Measuring Current in Coax Cable Using Clamp on Current Probe

Question

I once again find myself needing to measure the RF current in a coaxial cable.

The signal in question is a high power (100W-1kW range) signal which is a stepped waveform with a 5MHz fundemental with multiple harmonics (the 7th harmonic is highest of any interest). I have to hand an active oscilloscope current probe of the clamp-on style (specifically Keysight N2893A) which has sufficient bandwidth to do the job. The question is one of building some form of adapter to connect the probe and coax cable.

I did wonder whether the probe could simply be clamped around the coax cable directly, however I'm not sure this will work - AFAIK the magnetic fields of current flowing through the sheild and central conductor should cancel out (?).

In the past I have used an adapter we built which is nothing more than two BNC connectors attached to a metal box (connects the sheilds), and a length of single core wire connecting the signal lines which the clamp can then attach to (pictured below).

Unfortunately I don't currently have access to this adapter, so I am going to need to build a new one.

I can probably cobble together something similar to the previous adapter with bits to hand, but I was wondering if there is any standard or common approach to doing this, or whether the DIY adapter we made previously is the best way to go.

Answer 1

AFAIK the magnetic fields of current flowing through the sheild and central conductor should cancel out

Theoretically and practically they do - the external magnetic field (beyond the shield) produced by the inner conductor current is exactly opposite to the magnetic field outside the shield from the returning current in the shield.

The only magnetic field that exists inside a coax cable is between inner conductor and shield.

This is why coax is so sexy - the outer shield (being a tube) is unaffected by the internal magnetic field therefore, from that standpoint, the shield has zero inductance and, given that the fields cancel outside the shield, the shield can be well and truly regarded as having zero inductance because it isn't subject to any net flux changes that induce voltage in it.

Unfortunately I don't currently have access to this adapter, so I am going to need to build a new one.

That seems the only route given that (a) it works, (b) you've used one before and (c) the theory is sound. That would be my choice. The 7th harmonic of 5 MHz is 35 MHz and the wavelength is therefore around 10 metres so a few inches of extracted inner isn't going to make a whole deal of difference to impedance balancing or reflections.