The explanation of the Armstrong regens is that the signal is fed back to the grid and re-amplified. Thus regenerative. A secondary was placed in the main coil which inductively fed back the signal into the triode tube. In fact, the Super regen discussed the need for "quenching" and Armstrong solved this too. He place a second tube which was oscillating at just above human hearing(obviously so we won't hear it) which shuts down the main tube to "reset" it. The explanation is that the original signal does not disappear and "congest" the amp.
Recently came across a single FET FM RX which has a free running vibrator made of two transistors which also "quenches" the FET.
But more recent articles I came across (Talking Electronics) described many of its transistorized "regens" quite differently. It says that the transistor with a tank is made to oscillate very low power at a certain frequency. The feedback to oscillate it is capacitive, a cap between the C and the E of the transistor.
Once a signal comes in of the same frequency, the receiver oscillator gets affected ( becomes difficult to oscillate it says) and the voltage of the coil to the rail plus line changes in accordance to the varying strength of the incoming signal (AM). The changing voltage is amplified by an audio amp.
Nowhere does it talk about the received signal amplified and re-amplifed and thus does not talk about the need to quench the main osccillator. In a way, the signal does not enter the transistor at all as in the explanation of Armstrong regen. Sounds to me like when a stronger TX jams a weaker TX. The change in current draw reflects the signal but the signal is never re-amplified.
I came across another explanation that "noise" initiates the Armstrong regen to oscillate. Well, in that case it is also an oscillator. And a signal appears in the surrounding, it gets affected and is picked up and fed to an amplifier.
It is interesting cuz I am also reading on Direct Conversion Receivers using the famous SA602/612 originally NE by Signetics. Somehow similar. A local oscillator frequency is mixed with an incoming signal and any differential in frequency within the human hearing is heard (for CW). For SSB and AM, it says that when the frequencies are the same, known as zero beat, the IC extracts the sound. (See John Dillon's Neophyte DC receiver). The latter part on detecting SSB and AM sounds to me like the local oscillator gets suppressed when an incoming signal of the same frequency appears. The change in current draw is amplified instead of the differential frequency heard.
Would like to hear from Masters around here. Thanks.