Are you interested in an easy to build old time transmitter from the primeval days of tube rigs? You may want to try this one. It is based on the one tube power oscillator circuit presented in H. Lauer and H. L. Brown's "Radio Engineering Principles" McGraw-Hill Book Co., 1920, page 200. Identical circuits are also presented in J. H. Morecroft's "Principles Of Radio Engineering," John Wiley & Sons, 1921 and 1927, pages 487 and 569, and in other engineering texts and applied technology books of that period through to the 1960's (see references at end of article). Depending upon the tube used, power outputs from milliwatts to hundreds of watts could be realized in the early 1920's, and of course, with today's tubes.

As you can see from the schematic diagram, the basic transmitter contains only one resonant circuit (L1/C2/C3), a parallel-tuned plate circuit). A non-resonant "tickler coil" (L2) is inductively close coupled to the RF ground side of the plate tank circuit. This coil feeds the tube's grid circuit through a parallel combination of a 27k resistor and 300 pF capacitor. The resistor and capacitor bias the tube when it is oscillating. Shunt feed is used to keep the plate tuning circuit at DC ground. The earliest circuit used series feed.

I use a 50 ohm input antenna tuner to feed a parallel wire transmission line to an untuned semi inverted-V antenna. The tuner's input is connected to tank circuit's inductor near the RF ground end without a link. A series variable capacitor is used to control loading. (If you use series feed, the variable capacitor will provide the DC isolation to your antenna tuner.) The period transmitter used the antenna wire directly coupled to the plate tank circuit's inductance, and a substantial counterpoise connected to the transmitter's ground bus for the RF ground.

The antenna, or tuner, tap point on the plate circuit's inductor is dependent on the impedance of the load. Low impedance loads are connected near the bottom of the inductor; high impedance loads near the top of the inductor. Coupling an antenna directly to a series fed tank circuit is dangerous, and was against the law in the 1920's unless a d.c. blocking capacitor was used.

Most definitely this transmitter is one of the "basic primeval" tube transmitter circuits appearing at the close of World War 1. It's very easy to build, suitable for "file to fit-fit to file" construction, and very accommodating of whatever parts you have stashed in your junk box. Besides, it's fun to operate. Too bad you can't see the expression on the face of the amateur you're in contact with when you describe "stats" of the transmitter, but you can definitely hear the surprise in their reply.

Tube.  I use a 6CK4 low-mu triode (see the RCA RC-22 Receiving Tube Manual, 1963). The 6CK4 was used in some TV vertical deflection Amplifiers. The mu is 6.6, plate resistance 1200 ohms, transconductance 5500 micromhos, and the plate dissipation is 12 watts maximum. The cost is nominal, around $3-$4 at local hamfests, and that's for a new boxed tube!

I have not tried it, but I am sure that, with appropriate minor circuit changes, any low mu-triode can be used, including a triode-connected beam power tetrode such as the 6V6, 7C5, 6W6 or 6L6, or even a triode-connected TV sweep tube. Since no shielding is used, and there is minimal harmonic filtering, I would not recommend the use of high power level tubes!

Typical early 20's amateur triode tubes were the 205-D, VT-2, UV204, UV203, UV202 or even the small UV201. If you have a vintage tube, you may want to use it for this project.

Band.  This design is for the 80 meter band. It is interesting to note that the RCA Radiotron Designer's Handbook recognizes this type of oscillator circuit as being capable of operation up to 50 MHz. The circuit can also definitely be scaled for 160 meter operation. Scaling to 40 meters may not be advisable if there is any chance of trees near the antenna, antenna wire or the feed line swaying in the wind. Sway may make for an unacceptable amount of frequency pulling if the antenna coupling is not backed way out.

Voltage and Currents.  I use a brute force pi LC filtered power supply. The power supply is regulated by two VR-150's in series. Regulation was necessary to minimize keying chirp. The transmitter's loaded key-down plate current is about 22 milliamperes, unloaded about 8 milliamperes. The loaded grid current is about 2 milliamperes. The period transmitter may have used batteries, a DC generator, or even raw unfiltered AC for the plate supply.

Output Power.  This design produces about 2-1/2 watts into 50 ohm. More output is realizable, but with a degradation of tone and frequency stability.

Output Plate Coil.  The plate coil is 15 turns of space-wound 1/4 inch diameter soft copper tubing. The wound length is about 6 inches. The coil form is a 3.3 inch OD varnished cardboard tube. The coil ends are purposely left long to connect to the tuning capacitor, and ground bus.

Tickler (Grid) Coil.  This coil is 10 turns of space-wound #12 AWG solid bare copper wire, about 1.6 inches long. It slides over the plate coil form and is close-coupled to the grounded side (r.f. cold side) of the plate coil. The tickler coil and plate coils are wound in the same direction. Sections of popsicle sticks are glued to the inside diameter of the tickler coil to give it form and permit easy sliding when adjusting its coupling to the plate coil. The tickler coil is r.f. and d.c. grounded at the shunt fed plate coil's RF ground end.

RFC (plate choke) Coil.  The coil form is a varnished 15/16 diameter wood dowel rod (actually a section of an old broom handle). The coil consists of a four-section series winding of #28 AWG enameled wire. The sections are plain solenoid close wound, 50T, 70T, 70T, and 50T, with each section separated from the other by about 1/4 inch. All windings must be in the same direction. The period coil was close wound on a much larger diameter hollow form and, very likely, almost parallel resonant at the operating frequency. The period coil used single or double cotton or silk insulated wire.

