Path: utzoo!utgpu!ugw.utcs.utoronto.ca!CUVMA!SWL-L Date: Sat, 3 Feb 90 12:31:39 EST Reply-To: Don Liu Sender: Short Wave Listener's List Comments: Warning -- original Sender: tag was From: Don Liu Subject: Re: Selectivity vs. audio quality X-To: swl-l@cuvmb.cc.columbia.edu To: UofToronto LAN redistribution Message-ID: <90Feb3.125329est.57507@ugw.utcs.utoronto.ca> Newsgroups: bitnet.swl-l Distribution: ut Approved: devnull@gpu.utcs.toronto.edu In article <22115.25c84581@kuhub.cc.ukans.edu> arritt@kuhub.cc.ukans.edu writes: >I have a Sangean ATS-803A with a wide/narrow selectivity switch. On >occasion I have tried using the narrow selectivity setting, but this >makes the audio very muffled, often to the point of un-intelligibility. >Turning the treble all the way up and the bass all the way down often >helps, but the audio still isn't as good as with the wider selectivity >setting. (Turning the treble up seems to emphasize the interfering >station as well.) Usually I just go ahead and use the wide selectivity. > >Is this (1) a problem with my particular set; (2) a problem with the >803A; or (3) a natural consequence of narrow selectivity? Would a more >expensive set having even narrower selectivity have even worse audio >quality? Any suggestions for improvement? This has always been a challenge to radio engineering. Actually modern receivers is much better in selectivity/audio quality than those 20 years ago. The problem is to make a good frequency filter. In shortwave communications the audio signal is used to control the magnitude of the carrier frequency. Audio frequency is not a constant, it contains frequency componaents up to 5khz (in AM systems). Because the magnitude of the carier is altered at the audio frequency, the carrier itself is nolonger a pure frequency. Infact, there is a frequency "band" centered at the nominal carrier frequency. Higher audio frequency causes wider side band. The frequency can be expressed by f=f0+-fa If stations are spaced 2fa apart, then theoretically all 5khz audio can be recovered at the reciver without interstation interference. However, this requires a perfect filter which only passes frequencies from f0-fa thru f0+fa for a nominal station frequency f0. In reality there is no such filter. A typical filter has a quite flat passband and very sharp attenuation beyond. Although the attenuation can be vary sharp, the adjacent station still has a certain frequency amount that goes into the receiver's amplifier together with the selected station frequencies. One way to improve the seperation is to narrow the passband. This way the adjacent channel is further attenuated. At the same time the high audio frequencies of the selected station is also lost (see formula above). THat is exactly what Sangean is doing. This is not a problem with Sangean because it is already using state-of-art technologies by using multiple stage crystal resonator filters. Even with perfect filter you may also have interstation interference because the frequencies gets changed (off from its center) due to long distance and wheter/atmosphere conditions so as to invade the adjacent station frequencies.