Looking at Figure 3, one can see that is
the case. If the two waveforms are added in the scope, the result
should
be nearly zero, hence the name "null experiment." As obvious as this
seems, I have seen many current
probes where the two waveforms would look significantly different due
to electric field coupling through inadequate probe
shielding. Electric field, capacitive, coupling into a current
probe results in an output that would not depend strongly on which
direction the probe is placed on a cable. When this is added to the
probe current response that inverts when the probe is reversed, the
result would be that the
inverted waveform would have a different shape and the sum of the two
waveforms in Figure 3 would not add to zero.
The same effect would hold for
sinusoidal currents.
When measuring sinusoidal currents,
the output of two current probes could be added in a combiner with one
probe reversed on the cable. To the extent the output of the combiner
differs from zero, electric field coupling is likely affecting the
current
probes (assuming the probes are really matched and other links in the
measurement chain are working as intended). A spectrum analyzer can be
used in this case as well as a scope since the combiner provides a
single output. Spectrum analyzers have a much wider dynamic range than
an oscilloscope, so the outcome of the null experiment often is a
result of the combiner's accuracy.
The Folded Wire Null Experiment
A method of verifying current probe operation not requiring a second matched
probe is shown in Figure 4. In this case, the cable is
folded and inserted into the current probe. Doing this exposes the
current probe to the electric field from the cable, but no current
actually flows through the probe so its output should be zero. Figure 5
shows the result of such a measurement.
Figure 4. Folded Wire Null Experiment
Figure 5. Folded Wire Null Experiment Result
The current on the cable that produced the "null experiment" result
of Figure 4 had an amplitude of about one Ampere, so the ~100 mA error
in the measurement was on the order of 10%. The current measurement
itself used the current probe right on the ground plane so the electric field
and resulting error would likely have been less than shown in Figure 4.
The "folded wire" null experiment checks the current probe electrical
field response only. It does not check the probe cable or other
elements of the measurement chain the way using two matched probes
does. For the sum of the two probes (one reversed) to add to zero,
everything in the measurement chain must be working properly: the
current probes, cables, connectors, and oscilloscope vertical
amplifiers.
Being able to check electric field induced error with only the current probe used for the measurement is
important at frequencies over a few hundred megahertz. Above that,
matching of probes becomes increasingly difficult if not impossible.
In my experience, many current probes have significant electric field response.
The Fischer F-33-1 probes used for the measurements in this article are among the best current probes
available for rejection of electric fields.
I have seen other current probes with an error due to electric field effects on the order of 50%. A number of popular
commercial current probes have electric field induced errors on the order of 25% for common measurement configurations.