Search

Controls used in electric power distribution systems are generally tolerant of the high level of interference signals found in that environment. Recently, however, increased power demnds have required a new class of solid state controls which are more sensitive to RFI. In order to delineate the problem this report examines the magnitude and distribution of RF fields in a small power substation during an incident of radio transmission. RF electric field strengths were of the order of 10 V/m. Exposure of cormercial distribution relays to such fields are shown to produce considerable interference with normal operation. Radio frequencies from 20 to 500 MHz were employed

An experimental and theoretical study was made of the shape‐dependent demagnetizing fields in thick (∼1 μ) uniaxial Permalloy films. Measured values of the wall motion coercive field Hc and of the rotational anisotropy field Hk change by factors of 20 and 3, respectively, as the lateral sample dimensions vary from 20:1 through 1:20. The values of Hc agree with a theory in which magnetization switching occurs when the total internal field (applied field plus demagnetizing field) exceeds an intrinsic coercive field which is independent of sample shape. In films for which the demagnetizing effects are significant this leads to a well‐defined range of effective switching fields and to highly skewed hysteresis loops. The total internal field also gives agreement with the measured values of Hk, reflecting the fact that the remanent state of a hard direction hysteresis cycle is multidomain, with a consequent reduction of the easy axis demagnetization. The theoretical analysis approximates the rectangular film by...

The model of a magnetic film as one domain with a single anisotropy axis is often an inadequate approximation to true film behavior. A better approximation considers the film to consist of many uniaxial regions, with a range of anisotropy constants and of axis orientations in the plane of the film. The dispersion of values results in considerable confusion and awkwardness when using the single domain representation to understand magnetic switching. It is shown here that the total anisotropy distribution can be displayed on a polar plot in which each film region is characterized by a single point with the coordinates: Hk, its anisotropy field, and α, its axis orientation. In this (Hk,α) plane the switching effectiveness of the applied field is represented by a curve which is fixed in size and orientation relative to the field vector. Switching occurs in those parts of the film whose anisotropy points fall within the curve. The field figure is extended to include magnetization reversal by domain wall motion...

We consider a model of a magnetic sample with distribution of Barkhausen switch sites which reverse their moments at a constant rate. Switching is considered in the case that reversal is induced by the incidence of a domain wall. The induced current distribution J of symmetrically placed regions has already been determined. The loss calculation involves summing J2 over switching regions in the plane of the moving wall, and summing as the wall traverses a full cycle. Averaging techniques are employed and the expected loss over unit volume, per cycle is found, due solely to the Barkhausen regions. This is the classic static or ’’hysteresis’’ loss. In a later report the analysis is extended to determine the full losses.

Energy loss in magnetic steels is generally taken to consist of two parts: static loss, associated with Barkhausen switching and assumed independent of drive frequency, and eddy current loss due to wall motion. Little effort has gone into analyzing the Barkhausen contribution. We consider a model of a magnetic sample in which the Barkhausen volumes are represented as a distribution of sites which reverse their moments at a uniform rate and only along a single axis. Using Green’s function the current distribution of symmetrically placed regions is determined subject to the boundary conditions of periodic bar domains in a thin sheet. In subsequent reports this is used to calculate losses.

A model of a polycrystalline magnetic film is considered in which the film is made up of transverse strips of (N) grains, with the magnetization M everywhere parallel within a strip but varying in direction from strip to strip. It is shown that the effective crystalline anisotropy in each strip can be represented by a transverse field Ht which varies randomly from strip to strip, where Ht (rms) = K/[2M(2N)½] (K is the crystalline anisotropy constant). It is recalled that random number sequences appear to have a periodicity of about 3.5 units, a phenomenon now recognized as having led to the assignment of a periodicity in the population of arctic foxes. A similar apparent periodicity of the transverse anisotropy field leads to an apparent periodicity of the magnetization ripple. Exchange forces, however, modify the amplitude and wavelength λ of the magnetization ripple when λ

Studies are reported of the fields surrounding the T‐bar pattern commonly used in magnetic bubble propagation devices. Large, appropriately scaled, T‐bar circuits were used to experimentally determine the normal field patterns in the vicinity of the T‐bar elements. These patterns were determined with external drive fields having various in‐plane orientations. From these data the approximate stable position of the bubble was determined for each external field orientation. The magnitudes of the perpendicular (driving) field gradients were estimated from the changing field patterns and bubble velocities were calculated. The discussion of these results includes comparisons with available data on bubble velocity along the propagation path. Bubble oscillations normal to the propagation direction are noted.

This paper examines the dynamic effects of magnetization fluctuations in polycrystalline films on their uniform mode resonance behavior. The dynamic interactions of the non‐uniform components with the spatially averaged magnetization M⇒0 is first treated as a perturbation correction on the susceptibility χ(ω) when the interaction is small. When the interaction is large such that phenomenological damping is neglected it is shown that the fluctuations introduce regions of damped behavior. Illustrative numerical calculations are made on a model whose spatial variation is described by a single wave vector.

