The circuit in Figure 1 was given to me some while ago as a three-pole, active 1 dB Chebyshev lowpass filter. I never confirmed that its transfer function complied with the Chebyshev polynomial, but the SPICE simulation looks pretty Chebyshevian, so I guess I was told correctly. Just to mention, the -3dB point of this filter is 3 kHz.

This circuit is almost a zero offset configuration except for the effect of the op-amp's input bias current flowing in the three resistors.

Chebyshev lowpass filterFigure 1 This circuit was given to me as a three-pole, active 1 dB Chebyshev lowpass filter.

An important point to note is the passive RC pair of R1 and C1 in Figure 1. Recalling from part 2 of this series, this RC pair protects against placing a greater demand for speed from the op-amp than it might be able to provide.

Take note as well that this circuit is sensitive to a load impedance such as R8 going in parallel with C3 (Figure 2). This 10 Meg load shifts the Bode plot a little bit, but maybe such little shifting is okay for most purposes.

Chebyshev lowpass filter loadingFigure 2 This 1 dB Chebyshev lowpass filter includes loading.

If we want true zero offset, we can add a DC block to the op-amp input which yields the result of Figure 3.

Chebyshev lowpass filter DC blockFigure 3 This 1 dB Chebyshev lowpass filter has a DC block.

Again, we see a little shifting of the Bode plot from the original Chebyshev response, but the op-amp's contribution to DC offset really is zero.

Chebyshev lowpass filter loaded DC blockFigure 4 This 1 dB Chebyshev lowpass filter is loaded and with DC block.

If we do both, if we load and we add the DC block as in Figure 4, the Bode plot shifts a little bit more, but still not too much of a shift for most purposes.

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