Archaeology Resistivity Meter

Interesting ideas so far.  I meant to chime in earlier, but things have been pretty busy for me lately (building a deck and entertaining my Grandchildren again).

Several years ago, I was asked to sit in with some friends of my sister that work at a geotech company.  They were looking to build impedance measurement devices for soil surveys.  I came across this amazing looking chip from Analog Devices that looked like a great way to measure impedance (including a complex component).

Here is a snippet from the specification:

The AD5934 is a high precision impedance converter system solution that combines an on-board frequency generator with a 12-bit, 250 kSPS, analog-to-digital converter (ADC). The frequency generator allows an external complex impedance to be excited with a known frequency. The response signal from the impedance is sampled by the on-board ADC and a discrete Fourier transform (DFT) is processed by an on-board DSP engine. The DFT algorithm returns a real (R) and imaginary (I) data-word at each output frequency.Once calibrated, the magnitude of the impedance and relative phase of the impedance at each frequency point along the sweep is easily calculated using the following two equations:Magnitude = 22IR+Phase = tan−1(I/R) A similar device, available from Analog Devices, Inc., is the AD5933, which is a 2.7 V to 5.5 V, 1 MSPS, 12-bit impedance converter, with an internal temperature sensor, available in a  16-lead SSOP.

This might be of some help in your planning.

Good luck and let me know if you need any help on this project.

Hi Gene,

It's a super-interesting chip, I was keen to use it a few years ago for plant soil purposes, and for hydroponics - to try to see if the soil or liquid has nutrients. The idea being to have a signature of known good soil or water by sweeping through the spectrum. I never got to try it though sadly, the project moved on to something else.

It was felt that it could have had a lot of merit because then you could publish the signature, so others could try to replicate a yield (it wasn't going to be for farmers, more for home use), and to not waste nutrient. But, I have no idea in practice if the result would have been usable, or too inconsistent/variable.

I wish I'd done some work on it at the time, since it could have been useful for other purposes too.

The proposed design so far is one half of the impedance measuring system, but with digital processing. In theory it could be converted to an impedance measuring system with no additional hardware change, just a software upgrade, since the frequency will be know, and there will be some sync pulse from the FPGA, we just need to internally multiply with a 90 degree out of phase signal from that sync pulse too.

I was thinking about generalising the design a bit - basically it's an impedance analyzer but with a restricted bandwidth and much higher than normal resolution.

The chip that Gene mentions (AD5934) is let down by poor resolution (12 bit).

The other big difference is that in the Earth resistance design we force current rather than the more common voltage.

If we aim for 20mA RMS max current and 20VRMS excitation we can drive 400uA into 50k and 20mA into anything below 1k.

The same system can be used for 4 wire resistance measurement, the 24 bit ADC,with gain at max of 64, should have noise and resolution

well below 1uV, allowing us to resolve better than 50uOhms with 20mA excitation (possibly much better).

I spent a little time doodling a +/-15V powered differential current source:

For 20mA 20V RMS operation R3 and R8 need to be 50R, 1VRMS drive needed on input. Op amps are LM7321 in SOIC package (to cope with the power).

We would need to be able to monitor the voltage on the current amplifier outputs  to be sure it isn't limiting when driving into high impedance loads.

NB:

The circuit as simulated would not work with LM7321 due to its high bias and offset currents.

In real life it would be better to use composite op amps for U1 and U2 where a high impedance low offset amp is buffered by a an LM7321 wired as a unity gain buffer.

MK

Hi Michael,

I had difficulty making it work too with a few different op amps, the output would look reduced for half of the waveform, and shifted down. I searched for ideas and found a doc (called "The differential Howland current source with high signal to noise ratio for bioimpedance measurement system", I can the PDF across) that suggests a 3-op-amp method, and that worked with the first op-amp I tried (OPA192). Then, I tried adding the LM7321:

I've not checked in detail, but the output looks good at birds-eye-view (the input VG is 100 Hz, 1Vpeak):

However, OPA192 is only rated for supplies up to +-18V. Also I don't know how bad the phase shift will be (since that will affect the measurement).

