[Elecraft] OT: USB and LSB - How we got there

[Ian White]

In 2003 I researched the subject for my RSGB Q&A column, 'In
Practice', and was fortunate to be in contact with some amateurs who
were personally involved in the decision to switch sidebands at
10MHz. This decision was made in April 1952 and eventually became an
IARU standard - but its origins are surprisingly technical. The
standard came out of two totally unrelated design decisions, made by
different people on different continents, and at different times in
the late 1940s.

Those post-war years saw a rapid development in intercontinental HF
telephone links. These links used independent-sideband (ISB)
modulation to carry two separate voice channels on opposite
sidebands, and a major manufacturer of ISB equipment at this time
was the Marconi company. The ISB signal was created by up-converting
two separately generated USB and LSB voice channels to the same
suppressed carrier frequency, and the Marconi engineers made the
smart decision to generate the ISB signal on 10.000MHz (a frequency
on which they would never need to transmit, because it was already
occupied by beacons such as WWV). 

For transmitted frequencies above 10MHz, Marconi used a
crystal-controlled LO that was 10MHz below the output frequency; so
the IF frequency was added to the LO and the two independent
sidebands remained "the right way up". But for transmitted
frequencies *below* 10MHz, the LO frequency was 10MHz *above* the
output frequency; so the IF frequency was *subtracted* from the LO
and the opposite sidebands were *inverted*. In an ISB system, that
meant that the two telephone channels might very easily become
swapped, so station engineers all around the world needed to be sure
when to flip the appropriate switches. 

Out of these working arrangements between engineers, a worldwide
CCIR standard emerged that 10MHz would be the frequency where the
sidebands in ISB systems changed over. 

So what has this to do with amateur SSB? Amateur development in the
late 1940s quickly followed the developments in commercial world -
and sometimes involved the same individuals. A major influence was
the W1DX SSB exciter, published in 1949, which automatically
produced a sideband inversion between 80m and 20m. The W1DX design
used the phasing method which allowed easy sideband selection by
flipping a switch at AF, but by the early 1950s there were also many
filter-method exciters that were not so agile.

By 1951-52, experimenters in Europe and the USA were beginning to
talk to each other on 20m, and quickly realised that they were
heading for a mess. Most people were using USB on 20m, but there was
no international agreement on 80m... and what about the other bands?
>From eyewitness accounts, April 1952 was the moment when the
agreement crystallized as we know it today. 

The two key points in this history are: amateurs were *already
aware* of the commercial dividing line at 10MHz; and the popular
W1DX exciter was *already compatible* with the new proposed standard
[1].

And so it was that two entirely separate and obscure design
decisions - by Marconi engineers and by W1DX - came together to
create the standard that we have today.


[1] SSB exciters using 9MHz SSB generation and a 5MHz VFO are not
relevant to this history. They all came *after* the 10MHz standard
was already in place. 

73 from Ian GM3SEK