During the 1970's, a popular home construction project was the QRP transmitter built on a tuna fish can. Referred to as the Tuna Tin Transmitter, it was a simple, solid state design. Back a few years ago, a QST article appeared wherein the Tuna Tin Transmitter was re-designed using hollow-state devices - better known as vacuum tubes. The transmitter used a 6C4 triode as the crystal oscillator and a 5763 pentode as the final amplifier. The transmitter was capable of up to 8 watts but was to be used at 5 watts to stay within the QRP "legal limit". When the QST article appeared, I decided to take it one step further. I disposed with the tuna tin approach and designed a small PC board that added AM telephony capability to the basic CW design. To facilitate this, I added a few tubes to the design: A 6360 dual beam tetrode as the modulator stage (conventional plate modulation was utilized) and a pair of 0B2 gaseous voltage regulator tubes for stablization of the modulator screen and crystal oscillator. To keep things simple, off-the-shelf parts were used; one could built this rig today with little difficulty in sourcing parts. Finally, while a simple pair of 1N4007 diodes can be used in the fullwave rectifier circuit, I also left provision to use a 6X4 fullwave vacuum diode for the rectifier.
The circuit is straight forward: A fullwave, choke input power supply provides 300VDC for the plates of the 5763 PA and 6360-modulator stages. A series string of 0B2 regulator tubes are used to produce a 30mA, 216VDC supply for the screens of the 6360 as well as the 6C4 oscullator. In addition, a +/-18VDC supply is used for the audio driver which is a simple op-amp circuit. The -18VDC supply also provides protective bias for the PA and operating bias for the class AB1 modulator. Finally, the unregulated 21VDC supply provides operating voltage for the relay switching circuits. During PTT operation for AM work, three small reed relays sequentially allow the oscillator to come up to speed, bring up the PA and finally power up the modulator stage. When CW operation is selected, the oscillator operates fulltime when in the transmit mode and the cathode of the PA stage is keyed through a key-click filter. The modualtor stage is not keyed but the secondary of the modulation transformer is shorted to avoid inductive chirping of the CW note. The design is rather unorthodox: Op-amp stages drive the grids of the modualtor tube directly. This is possible since the tube is baised to -18 VDC and the op-amp rails slightly below this value; the 6360 is never driven into grid current since it is not possible to raise the peak grid voltage above 0VDC. A small amount of cathode degeneration is also used to improve linearity of the modulator stage.
Iron required to built the transmitter is available, new, today shoudl one decide to use off-the-shelf components. The plate transformer is a Hammond unit as is the filter choke. The low voltage transfomer is a torroid from Digikey although a conventional 36 volt CT transformer may also be used. The unusual item is the modulation transformer. It is also a Hammond device; in this case, an audio driver transformer with a 10KCT winding used as the primary and a 4K secondary used to match the PA stage. the transformer can easily handle to PA plate and screen current. The PA screen is fed from the modulated plate supply through a dropping resistor. This insures that +/- 100% modulation is possible with low harmonic distortion.
I am considering producing a small number of kits, including all iron, tubes, PC board, and components if there is interest. Please email me for more details. I need to commit to at least 10 PC boards before placing the order to reduce the cost of the PCB to a reasonable value. The kit will not include a chassis or cabinet. All components except the transformers and choke mount on the PCB. The builder will have the choice of 75 or 40M. It is also possible to build the rig for 160M as well as 20 and 10M. I have not constructed one for 6M but it should also work with a few mods; mainly in the oscillator stage.
To make the rig a bit more intersting, I also designed a small board that takes +/-300VDC plus sampled RF and high level audio displaying a trapezoid pattern on a 1EP1 1 inch CRT. The board will also accept a 1DP1 CRT, but they are quite rare these days. Of course, more common 902 or 2AP1 2 inch CRT can be used but the socket will need to be wired to the board rather than direct mounted.
I will post a few pictures of the protoype and finshed, final PCB soon. In the mean time, I am posting a copy of the schematic and assembly instruction. This is a work in progress; please excuse the unfinished nature of this page at current. If you decide to build the unit on your own or distribute the schematic and notes, please simply credit me with the initial concept and design.