[ARC5] Lopsided modulation

[Tom Lee]

There seems to be a danger of erroneously thinking that asymmetrical 
*time* waveforms somehow imply asymmetrical sideband *spectra*. I've 
only been talking about the latter, as that was the subject of the 
initial conundrum.

Arbitrary time waveforms will almost never be symmetrical, so asymmetry 
there is the norm, but that has absolutely nothing at all to do with 
whether the upper and lower *sidebands* of an AM signal will fail to be 
mirror images of each other. If I feed a perfect AM modulator with a 
horribly asymmetric time waveform (produced, say, by a limiter with 
asymmetric limits), the sidebands produced will still be perfectly 
symmetrical.

Cheers
Tom

-- 
Prof. Thomas H. Lee
Allen Bldg., CIS-205
420 Via Palou Mall
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu
650-725-3383 (public fax; no confidential information, please)

On 2/28/2018 11:31 AM, Richard Knoppow wrote:
> For the asymmetrical voice begin by visualizing a sine wave with a 
> line to indicate the zero volts going along the center. Now, imaging a 
> sine wave with one side flattened. That is asymmetrical. If the center 
> line is still in the same place the voltage on one side is larger than 
> the voltage on the other.
>    The human voice is generated by forcing air from the lungs past the 
> vocal cords. The vocal cords act as a valve. Since the pressure is in 
> one direction only the modulated air stream is mostly more positive 
> pressure than the average air pressure. However, since the voice is 
> also partially generated by a number of resonant chambers in the 
> throat and head, which are after the vocal cords, the pressures are 
> modified so that its not all more positive than the average air 
> pressure. The displacement from average depends on the individual 
> voice: some voices are very asymmetrical, some are more nearly 
> symmetrical. I don't think any are more negative going because the 
> voice starts out as a positive pressure from the lungs. The vocal 
> cords can vary or modulate this pressure but can not reverse it.
>     If one picks up the voice with a purely pressure sensitive 
> microphone the asymmetry is reproduced as an asymmetrical electrical 
> wave form. However, since microphones have some resonances the 
> symmetry or lack of it may be affected. Some microphones, namely the 
> figure eight type or cardioid type are sensitive to either air 
> particle velocity (figure 8) or both velocity and pressure (cardioid). 
> The figure-8 pattern results from sensing a pressure difference 
> between two points in the pressure wave so the microphone, in effect, 
> performs a first order differential on the wave, and thus looses any 
> constant pressure. That results in a waveform which may be 
> asymmetrical but not in the same way the original pressure wave was 
> asymmetrical. All this is made much clearer with a couple of drawings 
> than it is in words.
>     In any case, if you have an oscilloscope attach a microphone to it 
> and talk into it. You don't even need a modern scope, one that 
> responds to audio frequencies is adequate. The scope should be DC 
> coupled. It will show the asymmetry in the voice.
>
> On 2/28/2018 10:17 AM, Tim wrote:
>> Yep!
>> I'm waiting for Prof. Lee to hit us with the all-defining modulation 
>> math.  (leaving it as an exercise for THIS student would not be 
>> productive!)
>>
>> Also. I can imagine asymmetrical sideband power on either side of an 
>> AM radio carrier but I cannot imagine what an "asymmetrical voice" 
>> could be. ;o)
>>
>> Tim
>> N6CC
>