3rd order active filter

[franticEB]

Hi,
someone knows this kind of filter?



I've calculated transfer function with sapwin but now i want to obtain the center frequency of bandpass.
Anyone could help me?

 

[LvW]

Very unusual.
It is a classical lowpass with a series capacitor at the input that attenuates low frequencies and dc.
You are asking for the center frequency - but how do you define (in this 3rd order case) the center frequency?

Edit: Classical bandpass responses are of 2nd order (or n=4,6...).
In this (unusual) case with n=3, you have two possible choices for defining the center frequency: maximum of transfer function or phase shift of 225 deg (mid between +90 and 360)

 

[deepmak]

hi ,
except for the C2 , the filter looks like a second order multiple feedback low pass filter

 

[franticEB]

Very unusual.
It is a classical lowpass with a series capacitor at the input that attenuates low frequencies and dc.
You are asking for the center frequency - but how do you define (in this 3rd order case) the center frequency?

Edit: Classical bandpass responses are of 2nd order (or n=4,6...).
In this (unusual) case with n=3, you have two possible choices for defining the center frequency: maximum of transfer function or phase shift of 225 deg (mid between +90 and 360)

I were also thinking about low-pass + decoupling cap that acts as high pass.
So if i want to move the cut-off frequency pole with the general effect to enlarge the band of HighPass+LowPass filter, what i have to do?

 

[LvW]

I were also thinking about low-pass + decoupling cap that acts as high pass.

Your circuit is, in fact, a second-order lowpass in series with a coupling cap that introduces a highpass response.

So if i want to move the cut-off frequency pole with the general effect to enlarge the band of HighPass+LowPass filter, what i have to do?

I don`t understand what do you want to do. If you want to move the lowpass poles you must change component values correspondingly (look at the transfer function)

 

[deepmak]

Good day again
FranticEB , can you use other filter design like second order band pass filter (multiple feedback ) if you want to a band pass one ? or you are tied to this design ?

 

[Milad-D]

This circuit is just a cascade of a second order low pass filter with a first order high pass filter.

 

[LvW]

The plus and minus signs of the opamp are placed incorrectly. I mean, in this way, the filter can not produce complex poles. This circuit is just a cascade of a second order low pass filter with a first order high pass filter. View attachment 93043

Milad-D, check it again.
The multi-feedback topology always uses inverting amplification. Thus, the circuit as given in post#1 is correct!

 

[franticEB]

simulating the only low pass stage i noticed that @19khz the knee of the lowpass curve starts as in figure


where the red curve is highpass+lowpass and the blue one is the only lowpass.
First of all i don't understand why the knee is @19KHz, since the cutt-off freq is @86KHz, and then i want to move that 19KHz to 20KHz/21KHz(i.e. the filter has to begin to attenuate after 20 KHz)

 

[LvW]

First of all i don't understand why the knee is @19KHz, since the cutt-off freq is @86KHz, and then i want to move that 19KHz to 20KHz/21KHz(i.e. the filter has to begin to attenuate after 20 KHz)

How do you define the "knee"? This is your own definition - right? If you look carefully (magnify the curve with better resolutuion you will see that it "begins to attenuate already at lower frequencies).

Don`t confuse your "knee" with the cut-off that is defined at the -3dB point.