[GreenKeys] Teletype power supply design

[John Nagle]

    As a design exercise, I've been building SPICE models of possible
switching power supplies for Model 15 Teletypes.  The idea is to get
rid of the bulky and hard to find 120VDC supply and inefficient load resistor.

    The usual setup for a Model 15 Teletype is a 120VDC 60mA supply with
a 2K 10W resistor in series.  120VDC is needed to load up the selector
magnet (4H inductance, 55 ohm DC resistance) in under 1ms.

    But 120V is only needed for the first 1ms after a SPACE to MARK
transition to force 60mA through that huge inductance.  In about 1ms,
a steady state is achieved, and only 3.3V is needed to maintain 60mA.

    So I'm looking at a design where, during SPACE periods, a capacitor
is charged up to 120V.  On a SPACE to MARK transition, the capacitor is
discharged into the selector magnet.  A 3.3V source is also fed in,
so that once the capacitor is discharged, there's a sustaining source
to keep 60mA flowing.

    At 45.45 baud, we have a minimum of 22ms dwell time in SPACE for
charging.  Only 17.5ms at 56.9 baud (75WPM), which is the usual top speed
for a Model 15, and at 75 baud (100 WPM), only 13ms.  Beyond 75 baud,
60mA current loops aren't usually used, I think.  So we need to be
able to develop 120VDC in 13ms.

    How big a capacitor needs to be charged up?  About 1uF.  It takes
about 0.0072J to load up a selector magnet (I=0.060 amps, V=120 volts,
L=4 henries, W = 1/2 L*I*I = W = 1/2 C*V*V), and a 1uF cap at 120V
holds that much energy.

    Charging is done with an IC intended for photoflash capacitor charging,
the LT3484-1.  This is a specialized switching power supply in a single IC.
It takes few external components, the main one being a tiny transformer.

    The whole thing could be powered from a 6VDC 250mA wall-wart type power
supply.  No more need for a 120VDC supply and a big, hot current limiting
resistor.  There's also the nice property that the stored energy is well below
the 1-Joule safety limit for hazardous electric shock.  So the electronics
doesn't need to be protected as a high voltage device.

    Below is a preliminary SPICE model of the charging circuit, just
to show that this is possible.  This will work with the free version of
LTSpice IV (http://www.linear.com/designtools/software/).  If you
want to run this, download LTSpice IV, paste the text below into
a file and name the file something like "ttytest.asc", and run it
with LTspice.  Then you can look at the voltages and currents.
The transformer isn't a stock part; I have to find a suitable one and
redesign to it.  Nor is this a full design; it's just a proof of concept
circuit simulation.

     Is this something people are interested in, or are users content
with power supplies oversized by modern standards?

Version 4
SHEET 1 1028 680
WIRE -224 -112 -224 -160
WIRE 0 -112 -224 -112
WIRE 96 -112 0 -112
WIRE 224 -112 96 -112
WIRE 288 -112 224 -112
WIRE 384 -112 288 -112
WIRE -224 -48 -224 -112
WIRE 288 -32 288 -48
WIRE 0 16 0 -112
WIRE 96 48 96 -112
WIRE 224 48 224 -112
WIRE -224 64 -224 32
WIRE 384 80 384 -112
WIRE 512 80 432 80
WIRE 624 80 624 64
WIRE 624 80 576 80
WIRE 0 112 0 96
WIRE 48 112 0 112
WIRE 384 160 272 160
WIRE 624 160 624 144
WIRE 624 160 432 160
WIRE 48 208 -224 208
WIRE 624 208 624 160
WIRE 160 304 160 272
FLAG 160 304 0
FLAG -224 64 0
FLAG 624 208 0
FLAG 624 64 OUT
FLAG 288 -32 0
FLAG -224 288 0
FLAG -224 -160 Vcc
SYMBOL ind2 368 64 R0
WINDOW 0 -49 34 Left 0
WINDOW 3 -52 75 Left 0
SYMATTR InstName L1
SYMATTR Value 10µ
SYMATTR Type ind
SYMATTR SpiceLine Rser=2
SYMBOL ind2 448 176 R180
WINDOW 0 -47 76 Left 0
WINDOW 3 -64 43 Left 0
SYMATTR InstName L2
SYMATTR Value 140µ
SYMATTR Type ind
SYMBOL voltage -224 -64 R0
WINDOW 123 0 0 Left 0
WINDOW 39 24 132 Left 0
SYMATTR SpiceLine Rser=10
SYMATTR InstName V1
SYMATTR Value 6v
SYMBOL diode 512 96 R270
WINDOW 0 32 32 VTop 0
WINDOW 3 0 32 VBottom 0
SYMATTR InstName D1
SYMBOL cap 608 80 R0
WINDOW 0 43 31 Left 0
SYMATTR InstName C2
SYMATTR Value 1µF
SYMBOL res -16 0 R0
SYMATTR InstName R1
SYMATTR Value 100K
SYMBOL cap 272 -112 R0
SYMATTR InstName C1
SYMATTR Value 10µf
SYMBOL voltage -224 192 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V2
SYMATTR Value PULSE(0 5 5ms)
SYMBOL PowerProducts\\LT3484-1 160 160 R0
SYMATTR InstName U1
TEXT 344 184 Left 0 !K1 L1 L2 0.98
TEXT -128 336 Left 0 !.tran 25m startup
TEXT 288 328 Left 0 ;60mA Teletype power supply design test.\nPRELIMINARY\nJ. 
Nagle (nagle at animats.com)\nApril 2009


				John Nagle