interpreting magnitude of HSPICE's FFT-output of mixed signals

[prasfft]

I ran a time-domain simulation in HSPICE & I am trying to analyze the frequency content of the transient current draw(FYI- I have the power systems "impedance vs. frequency" curve, and my aim is to determine the noise voltage on my power rail given its impedance now)

So I run FFT (current-waveform,t_start, t_stop, NP) using the in-built calculator...

Now how does one interpret the resulting spectral density curve?

Is the magnitude (on Y-axis) the real amplitude of current at that frequency i.e. can I just take that amplitude and multiply it by impedance Z at that frequency to calculate the noise voltage...??

I'd highly appreciate if some one can lend insight into this...thanks a bunch

 

[zappedagain]

I've seen a lot of variation to that answer. You really need to set up a test case where you know what the answer will be (X amps at Y hertz, and then V amps at T hertz, where X,Y,V, and T are variables you control) and run an FFT. You can pull the scaling back from there.

dBA - ratio against one ampere

dBFS - ratio against full scale of your A/D (more real world)


Z