[BC] Re: SUPER MODULATION & AM STANDARDS

[Hershberger, Dave]

A couple of comments:

1. Super modulation. I recall a paper published by somebody from Gates in
the early 1970s about super modulation. At the time there were worries about
splatter, bandwidth, etc. So somebody at Gates throttled a transmitter way
down in power so that extreme levels of super modulation could be applied
without distortion. It was something like 300% or more on positive peaks.
Although you can put a lot more energy in the sidebands that way, Gates
showed that the occupied bandwidth was not increased through the use of
super modulation. I think that the paper was published in Broadcast
Engineering Magazine, somewhere around 1971 or 1972.

2. CSSB. CCIR did a series of studies in the 1960s that concluded that CSSB
was not a very good idea for European use. The problem was mainly
compatibility with receivers, but also bandwidth. The predistortion terms
would be truncated and/or phase rotated by IF filtering in receivers,
causing envelope detector distortion. Also, the presence of the
predistortion terms somewhat increased the bandwidth of the selected
sideband. I can get the references if anybody wants them.

The various attempts made at CSSB by Leonard Kahn and others were all
approximations to "true" CSSB where only a second order term appears in
single tone modulation. While many of these attempts at CSSB worked with
single tones, they would fall apart to varying degrees with complex
(multitone) modulation. However, now I know how to produce "true CSSB"
exactly, with complex multitone modulating signals, where there only the
minimum second order predistortion terms appear, and there are NO sidebands
at all on the unwanted side of the carrier, and the envelope (in wide
bandwidth) is perfect. Dr. Kerns Powers wrote a paper on single sideband FM
that establishes the relationship between envelope and phase functions for
complex (analytic) signals. SSB-FM imposes a "magic" envelope function on an
otherwise normal FM signal that makes one set of sidebands COMPLETELY
disappear. It's very interesting - but totally useless. The envelope
function is an exponential function of the Hilbert transform of the angle
(phase) modulation. Talk about "super modulation" - this results in extreme
envelope modulation, typically thousands of percent positive and very close
to envelope pinchoff. To restore SSB-FM to normal FM, you just run it
through a limiter to remove the envelope function. But because the SSB-FM
(for reasonable deviation levels) very nearly pinches off the envelope,
you'd get noise bursts from the limiter after demodulation where the
envelope is nearly pinched off. That, plus the very high peak to average
ratio, is what makes SSB-FM totally impractical. But interesting, since ALL
of the sidebands on one side disappear. This phase/envelope relationship
disclosed by Dr. Powers can be used to make CSSB, which is a whole lot more
practical and useful than SSB-FM.

For example, 5 kHz deviation at 500 Hz (modulation index of 10) would
require envelope modulation of -99.99546% and +2,202,747% to produce SSB-FM.
Those are some pretty wild numbers.

About 20 years ago I talked to Dr. Powers about his SSB-FM paper. He said to
me, "Oh, it's a TERRIBLE system, absolutely useless!" But interesting
nonetheless, with information that makes other systems possible (like
"perfect" CSSB).

Dave Hershberger 
Principal Engineer 
AXCERA
Tel: 530-272-HDTV (4388)
Fax: 530-272-4505 
dhershberger at axcera.com <mailto:dhershberger at axcera.com> 
AXCERA. Axcessing the new era of digital communications. www.axcera.com 
960 McCourtney Road, Suite C, Grass Valley, CA 95949-7423