Checking your RF Bridge Calibration

Many of you are using one or other of these devices which Gerd Janzen, DF6SJ refers to as an 'Active Standing Wave Ratio Meter'; a battery powered, comprehensive antenna measuring instrument all in one box. Gert has devoted a book [1] to whole range of RF measurements to which one of these instruments can be used. Most manufacturers of these devices call them Antenna Analysers The two main suppliers of commercial antenna analysers, are Autek and MFJ. I have the RF1 from Autek (1.2 to 35MHz) and the MFJ-249 (1.8 to 170MHz) from MFJ. These instruments do a fine job of measuring SWR but if you are into antenna experimenting then it is useful to measure R +/- j impedance. This impedance can be extracted from readings of SWR and Z from the RF1 and converted to R +/- j using TLW program that comes in the CD with the ARRL Antenna Book. However, as you are aware, the impedance at the antenna will not be the same as that measured at the other end any feeder connected to the antenna due to the transmission line transform effect.

There are two ways to find out what the real antenna impedance is: 1. Measure the electrical length of the feeder and use a transmission line calculator, such as the Smith Chart or the TLW program (or similar) to find the antenna impedance. 2. Measure the impedance of the antenna using a halfwave, or a multiple of a half wavelength, of coaxial. I use this method quite a lot but you have to bear in mind that, because the cable is resonant, it can result in antenna currents on the cable, which can give inconsistent impedance measurement results. However antenna currents can be minimised using an RF current choke. Remember that the cable is a half wavelength long on one frequency only.

I have recently acquired an Autek VA1. This remarkable little instrument has additional facilities compared with the RF1. It can measure R +/- j, and its polar equivalent. Furthermore it can measure equivalent parallel resistance and reactance and even the transform action of a length of feeder. My interest is the measurement of impedance so I was interested on how the good the VA1 was in this respect. One method is to use a set of accurate dummy loads. Assuming a 50ohms system, dummy loads for 50ohms, 25ohms and 100ohms are useful. However, these dummy loads only measure the resistive component of impedance. How would you go about measuring the reactive component? A novel approach, first described by W. N. Carron [2], uses the impedance transform effect of transmission line. A length of coaxial cable is used with a resistive load of which will produce an SWR of around 1.5:1 to 2:1. The impedance is plotted over a range of frequencies, which has the effect of changing the electrical length of the test feeder. The physical length of the test feeder is not critical because the accuracy of the instrument will be apparent by any deviation from the SWR contour when the results are plotted on an impedance diagram. I made a test using my prized HP 4085A vector impedance meter and the Autek VA1, the results of which are shown in below.

Above are the results of an impedance plot of a 22ohm resistor made via a length of coaxial cable over a frequency range 3 to 29MHz. The blue circle is the calculated 2.27:1 SWR produced by a 22ohm resistor. Ideally, the impedance plot should follow this circle with a slight spiralling towards the centre due to coax cable loss. The red and green plots are of the HP4085A and the VA1 respectively.

While neither of the plots look ideal you have to remember that this is a sensitive test. Both instruments are accurate enough to enable one to devise a suitable matching network for an antenna. There may be errors due to the measuring technique. I feel that it would be more accurate to use a Cartesian projection for plotting R +/- j. Plots on a Smith chart projection are only necessary if you need to calculate impedance transforms, which are much more easily done on a computer these days. An example of an impedance plot of a 88ohm resistor made via a length of coaxial cable over a frequency range 1 to 30MHz using a HP4085 on a Cartesian projection is shown below.

The apparent shift in resistance was found to be due to the resistance scale not being set up. 
 The inductive shift at the higher frequency is thought to be due 
 to the leads used to connect the instrument to the test cable.

REFERENCES
[1] RF Measurements with an Active Standing Wave Ratio Meter,  Gerd Janzen, DF6SJ. DARC 
Verlag GmbH, Baunatal. ISBN 3-88692-023.
[2] The Hybrid Junction Admittance Bridge’,W N Carron, Antenna Compendium Vol 3 ARRL 
1992.