0
\$\begingroup\$

Circuit is for 2nd order band stop filter...from 1.2 kHz to 6.2 kHz.... But the circuit is not working as shown in bode plot... Cutoff frequency is also not coming right at -3dB ....... What changes are required....?enter image description here I have tried by varying resistance at load....but nothing happens...

\$\endgroup\$
1
\$\begingroup\$

This is not how you design a band-stop filter.

  • First start with specs for Pass-band -3dB points and ripple
  • Next define acceptable Band-stop rejection thresholds at lower and upper f.
  • The slope of these above specs implies n required for nth order LC filter required or RC filter required. ( e.g. cauer, chebyshev active filter IC)

  • Alternative way is design n high Q, RC notch (twin-T)filters, cascaded, but this is suboptimal

  • you cannot simply cascade 1st order LPF and expect good band-stop results by trial and error then take difference of LPF and BPF
  • reason is; 2nd order filter is up to 12dB/octave and only 6dB in 1st octave and your f2-f1 spread is only 4 octaves
\$\endgroup\$
1
\$\begingroup\$

The idea behind this filter is that one op-amp provides +90 degrees phase shift at the cut-off whilst the other provides -90 degrees at its cut-off. Somewhere between those two points (different cut-off frequencies) there should be two sinewaves that are exactly antiphase and equal magnitude therefore they will cancel in the linear mixing stage. This is difficult to achieve with filters operating at different cut-offs FYI.

One low pass network of 1326 ohms and 100 nF gives a cut-off of 1200 Hz. One high pass network of 100 nF and 256 ohms gives a cut-off of 1608 ohms. That's for sure. However, when you cascade two passive networks you don't get double the phase shift at the cut-off previously calculated because the 2nd network loads the first and the numbers come out different.

Why don't you just convert both LP and HP filters to sallen key types. To make life easier make the cut off frequencies the same for the high-pass as the low pass then at one particular frequency both will be phase shifting by 90 degrees (in opposite directions) and have equal amplitude. This has to mean that there will be a significant cancellation and deep band-stop filtering due to equal amplitude but exactly opposing phase shift.

\$\endgroup\$
  • \$\begingroup\$ S&K filters will improve response but , user should learn to start with specs. \$\endgroup\$ – Sunnyskyguy EE75 Sep 21 '16 at 15:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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