Maker Pro
Maker Pro

Powering a 60mA current loop with less heat

J

John Nagle

I'm powering a Model 15 Teletype current loop. The Teletype has a selector
magnet resistance of 55 ohms and needs 60mA, so it only needs 3.3V. But
because of the huge inductance of the selector electromagnet (4H), a
huge voltage is needed to pull in the electromagnet fast enough to
follow the incoming data stream.

I'm using the traditional solution, a 120VDC supply through a 2K
10W resistor. It's really dissipating 10W, continuously. A 10W resistor
gets hot enough (250C!) that you can't mount one on a PC board. I'm
designing an interface board, and would like a better solution.

Basically, I need a constant-current regulator circuit that
takes in 120VDC and limits the output current to 60mA. Preferably
one that doesn't dissipate too much heat. Suggestions?

John Nagle
 
J

Joerg

Tim said:
Switcher. Possibly with an oversized coil to give your control circuit
enough time to _really_ keep things at 60mA. You'll need the voltage
overhead to supply that relay inductance, but you'll only be supplying
that power when you need to.

Of course, if you have to ask it'll take you some time to make it work,
but that ain't _my_ cross to bear (heh heh heh).

Perhaps a SEPIC switcher that has enough oomph to ramp it's voltage from
near zilch to tens of volts within a reasonable speed. AFAIR teletype
never ran more than 100bd so this could work. You'd have to regulate for
current and keep a cap on the voltage (in case the loop connection goes
open). 20:1 is a stretch but possible so you should be able to get to
60V or so.

Another option would be to look at what's available in the stepper motor
driver world. They operate in similar fashion, huge voltage overhead to
force current into the winding inductances.
 
S

Spehro Pefhany

I'm powering a Model 15 Teletype current loop. The Teletype has a selector
magnet resistance of 55 ohms and needs 60mA, so it only needs 3.3V. But
because of the huge inductance of the selector electromagnet (4H), a
huge voltage is needed to pull in the electromagnet fast enough to
follow the incoming data stream.

I'm using the traditional solution, a 120VDC supply through a 2K
10W resistor. It's really dissipating 10W, continuously. A 10W resistor
gets hot enough (250C!) that you can't mount one on a PC board. I'm
designing an interface board, and would like a better solution.

Basically, I need a constant-current regulator circuit that
takes in 120VDC and limits the output current to 60mA. Preferably
one that doesn't dissipate too much heat. Suggestions?

John Nagle

What's the shortest time that the coil is "off"?
 
S

Spehro Pefhany

I'm powering a Model 15 Teletype current loop. The Teletype has a selector
magnet resistance of 55 ohms and needs 60mA, so it only needs 3.3V. But
because of the huge inductance of the selector electromagnet (4H), a
huge voltage is needed to pull in the electromagnet fast enough to
follow the incoming data stream.

I'm using the traditional solution, a 120VDC supply through a 2K
10W resistor. It's really dissipating 10W, continuously. A 10W resistor
gets hot enough (250C!) that you can't mount one on a PC board. I'm
designing an interface board, and would like a better solution.

Basically, I need a constant-current regulator circuit that
takes in 120VDC and limits the output current to 60mA. Preferably
one that doesn't dissipate too much heat. Suggestions?

John Nagle

Just a thought here.. what if you had an inductor that had similar
characteristics to the selector magnet. And a 55 ohm resistor to
ground and a 6.6V supply. Now connect one side of the inductor to 6.3V
and use the data to switch (using something like a 400V SPDT optomos
relay with some TVSs) the other side from the 55 ohms to ground (0) to
the selector magnet (1).
 
W

whit3rd

     I'm powering a Model 15 Teletype current loop.  The Teletypehas a selector
magnet resistance of 55 ohms and needs 60mA, so it only needs 3.3V.  But
because of the huge inductance of the selector electromagnet (4H), a
huge voltage is needed to pull in the electromagnet fast enough to
follow the incoming data stream.

     I'm using the traditional solution, a 120VDC supply through a 2K
10W resistor.  It's really dissipating 10W, continuously.

How about a class G amplifier: use a sense resistor
and op amp to drive two output circuits, one with
a modest (12V) power supply, the other with your 120VDC.
Use an isolator diode on the 12V output, so it doesn't blow
up when 120V shows up on the Teletype...

On turn-on, the sense resistor shows less than 60 mA
current, so the op amp output rises; at 3V, the low-voltage
output circuit turns on, but (because of the inductance of the load)
current is still too low. The op amp output rises to 4V, which
turns on the 120V section of the output drive, and the output
current rises to 60 mA. A few microseconds later, the
current hits 60.1 mA and the op amp output drops, until
it's once again at 3V and only the 12V drive circuit is
drawing current.

Peak power in the 120V section is the same, but the duty cycle
is much lower, and the power efficiency of the 12V (carrying)
amplifier is better.
 
J

James Arthur

Tim said:
Switcher. Possibly with an oversized coil to give your control circuit
enough time to _really_ keep things at 60mA. You'll need the voltage
overhead to supply that relay inductance, but you'll only be supplying
that power when you need to.

