on band pass filters

Thread Starter

vj39

Joined Feb 16, 2012
17
while designing an active band pass filter using a lpf of of 6khz and hpf of 10khz cut off frequency the output bandwidth i got is around 8.5khz(graphical) instead of 4khz..the total gain will be the product of the two but how is the bandwidth affected??
 

Audioguru

Joined Dec 20, 2007
11,249
You have the filters backwards.
With a LPF at 6kHz and a HPF at 10kHz then there is a null at about 8kHz.
Instead you need a HPF at 6kHz and a LPF at 10KHz for a peak at 8kHz or 8.5kHz if your opamps work at frequencies that high.

The total gain depends on the sharpness of the filters. If the filters are very simple then the lowpass filter also filters out 8.5kHz and the highpass filter also filters out 8.5kHz.

If the filters are sharp then low frequencies and high frequencies are filtered out but 8.5khz is passed or amplified.
 

Thread Starter

vj39

Joined Feb 16, 2012
17
sorry abt that what i meant is a lpf of 10khz and hpf of 6khz...the bandwidth has to be the difference of the two frequencies(theoretically 4khz)..i'm using two 741 opamps for the circuit and the bandwidth i got is 8.5khz from the plot(not the frequency at which gain is maximum)
 

Audioguru

Joined Dec 20, 2007
11,249
sorry abt that what i meant is a lpf of 10khz and hpf of 6khz...the bandwidth has to be the difference of the two frequencies(theoretically 4khz).
You still have it backwards. You want to pass frequencies higher than 6kHz and pass frequencies lower than 10kHz. Then the HPF is 6kHz and the LPF is 10kHz.

.i'm using two 741 opamps for the circuit and the bandwidth i got is 8.5khz from the plot(not the frequency at which gain is maximum)
A lousy 741 opamp is 44 years old and has trouble above 9kHz.
You probably want to cut high frequencies up to 100kHz then you need an opamp that works well up to 100kHz. There are many modern opamps that work well up to 100kHz like the TL071 single or TL072 dual.

You forgot to attach your schematic so we do not know what kind of filters you have.
 

Ron H

Joined Apr 14, 2005
7,014
You have the filters backwards.
With a LPF at 6kHz and a HPF at 10kHz then there is a null at about 8kHz.
You won't get a null. You will still get a peak, but the peak amplitude will be attenuated. The amount of attenuation will be proportional to the order of the filters.
That wasn't intuitive to me. I ran some simulations to come up with this conclusion.
After thinking about it, though, it makes sense.The attenuation at zero and infinity (frequencies) will be infinite. It will be minimum at a point somewhere between the corner frequencies.
 
Top