[GreenKeys] ST-6's in this day and time (also old enough to remember all those tobacco ads)

[Jim Haynes]

On Thu, 26 Mar 2020, Ethan Blanton wrote:

> That's me!  I'm glad you enjoyed it; I have plans for a couple more
> videos with the ST-6, one figuring out what's wrong with the motor
> control circuit (if anything) and one comparing its operation and
> performance with and without the limiter engaged.  However, I have too

When you talk about the limiter engaged, the ST-6 is far from correctly
designed.  The late Don Wiggins W4EHU explained all this in some article
in RTTY circa 1960.  If you're going to use a limiter you want the sharp
filtering ahead of the limiter, to try to make the desired signal as
strong as possible compared to the interference.  The limiter improves
the SNR for whatever is the strongest signal at its input.

But for that to work, when you use a limiter it needs to be followed by
a linear discriminator that is linear far beyond the shift of the FSK
signal.  The reason is that the limiter, with interference, produces
spikes in the instantaneous frequency of its output.  If you have a
wide linear discriminator these will average out so that the resulting
signal is close to the desired signal.  Then the low pass filter that
follows the discriminator will do the averaging.  If the discriminator
is sharply peaked, by using narrow filters such as are appropriate for
two-tone detection, the spikes do not average out.  So there needs to
be more done that just switching the limiter on and off.

But it seems like in these modern times the two-tone demodulators without
limiting usually outperform the limiter-discriminator demodulators.

The author of the Two Tone demodulator software told me that when he
read the early papers on two-tone detection he built a hardware 
demodulator working on those principles and it outperformed the ST-6.
But he didn't go into any detail and I don't know if he published
anything about the design.  Nowadays if you want high performance you
are going to use DSP to get it.

Of course for ITTY the signals are nearly perfect to begin with, so
practically anything will demodulate error-free.  It's a different
matter when you go into HF radio in the middle of a contest.