2
\$\begingroup\$

I recently design an OTA based filter, and simulate it with LTspice IV. I was surprised to see that the gain vary with frequency (of course) but also with the input amplitude of the signal.

So I decided to simulate a basic RC filter to see how if the gain change with input amplitude, and the same thing happened.

Here is the schematic, with a variable AC amplitude (0.1, 0.2, 0.5, 1) :

enter image description here

And here is the plot :

enter image description here

I don't understand, the frequency response of a filter characterized by a transfert function V(out)/V(in) should vary only with frequency, and not with V(in), isn't it?

\$\endgroup\$
2
  • 4
    \$\begingroup\$ But you aren't displaying V(out)/V(in) in your plot; you're just displaying V(out). If it DIDN'T vary with V(in), I'd be worried! \$\endgroup\$
    – Dave Tweed
    Aug 2 '16 at 16:50
  • \$\begingroup\$ Sorry for this question but how do you display Vout / Vin ? Thank you \$\endgroup\$
    – Gab
    Aug 2 '16 at 16:52
3
\$\begingroup\$

If you measure V(V1) you will see that said node will also drop in amplitude, and of course remains constant over the frequency sweep being an ideal AC voltage source. Therefore the simulation is correct.

The graph you've posted in the question is actually not the transfer function V(out)/V(in); it is just the absolute output amplitude. Unfortunately LtSpice does not explicitly say the units are dBV.

To display the transfer function, you can e.g. measure V(out) and then right click the label (on top) in the graph. This opens a dialog that allows you to edit the line formula and color. You can than change it to:

V(out)/V(V1)

This should show the result you were expecting.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you, I tried it and now I got exactly what I expected ! \$\endgroup\$
    – Gab
    Aug 2 '16 at 17:07
5
\$\begingroup\$

The "gain" you see in the plot is always referenced to an AC amplitude of 1. Since you're changing the amplitude, it's not strange that the amplitude shown is different. This becomes more obvious if you open up the settings for the plot window and select Linear instead of Decibel.

enter image description here

Try measuring directly on the output of the AC source - it will show the same magnitudes as you see for 1 Hz in your filter.

Remember that the AC simulation mode is a bit special, in that it is doing a "small signal analysis". You won't get effects from non-linearities such as clipping due to excessive voltages.

\$\endgroup\$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.