FFT in LT-SPICE

Hello everyone,

I am simulating the following circuits:

I run the transient simulation and then I want to see the FFT of the signals V1 nd V2. The results are shown below:

I am quite sure that the noise that I see in the spectrum of v2 is introduced by LT-SPICE. Although, this noise is something unwanted is quite clear that is an artifact of LT-SPICE, because it is much lower than the noise the I would expect from the resistors in the circuit. HOwever, it does not bother me too much because it is so low that I can recognize the source of this noise. HOwever, for the signal v1 the noise is much higher. Why is that? The two circuits are identical except for the presence of a capacitor. Can a capacitor increase that much the noise floor? 

Can someone explain me why the presence of the capacitor increases the noise floor? thanks Slight smile

  • Hi master89,

    What happened here is that the voltage divider (right circuit) became an RC filter (left circuit) when a 1nF capacitor is added. The capacitive reactance (Xc) of C1 is now paralleled with R2 and affects the voltage divider voltage across frequency. Ideally, we expect v2 (orange trace) to be half of V2 (green trace) across frequencies. We expect that the addition of a 1 nF capacitor changes the frequency response of v1 (blue trace) and became attenuated after the cutoff frequency of the RC filter.

    What happened in LTspice simulation is strange as we expect that the noise floor will not increase that much. It seems that V1 is a fixed voltage across frequencies as shown in the red trace below that rolls off after the cutoff frequency.

    We will continue to look into this.

  • Additional notes I got about this topic.


    The circuit needs time to settle. Look at the beginning:


    This initial settling process must be excluded from the FFT. One way to do this is to say so in the FFT dialog window:


    Result: