Thesis Touch Down Detection at Spin-Stand Level Raksak111574

Touch Down Detection at Spin-Stand Level

by Touch Down Sensor

by

Raksak Rujipornkasem

A thesis submitted in partial fulfillment of the requirements for the

Degree of Master of Science in

Mechatronics

Examination Committee: Mr. Brent Bargmann (Chairperson)

Dr. Mongkol Ekpanyapong

Mr. Somen Choudhury (External Expert)

Mr. Jim Potter (External Expert)

Nationality: Thai

Previous Degree: Bachelor of Engineering in Industrial

Thammasat University, Thailand

Scholarship Donor: Western Digital - NECTEC, Thailand – AIT Fellowship

Asian Institute of Technology

School of Engineering and Technology

Thailand

July 2014

ii

ACKNOWLEDGEMENTS

First, I would like to thank WESTERN DIGITAL (WD), ASIAN INSTITUTE OF

TECHNOLOGY (AIT) and NATIONAL ELECTRONICS AND COMPUTER

TECHNOLOGY CENTER (NECTEC) for me a chance and opportunity to study the very

valuation special class at AIT.

I would like to thank AIT’s professors, Dr. Nitin V Afzulpurkar, Dr. Joydeep Dutta and

also Mr. Brent Bargmann for fantastic classes, advice and the recommendations.

I would like to thank WD’s colleagues, Saikumar B, Wimonwan P, Sanpong S, Sansanee

S, Watcharasak T, Sonklod and Production team to the great support, Jim Potter and my

boss Pradip Kuchibala for suggestions.

I would like to thank WD’ Fremont (California) team Tao P, Shaoping Li, Gopal Kote,

Eric S, David F. and Francis Liu for valuable training packages of magnetic recording

especially DFH training package.

I would like to thank AIT’s wave four classmate; they made a wonderful study class and

discussion during the hard time in lecture.

Finally I would thank my family for their great support and always beside me during the

hard time of study at AIT.

iii

ABSTRACT

According to the increasing trend of HDD areal density, there are many factors to

achieve higher areal density, for example magnetic write width reduction and magnetic

spacing reduction. Magnetic spacing reduction can be achieved by both mechanical fly

height and dynamic fly height (DFH). Dynamic fly height relies on the reduction of

magnetic spacing by applying heat to the device as to increase the protrusion.

Magnetic spacing reduction by DFH can increase read back signal and Signal to

Noise ratio. At the same time, too low magnetic spacing can cause HDD problems such as

reliability issue and on head disk interference. Device designer needs to find the optimum

magnetic spacing by performing spin stand level testing and drive testing in order to

characterize DFH to define actuation efficiency. Actuation efficiency is the device

protrusion rate per heater power unit, of which the unit is nanometer per mW.

Operating spacing can be defined as gap of “Touch down point” and “Back off spacing.”

Back off spacing is the term for reliability’s improvement. In addition, an important step

before read and write process is “Touch down point’s identification.” Accuracy of touch

down detection can boost up electrical performance and maintain reliability.

This thesis proposes new technology of touch down point detection by “Touch

down sensor” or TDS which is built in the device. TDS uses a material’s property called

“TCR”, thermal coefficient of resistance, to be the measured parameter. Heat is generated

during very low magnetic spacing or at very near real touch down point. The TDS can

detect the heat and report the change of temperature for touch down point identification.

This thesis compares the conventional detection (Phase modulation) and new technology

(Touch down sensor) detection. The result will be analyzed and considered before

implementation.

