Fig. 2. Lumped-line oscillator with tuning provisic MHz oscillator.

proximity of conducting surfaces, and mechanical vibration. In the Clapp circuit in particular, the juncture of the inductor and the resonating capacitor is at a very high impedance level and displays very high sensitivity to stray capacitance. Awareness of these matters leads to defensive measures such as rugged construction, shielding, short leads, etc. But would we not have a more desirable situation if "built-in" immunity against frequency change could be achieved? This seems to suggest a different type of resonant circuit. In what follows, the revival of an "old-hat" approach may suggest itself. However, if attention is focused on the mode of operation involved, it will be seen that the described oscillator embodies a combination of theoretical and practical approaches which makes it unique and, to the best of my knowledge, is not to be found in the technical literature.

Oscillators employing transmission Hues are well known. Chief claim to fame here

and buffer amplifier. Circuit values shown for 21

stems from the high Q attainable from lines, especially at the higher frequencies. The quarter-wavelength resonance generally utilized has precluded the construction of such oscillators for use below the vhf region, too, the general design approach of such oscillators has endowed them with familiar shortcomings. The obsession with push-pull configurations results in "hot" lines wherein the detuning effect of the immediate environment is much in evidence. The high impedance end of the quarter-wave lines is associated with the active device and thereby becomes very susceptible to frequency pulling. Moreover, such an association with a transistor would be even more detrimental than with a vacuum tube, On the basis of such facts, the usual substitution of a resonant line for an L-C tank circuit constitutes neither improvement nor innovation. But the transmission line has not been brought up for the sake of rejecting it! When one considers certain properties of such lines, it becomes apparent that conventional line-type oscillators do not advantageously use ihese properties.

The reader should by now be prepared for some kind of marriage between the Clapp circuit and a transmission line. This, indeed, is the destination of our discussion, but we shall direct our attention not to the conventional quarter-wave line, but to the full-wave line. Whereas odd quarter-wavelengths simulate the parallel resonant condition of an L-C tank, even quarter-wave lines simulate the condition of L-C series resonance. Let us further stipulate that our full-wave line be coaxial cable. For the moment, we will deal with the simple full-wave coaxial line shorted at the far end. Inasmuch as we will be interested in establishing full-wave resonance at least as low as the amateur forty-meter band, we can make use of one of the Microdot miniature coaxial ■cables, and wind it solenoid-fashion around a cylindrical form. Now the cat is out of the bag: We have a "coil" which behaves as a series-tuned tank circuit and is therefore suitable for use in the basic Clapp circuit. Yes, but what a coil this is! It possesses the following features:

It can be wound on either a metallic or insulating form.

Its resonance frequency is not influenced by turn spacing, turn diameter, or winding length.

Its resonance frequency is not influenced by surface dust or moisture, ft will not participate in undesired inductive or capacitive coupling to other parts of the circuit.

The entire length of the outer sheath is a rf ground potential and there are no frequency disturbances from proximity effects.

It displays virtually no microphonic effects and is very nearly immune to mechanical vibration.

Its temperature coefficient is relatively low, being a calculable function of its change in end-to-end length. There is no "nerve ending" in the form of a high impedance junction. Such a line is "happy" to work into the low impedance input of a transistor. Reasonable compactness is readily attained. Layer winding may be employed for lower frequencies. Nearly complete shielding with no reduction of "Q."

The circuit of the basic "limped hue" oscillator appears in Fig. 1. From the viewpoint of the active device, i.e., the transistor, a series resonant "tank" circuit is seen from points A—B. This being the case, oscillation occurs in the same manner as in a Clapp oscillator. Although the circuit has the same configuration as a conventional Colpitts with a resonant line operating in the quarter-wave mode (simulating a parallel-tuned L-C tank) such operation is here ruled out by the very






WRITE OR CALL LAMPKIN LABORATORIES INC. P.O. Box 2048 » Dept. L • Phone 813-746-4175 Bradenton, Florida 33506

large capacitors across the open end of the line. These capacitors tune quarter-wave resonance to such a low frequency that the circuit Q is too low to allow oscillation to occur. Indeed, the purpose of these inordinately large capacitors is to all but short out the transistor end of the line! This of course is perfectly agreeable to a line which operates in a mode simulating a series-timed L-C tank. And why not use the line in the half-wave rather than the full wavelength mode? Although transmission line operation is similar in all multiples of one half-wavelength, it so happens that the phase-relationship between voltage and current is suitable for oscillation in this circuit only for even multiples of one-half wavelength. So a full-wavelength line is the shortest one which will function in the simulated series-resonant mode in this circuit.

Digressing back to LC tank circuits, the usually cited formula for parallel resonance is: fo = l/27r-\/LC but the true formula is actually: fo = iWlC(VL-RL2C7L-RC2C) where R]_ and R(* are respectively the equivalent series resistance of the inductor and the capacitor. Although Rl and Rc can often be considered negligible in order to enable the use of the simpler formula, the true formula clearly shows that the frequency of parallel resonance is affected by tank circuit losses. Conversely, the frequency of a series resonant tank circuit losses. Conversely, the frequency of a series resonant tank is not per se directly detuned by resistance associated with the tank. Although the Q is adversely affected in both types of resonances, the series-tuned type is more likely to be superior in providing an oscillator with immunity against frequency

Calculation of Frequency of Oscillation

1. Wavelength - physical length/velocity of propagation where velocity of propagation is approx. 0.66 times the free-space velocity.

0 0

Post a comment