▼ Menu

Transitron Oscillators

T J Rehfish [★] Northampton Polytechnic, Wireless World June, 1943.
    
Extras ▼

 

A Letter to the editor.
The circuit presented by the author.

There are two remarks which I should like to make with reference to A G Chambers articles in your March and April issues.

The first is theoretical, and concerns the operation of the transitron oscillator in the metre region. It was noticed by Herold that a circuit which was expected to oscillate at 10 MHz did not; the expectation was based on the knowledge that the negative resistance of the valve (at low frequencies) was less than the positive dynamic resistance of the tuned circuit at 10 MHz; and the failure was explained in terms of transit-time effects, although a more modem explanation would blame feedback due to lead inductances rendering the negative screen-cathode resistance too high for transitron type oscillation. I pointed out in a letter to Wireless World (October, 1940) how the unusually low negative resistance of an AC/SP1 valve could be used to extend the frequency range. Nevertheless, the figure of 60 MHz claimed by Chambers (and since verified by students of this Institute) appears excessive, especially as he does not claim a particularly high G2-to-G3 transconductance. My doubts on this point were further increased by the second article, wherein feedback in the G1 circuit was mentioned. I suggest that there is little likelihood of the transitron mechanism accounting entirely for oscillations at 60 MHz; it is more likely that some other mechanisms come into play, such as positive feedback from other electrodes, Barkhausen-Kurz electron oscillations, etc. This could be checked by strapping screen and suppressor and observing the continuation of oscillation, the suppressor earth return circuit (R1) being disconnected.

Of greater practical importance may be a warning concerning frequency stability. While this is inherently high without question, Chambers Fig. 2 (April Wireless World) shows the 1 V output developed across a coil coupled to the tuned circuit L1C1, presumably rather tightly as the RF volts across L1 are of the order of 10 V. Hence the effective parameters of the tuned RF circuit will depend greatly on the load across the output, unless, of course, an attenuator of considerable step-down ratio is used; even then, the circuit would compare unfavourably with a more violently oscillating circuit, as an output of 0.1 V at least would be required in practice and still give rise to far greater pulling than if it were taken from the tank circuit of a Hartley, say, with a PD of 100 V RMS or so across it. Frequency modulation is another likely drawback of the circuit shown, especially at low values of dynamic resistance.

On the other hand, the absence of a high signal voltage is a good feature in a test oscillator, as less screening is called for. It is, therefore, suggested that the output be collected from a buffer stage, e.g., a cathode follower. Modulation could be carried out on that valve, e.g., by means of transitron action. The circuit suggested is shown in the figure; the values indicated are purely tentative, and the numbering follows Chambers diagrams.

The RF output from V1 is applied to the first grid of valve V2; R12 is the self-bias resistor and the cathode load is R12 + R14. Audio oscillation takes place in the circuit L2C2 which is in series with R12 + R14; the audio action is partly transitron type and partly positive feedback (R12 + R14). R10 is made variable and controls the modulation depth. The voltage across R14 contains the modulated carrier plus some audio component which, however, may be blocked off by C12, which feeds into the attenuator (not shown).

Another objection to the transitron oscillator has been mentioned to me. It concerns the effect of supply voltage changes on frequency stability. This is apparently very serious, rendering the performance of the transitron no better than that of the secondary emission (dynatron) type. The explanation is probably due to the effect of supply Volts on the space charge near the suppressor grid, which is equivalent to a capacity between screen and earth.

Use browser back button to return.
art-304b