[GreenKeys] FSK converter madness

[Jim Haynes]

On Fri, 5 Jul 2024, Harold Hallikainen via GreenKeys wrote:

> Following up, issues like this is why I really like having a terminal unit
> act like a Teletype. It should be able to key a floating loop and sense
> the current in a floating loop. An external loop supply with all equipment
> (KSR printer, TD, Reperf, and TU) have the same interface to the loop.
>
> Harold
> https://w6iwi.org/rtty/DspTU2/
>
And this has been one of the intellectual challenges of the last 2/3
of my life.

On the one hand, we would like to have things run on 60 ma loops so we
can just plug in standard TTY gear and have it work like it did in
history.

On the other hand, putting things in loops makes simple things more
complicated, like you have to use closed-circuit jacks for all the loops,
and you have to have dummy loops to keep machines from running open
when they aren't being used.

So if I were designing a set from the ground up I would have every
selector magnet use a solid-state driver, input voltages something
like RS-232 or MIL-STD-188, and selector magnet drivers biased so
that with the input not connected to anything the selector magnet
is held marking.  Every keyboard or tape reader would likewise be
hooked up to generate the low-voltage signals.  You have one or more
signal hubs, and everything can send into a hub and everything can
receive from a hub.  (Using gates so with multiple senders any one
of them can drive the hub to spacing.)  The low voltages can easily
be interfaced to electronic gear, so if you want to stick in a solid
state speed converter or a regenerative repeater it is duck soup.

I would insist on bipolar voltages like RS-232 or MIL-STD-188 because
in a radio setup there is RF floating around and it will be rectified
by all kinds of devices such that you don't get zero volts when nothing
is happening but the transmitter is running.

Many years ago I built my station along these lines, at first using
tubes since we didn't have any good high voltage transistors at the
time to use for selector magnet drivers.  Later I moved to a modified
RS-232 scheme when we got good transistors for the purpose.  But I
never got very far with it because I'm lazy and it was too much work
to build all those selector magnet drivers and I was constantly needing
more of them.  And I was somewhat more interested in having a historical
kind of installation where the 60m loops were part of the scheme.

Things were complicated by irv Hoff's insistence that you should always
be able to plug in a keyboard and printer in series into a loop and
have everything work.  I did some work on a "leg combiner" which could
take a keyboard and printer in series and tell the difference between
signals generated by the keyboard and signals coming from outside.