Hi Jose,

> I had never given too much thought to that idea before, but two I/Q ISB 
> signals should be used to be translated to the working frequency, using 
> the same two carriers in quadrature on the PC. That allows to use 
> existent hardware, and I believe it is advisable to use I-Q signals to 
> be able to take adventage of image rejection mixers as many designs I 
> have seen (Softrock, YU1LM, etc) do to keep spurious outputs low.
It is much simpler. We want two independent SSB signals at the antenna.
One LSB and another USB. They should be independent so the (suppressed)
carrier does not have to be the same. Both carriers just have to be
separated by less than the tuning accuracy needed at the receive side.
It is perfectly fine to use two conventional SSB transmitters with
separate antennas. On the RX side you would simply have the summed
electromagnetic field from the two transmitters.

It is self evident that one could use two independent SSB exciters
and combine the two signals in a -3dB hybrid and then use a common
power amplifier with a single antenna to achieve the same result.

The Softrock (or any other quadrature mixer) is a linear device. It
means that one can produce two independent SSB signals in the form 
of two I/Q pairs from two independent computers. Those signals can 
be summed in a resistive network to feed a single Softrock. That 
would produce identically the same RF signals as would two independent
Softrock exciters. One would use existing Softrock software and 
use two independent computers. The operator may send two independent 
signals. He could make one SSB and another SSTV. He might make 
one 10 dB stronger and 10 kHz higher in frequency than the other 
or he might make one USB and the other LSB and use approximately 
the same carrier frequency for both.

All this is easy with existing hardware and software:-)

It should also be possible to run two instances of the Softrock
software in a single computer and send the audio from both
transmitters to the same soundcard. That would give exactly the
same result as doing summation in a resistive network.

> Dreaming a bit, I believe that ISB could be something to consider for 
> the Linrad transmitter.  As I see it, usually the ISB modulator should 
> be disabled and the transceiver should work as a common SSB tranceiver. 
> If someone who knows what he/she is doing desires to use ISB, then that 
> could be enabled. Actually, the whole transceiver should be capable of 
> being toggled from SSB to ISB and viceversa, perhaps by some 
> configuration command.
ISB as such is trivial. The problem is the user interface. Currently
Linrad has CW with extremely well shaped envelope for extremely narrow
bandwidth (no keying clicks) There is also SSB with a rather complicated
speech processor. It would be possible to pick up another signal from
the microphone input (which would have to be stereo then) and to 
place it anywhere in the passband in any desired form.  

> Also, how much extra code is required must be weighed, as it is to be 
> expected to increase the CPU loading as well, so it may not be something 
> for the feebler CPU's. There are quite a few details to be worked out, 
> without losing sight of common sense.
> 
> About inputs, on Windows, some sort of "virtual audio cable" might prove 
> convenient, say, for SSB/SSTV. On Linux perhaps some juggling with Jack 
> might allow it.
The easy solution would be to run whatever SSTV software one would like
and send the output to the same soundcard where Linrad sends its I/Q.
There is no standard for receiving ISB as far as I know so there is no 
reason to use the same (suppressed) carrier frequency. It would be better
to move the passbands a bit closer. For ISB one would use something
like 300 to 2700 Hz on each side leaving 600 Hz of unused spectrum in
between. It would be better to move the SSTV signal closer.

All sorts of new modes are possible. The CPU load is no problem at all,
the baseband processing is at a low sampling rate. I will not introduce
these modes however since it is so easy to run two instances of Linrad
from the same hardware to receive two signals simultaneously. One could
be SSB, the other SSTV. It is almost as easy on the transmit side.

I do not think ISB is the proper way to go. I would rather think that
it would be better to extend the SSB bandwidth and place the SSTV
signal above the SSB signal. That would make the two "independent" 
transmissions easier to receive on a conventional receiver, an SSB
receiver with 5 or perhaps 6kHz banswidth.

As I understand it, ISB is a very special thing. The idea is that 
simple receivers use an AM detector to receive the sum of both 
sidebands while a more sophisticated receiver would receive the 
sidebands independently. That means that the sidebands are not 
really independent because they are designed to be received together 
by some users. There has also to be a carrier that conveys no 
information and doubles the transmitted power. That is to allow
the standard household receiver with its AM detector.

73

Leif / SM5BSZ

Reply via email to