Very good point...

of course it can't...

 - just saw the option in the design S/W and didn't actually try it out.

 

OOPs !!!!!!!

 

I've done this quite often with bandpass filters, where it does work very well indeed, and can sometimes be used to improve the shape over an equally terminated design.   Particularly for top coupled designs

 

Andy

 

 

On 16 December 2010 20:49, James Moritz <[email protected]> wrote:

Dear Andy, LF Group,

Unfortunately, this doesn't quite work... If you think about it, at very low frequencies, any low-pass filter will just be a "straight through" connection, so in your example will just give whatever voltage the 50ohm source delivers to a 500ohm load with no filter present. The filter synthesis algorithm then makes a circuit that extends a more-or less flat response at this level up to somewhere near the cut-off frequency. You could, however, design a "peaked low-pass" circuit that would give voltage gain over a relatively narrow frequency range just below the cut-off frequency, falling to unity at DC - a pi matching network does this, for example.

Cheers, Jim Moritz
73 de M0BMU

----- Original Message ----- From: "Andy Talbot" <[email protected]>
To: <[email protected]>
Sent: Thursday, December 16, 2010 8:25 PM
Subject: Re: LF: Sound Card mic Impedance

If its a passive L/C filter, why not do a version with unequal Zin and Zout
and take advantage of the voltage gain in going from 50ohms to, say, 500
ohms?   That will give you an apparant 10dB without trying.