R F Circuit

Since simplicity was a prime factor in the design of the instrument, a very elementary r-f circuit was chosen. Basically it consisted of a plate-modulated oscillator whose output was fed to an output connector or an antenna through a coaxial cable. The coupling between cable and oscillator was accomplished by means of a rotatable loop which served as the output control.

In the frequency range covered by the instrument, a butterfly circuit1.2,3 has several characteristics which make it a logical choice for use as the oscillator resonator. Some of these characteristics are:

1. A butterfly is capable of tuning over very wide frequency ranges at relatively high frequencies, thus minimizing the complexity of band switching problems.

2. With a butterfly of the proper design, wide-range oscillators usually can be made to operate up to about 80% of the natural resonant frequency of the oscillator tube.

3. The frequency of oscillation is varied simply by a single control.

4. A butterfly circuit has no sliding contacts carrying large r-f currents.

5. The unit is small and compact.

In order to keep the size of the oscillator at a minimum and to simplify the installation of a band changing switch, a semi-butterfly of the type shown in Figure 2 was selected.

A butterfly resonator, which is normally a two-terminal network, is well suited for use with a triode tube in a

*E. Karplus, "The Butterfly Circuit," General Radio Experimenter, Vol. XIX, No. 5, October, 1944. 2E. Karplus, "Wide-Range Tuned Circuits and Oscillators for High Frequencies," Proceedings of the I.R.E., Vol. 33, No. 7, July, 1945, pp. 420-441. 3U. S. Patent No. 2,367,681.

Figure 2. Diagram of the r-f oscillator showing general construction of the butterfly and the band switch.

Figure 2. Diagram of the r-f oscillator showing general construction of the butterfly and the band switch.

Butterfly ResonatorButterfly Feedback Diagram

Figure 3. Basic r-f oscillator circuit.

Figure 3. Basic r-f oscillator circuit.

modified Colpitis oscillator circuit in which the interelectrode tube capacitances form the feedback circuit. A basic schematic of the circuit is shown in Figure 3.

The Type 9002 Miniature Triode was selected as the oscillator tube because it was a preferred type and because of its relatively high natural resonant frequency, small size, and low power requirements. Since continuous coverage is not possible with the butterfly resonator at frequencies higher than about 80% of the natural resonant frequency of the tube, the upper frequency limit was fixed by this choice of tube.

A semi-butterfly resonator is actually a parallel resonant circuit which has a capacitive branch and an inductive branch. From Figure 2 it can be seen that the capacitive branch is made up mainly of the capacitance from one stator section to the rotor in series with the capacitance from the rotor to the other stator section, and that the inductance of the arm between points A and B constitutes the main portion of the inductive branch. As the rotor is turned in the counterclockwise direction from the position shown in the figure, the capacitance from the right-hand stator section to the rotor decreases, thus decreasing the total effective capacitance. The rotor also advances along the inductance arm, partially shielding it magnetically and hence reducing its inductance. As the result of the decrease in inductance with decreasing capacitance, a much wider tuning range is obtained than can be produced by varying the capacitance alone. In this butterfly a 2.25 to 1 change in inductance over the tuning range was obtained.

In spite of the advantage gained from the variation of both L and C in a butterfly, it was found to be impractical to cover the entire band from 40 to 500 Mc in one step, and it was necessary to break it into two bands. The high-fre-quency band from 115 to 500 Mc was covered with the resonator acting as a butterfly as described above. However, for the 40 to 115 Mc band, the butterfly inductance arm was open-circuited by means of the band switch which placed a low-frequency coil across the capacitance sections of the butterfly as shown in Figure 2. In the low-frequency band, the inductance remained fixed and the tuning was accomplished by means of the variation in capacitance alone.

The design of the band switch was important, as the inductance it introduced in its closed position had to Vie small compared to the minimum inductance of the inductive arm of the butterfly, or the tuning range would have been appreciably reduced. The losses it introduced in the circuit also had to be small to avoid reducing Q below the value required to sustain oscillation. Therefore, it was constructed of a multiple set of blades which interleaved the rings forming the inductance arm as shown in Figures 2 and 4. Each blade consisted of two spring leaves which were compressed when the blade was between the rings and made a rigid, low-loss, low-inductance connection.

In the type of Colpitis circuit used, the whole oscillator circuit, including the cathode of the tube, is floating with respect to ground and connections were made to the cathode, heaters, and plate through chokes. However, a choke is not an infinite impedance, and it was found that a "hole" in the oscillations occurred over a narrow frequency band, which appeared to be caused by a resonance in the circuit consisting of the cathode lead inductance, the choke impedance, and the stray capacitances of the stator sections to the shield. Although the frequency at which the hole appeared could be shifted by changing the reactance of the choke, a practical choke design was not found which would shift it out of the desired range. The hole was eliminated by connecting a resistor across one of the chokes as shown in Figure 2, to reduce the Q of the undesired resonance.

At the upper end of the frequency range, the operating frequency is largely dependent on the tube capacitances. Therefore, in order to compensate for changes in tube capacitance when the oscillator tube is replaced, a small adjustable trimmer capacitor was connected from one stator section to ground.

Mechanically, the butterfly stator was mounted on three insulating posts as shown in Figure 4, and was covered by an insulating plate that performed the dual function of supporting the oscillator tube socket and of clamping and aligning the two stator sections of the butterfly. The insulated rotor shaft turned in a set of ball bearings mounted in a casting below the butterfly. However, it was found that quarter-wave resonance occurred in the shaft circuit near the upper end of the tuning range, and it was necessary to ground the upper end of the metal shaft with a grounding strap as shown in Figure 4 in order to shift the spurious x'esonances out of the operating range.

The orientation of the output coupling loop was adjusted by a knob on the panel and was arranged to have its minimum and maximum coupling to both the butterfly inductance arm and the low frequency coil at approximately the same position. The power from the coupling loop was fed through a coaxial

transmission line to the modified type N connector located on the panel. This connector was constructed in a novel manner as it could be used as a standard connector or an antenna. For use as an antenna, the center conductor was pulled out to any desired length up to 6 inches, as illustrated in Figure 1. The center conductor was formed of a tightly coiled spring and hence was not damaged by bending.

Was this article helpful?

0 0

Post a comment