FWIW -- In the stop band - normal low pass filters (not of the type that
terminate out of band harmonics instead of reflecting them ) will be
anything but matched - Further , especially on broad band PA's like tx
coupled solid state - you can run into some unexpected problems (especially
poor efficiency and gain flatness ) - when you terminate even or odd
harmonics in un-determined impedances-that is why some of the solid state
guys are designing and building filters that terminate rather than reflect
harmonics-.( These are basically lowpass- highpass complementary filters
with the highpass section feeding a resistive load to terminate the
harmonics).
Also , a finer point of conventional filters is that the "designed"
attenuation will be different depending on whether the source looks like a
constant current or constant voltage generator - If you have a circuit
simulator you can prove this to your self pretty quickly.
Low pass filters with the first element as a series inductor rather than a
shunt capacitor will look entirely different at harmonic frequencies - and
affect the amplifier differently - of course the length of coax (if any)
between the amp and the filter can rotate the filter out of band impedances
all over the place as well.
Lots of fun --
Hank K7HP
>>>BUT -- the same cautions about impedance matching being necessary
>>>still apply, and that's the fly in the ointment. Not only must you
>>>be matched at the operating frequency for the filter specs to apply,
>>>you must also be matched at the harmonics! The chances of that
>>>occuring are slim to none unless you happen to be working into an
>>>antenna that is resonant at the harmonics. We DO use some of those,
>>>but it certainly isn't all that we use, and it would only apply to
>>>single band filters.
>>>
>>>Jim Brown K9YC
>
> http://www.minicircuits.com/appnote/filters8-2.pdf
>
> They are often used on the IF port of mixers, where it necessary to have
> constant impedance. The downsides are that they are pretty poor
> frequency response compared to filters designed without the requirement
> on constant impedance.
>
> I've never tried designing one, and know little about them, but I assume
> that since they don't reflect out of band signals and don't pass them
> either, then there must be resistive elements which absorb the unwanted
> power. There can not just be L's and C's, but it will need R's too.
>
> Without knowing the context in which the thread started, I can't make
> any further comment, but just wanted to point out you can make filters
> that present 50 Ohms over a wide range of frequencies.
>
> G8WRB
>
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|