[GreenKeys] HRS312 TTY Interface

[John Nagle]

From: "Duncan M. Brown" <duncanancy at earthlink.net>
Subject: [GreenKeys] HRS312  TTY Interface
 > To: "greenkeys" <greenkeys at mailman.qth.net>
 > Message-ID: <380-22009321005254760 at earthlink.net>
 > Content-Type: text/plain; charset="us-ascii"
 >
 >
 > I've been thinking about making a HV interface for a computer to drive a
 > TTY loop for some time, but never had gotten around to it. This would
 > mainly be for a RO demo system that could drive a M15 from ITTY or
 > HeavyMetal in the Museum.
 >
 > Then last month someone mentioned the HRS312 MOSFET optocoupler and that
 > looked like the heart of a very simple interface.  Jon Schlegel and I have
 > been playing with the HRS312 and  it seemed like the circuit was so simple
 > that it could be placed inside the 9-pin D-connector's housing.  I found I
 > had some AMP 745854-3 housings. There is room for a 5/8" x 3/4" board,
 > which is plenty big enough (point-to-point wiring).
 >
 > Hopefully this small picture will come through.
 >
 >  Jon measured 500v spikes coming off a M28 selector magnet, so we added
 > back-to-back, 200V zener diodes (1N5956) across the output of the HRS312.

     Right.  That's about what I did with

	http://www.animats.com/nagle/aetheric/connecting.html

That handles both input and output.  If all you need is output,
you just need D1, R5, IC1, C1, R1, and a 120V supply current
limited to 60mA with a 2K 10W resistor.  C1 and R1 are a snubber
to dump the kickback from the selector magnet. 	You could probably
omit D1 and get away with it.  Of course, you still need a 120VDC
supply and that hot 2K resistor, which won't fit in the connector
housing.

I've been chatting with some people on sci.electronics.design about
alternative drive circuits.  I'm trying to eliminate the need for
a 120VDC supply.  It only takes 3.3v to maintain 60mA through 55 ohms;
the 120VDC supply is only needed for the first 2ms or so of each
bit time to overcome the 4H inductance of the coil.  A design
that only requires a +5 supply with a boost supply to make the
small amount of 120VDC required would be a neat solution to the
problem.  The idea is to charge up a capacitor during the 95% of
the time that kick isn't required, then discharge it during turn-on.
I've been playing with SPICE models of this.

The drive circuits in later teletypes, like the Model 33, have a power
transistor, apparently to do something like this.  That's why they
don't need so much loop current.

				John Nagle