I would like to design an earth resistivity meter for use in archaeological surveying. This is essentially a current source placed in the ground and a ohmmeter to measure the resistance between the probes. This tells us something of the composition and moisture content of the soil, and whether there might be archaeological features beneath the topsoil.
An example of a commercial "res meter" is the Geoscan RM15-D, and John Becker wrote an article in Everyday Practical Electronics about 20 years ago with a design for one. The drawback with that one is it uses +/-5v on the probes which limits its usefulness in dry or very rocky soils. The Geoscan machine can output at either 40v or 100v (peak-to-peak I think) which makes it
a more versatile device. An AC-coupled waveform is needed to reduce the capacitive effects of the soil.
So what I would like to design is a circuit that gets me somewhere near the specification of the RM15-D:
- up to 100v p-p AC at <150Hz at a user-selectable constant current of 0.1mA to 10mA
- powered by a rechargeable LiPo/Li-ion battery
- be as lightweight as possible because frail archaeologists need to lug this thing around bramble patches all day
The circuit in the article uses a simple pulse train as its output waveform, and while the actual waveform shape doesn't appear to matter a great deal (as long as it is AC coupled), the potential for very high transient voltages seems to me undesirable. Assume for the purposes of this that I can generate a low voltage sine wave to be fed into the circuit. Presumably some amount of isolation will be needed as well.
What would be the recommended way of achieving this? Would using something as simple as a transformer be sufficient, or perhaps something like a DRV2700 as has been mentioned in this SE question?
Best Answer
This is a big project and cannot be answered in detail in a simple way. The crux of your question/problem seems to be the generation of the 40 - 100V 'AC' with a variable/preset 'constant current' so that's what I've concentrated on. There are other solutions but this is how I would approach it.
Taking an overview of the project gave me a block diagram like this.
I've expanded the blocks for the 'AC' generation. By generating a DC high voltage first and then converting it to AC using an H bridge it allows the output frequency to the probes be easily changed. The output is a squarewave. The probes need an AC signal to prevent polarisation but the waveform is not important.
Rather than trying to produce a constant current I went for the same approach as described in the EPE project - using switched current limiting resistors. This produces a very simple and easily constructed (bi-directional) circuit.
The high voltage (shown as 100V but could be 40V if needed) is generated using any suitable dc-dc converter (such as an LT8331). By adding an H bridge circuit, supplied by the high voltage, the bridge outputs can be continuously reversed by a bi-phase clock/driver circuit supplying current to the probes. (a more common use is to reverse current through a motor) There are lots of examples of this type of circuit on the net.
This H bridge could be driven by a simple 'clock' type oscillator (such as a 555 astable) or from a microcontroller output (with suitable interfaces to protect against the high voltage).
Looking at the rest of the circuit I think you'll also need to include a GPS module (so you know where the reading was taken), some form of data logging system (to record the data), an interface to an external computer if the data isn't portable (i.e. on a memory stick or card), a microcontroller to handle the data flow,timing readings, key input, display etc., some form of voltage measurement system that can be converted to digital data for the data logger, a keyboard or set of keys to operate different functions and some form of display to let you know what's happening. As I said - this is a big project - good luck!