Hi all,

I am looking into the effects of changing the capacitor has on a low pass filter.

The circuit being used is below, however C2 has been changed to 0.47nF in order to plot the graph.

I have read that the sharp gain increase at roll off is due to resonance, Id like to know where the break freq is measured from, before the or after when the output drops to within 3dB of the gain?

Thanks in advance

 

On the actual response, which is not linear, there is a frequency at which the response is down 3 dB. The corner or break frequency is actually the intersection of the 0 dB line and the tangent to the slope of the response in the transition from pass band to stop band. It is lower than the actual 3 dB point, but not by much. That is why it is a useful approximation.

 

I am surprised you understood "before the or after when the output" its fairly ambiguous

So here we take -3dB as the corner?

 

The corner or break frequency is an approximation to the -3 dB point. It is not exact. Look for example at the Bode Plot of a Butterworth filter to see the difference between the -3 dB point and the place where the slope of the rollof intersects the 0 dB level. It occurs before the -3 dB point.

http://en.wikipedia.org/wiki/Butterworth_filter

 

If you are trying to preserve the "Butterworth" response, you cannot just change one of the capacitors and expect the low-pass response to remain the same shape. You would have to change both capacitors by the same percentage, keeping the ratio C1/C2 constant.

 

Thanks for your help papabravo.

Mike, thanks also for the info, These caps behave almost like a voltage divider right? The difference being that caps are based on charge/discharge time?

 

Thanks for your help papabravo.

Mike, thanks also for the info, These caps behave almost like a voltage divider right? The difference being that caps are based on charge/discharge time?

It is not that simple. Read the "Sallen-Key" section of the Wiki linked above.

 

If you are trying to preserve the "Butterworth" response, you cannot just change one of the capacitors and expect the low-pass response to remain the same shape. You would have to change both capacitors by the same percentage, keeping the ratio C1/C2 constant.

View attachment 76342

His original plot was not a Butterworth response as indicated by the pre-rolloff peak. I used the Butterworth response as a clearer example of how the corner frequency and the -3 dB point are related. As mentioned in the wiki, a Butterworh response requires more restrictive values on the components.

 

My take was that the original circuit (before he changed the one capacitor) had a Butterworth response. By changing only one capacitor, the response became very un-Butterworth...

You can see that the original, unmodified circuit has a Butterworth response if you look at the sim I posted.

 

My take was that the original circuit (before he changed the one capacitor) had a Butterworth response. By changing only one capacitor, the response became very un-Butterworth...

You can see that the original, unmodified circuit has a Butterworth response if you look at the sim I posted.

You're right. I never saw the original response, only the one with the peak and did not properly connect the response plot to the actual circuit configuration.