Yes, there are differing opinions on the required bandwidth. And indeed
the required bandwidth depends on the nature of the communication AND
the voice. A kid or YL is going to need more top end to be understood
and have no bass while the golden tones radio announcer may need all of
the available spectrum to sound natural, and the survivor of cancer of
the voice box speaking with belches needs on a few hundred Hz starting
at 150 Hz. The necessary bandwidth is a variable.
Still, I like 2.1 better than 2.4 or 2.8.
Terman's selection is probably based on long time telephone research. In
their frequency multiplexed SSB telephone systems, narrowing the
spectrum means they can get more users on a conductor and so there's
plenty of push to make is narrow as possible, yet if we don't recognize
Aunt Nellie on the phone, we may treat her as a telemarketer. Today that
bandwidth limit also applies commercially to digital telephones. The
slower the sample rate the lower the data rate and the greater the range
or the more users can use a limited spectrum.
Vague rules are a good thing, they are hard to enforce to the nit and
they do leave room for experimentation. Without modern experiment we
wouldn't know that adding 3 kHz to the transmit and receive SSB filter
widths added just background noise (ambience?) to the signal which QRMd
near by stations and added to the intermod possibilities of the
transmitter.
Remember that some of our rules have been set by actions of amateurs a
long time ago. The anti pecuniary gain came from amateurs handling
message traffic in the spark days and charging while competing with
Western Union, who thought they had a monopoly on electrical message
communications. That was long about 1912. Then the anti music rules came
from amateurs experimenting with continuous carriers (1920 or 1921) and
modulation used records as tests and to entertain none amateur
neighbors. The broadcasters didn't like the unfettered competition,
though they were very happy to use amateur frequencies and equipment for
their own purposes. Even before modulation the ISU radio club station
broadcast farm market reports on CW and so the ISU radio stations
consider that ham station part of their long history.
Before the S-line Collins filter bandwidths for SSB tended to be wider,
often 3.1 KHz in the A4.
Me, I find the added noise from the wide bandwidths fatiguing, which is
often enhanced by wideband noise from all the IF and audio stages after
the filter to the point that I sometimes use a DSP filter or a passive
low pass on the speaker to clean up all the circuit noise. In one of my
radios I'm going to make the output stage into a simple low pass filter
to help some of that.
As we discuss bandwidth, be aware that there are petitions today to the
FCC to make bandwidths more restrictive, e.g. to put numbers in place of
"minimum bandwidth." We have had quite restrictive rules on bandwidth at
times, such as the first rules for HF RTTY that specified frequency
shift of 850 Hz (no tolerance specified) that were later relaxed to be
under 900 Hz shift and it was soon apparent that narrower shift tended
to work better.
It strikes me that determining transmitted bandwidth isn't trivial. With
multiple signals on a ham band a spectrum analyzer can be confused by
multiple stations. And since that transmitted spectrum depends on the
voice driving the radio that adds complication. Is the rule to be
enforced by examining the filter for its rated bandwidth? What if the
filter was mislabeled by the manufacturer and is 10% wider than
regulation and specification? Who gets pinched, the maker or the user?
What about user modification? At this point we aren't restricted to
never modify our radios. We are allowed to build and experiment AND
MODIFY without oversight.
What if we were to adopt a digital scheme that used 30% wider bandwidth
than ordinary voice bandwidth but allowed 2 dozen QSOs simultaneously in
that same bandwidth? Would numeric bandwidth rules prevent experimenting
with that digital voice mode? Our vague present rules do allow such an
experiment.
I don't know of any petitions to put numbers for received S/N on the
power levels. Typically amateur communications use lower powers than
commercial communications who demand far higher reliability on a path
and its well known that any significant path length has occasional deep
fades. We mutter and put up with the vanishing S/N expecting the strong
signal to return while commercial interests simply set the initial power
high enough to prevent noticing the fade as S/N. Cell phone systems do
send control signals to the mobiles to set their power so they can share
cells better on the same frequency.
I've suffered listener fatigue from Kenwood's narrow CW filter because
it rang on noise making a 200 Hz wide tone, maybe narrower at the same
pitch as received signals so I had to pick out keyed tone from tone. The
ringing characteristic of a filter system do affect ear fatigue.
In the wake of recent aurora events, we need to be sure that any rules
changes don't demand the received bandwidth be as narrow as the
theoretical, e.g. no key clicks and no aurora broadening where 2m CW
signals were at least 2 kHz wide, sometimes 4 or 5 KHz as received. 6m
were a little narrower. SSB was useless on 2m, and nearly so on 6m.
Signal amplitude variations were as diverse as the frequency spread.
73, Jerry, K0CQ
--
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|