Full resolution (JPEG) - On this page / på denna sida - A Method of Storing Binary Information in Ferrite Memory Cores Facilitating Non-Destructive Read-Out, by Alvar Olsson
<< prev. page << föreg. sida << >> nästa sida >> next page >>
Below is the raw OCR text
from the above scanned image.
Do you see an error? Proofread the page now!
Här nedan syns maskintolkade texten från faksimilbilden ovan.
Ser du något fel? Korrekturläs sidan nu!
This page has never been proofread. / Denna sida har aldrig korrekturlästs.
A Method of Storing Binary Information in Ferrite
Memory Cores Facilitating Non-Destructive Read-Out
Alvar Olssön, Telefonaktiebolaget L M Ericsson, Stockholm
I)et kan visas att utlösningen från en minneskärna
kan göras icke-raderande om det ena läget
representeras av remanensläget efter kraftig
uppmagnetise-ring av kärnan, t.ex. efter mättningsmagnetisering,
medan det andra lagras som en inre hgsteresslinga,
t.ex. en som erhålles efter avmagnetisering av
kärnan. Utläsning göres med magnetiserande fält, som
inte förmår driva en kärna i det förra tillståndet ur
dess remanensläge, men väl förmår driva en kärna
i det senare tillståndet runt sin inre hgsteresslinga.
Denna lagringsmetod medger en utläsningstid, som
är oberoende av kärnans ommagnetiseringstid och
som alltså kan göras mycket kort.
Ommagnetiserings-tiden begränsar däremot inskrivningstiden nedåt.
This article shows that the read-out from a ferrite
memory core can be made non-destructive if one
of the binary values is represented by the state of
magnetic remanence after having magnetized the
core strongly, for example to saturation, while the
other value is represented by the state of
magnetization in a minor loop, for example one obtained after
demagnetizing the core. Read-out is implimented
with a drive not strong enough to drive a core in
the first mentioned state out of remanence but strong
enough to drive a core in the latter state around its
minor loop. This method of storage permits a
readout time independent of the switching time of the
material to be obtained. Thus read-out time can be
made very short. However, the switching time limits
the speed of writing.
Destructive and non-destructive read-out
In a core in a conventional core memory the
information, 1 or 0, is stored as one of the two
remanent states. On reading, all the cores in a word are
driven to the state of remanence which represents
the information 0. This means that, from the point
of view of the cores, the read-out is destructive. If
the word is to be retained, however, it can be
selectively rewritten in the cores, a process which, from
the point of view of reliability, is more dependent
on external circuits than on the cores.
For certain applications the information must be
retained in the memory for a long period and be
subjected to a very large number of read-outs before
a change is required. The advantages and
disadvantages of non-destructive read-out in such
applications are not to be discussed here — we only state
that a fast memory with non-destructive read-out
and electrical means for writing would be preferable.
Several methods of non-destructive read-out from
magnetic cores have been proposed. Some of them
require special and more complicated memory ele-
681.142 : 621.318.042
ments1’3’3, such as for example the transfluxor. Others
depend on special read-out methods4,5.
The method of non-destructive read-out described
in this paper is based on the use of a minor loop
for storage of the binary digit 1 and the remanent
state obtained after having magnetized the core
strongly for the storage of the binary digit 0.
Readout is affected by a drive not strong enough to
change the magnetic state of a core in the 0-state
but strong enough to drive a core in the 1-state
around a minor loop.
Core data
The possibility of non-destructive read-out is
apparent from fig. 1, which shows the family of
hysteresis loops for a ferrite memory core. For simplicity
let us inspect the largest loop, the saturation loop,
and the smallest one, which can be obtained after
demagnetization. If the read-out drive is limited
to a value smaller than the value indicated by the
knee of the saturation loop, different sizes of loop
will be obtained according to whether the core has
been placed in the remanent state of the large loop
indicated by 0 or in the minor loop state indicated
by 1. It may be understood from fig. 1 that the
reversible permeability is greater in the 1-state than
in the 0-state and furthermore that a certain
irreversible change of flux takes place in the 1-state, a
phenomenon which is not possible in the 0-state.
Of course a smaller "large loop" and a larger "minor
loop" than indicated in fig. 1 can be used.
In order to investigate the suitability of different
materials for the proposed method, cores have been
tested in the following manner. The cores were
driven to the remanent state, the 0-state, by the applica-
Fig. 1. A family of hysteresis loops for a ferrite memory
core.
ELTEKNIK 1959 1 99
<< prev. page << föreg. sida << >> nästa sida >> next page >>
