Regenerative circuit
The regenerative circuit (or self-regenerative circuit') allows an electronic signal to be amplified many times by the same vacuum tube or other active component such as a field effect transistor.
Description
Regenerative circuits were employed in early radio receivers, offering selectivity and sensitivity far beyond that available from a crystal radio receiver. Regenerative circuits were a great milestone in radio history. The secret to the regenerative radio receiver operation was the carefully controlled positive feedback. This also allowed the receiver to oscillate, allowing CW (Morse code) to be heard as beeps. The regenerative receiver is theoretically as sensitive as any radio can be, however, the adjustments are critical, and must be continuously monitored during listening. If misadjusted the receiver can transmit interference.
Regenerative receiver
The inventor of FM radio, Edwin Armstrong, patented the regenerative circuit (invented while he was a junior in college, and patented 1914), the Super-regenerative circuit (patented 1922), and the Super Heterodyne receiver (patented 1918). Lee De Forest filed a patent in 1916 that became the cause of a contentious lawsuit with the prolific inventor Armstrong, whose patent for the regenerative circuit had been issued in 1914.
The lawsuit lasted twelve years, winding its way through the appeals process and ending up at the Supreme Court. The Court ruled in favor of De Forest, although the experts still disagree about whether the correct judgement had been issued. The regenerative radio made the most out of very few parts. When the parts became easier to obtain, the superheterodyne receiver replaced it for all serious work. In 1906, Reginald Fessenden had already used a version of the heterodyne principle for broadcasting. The superheterodyne receiver is the most common receiver in use today.
Operating limits
Quality of a receiver is defined by its sensitivity and selectivity. For a single-tank TRF (tuned radio frequency) receiver without regenerative feedback, bandwidth = frequency/Q, where Q is tank "quality" defined as Q=Z/R, Z is reactive impedance, R is resistive loss. Signal voltage at tank is antenna voltage multiplied by Q.
Positive feedback compensates the energy loss caused by R, so we may express it as bringing in some negative R. Quality with feedback is Qreg = Z/(R-Rneg). Regeneration rate is M = Qreg/Q = R/(R-Rneg).
Obviously, M depends on stability of amplification and feedback coefficient, because if R-Rneg is set less than Rneg fluctuation, it will easily overstep the oscillation margin. This problem can be partly solved by "grid leak" or any kind of automatic gain control, but the downside of this is surrendering control over receiver to noises and fadings of input signal, which is undesirable. Note that modern semiconductors offer much more stability than vacuum tubes of the 1920s.
Actual numbers: To have 3 kHz bandwidth at 12 MHz (short waves travelling all around Earth) we need Q=F/f = 4000. A two-inch coil of thick silvered wire wound on a ceramic core may have Q up to 400, but let's suppose Q = 100. We need M = 40, which is attainable even without AGC, if amplification/feedback control is smooth enough.
Super-regenerative receiver
The superregenerative receiver uses a quenching oscillator to produce very high positive regeneration of the radio amplifying stage, while quenching or keying the built up regenerative oscillation at an ultrasonic rate. This further improves the gain of the receiver while simplifying adjustment.
On the other hand, a superregenerative system has an inherent contradiction limiting its use to relatively free and clear bands. Due to Nyquist's theorem its quenching frequency must be at least twice the signal bandwidth. But quenching with overtones acts further as a heterodyne receiver mixing additional unneeded signals from those bands into the working frequency. Thus the overall bandwidth of superregenerator cannot be less than 4 times that of the quench frequency, assuming the quenching oscillator produces an ideal sinewave.
Patents
- Armstrong, E. H., U.S. patent 1,113,149, Wireless receiving system, 1914.
- Armstrong, E. H., U.S. patent 1,342,885, Method of receiving high frequency oscillation, 1922.
- Armstrong, E. H., U.S. patent 1,424,065, Signalling system, 1922.
- Braden, R. A., U.S. patent 2,211,091, Superregenerative magnetron receiver, 1940.