Upstream I was introduced to this 1968 PE article. In addition to the phase shift osc used there, I found the bias arrangement used in that project interesting.
Figure 5 and elsewhere in the same project, there is an an interesting biasing arrangement between Q3 and Q4 (Fig 5). Or Q1/Q2, which just adds a fuzz gain adjustment.
(
Using Q3/Q4:
Basically it looks like the current is split between R19=10k and R16=22k + R15=10k.
In parallel that amounts to Rp=10k||32k => 7.6k.
Supply is +60V.
Assuming the base of Q4 is approx 1/2 of supply, say 30V, then it's emitter is at about 30-0.6 = 29.4V.
Then Iq4 = 29.4V / 7.6k => 3.9 mA
Due to voltage divider, Ir15 = 29.4v/32k => 0.92 mA
Vr15 => R15 * 0.92mA => 9.2V
Since the base current in R14 is very small, then Vbe3 ~= Vr15 = 9.2v.
But if Vb3 = 9.2V then Vr18 = 9.2 - 0.6 => 8.6v
Iq3 => 8.6v / R18 = 8.6v / 10k => 0.86 mA
Vr17 = 47k * Iq3 => +40 volts
So my crude calculations are off: I had assumed the voltage at Q4 to be about +30v, but when I do the rough calculations I get +40v instead. The answer probably lies between.
I'd like to know if this is a form of "bootstrapping", since the bias is determined by the 2nd stage.
Warren