Measuring inductance from y-parameters in simulation

[allanvv]

I'm calculating 1/(w*imag(y11)) to view inductance. However, using two different equations gets different results at DC:

Red curve: -1/(2*pi*w*imag(Y11)
Blue curve: 1/(2*pi*w) * imag(1/Y11)

Why are they different? The blue one is correct at DC and doesn't have have huge spikes.

 

[albbg]

-1/imag is generally different from imag(1/y). In fact y is a complex variable, we can represents as y=a+j*b then:

-1/imag = -1/b
imag(1/y) = imag[1/(a+j*b)] = imag[(a-j*b)/(a²+b²)] = -b/(a²+b²)

-1/imag = imag(1/y) if and only if a=0

since L=imag(Z)/(2*pi*f) and y=1/Z ==> L = imag(1/y)/(2*pi*f)

 

[allanvv]

Thanks. Now I'm trying to use the regular definition of Q as -imag(y11)/real(y11) to get Q at resonance. However, isn't imag(y11) = 0 at resonance? How do you evalulate Q this way?

edit: Nevermind, I figured this out.