I am trying to take 4-terminal ground resistance measurements in such a way as to distinguish between shallow and deep anomalies. The basic idea is shown at:
By spacing the electrodes further apart, the depth of penetration is increased. In theory, by comparing the 'short' spacing reading with the 'long' one (proportionately weighted), the differences due to the deeper penetration should be measurable. (There ought not to be a difference if the ground is entirely homogenous, so anomalies will indicate the presence of deeply buried objects.)
In practice the system is made a lot easier if five electrodes are used, with the connections being changed as appropriate:
However, this raises a question: If the current path is the same for both the 'short' and 'long' readings, the sum of the short readings ought to equal the long one as long as the electrodes aren't disturbed when changing over the connections. The current doesn't know where the voltage is going to be measured and the rule of summation of voltages should apply. In practice this does not happen and there is usually a difference.
I have not yet tried exchanging the current and voltage leads to give different short and long current paths with fixed voltage measurement. One theory says this should give a different current penetration and show up deeply buried objects - but the the theory of reprocity suggests that there should be no overall difference between this and the previous set-up.
I'm getting in a tangle with this, is there someone who can see the wood for the trees and explain what is really going on?
[All measurements were made with A.C. at about 850c/s to avoid electrode polarisation. The current was around 60mA with a bridge-type measurement system to balance out variation in the current and a synchronous detector to reject mains hum. Voltmeter input impedance was around 9 Megohms with screened electrode leads using bootstrapped screens to avoid capacitance effects]