Of course, if you have to ask it'll take you some time to make it work,
but that ain't _my_ cross to bear (heh heh heh).

A hysteretic switcher could be pretty simple. Here's one for
inspiration:

http://www.members.shaw.ca/novotill/LedHystFeedbackPfm/index.htm

Cheers,
James Arthur
 
M

MooseFET

     I'm powering a Model 15 Teletype current loop.  The Teletypehas a selector
magnet resistance of 55 ohms and needs 60mA, so it only needs 3.3V.  But
because of the huge inductance of the selector electromagnet (4H), a
huge voltage is needed to pull in the electromagnet fast enough to
follow the incoming data stream.

     I'm using the traditional solution, a 120VDC supply through a 2K
10W resistor.  It's really dissipating 10W, continuously. A 10W resistor
gets hot enough (250C!) that you can't mount one on a PC board.  I'm
designing an interface board, and would like a better solution.

     Basically, I need a constant-current regulator circuit that
takes in 120VDC and limits the output current to 60mA.  Preferably
one that doesn't dissipate too much heat.  Suggestions?

The inductance of the coil can be the working inductor of the
regulator. A simple comparator can be your controller. You sense the
current in the coil and turn off a MOSFET when it passes some upper
limit. A diode lets the current continue. When the current drops
below some limit, you switch the MOSFET back on.

When the plunger pulls in, the current in the coil will attempt to
decrease suddenly. Using the comparator as suggested, makes sure that
the MOSFET is on and forcing the current back up while this happens.
 
R

Rich Grise

Basically, I need a constant-current regulator circuit that
takes in 120VDC and limits the output current to 60mA. Preferably
one that doesn't dissipate too much heat. Suggestions?

An NPN transistor with a regulated voltage at its base (like two or
three diode drops) and an emitter resistor to regulate the current?

Hope This Helps!
Rich
 
J

Joerg

Rich said:
An NPN transistor with a regulated voltage at its base (like two or
three diode drops) and an emitter resistor to regulate the current?

That would dissipate the difference, like the resistor does. Except now
you need a large heat sink ;-)
 
F

Fred Abse

I'm powering a Model 15 Teletype current loop. The Teletype has a
selector
magnet resistance of 55 ohms and needs 60mA, so it only needs 3.3V. But
because of the huge inductance of the selector electromagnet (4H), a huge
voltage is needed to pull in the electromagnet fast enough to follow the
incoming data stream.

I'm using the traditional solution, a 120VDC supply through a 2K
10W resistor. It's really dissipating 10W, continuously. A 10W resistor
gets hot enough (250C!) that you can't mount one on a PC board. I'm
designing an interface board, and would like a better solution.

Basically, I need a constant-current regulator circuit that
takes in 120VDC and limits the output current to 60mA. Preferably one
that doesn't dissipate too much heat. Suggestions?

The crux of the issue is current rise time, which is equal to L/R, hence
the higher series resistance, the faster the rise of current. If you use a
resistor (in your case 2K), you're going to need 120 volts to push enough
current through it. That's the reason for the "traditional" teletype power
supply, which goes back to the early 20th century.

An electronic current source mimics a high resistance, while allowing a
lower compliance (ie. open-circuit) voltage. The following LTSpice
simulations show firstly a "traditional" teletype circuit, and secondly a
suggestion for a simple current source using only a 24 volt supply.
Dissipation is less than 2 watts total.

The pulse source represents a TTY signal of 1:1 mark/space (RYRYRYRY)at
50 baud (AFAIK, the Model 15 did 45, 50, or sometimes 56). Compare the
rise times of current in L1 in the two examples, whilst not quite as fast
in the suggested circuit, it should prove adequate.

Traditional circuit:

Version 4
SHEET 1 880 680
WIRE 64 176 -64 176
WIRE 272 176 144 176
WIRE -64 224 -64 176
WIRE 272 224 272 176
WIRE -64 368 -64 304
WIRE 96 368 -64 368
WIRE 272 368 272 304
WIRE 272 368 96 368
WIRE 96 400 96 368
FLAG 96 400 0
SYMBOL voltage -64 208 R0
WINDOW 3 -429 54 Left 0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR Value PULSE(0 120 0 100n 100n 40m 80m)
SYMATTR InstName V1
SYMBOL res 48 192 R270
WINDOW 0 32 56 VTop 0
WINDOW 3 0 56 VBottom 0
SYMATTR InstName R1
SYMATTR Value 2k
SYMBOL ind 256 208 R0
SYMATTR InstName L1
SYMATTR Value 4
SYMATTR SpiceLine Rser=55
TEXT -494 506 Left 0 !.tran 100m
TEXT 320 264 Left 0 ;R (series) = 55 ohm

Suggested circuit:

