A Bode plotter limitation in SPICE

Article By : John Dunn

There is a SPICE booby trap waiting for the unwary user of the Bode plotter.

MultiSim SPICE has a variety of measurement instrument simulations, one of which is the Bode Plotter, which will yield Bode plots of whatever circuit you assign it to such as the following. Please note that the analysis result does not depend on the drive frequency of V1.

Figure 1
Bode plotter at work (two examples)

When we look at two examples of the plotter examining identical circuits, their Bode plots are exactly alike, independently of the different frequencies of V1, which is 1 kHz, on the left and 100 kHz on the right.

To make the Bode plotter work, you need to provide some source of excitation, such as the V1 source shown here. Note that it does not matter what excitation frequency you choose for that source. The Bode plotter will simply examine the circuit itself and come up with its Bode plot results with no regard to the chosen excitation frequency. In both cases above, the analysis frequency range is 0.001 Hz to 1 GHz.

This is great, but there is a SPICE booby trap waiting for the unwary user of the Bode plotter. Please consider this zero-frequency IF circuit from my Living Analog blog entitled, Zero frequency IF.

Figure 2
Bode plotter failing to work

For the input V3, at the same frequency as the local oscillator represented by V1 and V2 at 1 MHz, the signal output at the junction of R3 and R4 replicates the signal input. However, the Bode plotter does not respond to this.

The reason is that the balanced mixers A1 and A2 each deliver two outputs, one at the sum frequency of 1 MHz plus 1 MHz which equals 2 MHz and the other at the difference frequency of 1 MHz minus 1 MHz which equals zero Hz (hence the name zero frequency IF). The 1 MHz inputs that each mixer receives are totally suppressed at the mixer outputs.

Therefore, since the signal from V3 is not present at the outputs of A1 and A2, the attenuation from V3 at 1 MHz to the mixer outputs is infinite.

The fact that we later construct a replica of V3 at the junction of R3 and R4 means nothing to the Bode plotter and therefore we do not obtain a usable Bode plot result.

As far as the Bode plotter is concerned, the input V3 never makes it to the output and thus, there is no display on the screen, even though a 1 MHz output is actually being delivered in response to the 1 MHz input.

Forgive me, but I cannot resist saying that this does not bode well.

John Dunn is an electronics consultant, and a graduate of The Polytechnic Institute of Brooklyn (BSEE) and of New York University (MSEE).

Related articles:

Leave a comment