Hi Michael,

I had difficulty making it work too with a few different op amps, the output would look reduced for half of the waveform, and shifted down. I searched for ideas and found a doc (called "The differential Howland current source with high signal to noise ratio for bioimpedance measurement system", I can the PDF across) that suggests a 3-op-amp method, and that worked with the first op-amp I tried (OPA192). Then, I tried adding the LM7321:

I've not checked in detail, but the output looks good at birds-eye-view (the input VG is 100 Hz, 1Vpeak):

However, OPA192 is only rated for supplies up to +-18V. Also I don't know how bad the phase shift will be (since that will affect the measurement).

I can't live with that - I have to have symmetry

The problem is that the Howland current pump doesn't constrain the voltage on the load at all when perfectly balanced - and my LTSpice model is unrealistically perfect.

I made a model in TINA (eeek what a revolting interface - I bought it once, a very long time ago - it a has neither changed nor improved.)

This is a DC simulation of my model:

I did similar stuff in the LTSpice and it works there too.

It might need a little more tuning but it's basically OK.

R12 and R13 add a little tiny bit of voltage feedback, so slightly degrade the output impedance, but not by much.

But they do fix the output voltage.

I'll go on thinking about it, to see if I can come up with a better method.

MK

Hi Michael,

Oh, awesome. That in-a-row implementation was bugging me, I was concerned about the phase shift and the degradation. I'll try out this new iteration.

Here's my first stab at the input amplifier. We need differential inputs that can take +/- 15V common mode  and outputs differential drive for the ADC centered on +2.5V.

I can't do it without needing more power supplies - +7.5 and -2.5 for the ADC drive.

Analog have the perfect chip, the LTC6373, does everything we want, but it's not actually released for production (and I've no idea of the price).

This scheme will do, the traces in the simulation are the common mode signal from V8 and the two outputs. This is in x 0.25 mode, for x1 R3 and R4 become 1k,

a relay will be needed to switch, the common mode range drops to about +/- 9V.

If any one wants the LTSpice files for any of this stuff please ask.

V4 and V5 are driving the inputs at 100Hz. V8 is providing a big common mode voltage at 2 Hz.

U1 and U2 need to rail to rail IO, low noise, low bias current, OPA192 will do, there are other cheaper options.

R3,4,5 6 need to be 0.1% or better.

I haven't drawn input protection but it will be needed.

The power supply for this will be fun, +/-15V@30mA, +5V, +3.3V@200mA, +2.5V, + 7.5V, +1.2V, -2.5V.

MK

The low current positive supply values may be a zener plus resisitor?

Hi Michael,

I had a stab at finding a solution to the power supplies.. I wasn't sure what source voltage to target, but I took an assumption of 6V, which could be two LiPo in series (e.g.  two 18650, or better still, a single camera battery, since they are low-cost, for hopefully at least a full day of use).

I spent ages trying to find a transformer (and trying to shoe-horn stuff like audio transformers into the role unsuccessfully), but there's the 6-in-one coils, (all 1:1 ratio to each other) and two of those devices could provide enough windings to cover all the analog supply rails:

Here the orange boxed windings are all one transformer, and the red boxed windings are the second transformer. The voltage regulator parts in the diagram are just placeholders, they don't represent any particular device.

I'm wondering if something like LT3439 could drive them. Its also got a oscillator sync pin which could be useful.

At 200 kHz I think VPH1-1400-RVPH1-1400-R (datasheet PDF) might be a suitable transformer. Regarding 5V, 3.3V and 1.2V, these could be from some DC-DC converter directly from the battery maybe. If you think this is the way to go, I can try to prototype up the analog supplies, since i'm not sure how effective I could simulate this.

I just packed my house to move to Oregon yesterday and all my lab with it. So for the next few weeks I will mostly be working with computer, paper, and pencil only.

michaelkellett - I am going to try to maintain LTspice by myself but will ask for your model if I get in trouble. There are a few things I don’t understand but let me try to work through them first.

Regarding the following - I used to be a project manager. Forgive me

I think this string of comments will potentially become difficult to follow and it would be nice to see it well documented with up to date project description, diagrams, schematics, drawings, design files, etc. in one place. Is that best done on GitHub?  Anyway, I volunteer but request guidance. Do you have an opinion on using GitHub for project documentation other than just code Jan Cumps?  I am seeing this done quite a bit now.

We should update the block diagram from shabaz early on. In my previous life the block diagram was fixed as well as key project objectives early in the project and only changed after much snarling and gnashing of teeth.  I propose the following be clarified now:

* block diagram - key hardware like uC can be pencilled in for now, major things like whether to use a FPGA and architecture need to be fixed early

* cost - 50£, 100£, 200£?  There should be some idea of single board cost.
* component size and packages

* the level of expertise required to build one - this will determine how many of these actually get built and used by DIYers.
* Michael laid out a good summary of performance characteristics
* I also like Michael’s comment that this could a general impedance analyzer. That is the way I would use it. The use cases need to be defined. Any others so as to minimize scope creep?

Regarding the above I am happy to make something up to get it started but others may already know or have better thoughts.

kltm do you have thoughts?  It would be great if other archaeologists found this useful and built one.
Frank

Thanks for ideas.

I like the idea of a standard dual lithium cell - can you point at any camera batteries ?

I'm not thinking that hard about PSU yet, there's still a big Elephant in the shape of the display to resolve and it might use more power than everything else.

However I would normally use some kind of switcher for the +/-15V supplies, LT8471 would do, LT3463 might do but is at the limit of power.

The nice thing about using LT switchers is that they have models and example circuits in LTSpice.

Some of the supplies will use very little current but must be low noise and I was thinking of using a quad op amp to tap down from the +/- 15 V supplies - not staggeringly efficient but low noise and simple.

MK

Good luck with moving !

Totally agree that we will need to be  (a lot) more formal to turn this thread into a thing that gets built.

kltm , hope you don't feel I've/we've muscled in on your scheme, it is well intentioned.

Is resistance surveying allowed under social distancing rules - it ought to be OK (outdoors and in fields)

Do you survey in any particular area or all over - I'm thinking about when it gets to testing this thing I'd like to see what happens.

(Are dogs allowed ?)

MK

Hi Frank

I am in a group on Facebook called Geophysical Prospecting in Archaeology. I will post it on there. Should I add that a GitHub project will be set up yet?

I’m very excited to see the progress so far, but I’m afraid some of the concepts are beyond me. Still great to see. I am sure there are a number of groups that will welcome a DIY update to the current (expensive) kit available.

I did build the original 2003 Resistivity Logger and although it works, the RS232 interface is very hit and miss, so you can not rely on being able to download the data. I’ve emailed you a pic of my efforts.

Ken

Hi Michael

No problem ‘muscling in’ on this. I’m quite excited in seeing the progress so far. Your experiences are streets ahead of mine, so I’m looking forward to learning a lot. I did build the original 2003 Resistivity Logger a few years ago and although it works ok the RS232 interface is very hit and miss, so you can’t reliably download the readings. As backup we would write the readings down as we go, obviously very time consuming. I can send you pics of the unit I made if you’d like to see it.
We do a lot of our geophys here in the U.K. at a local National Trust stately home. It is possible to stay 2 metres apart when surveying. One person would operate the unit and the other would hold and guide the cable to the fixed probes. Two others would move tasked with moving the lines. Oh and yes dogs are allowed. We can easily get permission to survey at anytime as well.
Thank you for time and efforts so far

Ken