Version 4
SHEET 1 880 832
WIRE -96 96 -128 96
WIRE 16 96 -32 96
WIRE 144 96 80 96
WIRE 240 96 144 96
WIRE 352 96 320 96
WIRE 144 128 144 96
WIRE -128 192 -128 96
WIRE -128 192 -256 192
WIRE -64 192 -128 192
WIRE 96 192 16 192
WIRE 256 192 192 192
WIRE 400 192 320 192
WIRE 464 192 400 192
WIRE 400 224 400 192
WIRE 464 224 464 192
WIRE -256 240 -256 192
WIRE 400 304 400 288
WIRE -256 400 -256 320
WIRE 80 400 -256 400
WIRE 352 400 352 96
WIRE 352 400 80 400
WIRE 400 400 400 384
WIRE 400 400 352 400
WIRE 464 400 464 304
WIRE 464 400 400 400
WIRE 80 448 80 400
FLAG 80 448 0
SYMBOL ind 448 208 R0
SYMATTR InstName L1
SYMATTR Value 4
SYMATTR SpiceLine Rser=55
SYMBOL pnp 192 128 R90
SYMATTR InstName Q1
SYMATTR Value TIP32C
SYMBOL diode 16 112 R270
WINDOW 0 32 32 VTop 0
WINDOW 3 0 32 VBottom 0
SYMATTR InstName D1
SYMATTR Value 1N4148
SYMBOL diode -96 112 R270
WINDOW 0 32 32 VTop 0
WINDOW 3 0 32 VBottom 0
SYMATTR InstName D2
SYMATTR Value 1N4148
SYMBOL res 224 112 R270
WINDOW 0 32 56 VTop 0
WINDOW 3 0 56 VBottom 0
SYMATTR InstName R1
SYMATTR Value 10k
SYMBOL res 32 176 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R2
SYMATTR Value 10
SYMBOL voltage -256 224 R0
WINDOW 3 -429 54 Left 0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR Value PULSE(0 24 0 100n 100n 40e-3 80e-3)
SYMATTR InstName V1
SYMBOL diode 256 208 R270
WINDOW 0 32 32 VTop 0
WINDOW 3 0 32 VBottom 0
SYMATTR InstName D3
SYMATTR Value 1N4148
SYMBOL res 384 288 R0
SYMATTR InstName R3
SYMATTR Value 2k
SYMBOL diode 416 288 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 0 Left 0
SYMATTR InstName D4
SYMATTR Value 1N4148
TEXT -688 504 Left 0 !.tran 100m uic
TEXT 512 272 Left 0 ;R (series) = 55 ohm
TEXT -680 536 Left 0 !.model TIP32C pnp\n+IS=1e-09 BF=134.366 NF=1.29961 VAF=10\n+IKF=0.742988 ISE=1e-16 NE=1.40014 BR=0.1\n+NR=1.46599 VAR=100 IKR=3.21978 ISC=1e-16\n+NC=2.71657 RB=7.44433 IRB=2.41268 RBM=0.218936\n+RE=0.0152284 RC=0.0761421 XTB=0.1 XTI=1\n+EG=1.05 CJE=3.26474e-10 VJE=0.446178 MJE=0.464223\n+TF=1e-08 XTF=3.50642 VTF=8.2848 ITF=0.0305862\n+CJC=3.07595e-10 VJC=0.77548 MJC=0.476497 XCJC=0.799334\n+FC=0.8 CJS=0 VJS=0.75 MJS=0.5\n+TR=1e-07 PTF=0 KF=0 AF=1


LTSpice is free, if you didn't already know that, from www.linear.com.


"Electricity is of two kinds, positive and negative. The difference
is, I presume, that one comes a little more expensive, but is more
durable; the other is a cheaper thing, but the moths get into it."
(Stephen Leacock)
 
R

Ross Herbert

: I'm powering a Model 15 Teletype current loop. The Teletype has a
selector
:magnet resistance of 55 ohms and needs 60mA, so it only needs 3.3V. But
:because of the huge inductance of the selector electromagnet (4H), a
:huge voltage is needed to pull in the electromagnet fast enough to
:follow the incoming data stream.
:
: I'm using the traditional solution, a 120VDC supply through a 2K
:10W resistor. It's really dissipating 10W, continuously. A 10W resistor
:gets hot enough (250C!) that you can't mount one on a PC board. I'm
:designing an interface board, and would like a better solution.
:
: Basically, I need a constant-current regulator circuit that
:takes in 120VDC and limits the output current to 60mA. Preferably
:eek:ne that doesn't dissipate too much heat. Suggestions?
:
: John Nagle


Back in the days of relay based telephone exchanges it was often required to
lower the power consumed by a relay yet have it operate with a snap action. This
was achieved using a second relay to maintain a capacitor charged to exchange
battery voltage (-48V) in the normal condition. When the main relay was required
to activate the secondary relay would, via its contacts, arrange for the charged
capacitor to be connected in series aiding with the main relay coil which had
-48V behind it. Effectively, a double voltage pulse was applied to the main
relay coil which snapped in fast but thereafter remained operated via normal
-48V at reduced current.

A modernised version of this arrangement has been devised by Maxim and it may be
applicable to your application.
http://www.maxim-ic.com/appnotes.cfm/an_pk/956

You will probably need to tailor it a bit but it appears as if it would be a
neat solution.
 
Top