iv

TABLE OF CONTENTS

CHAPTER TITLE PAGE

Title page i

Acknowledgement ii

Abstract iii

Table of Contents iv

List of Figures vi

List of Abbreviations viii

1 Introduction

1.1 General background 1

1.2 Statement of the problem 2

1.3 Objectives 3

1.4 Scope of study 3

1.5 Limitation of the study 3

1.6 Thesis schedule 4

2 Literature Review

2.1 Dynamic Fly Height for PMR Recording 5

2.2 Head Protrusion profile by Heater design 9

2.3Study Defect detection by Dynamic Fly Height detector 11

2.4 Defect mapping by Touch Down Sensor 13

2.5 Magnetic Spacing Trends 16

3 Methodology

3.1 Head Gimbal Assembly (HGA) 20

3.2 Dynamic Electrical Test (DET) 24

3.2.1 Amplitude 26

3.3.2 Track Profile and Micro-track Profile 27

3.3.3 Reverse Overwrite 28

3.3.4 Signal to Noise Ratio 29

3.2.5 Squash Measurement 29

3.2.6 Track Profile 30

3.2.7 Dynamic Fly Height (DFH) 32

3.2.8 Touch Down Detection 34

3.2.9 Touch Down Sensor 36

4 Results and Discussion

4.1 Touch down detection from DET tester 38

4.2 Experiment flow and Result 40

5 Conclusion and Recommendations

5.1 Implementation of Touch down sensor 54

6 References 55

v

LIST OF FIGURES

NO. DESCRIPTION PAGE

1.1 Areal Density Trend by Year for Mobile product 1

1.2 Disk drive components 1

1.3 Disk Drive Component 2

1.4 Thesis schedule 4

2.1 Bit Error rate and Linear Density relation 5

2.2 Bit Error rate and Read clearance (nanometers) 5

2.3 Media Signal to Noise Ratio and Read Clearance 5

2.4 Bit Error Rate and Signal to Noise Ratio 5

2.5 Device structure 6

2.6 AFM image of interesting area 6

2.7 Acoustic signal and a portion of numerically demodulated read signal 7

2.8 Modulating envelope as the heater power 7

2.9 Heater power and Signal amplitude 8

2.10 Device structure 9

2.11 AFM image of interesting area 9

2.12 Pole tip recession profile by AFM 10

2.13 Fly Height and Distance from Slider Substrate (um) 10

2.14 Defect sample prepare for two types Bump and Crater 11

2.15 AE signal output over defects on the conventional glide test 11

2.16 AE signal output over defects on the DFH glide test 11

2.17 Touch Down contact Sensor cartoon 13

2.18 Wave form of Contact sensor voltage output at asperity contact 14

2.19 Standard Deviation Sigma as function of heater power. 14

a) Inner diameter b) Middle diameter c) Outer diameter

2.20 Sensor amplitude and Head clearance 15

2.21 Defect mapping by decreasing clearance 15

2.22 Historical trend of bit dimensions vs. AD 16

2.23 Historical trend of BAR vs. AD 17

2.24 Historical trend of HMS vs. AD. Green: actual HDD data, 17

Red: Earlier predictions

2.25 Historical HMS vs. bit length 18

2.26 Estimate of the annual HMS contraction rate vs. the AD growth rate 19

3.1 Disk Drive components 20

3.2 Slider body and Suspension 20

3.3 Head Gimbal Assembly (HGA) components 21

3.4 Slider size compare to a Dime 21

3.5 HGA process flow 21

3.6 Pitch Motion 23

3.7 Roll Motion 23

3.8 Pitch direction 23

3.9 Bonding strength 23

3.10 Schematic diagram of DET Tester 24

3.11 DET Tester 25

3.12 Architectures of Spin-Stand Tester 25

vi

LIST OF FIGURES

NO. DESCRIPTION PAGE

3.13 Amplitude on down track position 26

3.14 Amplitude Asymmetry 26

3.15 Track Profile and Micro-track profile 27

3.16 Track Profile 27

3.17 Micro-Track Profile 27

3.18 Recording system. Longitudinal magnetic recording (left) 28

and Perpendicular magnetic recording

3.19 PMR Differentiated Waveforms from Oscilloscope 29

3.20 Squash measurement 30

3.21 Measurement of Track profile 31

3.22 Track profile and Differentiated track profile 31

3.23 Head protrusion, DFH operation 31

3.24 DFH Trade offs 31

3.25 Bulge profile 32

3.26 Actuation curve 32

3.27 Origin of material protrusion 32

3.28 Phase modulation and Touch down sensor 33

3.29 Phase modulation and Touch down sensor cartoon 33

3.30 Conductors and Insulators free electron 34

3.31 Touch down sensor cross section view and ABS view 34

3.32 Touch down sensor structure 35

4.1 Touch down detection Phase modulation and Touch down sensor 36

4.2 Touch down detection Phase modulation and Touch down sensor 36

After Re Test

4.3 % Repeat Early Touch down 37

4.4 Experimental Flow 38

4.5 Phase modulation 39

4.6 Touch down sensor 39

4.7 Touch down detection of sample before adjust PSA 40

4.8 Touch down detection of sample after adjust PSA 40

4.9 Phase modulation’s touch down power distribution 41

4.10 Touch down power mapping between TDS and PM 42

4.11 OverWrite vs Magnetic Write Width 43

