May 1952 QST
Have you ever made a power or noise figure measurement and had to scratch your head over why the reported value made absolutely no sense? How about measuring a positive gain value through a passive device? I have experienced this sort of nonsensical phenomenon on more than one occasion, and to my recollection every time the cause was stray frequencies - often oscillations - mixed in with the intended signal. This app note addresses the issue in regard to impedance matching networks that are wideband enough to support RF energy that is outside the band being tuned - or even unexpected inband signals - with a simple method of avoiding such surprises.
In adjusting an antenna coupler to obtain a 1 to 1 standing-wave ratio in a coax link to the transmitter, it sometimes happens that the best possible adjustment does not bring the meter reading down to a null. This could be caused by insufficient matching range in the antenna coupler, or might even be chargeable to a faulty .s.w.r. bridge. However, there is a good possibility that the trouble is impure r.f.
One of the first things that should be done after building a resistance-type s.w.r. bridge is to check it with a noninductive resistor of a value equal to the impedance for which the bridge was designed. If a good null is obtained the bridge is OK. In making this kind of check it does not matter if more than one frequency is present in the r.f. applied to the bridge, because the test resistor has little or no frequency consciousness.
An entirely different situation exists when the bridge is used to check a transmission line that feeds an antenna, either directly or through an antenna coupler. The load in this case is definitely selective, and it can be made to look like a desired value of pure resistance only at one frequency or over a small band, in terms of percentage. If the r.f. applied to the bridge contains more than one frequency - for example, harmonics - the s.w.r. at the "off" frequencies usually will be very high, and a relatively small voltage will give a good-sized meter reading. Since adjustments to the coupler or antenna matching system have little effect at any frequency other than the operating frequency, the bridge gives a false indication.
Troubles of this sort are most likely to occur when the output stage in the transmitter is a frequency multiplier, or is a straight amplifier connected to a frequency-multiplier driving stage in such a way that either a submultiple or a harmonic of the output frequency can be amplified and applied to the bridge. A check with a crystal-detector wave meter at the final tank will show whether a frequency other than the desired one is present in appreciable amplitude. If so, a tuned trap can be inserted in the line between the final stage and the bridge to take it out. The wavemeter check should be made with the amplifier operating at the power level that gives a full-scale reading on the bridge.
Resistance bridges take so little power that it is sometimes a problem to cut it down enough. However, with a tetrode final there is a simple solution - disconnect the screen voltage, ground the screen, de tune the plate circuit, and reduce the excitation by detuning somewhere in the exciter chain if no other means is available. Then when the bridge is connected the full-scale initial setting can be approached by retuning the plate circuit toward resonance and, if necessary, increasing the excitation. This method usually gives ample control over the output without requiring a reduction in the amplifier plate voltage, since the plate input will be small with the screen at zero d.c. voltage. - G. G.
Posted June 17, 2016