Passive linear circuits

It is by definition not linear.

Also not linear.

John

"Steven" The last thing I want is to try to design something that has

Plus you want it to consume no power ?

Please return to your alternative universe immediately - Mr. Alien

..... Phil

As others have said, the response you require is NOT linear.

Consider a system with a transfer function H and input x and output y

y = H(x)

Say we know that the response to an input X1 is Y1 and the response to an input X2 is Y2. We now want to know the the output, Y, for an input, X, made up from inputs aX1 and bX2, where a and b are constants.

We want to know Y Y = H(X)

Y = H(aX1 + bX2)

If the system is linear then we can simply calculate the output for the two components separately and then add the outputs:

Y = H(aX1 + bX2) = aH(X1) + bH(X2)

Y = aY1 + bY2

For the response you require, this clearly isn't true so your system is by definition non linear.

As you have probably learned by now, no such "passive, linear" circuits exist. You might want to Google for "AMSAT Leila" and "AMSAT Transponder" for some practical advice on the state of the art regarding active RF signal level control and frequency translation.

Also this link describes hardware.

By definition something that outputs 3V for a 3V input and 0V for a 5V input isn't linear!

I'm coming into this quite late, though, so you've probably already been told this.

OK, are you sure this isn't a homework problem? Again, check the basic definition of "linear" -- there's no way to change frequencies in a linear system. This is usually the sort of thing discussed in the first handful of weeks of Signals & Systems 101 classes... I remember them giving you a bunch of input and output exercises where you had to determine whether the system was line and time-invariant... most are obvious, but a few are actually a bit tricky.

---Joel

A PLL comes to mind. of course, this would have to be a case were only one signal is present at a time.. The PLL could lock with in coming signal to track as a reference signal into a summing circuit where the circuit would be balanced to generate a NULL output or very little output when the amplitude and phase are correct.

Nope, that is not possible. One intrinsic property of linear circuits is superposition, which means the output corresponding to a linear summation of inputs is simply the linear summation of outputs corresponding to each component of the input summation. So for example your circuit wants the response to a 5V sine to be zero, but the response to a 2.5V sine to be a 2.5V sine, then linearity requires the response to 5V=2.5V + 2.5V to be 5V and not 0V.

Again, this is not possible, a violation of superposition, due to the fact that sin(A+B)=sin(A)cos(B)+ sin(B)cos(A) and not sin(A) + sin(B).