1

I am trying to make a notch filter which blocks 500Hz, I beleive it should look like the circuit below. Any idea how can I find the proper values to make it not to pass 500Hz? I tried to give random values in LTSPICE but no success yet!

enter image description here

Dumbo
  • 3,274
  • 14
  • 45
  • 69
  • If that one you show is a bandpass, swapping the feedback network with the input network should give you the band stop. That is because the feedback network inverts the transfer function. – clabacchio May 15 '12 at 15:49
  • Have you tried with the capacitor and resistor values from stevenvh's answer? – Federico Russo May 16 '12 at 09:19
  • Not yet, I'll do soon and report back. I know that the steven's circuit is what I need but I am forced to use my own circuit to achive a notch filter (this is an assignement) – Dumbo May 16 '12 at 12:11

2 Answers2

3

I've never seen this circuit before, and I haven't analyzed it. In any case your opamp doesn't have DC feedback, so any input offset will make the output clip to the rails. See also this question.

You don't need an opamp for a notch filter, they're often made as a twin T-filter:

enter image description here

The center frequency is

\$ f_C = \dfrac{1}{2 \pi R C} \$

The disadvantage of this is a low Q. The following active twin T-filter has adjustable Q:

enter image description here

Community
  • 1
stevenvh
  • 145,832
  • 21
  • 457
  • 668
2

TI offers an op amp design guide (pdf, link below). Ref Section 16-6 for band-rejection. As with most things, using an op amp gives you more design freedom than a passive circuit.

http://www.ti.com/lit/an/slod006b/slod006b.pdf

Art Brown
  • 471
  • 3
  • 7
  • 1
    References are nice, and the document you refer to is an excellent one. But I think you should add more information in the answer itself. – Federico Russo May 16 '12 at 08:03
  • @FedericoRusso, Sorry, I'm no expert on these applications. I just happened to know of the reference. – Art Brown May 16 '12 at 16:44