[GreenKeys] Polar Relays II

[Jim Haynes]

When I was a student and had a summer job at Teletype one of the little
jobs my boss had in his desk drawer was to look into a solid state
replacement for the polar relay.  At the time we didn't have any good
high voltage transistors that could handle a 120V loop; there were
some circuits that put two or more transistors in series.  It was
many years later that I realized I failed to ask the right questions,
"Why do we want a solid state replacement for the polar relay?  what
problems are we trying to solve?"

I was aware that auto radios of the period were using power transistors
for the audio amplifiers, operating with a supply voltage of 12 volts.
So I thought instead of 120V 60 ma why don't we use 12V 600ma where
the readily available power transistors were well suited.  It turned
out my boss had already thought of this, but he didn't say anything,
but helped me get some Model 28 selector coils wound for 1/10 the normal
number of turns.  Of course that worked fine on a local loop.

The other influence on my thinking was that one of the engineers had
worked out a solid state regenerative repeater.  If that had to be
interfaced to 60ma loops it would be a lot worse than going directly
at low voltage to the 12V selector magnet driver.  Other transistorized
accessories could likewise benefit from a low-voltage interface.
So I proposed an internal set design where every selector magnet would
have its own transistor driver and all the local switching would be
done at low voltage.  Then the interface to the outside world could
be a polar relay or anything else we could devise that would meet the
needs of the transmission scheme - this was again putting transmission
back on AT&T's turf.

I failed to sell that concept.  My boss thought the regenerative
repeater would be too costly to interest any customers, whereas with
my RTTY experience I thought it would be a popular item for the
military especially.  This was several years before Bell System data
sets and other things using RS-232 had come into existence.

A couple of years later I had a visit to Teletype and was informed that
the low voltage selector magnet driver was going to be used in the
Model 32/33/35 products.  Someone smarter than me had redesigned the
circuit from a simple saturated transistor switch to one using the
constant-current properties of the transistor, and had decided upon
20 volts 500 ma as the selector magnet loop level.  Perhaps this
was a good decision since the new products had to operate at 110 baud
instead of 45 or 50 or 75.  I've seen constant-current drivers for
60ma selectors that operate at a lower voltage than 120.

So, anyway, I didn't solve the problem of the polar relay, whatever
that problem was seen to be.  I was focused more on internal design of
the set rather than connecting it to the outside world.

I remember a passage from MIL-STD-188 concerning polar relay replacements.
It required that they have the same features as real polar relays,
including isolation between coil and contacts, multiple "coils" that
added their effects, and "side stable" which meant that with no current
in the coils the relay would sit in either mark or space condition
wherever it had been left.  Those are pretty tough requirements to meet,
so that is probably why polar relay development continued.  The
Automatic Electric relay which Western Union sorely needed to replace
its antique relays, several other competing makes of relays, the 314
relay to replace the 255A...

Just as an aside the AN/FGC-5 vacuum tube time division multiplex had
a really odd substitute for the polar relay.  There was a solenoid coil
in the line circuit, and it surrounded a gas diode.  With no current in
the coil the voltage in the diode circuit was sufficient to fire the
diode into conduction.  But with enough current in the coil the magnetic
field prevented the gas tube from going into conduction.

In the AN/UGC-1 and AN/UGC-3 transistorized time division multiplex
systems we used mercury-wetted contact relays.

And, finally, Western Union and the military and others all went to
low-voltage handling for TTY signals, converting to high voltage loops
only when necessary.  MIL-STD-188 and Western Union Technical Review
are good sources.

 	---

 	"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