[hfsig] Some basic FSK questions

[otc at gmx.net]

The best filter is a matched filter for both fsk frequencies. I guess that
is always the case because of:
You can see a fsk signal as a orthogonal frequency set of two amplitude
modulated signals!

The limiter at the input removes residiual amplitude modulation left in the
signal (of fading action and false line equalization). But all in the
demodulator depends on the other components! You lost approx. 1.5dB S/N if
you use a limiter. Using a limiter simplicates the circuit: AM is supressed
up to 6dB. You don't need a AGC unit. The detector can be even simpler.

Is it even possible to improve by making a chain of bandpass filter and
limiter several times. Somewhere/sometime I read about it.

regards -
Henry


----- Original Message ----- 
From: "Ben Mesander" <ben at dick.jymis.com>
To: <hfsig at lists.tapr.org>
Sent: Thursday, December 08, 2005 7:39 PM
Subject: [hfsig] Some basic FSK questions


> Hi,
>
>   I have a couple of questions related to a HF demodulator I am
> working on. If these questions are more appropriate for another forum,
> please let me know.
>
>   For an FSK demodulator that works on an FSK signal that was
> transmitted with unshaped pulses (rectangular in the time domain),
> what is the best filter to demodulate with to minimize noise/ISI?
>
>   For a synchronizer that uses a squaring nonlinearity, what is a
> good criteria for designing the bandpass filter?
>
> Thanks,
> Ben
>
>
>
> _______________________________________________
> hfsig mailing list
> hfsig at lists.tapr.org
> https://lists.tapr.org/cgi-bin/mailman/listinfo/hfsig
>