Just so you know, I breadboarded both circuits exactly as shown, over the Labor Day '99 weekend. Both circuits performed as designed when powered from a +/-12VDC filtered, regulated, and bypassed power supply, showing good agreement with the Excel spreadsheet. I used 15k DIP resistor networks along with 10nF capacitors to obtain a center frequency of around 1100 Hz. I used Texas Instruments TL082 dual opamps, a plug-in replacement for the LF353.
High-pass, band-pass, and low-pass outputs and inputs worked as designed. Disconnecting the 'b1' signal resulted in a notch filter with more than -40 dB of attenuation at the center frequency. Inverting 'b1' with all three signals connected resulted in an all-pass filter, which spins output phase through 360 degrees, while maintaining a flat amplitude response. I observed that DC gain is unity for the low-pass signal, while it is zero for the band-pass and high-pass signals. If you AC couple the input, then all of the opamp outputs sit nicely near the ground reference established at the opamp non-inverting inputs. I also listened to music through the filters, and found the sound a little bit peaky, indicating that filter 'Q' should be lowered for audio use.
Last updated 5/31/2000 by
