Usually C dividers, taps on the coils (esp. Which should be higher Z for reasonable (component) Q, so you end up needing impedance transformers coupling into them in the first place. In which case, a transformation to all parallel resonators can be profitable. In which, the series resonant tanks can be especially tricky. What type did you build, and with what components? If you chose a narrow (high Q) bandpass, you may find accidental impedance transformations or other effects, due to stray capacitance or inductance. Kind of the enterprise-grade solution here, but not actually as intimidating as it sounds. Best case, measure s11 and s22, dump 'em into a simulator and solve for the inbetweeeny-bits. May not be possible (you can't exactly filter that which is already reactive), in which case you'll effectively end up with some mix of the above. Tune the filter to the antenna and receiver, straight up. This is effectively the above option, with the stop band explicitly linked to a port, rather than to resistors spread within the network. Construct a constant-impedance or diplexing filter. (Obviously, bad here.) - Equalize the impedances, cancelling or dampening out-of-band reactances or resonances. Solutions: - Add attenuators to ballast the impedance(s) closer to 50 ohms. That being the key fact: if they're tuned, they're not flat 50 ohms! If they aren't the flat 50 ohms assumed by the filter prototype, you're going to have poor results. What impedance is your antenna and amplifier? At all frequencies I mean, not just whatever they're tuned to. My question is this - is it useful/practical to specify and realize a RF filter from a web page calculator? Thanks for any suggestions The filters have 50 ohm in/out impedance and I'm using a nanovna to evaluate the realized filters. The 5-pole bandpass filter had minimum (in-band) insertion loss of over 20 dB as if it's high pass and low pass properties badly overlap the high pass filter has insertion loss of over 20 dB at 1.8 MHz, about 25 dB at 1.7 MHz an undesirably gentle transition for a 9-pole filter. I've tried two different designs (the first was a bandpass filter, the second a high pass filter, both seeking to differentiate between AM broadcast band and 160 meter/1.8 MHz ham band signals) using the LC filter calculator at, and my (as-built) measured filters have far more gradual transitions from stop to pass band than the transmission response suggested by the web page's calculator. The filter would be prior to the first gain stage, so passband insertion loss is a consideration. I'm trying to build a high pass filter or bandpass filter to reduce strength of AM broadcast signals entering a homebrew 160 meter/1.8 MHz ham band receiver.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |