Two Tube AM Transmitter
Version 1 (November 2003)
After several trials, I finally completed a working AM transmitter. The transmitter's design is based on several ideas, but mainly based on Phil's Li'l 7 Transmitter design. It is not always easy to achieve oscillation in LC-based transmitters without the right combination of tubes and coils. The tubes used in this transmitter is a 6J5 and 6SA7 that operates at around 1400kHz on the AM band. Below is the schematic diagram of the transmitter.
The antenna coil was salvaged from the oscillator circuit in a typical old radio with a 455kHz I.F. frequency. The coil generally has two windings; one with a larger DC resistance than the other and is indicated as the larger winding in the schematic. If there is no oscillation present on the antenna of the 6SA7 then reversing the connections to one of the coil windings usually does the trick. The plate current by the 6SA7 causes the larger winding to generate a magnetic field that induces current in the other coil. The coil has to be the correct polarity to push the grid negative to reduce plate current. When the plate current reduces then the grid current reduces and allows more plate current to flow. The oscillations are controlled by the LC circuit consisting of the grid coil and the 150pF capacitor. The trimmer capacitor in parallel with the 150pF allows small range adjustment of the carrier frequency. If oscillations do not start, flipping around windings might help.
Modulation is achieved by feeding audio in the second grid of the 6SA7 tube so the plate current is controlled by the audio signal. As a result, the oscillator signal and the audio signal are superimposed upon each other resulting in an amplitude modulated signal, hence it is an AM transmitter!
The 6J5 is a triode designed for general purposes, but in this transmitter it serves as a low power half-wave rectifier in this circuit. The power supply provides about 140VDC loaded. The 6J5 was one of the few tubes I could find in my junk box at the time, but I would recommend using a tube that is actually designed for rectification purposes such as a 6X4.
BEWARE! An isolation transformer or some sort of power transformer is highly recommended because one side of mains is connected to the ground side of the audio input. The plug must be polarized so this side of the circuit sees the neutral. If the plug were accidently reversed and the ground shield of the audio cable were to become hot then it could potentially SHORT out with most, if not all, AC powered audio equipment. I only use a portable CD player or audio device with this transmitter to avoid conflicts with AC polarity.
Version 2 (August 2008)
Some fellow ARF (Antique Radio Forum) members offered to send me some parts to get started on a quite successful 6888-based AM transmitter. The 6888 is an oddball octal tube that goes by the description of "dual control pentode" because its first grid and supressor grid both are used as a control grid. In the kit design, a 6AB4 or 6C4 is used as an audio preamp and the 6888 as the transmitting tube. Below is a schematic of the transmitter as designed by members on ARF.
The circuit is fairly straight forward for a transmitter but there are some noteworthy features. The 1MHz crystal oscillator requires 5 volts and produces a nice square wave output at 1MHz. These oscillators are normally used in TTL or computer circuits because it is a self-contained oscillator rather than a mere crystal. The cathode current by the 6888 is sufficient to provide power to the crystal oscillator and a 1N4733 (5V 1W zener) is used to regulate the voltage. The oscillator output controls the first grid of the 6888 to create a clean 1MHz carrier on the plate. The inductor and trimmer capacitor above the plate of the 6888 serves as a tuned LC tank. The trimmer capacitor will adjust the output strength and thereby the range of the transmitter. The range is largest when the LC circuit resonates at 1MHz.
For use with stereo audio, another 10K is used from the second connector to the 50K pot to basically combine both channels into a mono channel before being transmitted. The 6AB4 serves as a very high gain preamplifier. The carrier is modulated by feeding the audio signal from the 6AB4 into the other "control" grid of the 6888.
The power supply is solid-state and uses a bridge rectifier to provide a very smooth B+ for clean audio quality in the transmitted signal. The original design used two back-to-back low voltage transformers to provide filament power and isolated mains voltage. The isolation allows this transmitter to be safe to use with just about any audio equipment without the hazard of shorts or shock. Although I was sent the odd PC-12-800 transformer, I did not feel like sending in a mail order for the Radio Shack transformer so I used a Hammond 262D6 transformer that I had left over from an experiment. Below are schematics of the original and alternate designs.
The PC-12-800 is a dual secondary transformer that is capable of producing 12V 0.8A with the secondaries in series, or 6V 1.6A in parallel. The Radio Shack transformer is 12VCT 1.2A and half gives sufficient power for the 6V 0.8A filament power requirements of the 6888 tube. The other half provides power to the 6AB4 and the two 6V windings of the PC-12-800 transformer to produce isolated 120V on the other side.
Another interesting thing to note is the use of an ordinary indicator LED in series from the B+ to the transmitter load. When the transmitter is first turned on, the LED remains off and slowly illuminates to full brightness when the tubes warm up completely. The LED is a good indicator of the load current and most LEDs are limited to 20mA of current so I figured the 120V 54mA winding of the Hammond transformer should have no problem. During construction I experienced some unforeseen short or miswiring in the circuit that caused the transformer to heat up quickly and give off the smell of fresh polyurethane. Not good. The bridge rectifier ended up ruined and caused a 500mA !!! load on the 120V 54mA secondary of the transformer. Now with the issues cleared up, the B+ runs at around 180V at 20 to 25mA as expected. Before determining the cause of the short on the 120V secondary, I assumed the overheating was due to the 50mA excess load on the 6V winding. The 6888 and 6AB4 filaments combined draw 0.95A at 6V while the transformer is only rated for 0.9A. However, inserting a 1.7 ohm 5W resistor reduces the 6888 filament current to a reasonable 0.75A to bring the overall current load to a tolerable 0.9A for the transformer, albeit a slight reduction in transmission strength due to a cooler 6888 tube.
The 50K pot in the circuit controls the amplification of the input audio signal and in effect controls the modulation strength of the output. For strong audio signals, too much amplification would result in overmodulation and distortion. In short, the pot is used to adjust the output to achieve 100% modulation for best audio quality.
The transmitter components, including the Hammond transformer, were quite literally crammed in a roughly 2x5x2 inch Radio Shack project enclosure box. When in operation, the transmitter was able to reach many radios in the entire house with a short 3 foot telescoping antenna. With a portable CD player as the audio source, the audio quality of the transmitted signal through many AM radio receivers were very good.