No varactors in AF circuit design.
OK. I am an extreme n00b to this forum (1st post!) and I'm just learning a bunch of stuff about electronics I never thought I'd ever grasp.
That said...
No to varactors in AF circuit design. That's an anti-yes. You cannot ever get the capacitance to ever be in a usable range. On paper? Sure. In reality? I want to see your device when you're done with it because it's gonna be like the biggest ball of twine in Minnesota and I want a picture of it with me and my g/f standing next to it for scale.
I'm interested in the same types of filters because I want to build a tube-based semi-para EQ for a tube-based preamp, and I want to build with 0 components invented after about 1960. Why? Because it's cool, and I'm sick to death of the companies that make them going out of business or taking cool stuff off the market JUST as I am ready to afford them. OK.. introduction complete.
phase angle for walter just once in terms of complex impedance:
XL (sub L) = 2*pi*f*L
Xc = -1/(2*pi*f*C)
XL = inductive reactance, always positive in the RLC plane.
Xc = capacitive reactance, always negative in the RLC plane.
vector for inductive reactance: <R, XL>
vector for capacitive reactance: <R, Xc>
Note that Xc is always negative, and don't bother yourself with phase angles greater than +/- 90. Why?
Add the vectors. That will give you the complex impedance at that frequency. If the impedance is purely resistive, i.e., you have 0 reactance (i.e., <100k, 0>) you have the resonant frequency of the circuit.
see also:
f = XL/(2 * pi * L) [someone check this, this should be the resonant frequency when XL is 0, do the same thing and solve for f in the capacitive reactance equation, ~or~ *combine* the equations and solve for f.. this is only algebra and not really that difficult]
At other frequencies where there is significantly positive or negative reactance, you have attenuation at that frequency, i.e., you have a phase shift that adds negatively to the amplitude of the waveform at that frequency. Attenuation. Doesn't matter if there's an active element in the circuit or not. Powered non-attenuation at any given frequency = boost at that frequency.
ok, the phase angle:
take the vector resulting from adding the XL and Xc vectors. Atan it.
If you have a vector that looks like <50, 25>:
phase angle (radians) = atan(25/50) (yes, that's Y/X. The resistance, X, can *never* be 0 enough to invalidate the atan. Ever. Such a thing does not exist in real life because you can never have 0 ohms resistance ~[EDIT: because you still have to have an R in an RLC circuit no matter what superconductor you use! (see below
so you'll have to add a resistor even at -90 Kelvin which makes the whole superconductor idea moot in the first place (ty I'm laughing) though the resistor probably just jumped from a 1/2W to a 50W without changing the physical device because it'll never get hot enough to smoke!].)
phase angle in degrees = a * (180/pi)
Ok. That will give you your phase angle at any f for any RLC circuit. Play with the numbers, and for god's sake get at least a TI-84. If you formulate your graph right, you can actually get a graph of your frequency response curve. See the calculator manual for how to do that, I'm not going to go into it here.
BTW... 180 degrees is pi radians. A circle is 2*pi.
OH YEAH... why no bother with greater than +/- 90 degrees? Because atan() only works on the front side of the unit circle, i.e., +/- 90. The left side is the susceptance plane and you generally don't need to worry or care about it... that's working the RLC plane inside-out. You CAN do it, but ... why? Any phase angle with greater than 90 degrees of lead/lag is severe attenuation and you will likely not get crap out of your amplifier there. 180 degrees of shift is 100% negative feedback (with the same amplitude, anyway) and will 100% cancel the signal at that frequency.
I would like, at this time, for anyone to (please, I need it for my own understanding!) poke a hole in this. I literally learned this on my own last week. Those calculus classes came in handy... lucky I haven't needed them yet!
Work out the same stuff with a varactor and you'll quickly see that it *just doesn't work* even on paper.