Hi Chuck and all,
At 10:13 AM 4/24/99 -0700, you wrote:
Ok, one step closer. The description of core in the
PDP-8/a miniprocessor
users manual shows the wires going straight through the cores. Is this
correct? I'd expect the wire to make one or two turns around the core so
that the magnetic field it induced would be "inside" the torroid. Comments?
Yes, the wires went straight through. With a closed circuit (driver
circuits) this is one turn anound the core. More turns would require less
current, but would be much harder to make. Look at the picture at:
http://www.wins.uva.nl/faculteit/museum/CoreMemory.html
and imagine if the wires had to make loops around each core.
What about driving voltages? I've got a +/- 36v @
3amps supply here, the
PDP-8 uses its 15 volt supplies. I'm building a simply push-pull direct
coupled amp out of a couple of transistors to send the signal that Allison
drew. I'm using a Parallax BASIC Stamp to generate the waveforms (I could
use the HP but then it wouldn't be portable)
It's the current that's important, the driver circuits would obviously
require some voltage to deliver that current. Maybe someone could give the
write currents for some example core, I haven't had a chance to read any
spec's.
Sense wire? Straight through the core or also with a wrap?
Straght through, but
usually the 45 deg. wires for less capacitive coupling
to the driver wires.
Remember that a "read" is a write + sense + possibly another write if the
core was switched to switch it back. I don't know if this logic was taken
care by the core subsystem or integrated into the CPU cycles, for any
particular computer.
-Dave
--Chuck