Autodyne

The autodyne circuit was an improvement to radio signal amplification using the De Forest Audion vacuum tube amplifier. By allowing the tube to oscillate at a frequency slightly different from the desired signal, the sensitivity over other receivers was greatly improved. The autodyne circuit was invented by Edwin Howard Armstrong of Columbia University, New York, NY. He inserted a tuned circuit in the output circuit of the Audion vacuum tube amplifier. By adjusting the tuning of this tuned circuit, Armstrong was able to dramatically increase the gain of the Audion amplifier. Further increase in tuning resulted in the Audion amplifier reaching self-oscillation.

This oscillating receiver circuit meant that the newly produced CW or continuous wave transmissions could be aurally decoded. Previously only spark, ICW (Interrupted Continuous Wave) signals which were produced by a motor chopping or turning the signal quickly on and off at an audio rate, or modulated continuous wave (MCW), could be decoded.

When the autodyne oscillator was advanced to self-oscillation, continuous wave Morse code dots and dashes would be clearly heard as a series of beeps, instead of an all but impossible to decode series of thumps. Spark and chopped CW (ICW) had audio tones which didn't require an oscillating detector to extract from the radio signal. Such a regenerative circuit is capable of receiving weak signals, if carefully coupled to an antenna. Antenna coupling interacts with tuning, making optimum adjustments difficult.

Early transmitters emitted damped waves, which were decaying sinewaves at the transmission frequency. Bursts were made at an audible frequency. The damped waves were a result of the available technologies to generate radio frequencies. See spark gap transmitter. The transmitters could be keyed on and off to send Morse code.

Receivers could be made with a tuned circuit, a crystal detector, and a headphone. The headphone would respond to the detected bursts, and the operator could copy the Morse code. The received signal was not a sinewave. Instead of a crystal detector, a Fleming valve (tube diode) could be used; it was a stable detector, but not very sensitive. Even better was a using a vacuum triode because it provided some amplification. The regenerative receiver supplied even more gain, but required careful adjustment.

...
Wikipedia