New features in version 885 is
Fixed “Refer to Antenna” function... so lets check it out.
The AIMuhf was calibrated using the supplied SOL components and AIM885.
This is pretty much as expected. Xs is negative and increases with increasing frequency, Rs is a realistic small positive value.
Refer to Antenna
The refer to antenna facility was activated using the above values. Note that the config file is set to metres, so the feed line length being ‘subtracted’ if you like is 0.05m, and the remaining load is 1m of Belden 8262 (RG58C/U) cable (with o/c termination). The DUT is now 5% shorter, the plots should be very similar.
Above is a scan with refer to antenna. The Xs curve is of the shape expected, and the value at 10MHz (149) is close to expectation.
Note though that phase is negative at low frequencies and the excursions in phase around 15MHz, this is unexpected.
The saved graph CSV file has the following record for 10MHz:
Freq(MHz),SWR,Rs,Xs,Zmag,Phase(deg),Rho,ReturnLoss,%ReflectedPower 10.000001, 999, 0.115373, 149.653009, 149.653053, 89.955986, 0.998000, 0.017389, 99.600400
Note that the saved Rs Xs are postive, in conflict with the plot.
Setting the Zmag scale to 10 and rerunning the scan gives an expanded vie of Rs.
Above, Rs is negative below about 15MHz which is unexpected (though small negative values can be the outcome of errors in the original input Z measurement and coax parameters). More concerning is that although the plotted value for Rs at the cursor is negative, the value shown in the right hand margin of the window is positive, likewise for phase. Despite the phase change, Xs remains off scale in the negative direction across the whole chart.
But wait, there is more. Setting the Zmag scale factor to 10 (yes, that IS what was used for the last plot, but it will force a redraw which should be identical.
Above, the redraw is NOT identical. Now Rs is always positive, and Xs is off scale in the positive direction where it should be negative.
Is this just a limitation of this class of instrument
Above is the same problem using a Rigexpert AA-600 in its normal or fixed calibration mode (results would probably be better with a custom SOL calibration). At 10MHz, X is close to expectation, R is a very small value (0.0, but not negative… though that may be fudged in Antscope), smaller than expected from the sample, but within expectations for the instrument. Rigexpert seem to have no problem handling the test scenario.
A simple test revealed serious shortcomings in the
Refer to Antenna function, and the changes on redraw might cast more general doubt over the integrity of plot drawing, saved plot data, and handling of complex quantities.