[Elecraft] Is the K3 capable of ESSB?

[David Woolley]

Fred (FL) wrote:
> The SSB bandwidth 2007 realities, sound familiar.

Yes, this sort of false reasoning is common in marketing.  They rely on 
a perception that anything new to the market must be better, and that 
the general public doesn't understand the true reasons for limitations.

> In 1978 - max bandwith over telephonic modems,
> and conditioned AT&T lines was like 6250 baud.

The maximum baud rate over telephone lines with the standard analogue 
telephony bandwith is about 2400, and still is.

> A "limit" everyone agreed.
> 
> Then some clever soul or group, came up with
> quadrature modulation - and rather quietly,

6250 bps is not possible without at least the use of quadrature 
modulation.  I think the limit for that is 4800 (although it might just 
be 2400).  I guess, if you pushed the bandwidth to the limit, you might 
get 3125 baud, and therefore 6250 bps with quadrature modulation.

> modems into the hot-copper telephone lines

The limits were not set by the copper, but by the SSB carrier systems 
used on the trunks between exchanges (central offices).  Their bandwidth 
was set to the minimum needed for the general public to consider the 
quality acceptable for speech.

> of 52kbaud became a reality.

What actually happened is that it became possible to  manufacture 
digital signal processors cheaply enough to use them in telephone 
modems, and that made it possible to use echo cancellation and advanced 
equalisation algorithms.  In addition, it required that the transmission 
path from the local office to the service provider be all digital.  The 
modems now attempt to select every possible quantisation level in every 
sample on the digital bearer - this relies on the remote end not having 
an analogue connection (they are only acting as a modem over the local 
loop in the downlink direction)!  The actual 56kbps limit is set by the 
characteristics of the PCM network, including  US robbed bit signalling 
and the need to avoid putting too much power into the equivalent 
analogue signal, thus not achieving the absolute limit of 64kbps.  The 
baud rate is actually 8k, but one gets away with that because the local 
loop analogue connections are not bandwidth limited.

Actually, because PCM phone connections are companded, some of the steps 
between quantisation levels are much smaller than others, so 56k modems 
                  don't actually take full advantage of the signal to 
noise ratio.  Assuming the telephone SNR and that equalisation doesn't 
impose too much of a problem, an analogue radio channel should be able 
to achieve a lot more than 56kbps in 2.7kHz.

Also, a considerable time before this sort of modem became possible 
(and, I think, significantly before the enabling technology of PCM 
bearers became common) it became possible to transmit  communications 
quality speech over a 2400 bps connection.  That suggests that the true 
technical advance would be the transmission of amateur radio speech in 
about a thirtieth of the 2.7kHz, SSB, bandwidth, assuming landline 
telephone signal to noise ratios.

By comparison, there is really no new technology at all in using laxer 
SSB filters.


> rethink its SSB modulation approach, to allow wider
> more naturally sounding voice comms, and still
> not take up unnecessary bandwidth, beyone which
> supposedly 2.7khz now consumes. (when do we

It's the essence of SSB that it takes up the same bandwidth as the 
baseband signal!  To get more natural speech in less bandwidth, you have 
to basically treat the channel as being a digital one, and send a signal 
that takes advantage of the actual nature of speech signals (generally 
these use some sort of voice tract model).



-- 
David Woolley
Emails are not formal business letters, whatever businesses may want.
RFC1855 says there should be an address here, but, in a world of spam,
that is no longer good advice, as archive address hiding may not work.