[Elecraft] Vocal Cord Bias + SSB PEP

[Bob McGraw K4TAX]

 From my AM broadcast days and my pro audio days, yes it is not unusual 
to find that a given mike has more positive output than negative output 
in terms of the voltage waveform.  Thus if phased correctly, and we are 
only using one mike, the positive modulation peaks will be greater than 
the negative modulation peaks.  There is advantage to this with regard 
to AM stations.    At the same time, when multiple mikes are used then 
they all must be "in phase" in order to prevent a comb effect of 
frequency cancellation.  It is standard convention that a positive 
pressure on the mike diaphragm, then amplified to the speaker system 
causes the cone of the speaker to move forward or compress the air into 
the room.  Positive pressure thus causes positive pressure.   Some 
audiophiles claim they can hear the difference.   The jury is still out 
on that point.

 From a SSB point of view, it starts basically a modulated carrier, i.e. 
AM, which then has the carrier suppressed by the balanced modulator, and 
then one sideband removed by passing the DSB signal through a sharp and 
steep side filter to produce SSB.   Hence, then one can theoretically 
attain more positive modulation than negative.   Don't confuse amplitude 
energy with frequency spectrum energy.    At the same time, in a SSB 
transmitter, excessive audio can over modulate the carrier, before 
carrier suppression and before the sideband filter, and the same effect 
as over modulating an AM transmitter will exist.  Distortion and 
splatter.    Look at the opposite sideband of a signal on the air while 
using a SDR or spectrum display.  Clearly some signals will show these 
artifacts while others do not exhibit the artifact.

Final rule, except in a SSB transmitter the positive output is limited 
by the available capacity of the amplification stages which must remain 
linear.  In the SSB world,  maximum PEP value capability is the 
ceiling.  On the other hand, in an AM transmitter, the modulation must 
be limited to 100% negative so as not to cut the carrier, but can be 
allowed to exceed 100% positive, in fact upwards of 125% to 135% is 
quite normal, as long as the transmitter is capable of handling the 
increased positive peaks without distortion.   Makes for "loud" AM 
stations.  And of course there are "processors" which control the 
negative going values while enhancing the level of the positive going 
values in order to enhance the modulation.   Of course if the hams AM 
transmitter is not capable of 135% positive modulation, splatter and 
distortion would be generated making for unfriendly neighbors near the 
frequency.

Oh my, this is a lot more than I care to recall.

73
Bob, K4TAX







On 9/14/2018 3:24 PM, Nicklas Johnson wrote:
> An audio engineering friend noticed in some voiceover recordings something
> that I've also noticed in my own waveforms, that a lot of the time, human
> vocal cords seem to produce a biased waveform (seemingly a positive bias
> most of the time).
>
> I vaguely remember reading either here on this list, or maybe in QST,
> something about this as it relates to setting microphone gain (and bias) so
> that the resulting modulated signal is making more efficient (I'm probably
> not remembering the right word) use of the available amplitude (ie, not
> hitting the peak early on the positive side, but falling well below it on
> the negative side).
> My Google-foo isn't working out for me today though; does anyone remember
> an article or posting on this topic and/or have a link to it?  I haven't
> been able to track it down again.
>
>     Nick
>
>