Browse Prior Art Database

Barkhausen Noise Quantification Using a Derivative Approximation

IP.com Disclosure Number: IPCOM000122289D
Original Publication Date: 1991-Nov-01
Included in the Prior Art Database: 2005-Apr-04
Document File: 5 page(s) / 126K

Publishing Venue

IBM

Related People

Kahwaty, V: AUTHOR

Abstract

Quantization of barkhausen noise in a magneto resistive (MR) magnetic recording head has been a major problem. We have found a new technique which uses the derivative of the MR response to quantify barkhausen jumps in MR elements. The method has shown very good correlation to observed time domain noise performance. It provides a method of separating bi-stable MR element performance which is usually not a concern and barkhausen jumps which can be a head performance limiting factor. INTRODUCTION

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Barkhausen Noise Quantification Using a Derivative Approximation

      Quantization of barkhausen noise in a magneto resistive
(MR) magnetic recording head has been a major problem. We have found
a new technique which uses the derivative of the MR response to
quantify barkhausen jumps in MR elements. The method has shown very
good correlation to observed time domain noise performance. It
provides a method of separating bi-stable MR element performance
which is usually not a concern and barkhausen jumps which can be a
head performance limiting factor.
INTRODUCTION

      Prior art has used an amplitude characterization method to find
instability in MR sensors. We attempted to use this method for MR
sensors and found poor agreement with instability observed in time
domain and channel measurements. We have found dramatically improved
Barkhausen noise identification with the use of an approximate
derivative of a sine wave response signal normalized to the signal
amplitude.  Repeating this process for a series of amplitude levels
will quantize the magnitude of barkhausen jumps in a selected MR
sensor.

      One method of implementing this concept is shown with the use
of the following tester shown in block diagram form in Fig. 1a. A
MnZn ferrite substrate is wound with 400 turns of magnet wire and is
connected in series with a capacitor to form a l khz resonant
circuit. A power amplifier controlled by a programmable signal
generator drives this circuit. A 2-ohm resistor is placed in series
with the circuit to monitor the current and provide a x-input drive
for an oscilloscope.  The x-input drive is phase shifted to
compensate for other phase shifts in the network. The head under test
is biased via a 9 x 1 multiplexer which can provide single ended or
differential resistor bias. A multiplexer network provides selection
of 1 out of 9 MR elements. The absolute bias level is provided by a
programmable bias power supply. The head signal output is the y-input
to the oscilloscope.

      A bipolar DC drive field is provided by a DC power supply and a
polarity multiplexer. This is necessary to provide the DC field
excitation required for the stability test.

      Examples of single ended and differential signal responses vs.
a magnetic driving field (H) are shown in the insets of Figs. 1b and
1c for the same MR element. Note that the transfe...