I've been searching for a low cost method of employing my venerable HP spectrum analyzer/sweep generator as a scalar network analyzer. Commercial return loss bridges are quite expensive. After a bit of research I settled on trying to homebrew a resistive return loss bridge. The target was to realize a bridge that would give reasonably accurate results when used over the 30m-70cm bands. Figure-1 below shows a diagram of the circuit presently under evaluation.

After cranking a bit of algebra, I find the voltage at the spectrum analyzer input to be Vd = (R-Rt) / (3R + 5Rt). In this case, R=50 Ohms. The general analysis of the circuit is may be accomplished by splitting the brige in half with two identical signal sources and then converting each half to their respective Thevenin equivalent. The resulting series loop circuit can then be analyzed by Kirchoff's voltage law.

The transformer T1 employed in the circuit of Figure-1 is necessary because the spectrum analyzer input and output share a common ground. Without T1, the spectrum analyzer input ground connection would short out one side of the bridge.

A readily available 75-Ohm:300-Ohm VHF/UHF television transformer was scavanged for T1. The TV transformer comes typically as a plastic cylinder with a 75 Ohm type-F female connector on one end, and 300 Ohm twin-lead pigtails coming out the other. These transformers are used for the broadcast VHF/UHF TV bands (in the US, 54-216 MHz VHF, 470-806 MHz UHF).

On the transformer I used, a shrink-wrap label held the cylinder halves together. Once the label was peeled off, the cylinder was easily opened to reveal the binocular ferrite balun inside. There should be enough enameled magnet wire wound on the balun to allow rewiring as T1 as per Figure-2 below. Note, I define one "turn" on the transformer to to be each instance of a wire passing through the core.

Figure-3 below is a block diagram of the measurement setup. In-practice, the sweep generator is enabled with -10dBm output and the frequency range of interest set. A short circuit is connected to the Device Under Test (DUT) port, the trace is subsequently stored and normalized. Finally, the device under test is connected and return loss measured.

The figure below shows a measurement made on a Daiwa DAX-750B 2m/70m portable vertical antenna. The measurement shows return loss for the 2m band. The marker is set to the VSWR=2:1 (RL=9.5 dB) point to the left of center. The center frequency is 146 MHz. The total sweep is from 143.5 MHz to 148.5 MHz. The 2:1 VSWR bandwidth is approximately 1825 KHz.

The first realization of this bridge was done on a spaghetti sauce jar lid using bulkhead mount BNC connectors and point-to-point wiring (see figure below). I know... it's ugly, but it works and it allows me to easily modify the circuit. The jar keeps the circuit protected and anchors the bridge down somewhat against the pull of the test cables.

Here is a picture of the "Spaghetti Jar Bridge" assembled. The final construction will probably be done in a scavanged cast metal 2-way CATV splitter enclosure. Unfortunately, the F connectors will be 75 Ohms, but I'm hoping this won't matter much provided I transition immediately to 50 Ohm cables. If it does matter, I'll isolate the bridge with homebrew 5.7 dB minimum loss 75:50-Ohm pads. In addition, the ferrite bead transformer scavanged from the splitter may have broader bandwidth than the 75:300 Ohm VHF/UHF balun used initially.

In conclusion, this is a work very much in progress. I certainly welcome and appreciate comments and suggestions. You may email me by clicking the link at the bottom of this page

73's,

YB0ART/WB4ONA, Dave in Jakarta