Post subject: LC Low Pass Filters
Sat Mar 19, 2005 3:32 pm
Joined: Sat Mar 19, 2005 3:20 pm
constructed a 5th order LC low-pass filter to cut-off at 108 MHz (the
upper frequency of the FM radio band) with a minimum attenuation of
40 dB at 216 MHz. I connected my antenna to the input of this filter
and connected the output to the spectrum analyser. I found that I am
still able to receive substancial frequency components well above 108
MHz. Should these not have been attenuated?
Unread postPosted: Sun Mar 20, 2005
Check the filter by itself ona network analyzer. I fyou
do not have acces to a network analyzer then sweep the input of the
filter using a sweeg generator or analyzer and see the response. If
the filter has the rejection as designed then the filter passband may
be getting "pulled" by the mismatch of the antenna.
Post subject: Low-pass Filter
Sun Mar 20, 2005 10:27 pm
Well, it might be that things
are working as designed - even if not as intended.
filter has 3 bands: the passband, the transition band, and the stopband.
The edge of the passband is generally defined as the 3 dB point for
Butterworth filters, and the frequency where the attenuation first exceeds
the passband ripple for Chebyshev and Elliptic-function filters.
(Those last two would have either a parallel tuned circuit in the forward
path, or a series tuned circuit as a shunt path.)
So it sounds
like you've designed a filter with the top edge of the passband at 108
MHz, and the bottom edge of the stopband at 216 MHz. In between, you'll
have something that varies between 3 and 40 dB of attenuation, depending
on the design and construction. You'll still see signals at those frequencies
on a spectrum analyzer!
Do also note that circuitry at theese
frequencies can act as antennas, especially inductors, so you might
have to shield the filter from the environment to get the desired performance.