4.12 SNR vs Magnetic Write Width 43

4.13 Magnetic Write Width distribution 43

4.14 Signal to Noise Ratio and Touch down power 44

4.15 Overwrite (RevOW) and Touch down power 45

4.16 Touch down binning to average of SNR and RevOW table 45

4.17 SNR Sigma and touch down power 46

4.18 Reverse Overwrite Sigma and touch down power 47

4.19 Touchdown binning to SNR and RevOW sigma table 47

4.20 High-volume Touch down data Phase modulation 48

4.21 High-volume Touch down data Touch down sensor 48

4.22 % Early touch down of PM and TDS in high volume production 49

4.23 Touch down power sigma of PM and TDS in high volume production 49

vii

4.24 Touch down power average of PM and TDS in high volume production 51

5.1 % Early Touch down 52

5.2 Touch down power sigma 52

viii

LIST OF ABBREVIATIONS

A/D Areal Density

AFC Anti-Ferro magnetically Coupled layer

AlTiC Alumina Titanium Carbide

BER Bit Error Rate

BEST Basic Environmental Stress Test

BHV Buffer Head Voltage

CIP Current In Plane (GMR reader)

CIPT Current In Plane Test

CMP Chemical Mechanical Planarization

CMT Coercively Margin Test

CPP Current Perpendicular to Plan (TMR)

CSS Contact Start Stop

DDP Drive Development Process

DET Dynamic Electrical Test (aka MT)

DFH Dynamic Fly Height

DL Domain Lock-up

DLC Diamond Like Carbon

DTR Discrete Track Recording

ESD Electro-static discharge

FIB Focused Ion Beam

FL Free Layer

GBB Gold Ball Bonding Process

GMR Giant Magneto-Resistance

HAMR Heat-Assisted Magnetic Recording

HBM Human Body Model

HFAT Head Failure Acceleration Test

HGA Head Gimbal Assembly

HSA Head Stack Assembly

L/UL Load/Unload

LMR Longitudinal Magnetic Recording

MAMR Microwave-Assisted Magnetic Recording

Mgap Metal gap aka Write gap

MRW Magnetic Read Width measured at HGA test

DET Dynamic Electrical Test

MEW Magnetic Erase Width

MWW Magnetic Write Width measured at HGA DET

NLTS Non-Linear Transition Shift

PBERT Partial Bit Error Rate

PCRT Process Characterization & Readiness Team

PLR Pinned Layer Reversal (ESD/EOS Related )

PMR Perpendicular Magnetic Recording

POR Plan Of Records

PORM Product On-going Reliability Monitoring

PTD Pre Touch Down

PTP Pole Tip Protrusion

QST Quasi-Static Test

RDT Reliability Demonstration Test

RevOW Reverse Overwrite

ix

RFPE Return Field Partial Erasure

SJB Solder Jet Bonding Process

SMAN Spectral Max Amplitude Noise

SNR Signal to Noise Ratio

SpSNR Spectral Signal-to-Noise Ratio

STW Servo Track Write

SUL Soft under Layer

SWA Side Wall Angle

TMR / TuMR Tunneling Magneto-Resistance

TPI Tracks per Inch

VDT Variable Data Tracks

WATER Wide Area Track Erasure

1

CHAPTER 1

INTRODUCTION

1.1 Background/Rationale for the thesis

Hard Disk Drive technology tends to increase the Areal Density (AD). Figure1.1 is

the Areal density trend by Year when the product is available. It shows that

cumulative growth rate (%CGR) is 60% during years 2006 to 2010 and 20% to

60% CGR predicted for year 2010 to 2015. There are major technology changes

since the 90’s period to improve areal density.

MR head thin film was introduced during 1990 to 1997 with 60% CGR, Giant

Magneto Resistive (GMR) during 1997 to 2003 with 90% CGR, Perpendicular

magnetic recording (PMR) demonstrate 60% CGR between 2005-2010 and future

products with 20% to 60% CGR from 2010 to 2015, Discrete Track Magnetic

recording, Bit Pattern Magnetic Recording and Thermal assist magnetic Recording.

Those technologies require low magnetic spacing as possible and maintaining

Reliability performance both component level (Head) and Drive level. Lower

magnetic spacing requires lower mechanical fly height. On top of low mechanical

fly height, head can get closer to media magnetic layer by applying heat to make

head the temporary protrude.

To achieve high areal density by optimizing magnetic spacing while maintaining

Head and Drive reliability performance, Dynamic Flying Height (DFH) technology

was introduced around 2005. This technology improves electrical performance by

reducing magnetic spacing (uses Thermal expansion property).

Lower magnetic spacing also requires accurate touch down point detection.

Touchdown point identification is important because there will be a trade off

between electrical performance and reliability performance. We know that lower

magnetic spacing can boost up electrical performance but it is a risk for reliability

performance with increased probability of head disk interference problem (HDI).

