On Dec 13, 8:27�pm, (Mark Kramer) wrote:
In article
om,

The story I heard is that this started because Collins radios had
an IF of 9 MHz and only needed one set of (expensive) sideband
filters to have
LSB below 9 and USB above. Mixing to get the final output:
F1+F2 gives
same sideband you start with, F1-F2 inverts.

That's what I was told.

You were told wrong. That mixing scheme does not invert the
sideband.

This amateur radio urban legend has been around a while, but that's
the first time I heard it attributed to Collins.

Here's what really happened:

A few hams were using SSB in the 1930s. W6DEI, Ray Moore, was
on the ham bands with SSB in 1934 or earlier, and the use of SSB
is promoted in "200 Meters And Down", first published in 1936. But
it wasn't until after WW2 (late 1940s) that SSB began to become
really popular with hams.

The HF ham bands in those days were just 80, 40, 20, 11 and 10 meters.
US hams didn't get 15 meters until 1953 or 54, and didn't
get 30, 17 and 12 meters until after 1979. 160 had been a popular
band before WW2 but it was reallocated to LORAN during the war
and we got it back a little at a time over many years, with all sorts
of restrictions, until LORAN was finally phased out.

40 meters had no 'phone segment until 1953/54, either. 11 and 10
were great as long as there were sunspots.

So the two most popular 'phone bands were 20 and 75, which were
much narrower then than today. For some reason, the 75 meter SSB folks
chose LSB, while the 20 meter gang went for USB. Most rigs of the time
could do either sideband on either band, but the convention took hold
early on and never changed. When US hams got 15 meters, it seemed
natural for 15 to be USB, and when 40 got a 'phone band, it was
equally natural for 40 to be LSB. It had nothing
to do with Collins or 9 MHz IFs.

Now about the mixing scheme:

None of the early Collins ham rigs had a 9 MHz IF. Collins made
mechanical filters, which were only practical up to about 500 kHz
or so, and sideband choice with them was a simple matter of
switching a BFO crystal.

Some early SSB transmitters did indeed generate SSB at 9 MHz, and then
heterodyned it to 75 or 20 by mixing with a VFO in the
5 MHz range. But that mixing scheme *does not* invert the
sideband! If you take a 9 MHz SSB signal and mix it with a 5 MHz
VFO signal, you'll get the same sideband out of the mixer as you
put in, regardless of whether you add for 20 or subtract for 75.
The VFO will tune backwards on 75, but the sideband will not
invert. Those early rigs all had USB/LSB switches in order
to match the convention.

The only way to get sideband inversion by mixing is if the local
oscillator frequency is higher than both the input and output signal
frequencies. The 9 MHz-IF story isn't just an urban legend - it
flat out doesn't work.

It *is* true that if you generate SSB in the 5 MHz region and mix it
with a ~9 MHz VFO, you'll get the desired sideband inversion. Some
1960s SSB rigs used that scheme - but they were long after the
convention was in place. And the urban legend doesn't mention
those rigs.

If you want to see the math on how all this works, I can post it.

73 de Jim, N2EY