How to choose -3dB filter bandwidth for an input signal frequency?

[sudeep_]

Hi,

I have one doubt on choosing filter Bandwidth. My input signal frequency is 40 MHz for an active RC filter. I don't know how to choose value of the -3dB BW. Is there any relation between the input signal freq versus BW?



Thanks!

 

[FvM]

there any relation between the input signal freq versus BW?

I suppose it is, but dedepending on both, the filter and signal details. Are you talking about a low-pass filter?

To ask differently, what's the intended filter effect? What's the signal waveform, do you expect it changed or maintained after passing the filter.

 

[D.A.(Tony)Stewart]

It depends squarely on Nth order rolloff for -3dB, ripple skirt shape etc.​

But for 1st order approximation, 10ns rise time = 30MHz

1ns = 300MHz etc.

so tR = 3/10f when defined as 10~90% ,
not 0~ (1-1/e)=63%

Thus 40MHz= 7.5ns

- - - Updated - - -

FWIW, learn to specify RF filters as bandpass ripple f and bandstop rejection point.
Eg flat to 40MHz +/- 0.x dB and reject y dB at f2, then add phase error and group delay specs as needed

 

[crutschow]

It's very difficult to make an active filter that performs properly at 40 MHz.

 

[sudeep_]

If I set the BW at 40 MHz, the SR is dropped. If SR is setteled then BW of the Filter show at higher frequency.

Moreover, the currents at Compensation Capcitor in Opamp is very low (swing < 80uA). But Input stage and output stage currents are 300uA, 800 uA.

How to increase the Current swing at Cc in Opamp.

 

[D.A.(Tony)Stewart]

for a 1V step you can figure out the SR at 40MHz
Vstep*f3dB/0.35 > SR for 10~90% or SR>2pi*f/Vpk

Since you know dV/dt= I/C thus your limit determined by that stage.

https://www.ee.bgu.ac.il/~angcirc/History/Solutions_2003_2004_B/SomeStuff/History18opamp.pdf

 

[Borber]

Only transfer function can give you -3dB bandwidth from allowed attenuation at max. frequency of interest.

 

[RCinFLA]

Beside knowing what you need to reject, you have to know what you need to pass.

For example, passing a bunch of narrow band RF channels below 40 MHz is a different filter then passing a video signal that is critical to group delay.

An active LPF filter with corner at 40 MHz can be very tough to keep stable.