filter order

Actually, it's passive if it doesn't spit out any more energy than what you put into it. Or, closer to what you're saying, it can't amplify *power*. But plenty of passive circuits amplify voltage or current (built using just RLCs even -- no transformer needed).

"The usual "passive" components of catalogs, however, are exactly the linear components."

Well... transformers and inductors are purposely used outside of their linear ranges often enough that I'm not sure I'd write it that way.

The high-voltage guys occasionally use resistors outside of their linear ranges too.

I suspect you could make a (relatively-cruddy-but-largely-HEMP-proof?) mixer out of magnetic amplifiers, if you really wanted to.

---Joel

ts

s the

But the limit is on the order of the filter. You can play games and get a lower order. All the theorem does is tell you the order can be no higher than a limit, but it could be lower. This is a case where you really need to read the statement carefully.

That's not correct, it's correct for an R but not for an L or C, e.g. if I double the voltage over an L the current will increase linearly for L and C there is a first order differential equation which describes the relation between voltage and current. An electronic network is linear when it obeys the rules of homogenity and superposition, when you only use ideal Rs,Cs and Ls you always get a linear network.

regards, nukey

Look at that equation carefully: it specifies frequency dependence, NOT amplitude dependence. Double the amplitude of a voltage-source AC input signal, and the current through your L or C is going to double. The phase of the current remains the same as before the source changed.

But that's of course something different, in the preceding post you did't mention you were talking about an AC-source, you just mentioned doubling the voltage, that is a Heaviside step function. What you tell now is indeed another characteristic of linear circuits when excitated by a sinusoidal function (once the transient phenomena have disappeared).

regards, nukey

But that's of course something different, in the preceding post you did't mention you were talking about an AC-source, you just mentioned doubling the voltage, that is a Heaviside step function. What you tell now is indeed another characteristic of linear circuits when excitated by a sinusoidal function (once the transient phenomena have disappeared).

regards, nukey

That would make all the statement nonsensical; there's no well-defined number for the response to a step function. There's also no well defined answer to the DC current source into a capacitor or DC voltage source into an inductor; sources are always AC for R-L-C circuit analysis, in my experience.

Of course there's a well-defined answer. It's y=e^-t or something like that. With a cap, the voltage increases to the power supply; with an inductor, the current increases until it reaches the power supply current limit, or until the DC resistance of the coil limits it, whichever happens first.

Hope This Helps! Rich