Some super-regeneration circuit questions

I was obsessed by this circuit for a long time, I studied the super-regeneration ciucuit a lot but still have the below questions, I really hope you can help me out from them; thanks a lot!

Please refer to the attachment for the circuit details



l What?s the purpose of C7? Give the feed back from E to B so that to decrease the loop gain? How about the function of the L2 and R6 then?

n Recently, I found an issue that when using a new inductor L1, the sensitivity will be 10 dB less than the spec (L1 was adjusted to the best status); the test date of the inductor(inductance and Q) between this one and the previous one are almost the same, all are within the inductor?s spec; then, if I change the C7 to 82p, I can then get the sensitivity increased to meet the RSC spec; or, if I change the inductor from the new one back to the previous inductor, I can also get the sensitivity increased to meet the RSC spec?I can?t imagine the possible reason for the inductor L1!

l Why the R7 didn?t connect to the point between L2 and C7?

l Why use an extra R7 and C18? Why not just connect a capacitor in parallel with R6 to GND?

I am sorry- even if I wanted to try, there is no attachment.

R7 and C18 are a low pass filter to recover the modulation. L1, C7 and C8 are an oscillator circuit, only its not quite oscillating, I suspect that with L2 they might be oscillating at some lower frequency say .1 -1 MHz. The purpose is to allow the apparent "Q" of the L1 to be several thousand, with its increased voltage multiplication. The trouble is, that this is not a stable position, its liable to burst into oscillation at its tuned frequency, so often the oscillations are "quenched" that is they are stopped periodically by making the circuit oscillate at a much lower frequency BUT much higher then the modulation frequency.
Frank

Dear Vampire:

After a while the attachment appeared. The circuit around Q2 is the superregenerative oscillator which should oscillate at two frequencies.
One is at VHF, determined by L1 and C6, with a VHF feedback by C8.
The second, interrupting frequency, is determined by L2 and C7/R6.
R7 and the following capacitor are a low-pass filter rejecting the interrupt frequency while passing audio to output.
I do not exactly understand the purpose of Q3 the circuit of which is not shown.

You certainly can connect R7 to the other side of L2. If the circuit does not operate as you wish, I would try to adjust the transistor bias by the base resistor divider, or maybe better, by varying R6.

In some cases it may be difficult to force one poor transistor to oscillate at both frequencies. Then you can make a separate ~100 kHz oscillator and couple its output e.g. to Q2 base, to force it to interrupt.
The superregenerative detector, due to interruptions, starts to oscillate whenever an external signal appears. At R4 you should obtain pulses at interrupt frequency with a varying width due to the input signal. Usually the RF/audio gain is up to one million; this is why it is liked.
I would also rather connect C5 to Q2 base for a better operation.
Good luck!