output filtering.

[Crusher]

I am currently playing with some low pass Chebyshev filter designs. From the .01dB, .1dB, .5dB and 1.0dB of ripple. I was wondering what the advantage of using a higher or lower ripple over the others. I know the chebyshev has a steeper drop verses say the butterworth. From what I have read and understand, the more ripple, the steeper the initial decent or cutoff. Any comments or suggestions, please let me know.

 

[Farm Boy]

RE: output filtering

Ripple in a filter is seen as reflected power to the transmitter. (SWR) The greater the amount of ripple, the greater the variation as you tune across the filters passband. The same is true for a reciever. The same variations that caused reflected power to the transmitter will be seen as loss to a incomming signal.

Now for the hard part,,,,,
You can't get something for nothing. The tighter the filter, the more loss it is going to have. How much filter need, verses signal reduction (attenuation) is the question only you can answer for your application.

Steve / K6SCA

 

[Crusher]

Thank you for the reply. i have alot of books and such that get into some detail, but when it comes down to it they are at the same time very fague if you know what I mean.

 

[bushwacker]

The tighter the filter, the more loss it is going to have. How much filter need, verses signal reduction (attenuation) is the question only you can answer for your application.

That says it all....

 

[freecell]

"the more ripple, the steeper the initial decent or cutoff." you answered your own question. the chebyshev filter provides a steeper cutoff than either the butterworth, bessel or cauer elliptical types. the question is do you need that steep of a rolloff at the second harmonic?

you can't design any filter if you don't understand what is required of it related to your specific application. the details will appear less "vague" if you brush up on your polynomial and quadratic (algebra) equation expressions. to design a low pass filter these parameters must be defined:

Cutoff Frequency
Stopband Frequency
Passband Ripple
Stopband Attenuation
Filter Order
Input Impedance
Output Impedance

futhermore, ripple in the passband at values up to 1.0db. is insignificant in terms of loss at the fundamental. no one is going to miss it and no one cares. depending on the amount of second harmonic energy present (not a big issue in push-pull amplifiers) a steep rolloff immediately above the passband may not be required in the first place.

 

[Crusher]

After i reread one book I have, and studied it I have most of it figured out. Freecell I was using computer generated charts so that all i had to do is figure Rs and Rl and figure out how much attenuation (dB) I wanted or needed at a given frequency say at like 3rd,5th or whatever order harmonics that may be most detectable. Of course I added 3MHz to the frequency I was using and used that as my cutoff. Yes this is for a push-pull configuration.

Thanks and 73's