[GreenKeys] Constant current selector magnet driver

[jhhaynes at earthlink.net]

I haven't yet seen the schematic we are talking about.  There was an
article in QST some years back, and of course I can't lay my hands on
it right now, nor find it in the index, but was a constant-current
selector magnet driver.  As I recall it used about 60 volts.  I wish
I knew how to simulate this thing on a computer so as to satisfy
myself that it really does operate faster than the simple saturated
transistor circuit with high loop resistance.

You might be interested in a somewhat related story about the DRPE
driver.  DRPE was the last high speed tape punch from Teletype. The
mechanism consisted of a pretty stiff steel reed that was held up
by an electromagnet.  The idea was to cut off current through the
magnet, allowing the reed to swing downward at its resonant frequency
and push the punch pin through the tape.  Then the reed would swing
back up and be grabbed by the electromagnet at the top of its stroke.

One drawback to this concept was that when power was turned off the
reeds would go to a neutral position, with the punch pins dragging
on the tape, so it was necessary to make the electromagnet strong
enough to pull the reeds up when power was first applied.  Also the
process of punching tape was fairly energy-dissipative, so the magnet
needed to pull the reeds up rather strongly at the end of a punch
stroke.

There were three generations of drivers for the DRPE.  The first,
used only in a demonstration model, was a lot like our selector
magnet loops: high voltage and high resistance to overcome the
inductance, and enough current to pull the reeds up from rest.
This model dissipated a lot of power in the resistors.  The second
model was a clever circuit involving some big inductors.  Energy
was stored in an inductor, and then at the time to pull the reed
up the inductor was switched to the punch magnet.  The inductive
"kick" generated a voltage spike to pull up the reed, and then the
rest of the driver just had to supply enough current to keep it
pulled up.  The idea was to shuttle energy between the punch magnet
and the inductor so that the power supply only had to make up the
losses.  The third model used a high voltage switched on by a
transistor to pull the reed up, and then switched to a much lower
voltage to keep the reed against the electromagnet.  This used
high voltage and high resistance, and hence high dissipation, but
only during the time it took to pull the reed up.  The power to keep
the reed up was much lower.

I'm reminded too of a very simple circuit that was sometimes used
to make a relay operate fast.  The relay would be operated by a
high voltage through a resistance, but that supplied only enough
current to keep the relay operated.  The resistance was shunted
by a capacitor so that initially the full high voltage was applied
to the relay coil; then with the relay pulled in the capacitor
charged up and left a much lower voltage holding the relay operated.

While I was searching the QST index I noted from the May 2000 issue,
page 36, "New High Voltage Constant Current Regulators (New Products)"
I haven't looked that up yet, but wonder if something like this
might provide a ready-made selector magnet driver.



jhhaynes at earthlink dot net