A common task is an overall assessment of an antenna system, this article looks at NanoVNA display configuration that will often suit stand alone:
- measurement at any point on the feed line; and
- measurement with the reference plane, either by direct connection, fixture calibration or approximate calibration using e-delay.
Caution
When measuring an antenna system with the NanoVNA:
- drain any static charge at the coax connector before offering the connector up to the NanoVNA; and
- do not leave the instrument attached any longer than necessary to make the measurements.
Case 1: measurement at any point on the feed line
Since the phase relationship of the reflected wave at the point of observation relative to that at the feed point is unknown the only meaningful statistics are those based on the magnitude of s11 (|s11|), |s11|, ReturnLoss, and VSWR.
My NanovVNA does not offer a ReturnLoss plot natively, you could use |s11|dB remembering to multiply all values by -1 (ie ReturnLossdB=-|s11|dB.
Otherwise, the VSWR plot is most useful.
A Smith chart plot of s11 is sort of useful, but there is an unknown rotation from the feed point.
Since they are of no real value, you could disable traces 1, 2 and 3 to make the display less cluttered.
Case 2: measurement with the reference plane at the feed point, either by direct connection, fixture calibration or approximate calibration using e-delay
In this case, the phase of s11 is meaningful which means:
- the Smith chart plot is properly presented wrt the chosen reference plane; and
- R and X components of impedance can be properly calculated and presented.
An example
Let’s look at an example antenna sweep where the NanoVNA measurements are wrt the feed point (e-delay has been used as an approximate correction for a short feed line tail). The examples are from NanoVNA.H.v1.2.20 firmware.
Above is a screen capture, the colours are inverted for printing.
The e-delay adjustment does not correct for ~0.27dB of feed line loss, so all measurement of |s11| and derived quantities have a ~0.54dB error. which is not of great significance in this example. Calibration of the 6.3m odd length of feed line as part of the fixture would eliminate this small error, and is advised if the measurement is made through a long section of feed line.
The blue VSWR (SWR) plot is meaningful. One can see that the VSWRmin=1.014, and VSWR=2 bandwidth is ~210kHz.
The red Smith chart plot is meaningful, the phase orientation of the plot is correct.
The R and X plots (magenta and green) are meaningful.
What if the measurement was case 1, remote from the feed point?
Above is a measurement of the same feed point through a length of coax with loss ~1.5dB and Zo around 50-j1Ω.
We would normally expect that observation through a lossy feed line section will reduce VSWR. That would be exactly true if the measurement reference Z (50+j0Ω in this case) was exactly the same as feed line section Zo (50-j1Ω in this case), but due to those errors, VSWRmin is slightly worse than in the previous plot, but the astute reader will note that the VSWR curve is shallower (VSWR at the graph limits is lower in the second graph).
The frequency for VSWR is shifted very slightly, again a result of error in the actual feed line Zo as mentioned.
The Smith chart is rotated, ~230°, a full turn plus 50°. The effect of the line loss has reduced the magnitude of s11, so all points on the curve are closer to the chart centre.
The R and X curves are also affected by the rotation and line loss, and are meaningless. Note the very different shape of the R and X curves.
Optimisation target
Optimising for VSWR50 remains the best choice for a lot of common antenna systems, particularly coax fed antennas (eg Optimising a coax fed half wave dipole).
Conclusions
- The out-of-the-box NanoVNA display configuration is not particularly suited to most common antenna system optimisation / tuning tasks.
- Display options allow better suited displays.
- Display configuration is saved as part of a calibration save.
More at NanoVNA setup for shunt matching task – scripting the setup.