Okay, so l want to design a band pass filter that doesn't attenuate my signal. Is that possible. I want it to allow frequencies between 1khz and 3khz. I have tried to make one, but it ends up with my signal getting attenuation. Any hints on how l can achieve this ?

 

Do you mean an active filter design?

 

if not active, then tank circuit is usually better than regular passive filter implementation in terms of attenuation.

http://www.allaboutcircuits.com/vol_2/chpt_8/6.html#02124.png

 

If you're going to make a passive filter, you'll have some loss. The better you tune the filter, the less loss you'll have.

Here's a pretty decent one I threw together; it's a modification of a Cauer.

See the attached for a plot of the transmission (S21) and reflection (S11).
At 1kHz, the signal is about -0.443dB, at the center, around -0.223db, and @3kHz, -0.288dB

 

If you are trying to design an active filter, I uploaded this a few days ago:

http://forum.allaboutcircuits.com/showthread.php?t=61224

Maybe it will help you.

 

if not active, then tank circuit is usually better than regular passive filter implementation in terms of attenuation.

http://www.allaboutcircuits.com/vol_2/chpt_8/6.html#02124.png

I think the tank filter which shorts out signals within the frequencies l do not like is what l will try to design given the parts l have which include only one 10mH inductor. I will post my multism circuit for you guys to critique. Thanks very much for the hints.

 

An active filter is a good fit for this application and requires only resistors and capacitors. You can easily design one using TI's free software.

 

An active filter is a good fit for this application and requires only resistors and capacitors. <snip>

You missed the part about the opamp.

 

well then, active is the way to go, does not sound like you would need a fourth order one (although it is impressive)

 

You could always put an adjustable amplifier in front of or after your filter. This is called buffering and prevents one filter from loading a previous or subsequent stage. Most engineers, realizing the benefits of buffering, do it without giving the insertion loss a second thought.

 

You missed the part about the opamp.

True enough. I should have said the only passive components required are resistors and capacitors. Of course it's hard to have an "active" filter circuit without some sort of amplifier.