A recent online posting provides content for learning. K3EUI posted a NanoVNA-Saver screenshot of his antenna described as:
Set out a horizontal loop wire antenna for possible NVIS paths
Wire is about 140 ft length with an outside CLC tuner, fed with 50 ft RG213. …Here were the Nano VNA graphs of this new loop antenna, measured from inside the shack (50 ft RG213)…
Can we learn something from this?
So, noting particularly that these measurements were from the source end of the coax, no mention of preservation of the common mode current path (a common oversight), and that there is an ATU at the load end of the coax, the following analysis is offered.
Since there is an ATU at the load end of a length of 50Ω coax, it would seem logical that the purpose of the ATU is to achieve low VSWR on that coax, and the optimisation objective therefore is VSWR close to unity at the desired operating frequency.
The charts
The first thing to note is that there is no discussion of preservation of the common mode current path as would be seen by a transmitter, so the measurements discussed here may be of something different to that seen by a transmitter. If a poster doesn’t mention preserving the common mode current path, they probably didn’t do it, and probably don’t understand the relevance.
There are six charts, all derived from a swept s11 measurement so they are different and possibly partial presentations of s11.
Since the sensible objective should be low VSWR, the first plot is directly relevant, it shows VSWR vs frequency and allows direct comparison with any objective (the creator has shown a reference line at VSWR=2, perhaps that is his acceptance criteria).
Next let’s consider the Return Loss plot. Return Loss by definition must be a positive dB quantity, this plot does not use industry standard meaning for the term Return Loss… so what else about it is non-standard. I dismiss it.
Let’s jump to the Smith chart. To a sophisticated Smith chart reader, it can be informative… but not very much in this case. It lacks a frequency axis, so direct comparison with a VSWR objective is not easy… unless you drop frequency markers on the plot… and the guy has… but he has not shown a VSWR objective or the frequency markers at that VSWR. Interesting, but not very interesting since the VSWR plot tells all and there is no reason to dig deeper.
The plot entitled reflection coefficient is actually just the magnitude of the complex reflection coefficient. It is not very interesting for this scenario. VSWR is calculated from the magnitude of the complex reflection coefficient, so it is a bit redundant when the VSWR curve suits so well.
The plot entitled phase is actually the phase of the complex reflection coefficient. It is not very interesting for this scenario… but to a high priest of phase, it provides an interest that is usually misguided. Don’t forget this is phase as observed at the source end and phase at the load end is likely to be significantly different (frequency dependent) and in this case it is not the antenna feed point anyway, so also a somewhat esoteric point. More on ‘phase’ and the Smith chart later.
The S11|Z| is a plot of the magnitude of complex impedance… it is part only of the complex value and of very limited value… unless you think that |Z| can be used like R in DC circuit analysis. Wrong.
So of the six plots, one and only one is directly relevant to a VSWR objective, and it is a good interpretation and presentation of s11 for that purpose.
The measurements are about provision of a suitable load for the transmitter, it says nothing of the radiation efficiency of the antenna system.
For high priests of phase
Since the NanoVNA appeared and became so popular, there has been an interest in the “phase” plot, which is usually phase of s11 or complex reflection coefficient Γ.
Above is a point plotted on a Smith chart (the asterisk). From the legend we can see that VSWR=1.019… this would be a stunning match to almost any practical VSWR matching objective.
But, look at the phase of the reflection coefficient Γ…. it is -90°. Woe betide us?
Nonsense, in most practical cases, if the magnitude of Γ is very low (ie VSWR is close to unity), the phase of the reflection coefficient is unimportant. Not only is it unimportant, but measurements of very low magnitude reflection coefficient are usually affected greatly by measurement noise, and with relatively high measurement noise, the phase of the reflection coefficient becomes a random value.
Is the NanoVNA useful?
Whilst the VSWR plot was the best presentation and only useful presentation, it might seem that the NanoVNA is no better than a conventional VSWR meter.
That is not so, the NanoVNA brings some advantages to bear:
- it is possible to move the reference plane;
- a sweep captures a whole of band picture in a very short time;
- low power allows the sweep to extend out of band and across in-use channels without causing interference;
- it is easy to create a record for archival or to share with others (save the .s1p file along with png of the chart); and
- should something be untoward, other presentations of the measurement may provide insight into what is wrong.