Keying.  The transmitter is cathode keyed. Key clicks suppressed using only a 0.04 and 0.1 microfarad capacitor. A 47 ohm resistor in between the capacitors is used to minimize the effect of key-down resistance changes. These resistance changes were found to cause sudden small frequency jumps. I frequently use my keyboard keyer with this transmitter via a solid state interface circuit. But for a reason unknown to me, the transmitter develops frequency stability problems when keyed over 15 WPM.

Filament Supply.  A transformer-operated 6.3 VAC supply with two series 47 ohm half-watt resistors across the filament windings. Neither leg of the filament supply is grounded. In parallel with each resistor is a 0.04 microfarad ceramic disk capacitor. The junction of the resistors is connected to the key side of the tube's 47 ohm cathode resistor. This was done to eliminate the possibility of causing a cathode-to-filament short on key-up if one end of the filament had been grounded. The period transmitter used a direct heated (no cathode) tube powered from batteries, or a DC generator. Christmas lamps, each bypassed with 0.01 uF mica capacitors, were used in place of the 47 ohm resistors. The junction of the two lamps was brought out to the keying line.

Base Board.  Use half-inch heavily varnished AC exterior grade plywood, 12 by 16 inches (size is not critical). Alternatively, you may want to use a "choice" piece of wood for your base board.

Mechanical Considerations.  A few words of caution are in order when using "basic" transmitter circuits as this one: All connections, antenna change over relay(s) or antenna switches must be solid and clean, the variable tuning capacitor wipers very clean and positive, the plate and tickler coils, as well as all other components solidly supported, and no wind swaying the feed line, antenna, or trees near the antenna! Remember, the equivalent of a 1/2pF change out of about 300pF in the plate circuit can pull a 3550 kHz transmitter off frequency by 2.9 kHz!

This transmitter's frequency and tone, as with all simple non-buffered LC tuned self-excited oscillator transmitters, is very susceptible to dirty connections, components bouncing in sympathy with your keying, and feeder/antenna sway.

1. I had to put a few daubs of silicone rubber sealant (RTV) here and there on my plate coil to eliminate even the slightest bouncing of the coil on the cardboard form. Bounce caused sudden frequency shifts, or FMing, of the transmit frequency.

2. A long 1/4-inch diameter fiberglass "bicycle safety flag" dowel rod was used to couple the rotor shaft of the vernier tuning capacitor to the tuning knob. This was necessary to minimize hand capacity when adjusting frequency. (A 1/4-inch wood dowel works, but it is more springy, which makes fine frequency changes difficult).

3. You may have to carefully adjust your antenna tuner with a noise bridge or QRP SWR bridge to get the coax feed resistance as close as possible to resistive with minimal reactance. For my tuner and antenna setup, this minimized a slight tendency for sudden frequency jumps of about 500 to 800 Hz. A perfect 50 ohm "on the nose" resistance was not required.

4. Do not over-couple the load to the tank coil or use too high of a grid bias resistor. Squegging and other instabilities will result.

5. Yes, a big 200+ watt vintage or modern single or parallel jug version of this transmitter can be built. Remember, everything is exposed, the harmonic filtering is poor, and there is no metallic chassis to minimize radiated harmonics! Safety first; then, second, "harmonic" piece in your home and in the neighborhood!

6. Don't forget the little plate parasitic suppresser resistor, R4. Even in the early days the knowledgeable op was aware of parasitic oscillations that robbed power, possibly caused a shorter tube life, and could invite an unannounced visit from the radio inspector. The use of these resistors in plate and/or grid circuits was known early in the 20's.

7. Note the schematic's heavy lines. These indicate the ground buss and the main tuning coil. Make sure you use:

• a substantial ground buss (I use tin-plated 3/8 inch copper braid).
• short leads on all by-pass capacitors.
• long enough tubing at the ends of L1 so that the coil can be directly
  connected to C3, the main tuning capacitor C3.
• correct phasing on L1 and L2, with both wound in the same direction.

8. Grid current can be measured as follows: Unground (disconnect) the end of L2 marked with a dot on the schematic and install a .04-.05 mfd ceramic capacitor from the free end to the ground bus. Connect a milliammeter across the capacitor, positive terminal to ground.

Have fun, and don't be surprised when you make a whole bunch of contacts with "tail gaters" and "read the mail" ops who are curious to know more about your OT rig!

Radio Telephony For Amateurs by S. Ballentine. David McKay Co., Philadelphia, 1922, pages 114-117

Radio Engineering by F. E. Terman. McGraw-Hill, NY, 1932, First Edition, third printing, starting page 228

Theory Of Thermionic Vacuum Tubes by E. L. Chaffee. McGraw-Hill Book Co., N. Y., 1933, first edition, pages 322-329, with delta-Eo set at zero, very heavy academic analysis

Principles of Radio Engineering by R. S. Glascow. McGraw-Hill Book Co., NY, first edition, tenth printing, 1936, starting page 261, very heavy academic analysis

Applied Electronics by T. S. Gray. John Wiley and Sons, NY, second edition, 1954, seventh printing, pages 664-670, very heavy academic analysis