast enough with an ADC of sufficient resolution (SD-ADC chips like MCP356x got relatively cheap) and than do the math.

[jbb]

Looking for old structures is an interesting case. I have previously used soil resistivity data as inputs into electricity substation earthing grid design.

What will your deployment look like?

Are you planning to attach four electrodes to a fixed-length stick? (This is very convenient to carry around, but won’t penetrate very far into the ground because electrodes are close together.)

Or are you planning to have a field box to which you wire up some test electrodes? (This lets you put the electrodes as far apart as you like, which means you can ‘look’ deeper into the ground.)

How precise do you want to make things? A really consistent option for making a controlled AC test current would be an Enhanced Howland Current Pump. This is an opamp-based circuit which converts a reference voltage into a controlled output current (and yes, it can do AC).

(Note, may also be called an Improved Howland Current Source or some mixture there of.)

References:
https://www.ti.com/lit/an/sboa437a/sboa437a.pdf?ts=1722047338395&ref_url=https%253A%252F%252Fwww.google.com%252F

https://www.analog.com/en/resources/analog-dialogue/articles/a-large-current-source-with-high-accuracy-and-fast-settling.html

[Overspeed]

If the interfering signal is not very large one can do the synchronous detection in the digital domain. So read the AC signal fast enough with an ADC of sufficient resolution (SD-ADC chips like MCP356x got relatively cheap) and than do the math.

Hello
Yes that can a solution ''easy'' and flexible too

Regards
OS

[Overspeed]

Looking for old structures is an interesting case. I have previously used soil resistivity data as inputs into electricity substation earthing grid design.

What will your deployment look like?

Are you planning to attach four electrodes to a fixed-length stick? (This is very convenient to carry around, but won’t penetrate very far into the ground because electrodes are close together.)

Or are you planning to have a field box to which you wire up some test electrodes? (This lets you put the electrodes as far apart as you like, which means you can ‘look’ deeper into the ground.)

How precise do you want to make things? A really consistent option for making a controlled AC test current would be an Enhanced Howland Current Pump. This is an opamp-based circuit which converts a reference voltage into a controlled output current (and yes, it can do AC).

(Note, may also be called an Improved Howland Current Source or some mixture there of.)

References:
https://www.ti.com/lit/an/sboa437a/sboa437a.pdf?ts=1722047338395&ref_url=https%253A%252F%252Fwww.google.com%252F

https://www.analog.com/en/resources/analog-dialogue/articles/a-large-current-source-with-high-accuracy-and-fast-settling.html

Hello

To calibrate and make repeatability tests  I will use the Wenner method which is a 4 wires setup same distance between contact ratio is distance / deepness < 0.1 so 1meter spacing and 100 mm deepness , I can use metallic rods with a first setup of 4 and increase to 8 and 12 and change Current / voltage wiring connection make in line measurement as the purpose is to locate buried wall / foundation. 
 
Some new methods goes with liquid coupling with plates ( water jelly under metal plate ) to avoid mechanic penetration and problem with the law ( local or gov law ) on archaeology sites searches which forbid any king of excavation / ground penetration / drilling

Amp Op current source are quite limited in current and voltage level

I link an interesting doc on a conductivity cell base on AC current they use synchronous detection for voltage measuring

Regards
OS

[Overspeed]

Hello

Some useful data on current and voltage used for soil resistivity , value fron IRIS Instrument a professional soil resistivity equipment

Voltage are quite high

Regards
OS

[jbb]

That table lists some … worryingly high voltage and current levels. Bluntly: that could kill someone so adult supervision required.

However, the caption mentioned an array of 48 elements at 5m (presumably meter) spacing. That suggests a total length of 240 meters.

Overspeed: you mentioned a 1m electrode spacing, and may therefore need much lower drive voltages. Do you have the facilities to simulate a measurement scenario (ie model the soil, drive electrodes and sense electrodes)?

If the performance is satisfactory, the device should operate at Safety Extra Low Voltage (SELV) levels. (I think these are 42 V AC RMS / 60 V DC, but I’m not a compliance expert.) That way you could even let undergraduate students use it.

Edit: fix autocorrect

[Overspeed]

That table lists some … worryingly high voltage and current levels. Bluntly: that could kill someone so adult supervision required.

However, the caption mentioned an array of 48 elements at 5m (presumably meter) spacing. That suggests a total length of 240 meters.

Overspeed: you mentioned a 1m electrode spacing, and may therefore need much lower drive voltages. Do you have the facilities to simulate a measurement scenario (ie model the soil, drive electrodes and sense electrodes)?

If the performance is satisfactory, the device should operate at Safety Extra Low Voltage (SELV) levels. (I think these are 42 V AC RMS / 60 V DC, but I’m not a compliance expert.) That way you could even let undergraduate students use it.


Edit: fix autocorrect

Hello

Yes that extremely high voltage , during testing nobody is allowed in the test field same for seismography testing , I just search wall in an area of 100 x 300 meter so I can use 1 meter pitch with a line of 12 meters and move the contacts .... to cover all the range ( 12x12 m )

High voltage allow deepness over 30 meter which is not the goal / purpose of my trials , some wheeled system . portable system use low voltage level , take in consideration than procedure ( Wenner / bi-pole / Schulmberger ) can cover big surface put pitch stau in 10/12 meters I think High voltage is a solution to solve variation of soil from dry to wet , stone to sand or clay ... and also the current generated by the voltage injection ( 4 wires )


Regards
OS

[jbb]

I couldn't help myself... so I knocked up an opamp circuit.

For safety reasons I went for a 48V DC supply -- because 48V * 120% (heavy overvoltage) = 57.6V, which is under the 60V threshold.  To maximise output voltage, the design is a full-bridge driver.

U1, U2 and U3 form a Howland Current Source -- they convert a reference voltage at ground potential into a controlled current. Note U1 and U2 are a composite amplifier; U1 does the precision analog work and U2 does the power handling.  The output of U2 is current limited by design, which is nice.  However, it may need to dissipate quite a bit of heat...

U4 forms the other half of the full bridge. Unfortunately, it's not current limited - if the node 'Vb' shorts to ground somehow then a lot of current could flow.  I'd recommend getting a power opamp with output current limiting and over temperature protection.

The load RL is connected between the two sides of the bridge.  As designed, the average voltage of Va and Vb is set to half the supply rail (i.e. 24V) with very little AC content.  To put it another way; the AC common mode voltage is nice and low.

Note: I'm not up to date on safety specifications, so saying 'it's a 48V supply' may not cut it in terms of user safety.

[Overspeed]

Hello

Thanks , interesting

'' if the node 'Vb' shorts to ground somehow then a lot of current could flow.''

From my point of view current control is a mandatory parameter as following the soil resistivity that possible to be close to short

Regards
OS