From: xpyttl on Feb 6, 12:19 pm

wrote in message

Bandpass filters can be designed for specific bandpass without
using inductive coupling.

hmmm ... I had lept to the conclusion that OP was concerned about undesired
coupling, rather than desired coupling.

To make an inductively-coupled BPF one IS concerned about DESIRED
coupling. That's how all those "IF cans" have been designed for
about the last 65 years (give or take). EXACT desired magnetic
field coupling.

Two-resonator bandpass filters (narrow bandwidth relative to
center frequency) can be done with "inductive" (magnetic field
coupling), capacitive coupling in two ways, or inductor coupling
(without magnetic field coupling) in two ways. Kind of old hat.
I prefer the non-mag-field approach since I LIKE using toroid
inductors and do NOT want any magnetic field coupling in there.
Makes the packaging more flexible; i.e., no sweat.

I think SMD inductors are typically designed so as to not maximize their
susceptibility to objects in their vicinity, although not necessarily
minimize it.

I've never done a mind-meld with any SMDs but I have looked in
a lot of databooks, seen a lot of datasheets. SMD inductors
are designed for inductance, Q, and current-carrying capacity.
Unless they are SPECIALS, shielded, or some other thing stated,
no one knows what their magnetic field is unless they try it
out. Or discuss it with the SMD makers.

If you want to find out the magnetic field coupling, the setups
have been described in the better texts for decades. It is a
lot of trouble to do but that time can be amortized nicely if
one is making thousands of identical IF cans (narrow percentage
bandwidth bandpass filters).

Another issue of concern is self-resonance.

NS? :-) That's a built-in concern every time one designs
filters of any kind. Try a highpass filter and see where
inductor (or capacitor) self-resonance messes up response.

A two-resonator type of narrow bandwidth bandpass filter
DEPENDS on resonance. Self-resonance implies a parallel
capacitance component equivalent on the inductor. That, in
turn, dictates the exact choice of fixed or trimmable
capacitor one adds across the inductor.

Some SMD inductors have rather
surprisingly low self-resonant frequencies.

Quantitatively define "surprisingly," sir. That is
a null statement. Show some examples of SMD inductors'
self-resonance versus other types of inductors not SMT.
Crib from existing data if you like, some does exist.

Obviously, some designed for
cell phone service typically won't, but it is yet another parameter to fret
over.

Are you designing cellular telephones? There is PLENTY of OTHER
applications at UHF where some small, two-resonator BPFs can be
used. Or at VHF. Or at HF. Even at MF and LF.

"Designed for cell phone service" implies SMT, obviously. Cell
phones DEMAND smallness in the market. HOW they eventually get
so small doesn't depend solely on physical package sizes or the
other factors you allude to. I like jumping but not doing so
on conclusions.

I've been sharp and nasty in this reply. I've also skim-read a
lot of inane and time-wasting blather on the "FAQs" cross-
posted to a bunch of newsgroups...that should not have been.
I'm tired of a bunch of self-righteous arguers busy, busy
arguing for the love of arguing. :-(

In HOMEBREW the replies ought to concern themselves with more
factual responses, not cut-and-paste generalities nor inability
to understand the original question. That's the best help for
anyone who asks questions.