I was using the Analog Devices design Wizzard in order to design a band pass filter with theese characteristics:

Gain : 23 dB

Passband: -3dB at 15 KHz

Stopband: -40dB at 40 KHz

The filter has 4 stages (order 8), the first two stages look like a common Multiple feedback Band-pass Filter, the problem is that the last two stages look diferent:

enter image description here Does anyone knows what is this topology?, for me looks like a sallen–Key topology for low pass filters, but I was not able to find any reference on Internet about multiple feedback band pass filters that use resitors (R4C R3C and R4D R3D) on the positive input of the operational amplifier.

  • \$\begingroup\$ Depends what kind of BPF shape . Try TI’ filter wizard \$\endgroup\$ Jun 18, 2018 at 23:08
  • \$\begingroup\$ A discussion is found on page 479-480 of "Analog Electronic Filters: Theory, Design and Synthesis" by Hercules G. Dimopoulos, 2012. For example. \$\endgroup\$
    – jonk
    Jun 19, 2018 at 5:01
  • \$\begingroup\$ In case you don't have the book @jonk suggested, it's a modified multiple feedback. \$\endgroup\$ Jun 19, 2018 at 6:06
  • \$\begingroup\$ Or just go here. \$\endgroup\$
    – jonk
    Jun 19, 2018 at 6:22

2 Answers 2


The shown circuit is one of the classical bandpass topologies - multi-feedback (MFB), however, with a modification proposed by Deliyannis. This modification consists of some additional positive feedback. The circuit can never "latch" because for DC we have 100% negative feedback.

Advantage: Without positive feedback relatively high Q-values (high selectivity of the bandpass) require a high component spread and/or a very large midband gain. This can be avoided using the shown positive fedback path. This modification was proposed by T. Deliyannis in 1968 (Ref. Electronic Letters, vol. 4, page 577).

Remember the Sallen-Key principle, where a Q-enhancement is achieved solely by using posive feedback (using a fixed positive gain amplifier).

Comment: It can be shown that this Deliyannis-modification leads to the best trade-off (for high Q values) between moderate midband gain and low component spread. A very large component spread is to be avoided because it either leads to an unwanted low input impedance or very large resistor values in the vicinity of the idealized opamp input resistance. Furthermore, excessive midband gain would challenge the slew rate of the amplifier as well as the input drive capability of the succeeding stage.


The positive feedback will cause the opamp output to latch to one of the supply voltages unless the input signal level is very high. Then the output is a squarewave. If you want gain then add another opamp with negative feedback to give the gain.

  • \$\begingroup\$ It won’t latch up with the ratios given, only boost the Q \$\endgroup\$ Jun 19, 2018 at 3:09
  • \$\begingroup\$ It looks like you judged the circuit "at a glance". Things can go wrong this way. \$\endgroup\$ Jun 19, 2018 at 6:06
  • \$\begingroup\$ No - the output will not "latch". The shown circuit is a MFB structure with a marginal positive feedback. That`s all. See my detailed answer. \$\endgroup\$
    – LvW
    Jun 19, 2018 at 6:39
  • \$\begingroup\$ The circuit can NEVER latch because it has 100% negative feedback for DC. \$\endgroup\$
    – LvW
    Jun 19, 2018 at 8:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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