For an FM discriminator of the type John is describing, all you need is a frequency-dependent phase shift and a phase detector to turn that into a voltage change.
You can do it with an RC lowpass, even, and RC lowpasses don't have any delay at all. (One way to see this is that you need a time machine to undo a time delay, but an RC highpass can undo the effect of an RC lowpass.)
Cheers
Phil Hobbs
See my patent...
Where I used an active filter to accomplish the phase shift, and a multiplier for the "phase detector". ...Jim Thompson
If you use a pure multiplier type mixer, you do need a phase shift like 90 +-delta, not zero +-delta. That works out, technically, to be a peaky lowpass filter, not a bandpass filter. Maybe that's the reference to needing a delay.
That's just a detail. If you have a resonator with the appropriate Q and stability, a little phase shift is easily added somewhere.
John
Sure, a real resonator does have group delay, but IIUC Brent thought that the delay was intrinsic to the problem, i.e. that a resonator-based discriminator was a variety of delay line, and it isn't--all you need is a phase shift.
And I could imagine applications where the zero-delay property might make the RC approach handy, mostly in really wideband things like countering jamming.
Cheers
Phil Hobbs
Have something to add? Share your thoughts — no account required.
Ask the community — no account required