In article , "Richard"
writes:

Steve Nosko wrote:
I believe he is talking about adjacent channel signals which are much
stronger than the desired and trying to see if there is a way to "fix"
that. Capture does not apply for this situation. Capture is a co-channel
effect and is better the wider the whole system is, not just the IF.

Capture effect:

http://en.wikipedia.org/wiki/Capture_effect

It's an intersting phenomena, and potentially useful. If I could phase null
(at the antenna) the local stronger signal then it would dissappear as
inteference if it was a few db below the wanted signal.

Only problem is, is that if you cannot get the inteferring signal below the
wanted one, it's works very bad for you, because you cannot listen to an
inteferred with signal. Which is better than nothing I guess in some cases.

Simplistic explanations of "capture effect" don't help here.

The main, major cause of such strong signal interference is due to
the FM receiver's LIMITER stage. That stage (usually there are 2 to
3 of them) is biased such that a strong signal is going to clip or that
it will actually develop enough self-bias to clip even further. The end
result is that a limiter output is primarily composed of the strongest
signal's RF energy.

The action of a limiter is basically that of accepting only the strongest
signal and "pushing down" the energy of weaker signals, including
noise at the noise floor of the front end. THAT is what insures the FM
"clean signal" quality. It is also a sort of automatic AGC so that all
signals of the same modulation index will be demodulated to the same
audio level...even though the antenna input signal can vary considerably.

Such limiters have been used for some pulsed RF reception with the
ability to "push down" offending pulsed carrier frequencies' envelopes.

Pity that somehow you cannot design circuitry which recognises a 100Khz
difference between the wanted FM station and the unwanted FM station and
proceed to demodulate the wanted one only. If it were possible, it would
have been done by now. Maybe you could do it digitally,I dunno.

Not with "phasing" alone and not with limiters still in the IF chain. But,
it has been done by other means -

The brute force method is to use a many-pole crystal bandpass filter
ahead of all limiters. Even Heathkit did that. This cuts out very strong
adjacent signals and the limiters will be less affected by such. With
steep skirt selectivity, offending adjacent channel attenuation can be
60 db or greater.

A second method is to discard the limiter stages and try for a Ratio
detector instead of the Foster-Seeley discriminator variants plus AGC
on the non-limiting IF stages. Not a perfect solution but does work to
some extent. Most NTSC TV receivers use a variation of that for TV
sound, originally started many years ago as the "intercarrier sound"
system to replace a separate limiter-discriminator IF strip; the first
TV receivers actually had TWO IF strips...:-)

The most obvious way, and least familiar, is to use a large-bit digital
A-to-D system and high-speed microprocessor to (1) Do the channel
bandpass filtering with a "brick-wall" filter routine; (2) Do the FM
demodulation directly after digital bandpass filtering by summing the
sideband components of the FM signal; (3) Converting the demodulated
IF in digital form back to analog audio. That's been done but the
production costs for consumer electronics systems has only lately
come down enough to warrant such use.

All of the methods I've given are prior art.

Len Anderson
retired (from regular hours) electronic engineer person