k4kyv
Contributing Member
Don
Member
Offline
Posts: 10037
|
|
« on: March 11, 2011, 02:34:25 PM » |
|
Q I can see how you would get distortion on a DSB signal from a *frequency* difference between the original carrier and the Fbfo because signal elements represented by Fcarrier-Faudio-F(bfo+phi) and Fcarrier+Faudio-F(bfo+phi) would not be the same. But I'm not understanding why the phase difference would be important.
A When you are demodulating a DSB signal, the resulting audio is a composite of both sidebands, USB and LSB. The audio from each sideband is a function of the phase relationship between the carrier (or BFO) and that sideband.
With DSB, coherent detection of the sidebands is required. When a carrier is modulated, the two resulting sidebands have a defined phase relationship with each other. This is explained in most handbooks using simple vector diagrams. Dig up one of the early ARRL publications such as "Single Sideband for the Radio Amateur," where they compare the detection process of AM with that of SSB.
With DSB, the detected (demodulated) audio from the two sidebands must be in proper phase relationship so that they are vectorially additive. The only way that this can be accomplished with inserted carrier from a BFO is for the BFO to be precisely on frequency and in phase (or 180 degrees out of phase) with the original carrier.
One way that this can be accomplished is to use a phase locked loop to cause the BFO to lock onto the original carrier, which may be reduced up to 20 dB or so, making the DSB reduced-carrier transmitter more power-efficient. It is also possible, using a more complicated loop configuration, to use the sideband information only to re-establish the frequency and phase of the injected carrier. Then it is possible to fully suppress the DSB carrier and transmit sidebands only. The detection circuit is known as the Costas Loop. Without some kind of a phase-lock configuration, the only way to properly demodulate a DSB suppressed carrier signal is to use sufficiently narrow selectivity at the receiver to filter out one sideband and copy the DSB signal as SSB. But in the process, 6 dB of signal, and 3 dB of s/n ratio is lost. (We regain 3 dB of s/n ratio with the narrower receiver bandwidth filter.)
With SSB, the phase of the BFO is of no concern, but for the audio to come out right, the frequency still must be precisely the same as that of the suppressed carrier, although a few Hz of frequency error is usually of little or no consequence with speech transmission (but not the case with music). The phase of the demodulated SSB audio is still a function of the phase of the BFO, but there is no opposite sideband that it has to be in proper phase with, so the phase of the BFO signal is unimportant.
With normal AM reception using an envelope detector such as the conventional diode detector, this is not supposed to be a problem, since the carrier transmitted as part of the AM signal is used to demodulate the sidebands, and this carrier is inherently on frequency and in proper phase relative to the sidebands - unless the amplitude and/or phase relationship between sidebands and carrier is altered by the ionosphere; then we get the familiar selective fading distortion. That's one of the reasons why a synchronous detector designed to lock the local BFO onto the transmitted AM carrier enhances AM reception even with full carrier AM.
Interestingly, if you take a DSB AM signal, and rotate the phase of the carrier exactly 90 degrees, relative to the sidebands, you end up with pure phase modulation. The resultant amplitude variations of the carrier, caused by the AM USB and LSB components, are exactly out of phase and cancel each other. So the amplitude modulation nulls out, leaving only the phase modulation of the carrier.
That's how Armstrong generated his first FM signal in the 1930's. He generated a double-sideband signal at an extremely low carrier frequency, right at the upper limit of the range of human hearing. Then he rotated the carrier 90 degrees, and multiplied the resulting phase modulated signal many times, as necessary to reach the VHF transmit frequency. He then converted the phase modulation to frequency modulation by imposing the appropriate pre-emphasis (frequency response) curve between the audio source and the modulator.
Getting back to the AM signal, the other reason why the synchronous detector enhances full carrier DSB AM reception is that it allows us the full advantage of the product detector for receiving AM. The vast majority of the "advantage" of SSB over AM is due to the advantage offered by the product detector in the receiver. With properly functioning product detection, the only demodulated audio we hear is the heterodyne products between the local BFO and the sidebands, plus noise products within the receiver passband that heterodyne against the BFO. The very principle of operation of the envelope (diode) detector is intermodulation amongst all signal products that exist within the receiver passband, including the desirable signal as well as all the unwanted interference and noise.
With the envelope detector we have, in addition to the heterodyne products between the carrier and the signal and noise as mentioned above, heterodyne products between every element of noise and interference falling within the passband and every element of signal that make up each of the sidebands, plus heterodyne products between each element of noise/interference and every other element of noise/interference falling within the passband. All these additional extraneous intermodulation products resulting from noise and interference products heterodyning with each other and with each element of the desired sidebands, produces an additional jumble of noise that degrades the signal-to-noise ratio of the received signal.
|