For clarity (and perhaps for interest as well): The device I’m intending to build will measure the electrical resistance of the soil using the standard four-electrode arrangement. Two electrodes put a (predetermined or measured) current into the soil and two others measure the potential difference this produces to calculate an “apparent resistance” value using Ohms law and a geometric factor to deal with the electrode positions. The concept of electrical resistance in soil is not simple - there’s a degree of capacitance, as you say, and (a little) inductance from a variety of physical-chemical effects the spatial distribution of which is immensely complex and varied.
So when I say I want fast components I’m aware that the capacitance at the electrodes will slow things down - and I want fast components elsewhere so that when I measure the time it take for the voltage to rise in the soil (so I can explore the soil physical chemistry which determines this “soil capacitance” rise-time) I can discount the rise time of the other parts of the circuit, as far as possible. Typical rise-times for soil-electrode systems like this are a few mS, depending very much on the soil and the electrode configuration, so it would be helpful if the rest of the components do their job in nS to a mS or so. It’s not critical. For the simple purpose of measuring the DC soil electrical resistance I can manage without knowing the rise-time of the soil electrode capacitance but it might be helpful if I can make it possible to measure the complex impedance of the soil (which determines the rise-time) by design from the beginning (as some of the more sophisticated commercial soil electrical resistance instrument do).
I started out my thinking with the simple brute-force approach using a 40v supply into two electrodes in the soil and a shunt and an ADC to measure the current this produces, as well as the potential difference created by this 40v through the soil nearby. I then realised that it might be easier if I controlled the current instead so I didn’t need to measure it. But it can’t be a simple constant DC current because of polarisation at the soil-electrode contact. It needs to be switching polarity every so often so my intention has been to create a constant current and simply switch its polarity back and forth - and that’s what the relay-based version achieved. Geological ER systems use frequencies as low as 1Hz or less because of the time it takes for a stable current to get established in the ground over distances and depths of kilometres (and they use reed relays to do so to handle the much higher voltages they use). Typical instruments to measure building earthing (such as the various Megger earth resistance instruments) work by periodically reversing a constant current, thus creating as a square-wave at frequencies in the 50-150 Hz range. And that’s what I’m trying to achieve.
As a dumb soil physicist I have no reason to know the various parameters and capabilities of electronic components so this is a very interesting journey - and the advice I’m finding here, and the reading I’m being prompted to undertake, is much appreciated.
As I understand your last remarks, however, I won’t be able in any simple way to achieve the square, constant current (constant at each half-cycle, obviously) waveform I’d intended using a Mosfet H-bridge and I’d have to go back to my clumsy solution using solid state relays.
