Interferencia electromagnetica(emi)

Electromagnetic Interference (EMI) Damage to Giant Magnetoresistive (GMR) Recording Heads
A1 Wallash (1) and Doug Smith (2)
( I ) Quantum Corporation, 500 McCarthy Blvd., Milpitas, C A 95035 4 08- 324 - 75 39 ; 894 - 3207 ( fax ) ; a w a11ash @ q n t m .c o m (2) Auspex Corporation, 2300 Central Expressway, Santa Clara, C A 95050 408-566-2 157; dsmith@auspex.com

Abstract-It is shown thatgiant magnetoresistive (GMR) recording heads can be damaged by electromagnetic interference (EMI) from a remote electrostatic discharge (ESD) event. An EM1 test board is described and used to quantify EM1 damage thresholds for a dipole or loop antenna attached to a GMR head. A practical example of EM1 damage during resistance measurement is studied. SEM failure analysis after EM1 testing showsmelting damage to the GMR sensor that is similar in appearance to the damage caused by direct-contact ESD damage. It is concluded that it is important to control EM1 during handling operations that involve any wires connected to the GMR recording head.

INTRODUCTION
It is well known that a spark associated with an electrostatic discharge (ESD) can induce tens of milliamperes of current in nearbywires due to the radiated fields. [ I ] Usually the only effect of the radiated fields is electromagnetic interference (EMI) in the form of a “noise” spike in electronic circuits. It is generally believed that indirect EM1 cannot cause physical damage to static sensitive devices. However, it has recently been reported that a magnetoresistive (MR) recording head can be physically damaged by the EM1from a remote spark. [2,3] While the effects of a direct electrostatic discharge (ESD) to an MR head has been studied in some detail, the understanding of indirect EM1 damage to MR heads is still in its infancy. The goal of this work is to further the understanding of EM1 damage to MR recording heads. When wires are attached to an MR head, they act as “antennas” and pick up the high frequencyelectromagnetic radiation produced by a remote spark. Four factors that determine whether the EMI-induced current transient damages the MR sensor are the > failure current of the MR sensor, > type and length of the “antenna” connected to the MRhead, > intensity of the radiation from the EM1 source, and P distance between the head and EM1 source.

Figure I shows two different ways that attached wires tothe MR head can form an antenna. Figure 1 (top) shows two wires connected to the MR inputs so as form a dipole antenna. The bottom figure shows a single wire attached across both MR sensor inputs so as to form a loop antenna. In each case, the length of

Input

GMR Sensor

Input

Loop

@
Input GMR Sensor Input
Figure I . C M R sensor with wires attached to the inputs to form either adipole (top) o r loop (bottom) antenna.

4B.6.1
EOWESD SYMPOSIUM 98-368

the antenna or loop area will affect the magnitude of the EMI-induced current transient through the MR sensor. One common operation that requires wires to be connected to the MR sensor is the measurement (of MR head resistance using an ohmmeter. In this situation, a nearby spark could result in a damaging EMI-inducedcurrent transient while the ohmmeter leadls are connected to the MR head inputs. In this study, the effects of both dipole and loop antennas connected to MR heads are studied using a special “EM1 test board”. In addition, EM1 damage during resistance measurement is reported and studied. Finally, SEM failure analysis of EM1 damaged heads is compared and contrasted with ESD damaged heads.

(, I

0v) .v)

m c

c

45 (, I U 44 K I
(3
43

= !
Cl

I
I
Magnetic/ FaiIure

Melting Failure\
T

J-

20

40

60

80

HBM Voltage (V)

.0 0.8
I

3

Magnetic Damage

HBM ESD TESTING REVIEW
In order to understand the results from indirect EM1 testing, it is helpful to review the behavior of MR heads during direct contact ESD testing. Of the two types of MR...