The foster seeley FM demodulator is a complicated circuit , and I've been trying to understand its working in detail. Most explanations are based on the phasor representation, but they have'nt given the proper conditions under which they assumed. My question is , can phasor diagrams be applied to a non linear circuit with diodes ? The circuit resembles a full wave rectifier. Also can anyone elaborate on which all capacitors/inductors are short or open or reactive to these three frequency components? 1. DC 2.Message frequency (kHz) 3.Radio (modulating) frequency. (MHz)
If you look at the expressions for inductive and capacitive reactance you can use the process of taking limits to answer your questions: Example: DC is equivalent to saying that ω→0 Inductive reactance is given by ω⋅L, and the limit as ω→0 is 0. So an inductor looks like a short at DC Capacitive reactance is given by (1 / ω⋅C), and the limit as ω→0 is ∞, so a capacitor looks like an open at DC
A Foster-Seely discriminator was designed to output a DC control voltage. "Electronic Communication" second edition by Shrader, has an excellent section and analysis of this circuit. A 1920s version of industrial control wifi. There are different ways to look at circuits. That choke might look like a parallel winding of the primary coil. That might make the center of the secondary in phase with the primary. If the carrier goes off frequency, we'll get a difference. I highly recommend this textbook. It's old and most of the analysis is tube circuits, but I have not found a better text for electronics.
I had the privilege of talking to Bob Shrader a couple of times before he passed away. His book was definitely a standard. eric
One point I've noted in playing with the design, that is not immediately obvious from Foster-Seeley schematics, is that the mutual coupling between the primary and secondary tuned circuits must be relatively small for the detector to work. A unity coupled transformer will definitely not work. Coupling coefficients might need to be as low as 0.1 for instance. I'm not sure if there is an optimum design value for the coupling. A useful analysis can also be found in the attached pdf as Appendix A5.2.