Memory Devices Using Cores Made of Tape Permalloy
Abstract Generated abstract
This paper reports an experimental study of tape permalloy magnetic cores as memory elements for measuring, storing, and reproducing time intervals. A 79NM permalloy tape core with two 1800 turn windings was tested in a pulse generator and multivibrator circuit, where magnetization and demagnetization processes were used to record and reproduce pulse durations. The authors found that intervals from about 140 to 5000 microseconds could be reproduced, with a roughly constant error of 20 microseconds attributed to material properties, core geometry, and tape thickness. They suggest that accuracy and range may be improved by changing core dimensions, using thinner tape or higher squareness permalloy, and adjusting winding turns or voltage.
Full Text
Memory Devices Based on Tape Permalloy Cores
Yu. K. Petrov, Yu. G. Egoshev
(Presented by the scientific seminar of the Department of Mathematical and Computing Instruments and Devices)
In some devices it is necessary to measure and then reproduce time intervals. We carried out an experimental study of the possibility of using tape permalloy cores for these purposes.
A core made of 79NM tape permalloy, \(0.02\) mm thick, was tested. The dimensions of the core were \(16 \times 5 \times 2\ \text{mm}^3\).
The setup for testing the core is shown in Fig. 1: \(MGI\) is a rectangular-pulse generator, \(MV\) is a multivibrator, \(T\) is a trigger, and \(O\) is an oscilloscope. Winding \(W_1 = W_2 = 1800\) turns was wound on the core. The voltage was \(8\) V. The multivibrator \(MV\) was assembled according to an asymmetrical circuit with such a calculation that, during one of the half-periods of multivibrator operation, the core was completely demagnetized under the action of the constant voltage.
Fig. 1
In such a device, the time integral is reproduced in exactly the same way as it is stored; therefore, the reproduced interval must be equal to the stored one, regardless of changes in the core remagnetization conditions: the transistor parameters and the magnetic properties of the core, provided only that these parameters do not change during the time of storage and reproduction.
The results of the reproduction of several intervals are given in the table. The measurements were made from the oscillogram by means of time marks.
The constant reproduction error of \(20\ \mu\text{s}\) is associated with the properties of the material, the shape of the core, and the thickness of the magnetic tape. Apparently, this error can be reduced by changing the thickness of the ring, for example, from \(2\) to \(1\) mm, by reducing the thickness of the tape.
The error can be reduced by using permalloy with a very high squareness ratio of the hysteresis loop, although for large cores the shape of the core has the principal effect.
Using this ring, intervals in the range \(140 \div 5000\ \mu\text{s}\) were reproduced. The value of the interval to be stored can be
| \(t\) measured | 140 | 300 | \(900\ \mu\text{s}\) |
| \(t\) reproduced | 120 | 280 | \(880\ \mu\text{s}\) |
increased by increasing the number of turns \(W_1\) and \(W_2\), or by reducing the voltage \(U_0\). In this case the output signal will decrease somewhat. Tests have shown that the reduction of the output signal is permissible.
Fig. 2
The character of the oscillogram for a pulse of duration \(1000\ \mu\text{s}\) is shown in Fig. 2. The lower pulse corresponds to magnetization of the core, and the upper one to natural demagnetization.