This paper presents a semi‐classical model for uniform mode FMR and relaxation in a planar magnetized film The model involves a systematic treatment of surface dipolar and volume dipolar terms in addition to exchange and a uniaxial anisotropy term. M is assumed to be constant across the film thickness with planar spatial variations functioning as the spin wave coupling mechanism. The resulting dispersion relation includes surface and volume dipolar terms with thickness dependent attenuation factors. The density of states degenerate with the uniform mode is dependent on field and thickness, particularly at low field values. The consequences of these dependencies are discussed. For fields

This paper examines the effects of static magnetization dispersion in polycrystalline films on their low field, uniform mode resonance behavior. The quasistatic interaction of the disperse local magnetization with the average magnetization M⇒0, enters the description of FMR through modified initial conditions and through the internal field. The initial nonuniform state alone modifies the low field resonance behavior. In addition, quasistatic free energy minimization yields a dependence of the mean planar equilibrium angle on nonlinear interactions. This causes changes of the resonance frequency and susceptibility magnitude and, under certain conditions, the occurrence of a double resonance.

The use of scaled, field access bubble propagation systems has been proposed for examining the details of bubble conduction schemes. It has been argued that such scaling cannot represent size dependent effects, cannot include the influence of bubble dynamics, and does not lend itself to including interactions between bubble and overlay. In this report we describe a large,appropriately scaled T‐bar model which includes bubble interactions. It is shown that scaling effects are not significant, that bubble position is essentially determined by magnetostatic equilibrium and that the scale model can serve as a valuable tool in studying the design and extension of field access devices.

A short section of magnetically coated wire was used as center conductor of a section of coaxial transmission line. The complex reflection coefficient from the section was measured, at 110 MHz, as a function of an axial dc field. By relating the reflection coefficient to the surface impedance of the wire, and the surface impedance to the magnetic state of the film, the rf susceptibility χ could be determined. This technique yields absolute values of χ, so that only limited scaling can be used to obtain a fit to the theoretical curves. Two film types were used: high‐dispersion thick films (∼8000 A) and low‐dispersion thin films (∼1000 A). No fit to the experimental points could be obtained unless a complex susceptibility was assumed. The real and imaginary parts of χ and of the effective field have been calculated during the entire magnetization reversal process.

Faraday rotation measurements on ferrimagnetic oxides have so far been limited to single crystals which are rendered transparent by grinding. In the present studies films of NiFe2O4 and CoFe2O4, of thickness 1000–8000 A, suitable for optical studies are prepared by oxidizing films of the alloys NiFe2 and CoFe2 sputtered on silica substrates. The Faraday rotation is determined with the film magnetized along axes both normal and parallel to its surface. Using the latter arrangement, optical hysteresis loops are recorded and checked against those obtained with a vibrating sample magnetometer. The dependence of loop shape and coercive force on film thickness and oxidation temperature is studied.

Magnetoresistance measurements have been used to study the magnetic properties of Permalloy films. Resistivities parallel and transverse to the current were measured simultaneously as the applied magnetic field was varied in magnitude or direction in the film plane. The results are discussed in terms of a representation in which the resistivity components are regarded as Cartesian coordinates. This representation is independent of the current direction relative to any magnetically preferred orientations in the film. The statistical, domain-wall, switching model of Conger and Essig is extended to include the situation in which the field and easy magnetic axis are not parallel, and a switching profile is constructed which gives a unified representation of the domain wall and coherent rotational switching response of a uniaxial film to an applied field. The incidence of moving domain walls under certain field conditions is discussed with the help of this profile, and it is shown that the magnetoresistance cu...

Measurements of the transverse differential susceptibility χ were made on cylindrical NiFe films over the range 10–400 MHz. The field dependence of the real and imaginary susceptibility components were determined; both display regions of positive and negative values. These curves were analyzed using a dynamic equation of motion with Gilbert damping. It is found that the effective ripple field model alone can not explain the observed variations but that one must include coupling to higher‐order spin modes. With appropriate assumptions this analysis leads to a quadratic frequency dependence of the effective field at the zero of χ(imaginary).

Hall effect measurements on a ferromagnetic film can be used to examine the normal anisotropy and the distribution of stresses in the plane of the film. The measured Hall voltage exhibits a smooth transition from low to high field regions rather than the sharp crossover, at technical saturation, which is expected of a single domain. Since the ferromagnetic Hall effect depends upon the component of magnetization perpendicular to the plane one can attribute this difference of behavior to a domain structure describable by a distribution of normal anisotropies. The difference between the single and multi-domain Hall plots depends upon the nature of this distribution and from certain features of the experimentally obtained difference curve one can infer the type and extent of the distribution. If the anisotropy distribution is due to a corresponding range of strains superimposed on the geometric anisotropy of the film one can also estimate the form and extent of the function describing the planar stresses. It is shown that the existence of the anisotropy distribution results in a difference between the field at which the Hall measurements indicate technical saturation and that at which the normal anisotropy distribution is peaked.