The input of the band pass filter is 30Khz triangular wave of 1Vpp. I want the band pass filter to allow 27k-33k Hz . I have used these values based on the formulas I found online for this band pass filter.

enter image description here

Unfortunately the problem seems to be , that at high frequencies eg. 60kHz it does not cut them off and the second problem that it reduces the 1Vpp to 500mV Vpp which I don't want either.

This is with input of 30khz(in green ) and band pass output (in blue) enter image description here

And this is with 60kHz input (same colours)

enter image description here

To me it seems they are the same, which is bad. What I am doing wrong , which values should I use

ps. the formulas for the values I used are on this site http://www.electronics-tutorials.ws/filter/filter_7.html

  • 1
    \$\begingroup\$ You're going to need a much higher order filter than what you have if you want to have good attenuation of a signal only 1 octave above the LPF cut-off frequency. \$\endgroup\$ – The Photon Jan 9 '18 at 19:00
  • \$\begingroup\$ @ThePhoton I am using a second order low pass filter after this band pass , but I am specifically told to use a band pass filter that does the things I mention in my question . There has to be a combination of capacitors and resistors that accomplish this \$\endgroup\$ – user170589 Jan 9 '18 at 19:06
  • \$\begingroup\$ Are you not allowed to use inductors? Remember we only know as much about your problem as you actually tell us. \$\endgroup\$ – The Photon Jan 9 '18 at 19:17
  • \$\begingroup\$ @ThePhoton No its only RC, I know I am trying to provide as much information to be clear \$\endgroup\$ – user170589 Jan 9 '18 at 19:22
  • 1
    \$\begingroup\$ "Low" is not an engineering specification. You need to have a quantitative spec like "at least 10x attenuation" or "at least 4x attenuation" before you or anybody else can design this filter. \$\endgroup\$ – The Photon Jan 9 '18 at 19:46

The filter you have is the simplest kind of bandpass filter. Really it's just a high-pass filter cascaded with a low-pass filter. So long as you keep within the op-amp's operating frequency range, there will be little interaction between the low-pass components and the high-pass components.

So to block the 60-kHz signal, you're relying on a first-order low-pass filter. A first order filter can only achieve 6 dB per octave roll-off, so you shouldn't expect more than a factor of 2 difference in the output voltage between 30 kHz and 60 kHz with this design.

If you want to provide more attenuation in the stop-band at one octave above the pass-band, you will need a higher order filter.

First, you need to decide your specification: What is the minimum acceptable attenuation at 60 kHz.

Then look for a design that can meet your specification. This might be a higher-order low-pass filter, or it might be a (higher-order) band-pass filter actually designed as a band-pass rather than low-pass plus high-pass.

| improve this answer | |
  • \$\begingroup\$ Honestly, the OP may need to use all 4 sections of a quad op-amp. 24 dB per octave may have to be good enough with analog filters. \$\endgroup\$ – user105652 Jan 9 '18 at 23:13
  • \$\begingroup\$ @Sparky256, now they say they only need 4x attenuation at 60 kHz, so they might not need to go that far. \$\endgroup\$ – The Photon Jan 10 '18 at 1:48
  • \$\begingroup\$ Well at least the quad package gives the OP that option, if needed. \$\endgroup\$ – user105652 Jan 10 '18 at 2:02

Your Answer

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