[GreenKeys] Keyboard Reperf Wiring

[Jim Haynes]

If you want to consider something completely different...

Back in 1959 as a summer student at Teletype I was given a project of
a solid state replacement for the line relay.  At the time there were
practically no transistors that could handle high voltages like 120V
loops.  I didn't come up with a replacement for the line relay, but I
did propose a low voltage selector, 12V, 600 MA using the power 
transistors that were then being made for the audio output stages of
car radios.  I wasn't the first to think of it - my boss at the time
Bob Reek had thought of it too, but I didn't know that at the time.

What I considered more important, but failed to sell, was the idea that
the interior of a set should not use current loops but instead should
use voltages, maybe plus and minus 6 volts.  This was before RS-232
and for a different reason, but there is some similarity there.

You know if you have a reperf inside a set and it is not going to copy
anything then you have to provide loop current to the selector magnet
to keep it from running open.  And in general with current loops inside
a set you have to maintain current in the loop, and to cut something
into a loop you have to open the loop and insert the something (2 wires).

With the voltage proposal every selector magnet would have its own
transistor driver, and every signal generator would have a circuit
to develop +/- voltage from the contacts.  It would be simple to bias
the selector magnet drivers to maintain a closed current loop to the
magnet with no input.  Then it's just a simple matter to switch signal
generators and selectors using normally-open switches or diode gates.
If you have to connect to an external 20 or 60 ma loop, use a relay or
whatever (we didn't have optoisolators yet) to convert the signals between
the internal voltage levels and the external loops.  And it would be
simple to put in things like transistor electronic regenerators if
that's what the customer wanted.  (Teletype folks thought the electronic
regenerator would be too costly and nobody would ever buy it.  With my
RTTY experience I figured organizations using a lot of RTTY, like the
military, would gladly pay for it.)  Patching jacks are as simple as
those for audio circuits - you don't need the extra contacts to keep
loops closed.

As we know the low-voltage selector magnet and transistor driver did
go into effect for the 32/33/35 line, but not the concept of voltages
rather than currents.  The driver circuit is a lot more complicated than
I had in mind, but maybe to achieve 110 baud it was helpful to take
advantage of the transistor constant-current ability for faster response
of the selector magnet.  Bell Labs for some reason decided the modem
for dial TWX should use loop currents rather than voltages.  Though
of course when RS-232 came along it was necessary to add 
voltage-to-current converters so the Teletype sets could work with the
later modems.

Some years later I built my own 35 ASR set the way I thought it should
be done.  Power switch for the motors.  There are two circuits in the
machine, one local and the other line.  For the printer there is a toggle
switch between the two circuits so the printer is either on line or local.
Same for the keyboard.  For the tape reader and reperforator each has
three pushbuttons - line, local and off.  Pushbuttons because there are
relays to control the actual switching.  So that if desired the stunt
box contacts can put the reader and punch into either circuit.

None of this K-KT-T confusion.  K-KT-T suits exactly the way the Model 19
keyboard and punch work.  It doesn't make much sense when the punch is
a reperforator, and when the tape reader is included in the scheme.
(In Model 19 the XD has a separate loop and there is additional switching
needed to control just what it is connected to.  And there is only the
tiny excuse for a stunt box if you want a reperforator to be turned on
and off by incoming signals.)

One reason for K-KT-T is that in T position the keyboard is not hobbled
by the 60 wpm signal generator, so the operator can punch much faster
than 60 wpm.  In Model 35 the machine is running at 100 wpm so there is
a lot less need to accomodate a fast keyboardist.

Sorry to be so long winded, but please think about how much better off
we would be if we didn't have to deal with all those current loops.

Jim W6JVE

 	---

 	"Ya can argue all ya wanna, but it's dif'rent than it was."
 	"No it ain't! No it ain't!  But ya gotta know the territory."
 		Meredith Willson, The Music Man