It takes quite a bit to read one bit (the same amount to read
all the bits on one sense line).
You need to in order:
Select the specific word (or bit) and drive it with the X and
Y lines, each will be 50% plus a tiny bit extra current
needed to switch a bit. And yes that is a regulated
current or wire melting may occur! Typically it's a pulse
of several hundred milliamps.
Then at the right time after the select pulse is applied
you will see a big or smaller pulse on the sense line.
the difference between a 1 or 0(zero) is the timing and
the height of that pulse.
That was the nasty item as both timing of the XY select
pulse and the time to the read window are current,
temperature and specific to the cores used. The resulting
signal is quite small and is measured in tens of millivolts
for a core that switched and maybe a millivolt for one that
did not. Yes, the sense line does have a lot of noise!
To do that again or write a 1/0 you need to reverse or
leave the XYselect polarity. The inhibit line (for a 4 wire
core mat ) can be the sense line for write (three wire mat).
To write the opposite data you select the core via XY with
enough current to flip the core, to not write that value you
apply enough current of the reversed polarity to the inhibit
line to reduce the field(from the XY selects) in the selected
core to less than the half select value.
The basic operation of reading implicity sets all cores in the word
line to the same state. To write (a word) along a word line you
reverse the currents and also apply inhibit current on a bit by
bit basis. The end result is some cores in the word flip (reverse
the magnetic field) and those inhibited do not.
Now... if you want simpler try this...
Wind about 20 turns of wire on a nail (iron, or better yet steel).
Wind another 20 turns. We use a lot of wire to get a BIG signal.
Now if you hook a simple 1-3V DC meter to one winding and
then hook the other winding to say a 1.5V battery you should
see the meter pulse (basic transformer action). If you interupt
the connection (off on off) you will see another pulse of differing
polarity (needle will go the other way for a moment). Do that a
few times to note the reaction and then reverse the battery and
repeat. When you do that you will not the meter action indicates
a bigger pulse the first time and back to like before if you do it
again. Now the majik of core. When you magnitize something
like that the nail/washer/core takes a magnetic polarity and
holds it. If you reverse that magnetizing field you not only get
the field but also the change in magnetic polarity. That change
significantly greater. The other peice of majik is that while you
didn't measure it in that simple rig the current required to magnetize
the nail in the opposite direction is greater than if the magnetic
field in the in the nail was random(zero). This is the characteristic
called hysteresis, IE: to change a field takes more than the amount
needed to maintain or initially estabilish it.
I might have fluffed the grammer and spelling in my hurry to write
this but, thats about it.
Allison
-----Original Message-----
From: John Allain <allain_at_panix.com>
To: classiccmp_at_classiccmp.org <classiccmp_at_classiccmp.org>
Date: Sunday, December 16, 2001 2:02 PM
Subject: Re: [PDP8-Lovers] how to clean a PDP8/A, dishwasher?
>> No need to trace, they are quite regular, you can ohm them. You still
>> need core drivers, sense amps and a dozens of diodes for current
>> steering.
>
>We were thinking what could a person do to see the output from
>just one core bit, not too much more than that... and then work up
>from there.
>
>John A.
>
Received on Sun Dec 16 2001 - 14:25:04 GMT