The impedance of the parallel LC goes to infinity...

[T-Dogg]

"...at the resonant peak."

I'm flipping through H&H, trying to "get my learn on", and this statement simply seems counterintuitive to me...

I see a bandpass circuit like that and think to myself, "well, the impedance must be highest at the points farthest away from the resonant frequency, since their output values are significantly attenuated, and lowest at the peak, since it passes the most signal at this region." Resistance being a close kin of impedance, I'd think if resistance was higher within a particular frequency range, the volume of that frequency range would be quieter... Anyone mind straightening me out here? :roll:

 

[bcarso]

Energy is being stored in electric fields and magnetic fields, and at resonance exchanged back and forth. No real inductor or capacitor is free from loss---that is, something that converts the electromagnetic energy into heat---so the impedance never really reaches infinity. But the losses can be quite small.

There is probably a web animation somehwere that will show the way the exchange works on a moment-by-moment basis.

There are also mechanical analogies that may help, like a pendulum swinging back and forth in a uniform gravitational field, exchanging kinetic energy for potential energy.

 

[PRR]

> I'd think if resistance was higher within a particular frequency range, the volume of that frequency range would be quieter...

A part can be in Series or in Shunt.

If we put an infinite impedance between a preamp and a power amp, indeed we get a quiet signal.

If we put an infinite impedance in shunt with the preamp input, there is no effect; whereas if we stuck a low impedance in shunt the signal would be very weak.

In the classic bandpass, an AM radio IF filter, the tank is in shunt between a high-high impedance source (pentode plate) and a high-high impedance load (pentode grid). At low frequency, the 1mH coil has low impedance, shorts-out the signal. At high frequency the 100pFd cap is a low impedance, shorts-out the signal. At 450KHz, either coil or cap alone would have about 3K impedance, stage gain about 3. However coil and cap together slosh energy back and forth, rejecting external energy.

Impedance ideally rises to infinity, though 30 ohms copper loss in coil will limit it to only 300K peak impedance. Stage gain is maximum and would be 300, though other losses make it more like 150 from 445KHz-455KHz, falling fast off-resonance.

 

[T-Dogg]

Thanks! Need to digest that a bit, but I think I some of the confusion has been cleared up.

For one, I always visualize in series... but LC is not in series with signal (almost had to ask what shunt was, lol), so the signal takes the path of least resistance when within LC peak resonant frequency range...

The above scenario would seem to infer that energy is being dissipated when a signal is within the LC's low impedance range -- but it is stored in the electrical field of the capacitor or in the magnetic field of the inductor... The exchanged back and forth thing is still fuzzy, but I'm ok with that, lol...

For better or worse, that's where I'm at now!

 

[clintrubber]

Don't have the info I'd like look now in over here, but IIRC it goes to Q^2*R.

 

[PRR]

> almost had to ask what shunt was

The absolute value of impedance is almost never very important; most real work gets done with ratios of impedances. Everything is a voltage divider! Or at least, most things can be understood as voltage-divider action.

> energy is being dissipated when a signal is within the LC's low impedance range

Yes. The tank on a 50KW transmitter gets HOT.

> Q^2*R

I may have lost a factor of 2; good to know but not essential to the basic understanding. (I also used thumb-math for the reactance values, so they are off by many percent.)