Build a 1934 Transmitter

From the Column "A.W.A. On the Air"
Conducted by Ken Gardner
From OTB Vol. 20, No. 3 (December, 1979)

The last OT Transmitting Contest aroused much interest which in turn brought several requests for old time transmitter circuits. Early self excited transmitters (Hartley, TPTG) are not recommended for present day operation unless very low power and in the hands of experienced operators who have modern monitoring equipment.

To qualify for OT AWA operation the circuit and tubes must be 1939 or earlier. The easiest approach of course is to build a single tube transmitter using a 6L6 as xtal oscillator. To increase power, add 807, 811 or 812 amplifier and one is in business with a 1937-1939 vintage set. But how about something earlier?

Thumbing through a pile of 1930-35 QSTs we found a circuit that may be of interest to members: a transmitter that can fit anywhere between 1932 to 1935--depending on vintage parts and tubes used (Figure 1A, B). The set is fully described by AWA member George Grammer, WIDF in the October, 1935 issue of QST. Following is a brief description and a few comments.

First, note the set is designed for 20 and 40 meter operation--a 47 tube as 40 meter xtal oscillator with a 203-A as PA on 40 or doubler on 20. The same tank coils are used on BOTH bands (the PA coil tapped for higher frequency operation). A UX-210, 211 or other early tubes may be substituted for the 203-A by changing filament plate and bias voltages as required.

Next, in all probability most members will want to change the coils for 40 and 80 meter operation. In doing so, it will be necessary to increase the capacity of plate condensers Cl and C2.

And lastly, the new design will obviously have to include a modern antenna coupling system.

It is advisable to follow the circuit as shown since WIDF did much experimenting to come up with a workable transmitter. Relatively new amateurs building this transmitter are advised to read the original QST article. An earlier version with 210 buffer can be found in Feb. '34 QST.

For the benefit of those who may wish to build an outfit along the same lines, the complete circuit diagram as finally worked out is shown in Figure 2. Parallel plate feed is used on both tubes to keep the high voltage off the coils, which naturally have to be handled when taps are changed. The antenna tuning circuit, C3, L3, C4, is arranged for series tuning of Zepp feeders, the coil L3 being proportioned so that the condensers will hit resonance on both 7 and 14 mc. with feeders approximately 45 feet long.

The circuit shown in Figure 2 is arranged so that bias for the amplifier can be obtained from a voltage divider across the plate supply for the oscillator. The condenser C11 serves both as an insulating condenser between the two filament circuits and as a by-pass for the amplifier grid bias.

Those not familiar with the connections used in this type of biasing will find the details of the arrangement in Figure 3. When this method of obtaining bias is used, there must be no direct connection between the negative terminals of the two power supplies nor between the filament transformers. The proper connections for keying and for filament returns will be found in Figure 2.

The transmitter illustrated here is a case in point. The job on hand was that of constructing a transmitter relatively simple and compact--since it had to fit into a space of limited dimensions--and having a moderate amount of power output for c.w. work on 7 and 14 megacycles. Quick and convenient band-changing was considered to be an especially desirable feature, indicating the use of tapped coils. In view of the fact that a miscellaneous collection of parts already on hand had to be used, the available space did not permit using more than two stages, with antenna-tuning equipment.

While the circuit changes were being made, another problem, that of keying, came into the picture. It had been intended originally to key the center-tap of the amplifier, leaving the oscillator to run continuously. A trial of this system did not give very pleasing results, however. Keying the amplifier center-tap opens and close its grid circuit, with the result that the load on the oscillator changes with keying. This in turn causes the oscillator frequency to shift so the frequency under key-down conditions differs from that under key-up conditions by a kilocycle or two. Thus the main wave and back wave are on two different frequencies, and although this may do no harm since the back-wave should not get through the neutralized amplifier, yet it does represent an undesirable condition. Furthermore, since the installation of some sort of thump filter usually is necessary with center-tap keying, the time lag introduced by such a filter easily could cause a chirp. The alternative of keying the negative amplifier plate return alone, thus maintaining a fairly constant load on the oscillator, was discarded because of its still greater tendency towards thumps and sparking at the key contacts.

It was therefore decided to give oscillator keying a trial, so the key was put in the oscillator filament center-tap. A quick check showed that the oscillator would follow rapid keying readily enough, but a very pronounced chirp developed. At this time the oscillator screen voltage was being supplied through a 50,000-ohm dropping resistor. The scheme of using a voltage divider was therefore introduced; the chirps disappeared and keying was no longer a problem. Keying the oscillator brings with it all the advantages of break-in operation.

 BACK TO ARCHIVES