Fifth order passive filter - band pass

[Andrew_it]

Hi everyone,
I have to design a fifth order passive filter for my circuit, but I'm a little newbie in that matter.
I know well the classic RLC circuit and higher order but ACTIVE filter.
Can someone recommend me some software tools, documentation or something like that to design the filter?
I hope for your support,

Andrea

 

[pancho_hideboo]

https://www.changpuak.ch/electronics/butterworth_bandpass.php
https://www.changpuak.ch/electronics/chebyshev_bandpass.php

 

[schmitt trigger]

Just remember that, when you are applying a passive filter, both the source impedance and the load impedance have a significant effect on the actual filter response.

 

[Andrew_it]

Thanks for the support, the two links are perfect!
I'd like to know if there are some software program used to design filter?

schmitt trigger: you intend the "source impedance and load impedance" of every stage or of the entire filter circuitry?
Thanks

 

[pancho_hideboo]

https://edadocs.software.keysight.com/display/ads2009U1/Filter+DesignGuide

schmitt trigger: you intend the "source impedance and load impedance" of every stage or of the entire filter circuitry?

entire filter circuitry

 

[Andrew_it]

Ok!
The idea therefore is to use an op-amp buffer stage configuration placed ad the source and at the load in order to "fix" the impedence at 50 ohm. Can it work?
Can someone recomend me some circuits?
Thanks!

 

[LvW]

Ok!
The idea therefore is to use an op-amp buffer stage configuration placed ad the source and at the load in order to "fix" the impedence at 50 ohm. Can it work?
Can someone recomend me some circuits?
Thanks!

There are many "working" filter topogies. However, why do you want to have input/output impedances of 50 Ohms?
More than that, according to the commonly applied convention, the filter order is determined by the order of the polynominal which appears in the denominator of the transfer function.
For a bandpass only even orders are possible: n=2,4,6...

 

[Andrew_it]

Because I need impedance adaptation and I want that my source circuit and load circuit "see" 50 Ohm.
Anyway, can you suggest me some circuits to place with my filter in order to uncouple the circuit?
Thanks

 

[pancho_hideboo]

The idea therefore is to use an op-amp buffer stage configuration placed ad the source and at the load in order to "fix" the impedence at 50 ohm.
Can it work?

It depends on operation frequency, amplitude and waveform of input signal.

Generally simple source follower or emitter follower are useful for input buffer.
And common source or common emitter stage are useful for output buffer.

 

[Warpspeed]

A bandpass filter is best if the required bandwidth is quite narrow.
If its fairly wide, a separate low pass and high pass will probably work better.

 

[Andrew_it]

I'm working on 35KHz and 100KHz and I need a fifth order band-pass filter.
So I can use an easy transistor configuration?

Thanks

 

[FvM]

More than that, according to the commonly applied convention, the filter order is determined by the order of the polynominal which appears in the denominator of the transfer function.
For a bandpass only even orders are possible: n=2,4,6...

I notice that some sources are counting bandpass order differently as number of pole pairs. A possible reasoning can be seen in the construction per lowpass-bandpass transformation.

 

[LvW]

I notice that some sources are counting bandpass order differently as number of pole pairs. A possible reasoning can be seen in the construction per lowpass-bandpass transformation.

Yes - as far as the first sentence is concerned - I agree. I know that some (onmly a very few) knowledge sources consider the lowest bandpass order as n=1 (for example: filter_free from Nuhertz.com).
However, I disagree with the second sentence. For example, applying the lowpass-to-bandpass transformation to a first order lowpass (first-order denominator) gives a bandpass transfer function having a second-order denominator..

 

[Andrew_it]

Ok! I'm trying to design my Butterworth band-pass filter with the aid of Nuhertz filter solutions and seems to be very easy and quick.
Now I need the two buffer at the input and at the output of my filter circuit so that the next stage do not change the impedance.
Can you suggest me some literature or circuits?
Thanks again.

 

[BradtheRad]

Anyway, can you suggest me some circuits to place with my filter in order to uncouple the circuit?
Thanks

LC series makes a bandpass filter. And it blocks DC.

Bandwidth is influenced by L:C ratio, and by neighboring ohmic resistance.

 

[Andrew_it]

Ok, I realized the fifth-order passive band-pass filter. Now the problem is to adjust the 50ohm impedance at the input and at the output of my circuit.
Can someone suggest me some OP-amp configuration?
Thanks

 

[FvM]

I would dimension the passive band pass for the external circuit impedance instead of using buffers. 50 ohm seems inconveniently low for 100 kHz frequency anyway, exceeding the drive capability of a general purpose OP.

Buffer amplifiers are however needed if the external circuit impedance is varying respectively unknown. A non-inverting amplifier with G=1 or G=2 (the latter to compensate possible impedance matching losses) should help. It has about infinite input impedance and zero output impedance. Adding an output series resistor respectively input parallel resistor allows matching to arbitrary impedances.

 

[Andrew_it]

So I can use a simple non-inverting amplifier?
But I do not understand how to set the 50Ohm at the input of the circuit.

 

[schmitt trigger]

Google impedance matching networks.

Now, as FvM mentioned, most general purpose opamps won't drive such a low impedance. Your best bets when looking for a suitable device might be video amplifiers.

 

[Andrew_it]

Impedance matching network - bandpass passive filter

Hi everyone,
I designed and realized a passive fifth-order band-pass filter.
I'm working with 30KHz-100Khz.
I'd like to know how to uncouple the filter with the circuitry that is placed at the source of it and how to let the filter "see" the 50Ohm. I thought that an op-amp in a buffer configuration can be the solution but when I use it the signal is drastically lowered.
Can someone help me in impedance matching matter or suggest to me some solutions?
Thanks