[GreenKeys] Polar and Neutral loops

[Phil Schelinski]

Don Robert House <drhouse at nadcomm.com
In the steel mills and at FORD I believe polar subsets were used but they had a vacuum tube rather than the original polar relays .
SOme of the early 40a subsets used a combination of tubes and mechanical relays.
Which reminds me , I stumbled on my AT&T TTY Testboard book that helps explain the varous transmission schemes used, and need to forward it to NADCOM .
I promise this time to put my books in a box and forward it to ????
 
Phil Schelinski
Excellent explanations Dave and Jim. Thank you very much.
Don

Don R. House
Curator - NADCOMM

>
> Polar loops were pretty much the standard method of transmitting
>tty and other telegraph traffic over long distances on wire lines. 
>The standard was plus and minus 20 ma over twisted cable pairs that
>could run into tens of miles.
>
> Polar transmission is symetrical and aviods a lot of kinds
>of distortion that can happen with on off keying, especially if the
>loop is actually open during the current off interval.
>
> Generally, of course, there was a polar relay between the line
>and the Teletype machine that converted the plus and minus current
>swings to contact openings and closings of a local neutral loop
>with the actual selector magnets on it. I have heard of very few
>if any machines that actually used selector magnets driven off a polar
>20 ma source.
>
> Back in the early days of minicomputers when ttys were common
>as console devices, plus and minus 20 ma signalling over a current
>loop was the prefered way of interfacing to the model 33 or 35.
>In theory this signalling would work over much longer distances than
>voltage driven relatively high impedance RS-232.
>
> But 33's and 35's all had transistorized magnet drivers that
>actually drove the selector magnets - they did not have mechanical
>polar relays or high voltage neutral loops. And the magnet drivers
>had resistors in series with the loop to sense current, not inductive
>relay coils so they could actually be driven by simple low voltage
>interface circuits without worry about inductive spikes.
>
> Dave Emery N1PRE, die at dieconsulting.com DIE Consulting, Weston,
>Mass 02493

>From Jim Haynes:

I suppose the traditional way to convert from neutral to polar would be to
use a polar relay, with a bias current, in the neutral circuit and have
the contacts generate a polar signal. As is discussed in lots of places,
you put a constant 30ma current through one winding of the polar relay
in a direction opposite to the 60ma line current in the other winding.

Long ago I had a W.U. 401A tape printer, a design derived from the stock
ticker. I'm pretty sure it worked on a neutral circuit, but I don't
remember how. Perhaps there was a spring attached to the armature so
that it biased the armature to spacing and the neutral current moved
it to marking.

The polar circuit as you suggest is used because it gives a cleaner
signal - variations in the strength of current don't matter much, because
the polar selector operates over a wide range of currents. And it
is symmetrical about the zero axis. The armature is pulled in one
direction by mark and in the opposite direction by space, so in a true
polar selector you don't have the complication that the magnetic pull
increases with decreasing air gap as the magnet moves in one direction
and decreases with increasing air gap in the other direction. If the
positive and negative currents are equal, as they should be, the magnet
behavior is symmetrical in the two directions.

I suppose nowadays it is passe to use a polar relay and you'd like to
do something electronic to generate a polar signal.

-- 
---------------------------------------------------
Don R. House
4716 Patty Lane, Ringwood, IL 60072
Tel: 815-653-0683
FAX: 815-653-0684
*****************************************
North American Data Communications Dept. of CMA
URL: http://www.nadcomm.org
Computer Museum of America (CMA)
URL: http://www.computer-museum.org