VSWR ripple

Having seen some recent discussion where the online experts opined that an example given of a VSWR plot that contained a fairly consistent ripple was quite normal, this article suggests there is an obvious possible explanation and that to treat it as quite normal may be to ignore the information presented.

Above is a partial simulation of a scenario using Rigexpert's Antscope. It starts with an actual measurement of a Diamond X-50N around 146MHz with the actual feed line de-embedded. Then a 100m lossless feed line of VF=0.66 is simulated to produce the plot that contains a ripple apparently superimposed on an expected V shaped VSWR curve.

This is the type of ripple that the expert's opine is quite normal.

Let's apply some thinking to this.

The ripple is periodic with frequency, the pattern repeats about every 1MHz around 146MHz, or about 0.68% of frequency, or 1/146 of frequency.

The question that should be asked is, is this periodicity related to the electrical length of the feed line? Recall that VSWR patterns repeat every half wavelength along a transmission line, so we are interested in the feed line length in half wavelengths.

The feed line is 73.8λ at 146MHz, or 147.6 half wavelengths. This means that with a frequency change of 1/147.6, we will increase the electrical length by one half wavelength.

In this case, the frequency periodicity of the plotted VSWR curve is approximately the same as the frequency periodicity of the electrical length of the line, and is a likely explanation of the VSWR curve if the reference impedance for the VSWR plot is significantly different to the characteristic impedance of the transmission line section.

Since Zref of the instrument is 50Ω we might properly suspect that the VSWR plot ripple is caused by depart of the transmission line from 50Ω and using Antscope, explore other Zrefs to minimise the ripple.

Above is a plot with Zref=45Ω, it is more like what might be expected, and it suggests that the transmission line has Zo=45Ω and that is the cause of the ripple.

If the transmission line was specified to be 50Ω, analysis of the VSWR ripple suggests the transmission line does not meet specification and warrants a more conclusive test by placing a known frequency independent load at the far end and measuring VSWR, observing ripple and finding Zref that minimises the ripple.

Of course, the VSWR ripple may be caused because Zref of the instrument is not correct, so using an instrument that is incorrectly calibrated might produce similar ripple.

Above is a simulation where the 100m transmission line Zo=50Ω and Zref=45Ω. Check that instrument Zref properly suits the transmission line before making measurement.