Hello everyone

I'm working on a project where I need to design a few bandpass filters, however their bandwidths are really small.... like around 20Hz.
I tried simulating a passive bandpass filter by cascading a lowpass filter and a highpass filter.........
but it didn't really work even though my cutoff frequencies were set to 420 and 460hz respectively..... signals with frequencies between 400 and 500hz are produce almost the same output no attenuation happens.
Furthermore my input is a 5V signal no more, is there any way to strongly attenuate frequencies other than 420< f <460.....
I mean is it possible....... by making a filter.... if so what would the easiest way be?

Thank you all

 

Your expectations exceed your grasp of filter design basics. As you have discovered it takes a filter with more poles than you had in your experimental design to achieve a steep rolloff in the transition band. The proper approach is to specify the minimum allowable attenuation in the passband, the maximum allowable attenuation in the stopband and the width of the transition band. Only by taking this approach will you be able to match your experimental results with your expectations. I recommend:

https://www.amazon.com/Analog-Filte...qid=1493570616&sr=8-4&keywords=Van+Valkenburg

A bargain at $9.61 plus shipping.

 

https://forum.allaboutcircuits.com/threads/4th-order-mfbp-filter-design.61224/

 

Below is the LTspice simulation of a single-stage multiple-feedback bandpass active filter.
I calculated the values here for a Q of 40 and a center frequency of 450Hz.
For some reason I had to put in a center frequency value about 10Hz higher than the desired center frequency of 440Hz.
Pot U2 is to tweak the center frequency to the desired value to compensate for component tolerances.

The simulated rolloff is nearly -12dB at 420Hz and 460Hz with a peak of +1.2dB at 440Hz.
Is that sufficient for your purposes?

Edit: If you need more rolloff you can cascade them as Joey posted.