This thesis studies the new technology for Touchdown detection, Touch Down

Sensor (TDS). The TDS will be compared with the conventional technique, “Phase

modulation” or “PM”. Electrical performance, Dynamic Electrical Test (DET), will

be used as the factor to judge whether TDS is good enough to implement rather

than PM.

Figure 1.1 Areal Density Trend by Year for Mobile product

Figure 1.2 Magnetic Spacing Road map.

2

1.2 Statement of the Problems

Low fly height is less than 10 nanometers. Conventional Touch Down detection

method (PM) has the limitations of both hardware and software. Production data

indicates that PM creates high “Early touch down” problem (ETD).

Early touchdown is an irregular phenomena of head touch down detection. This

problem occurs when the head flies too high, contamination sticks on the element

area or comes from tester module itself. Early touch down problems can cause poor

electrical performance by impacting magnetic spacing, Too high a magnetic

spacing induces low amplitude and high noise. The major electrical parameter that

is impacted from ETD is SNR, “Signal to Noise Ratio”.

Touch Down Sensor detection method or “TDS” was proposed to improve the

detection performance and reduce early touchdown failure.

Figure 1.3 Disk drive components. Source: Western Digital Company

3

1.3 Objectives of the Research

This thesis is studies and compares Touch down detection performance of new

technology, “TDS” and compare to conventional detection technique, “PM”.

The electrical parameter of interest at spin-stand level is ETD. The sample is

prepared in “Head Gimbal Assembly” form or HGA form. The sample is tested

with two detection methods. ETD rate and electrical performance, especially Signal

to Noise ratio (SNR), are the major parameters to consider.

1.4 Scope

I. Study conventional touch down detection technique, PM by collecting high

volume data from spin-stand level, Dynamic electrical Test (DET) to see early

touch down rate.

II. Study new touch down detection technique, TDS, at a spin stand level to see

electrical performance and early touchdown.

III. Compare the result between conventional and propose technique to identify

which technique should be implemented.

1.5 Limitation of the study

I. The limitation of this thesis defined as below.

II. Samples (HGA) to be evaluated have to be degradation free or clean.

III. Tester reservation, Spin-stand tester to be prepared and reserved to test the

sample.

IV. Decision making to implement new techniques, high volume data required.

Need to collect high volume data about 2 weeks.

4

1.6 Thesis Schedule

Figure 1.3 Thesis schedule

5

CHAPTER 2

LITERATURE REVIEW

2.1 Dynamic Fly Height for PMR Recording

The magnetic spacing control through a device by increase a temperature has many

techniques. Properly applying the spacing control can increase electrical performance,

especially Bit Error Rate and Signal to Noise ratio, (SNR). A good touch down detection to

define a magnetic spacing is needed for HGA or Drive operation. The study try to

evaluates the electrical performance by magnetic spacing optimization when the writing

and Reading process.

In Hard Drives, Magnetic spacing is the critical parameter for higher areal density.

Reducing magnetic spacing is important and needed. The technology to reduce magnetic

spacing is heating the element. Apply power then sensor area is protruded and magnetic

spacing can be reduced. Magnetic spacing reduction improves electrical performance, SNR

and BER both HGA level and Drive level.

How to control the magnetic spacing by apply voltage and heat up the device, The good

detection method is required. Spacing change during the head fly or operate can also be

calculated by the Wallace Equation. Magnetic spacing define during the head operation,

we need to know the reference point or touch down point.

Signal and Noise in perpendicular magnetic recording improves with Clearance reduction

expected to improve Bit Error Rate. Improvement comes from various factors.

Write’s operation can generate both amplitude signal and noise signal and then Reader will

be detect both signal and also noise from media to convert to electrical signal.

BER improve as magnetic spacing reduces (R refers to Read Clearance in Nanometers

W refers to Write Clearance in Nanometers.)

Figure 2.1 Bit Error Rate (BER) improves when Read and Write spacing is reduced. The

Write spacing operation and Read spacing operation delta by 1 nanometer per step. BER

improves by approximately one order when head magnetic spacing reduced in both reading

and writing by 3 nanometers. The electrical performance expands over a wide range along

a bit length Kbpi, Electrical performance can be improved on high Kbpi.

Evaluation of writing and Reading impact, one Kbpi number was selected and the

combination Read-write magnetic spacing was settled for BER measurement. Achievement

Figure 2.1 Bit Error rate and Linear Density relation

Figure 2.2 Bit Error rate and Read clearance (nanometers)

6

of Bit Error Rate with read and write magnetic spacing optimization, SNR improvement is

different in these 2 procedure. The Write spacing minimization induce high to MF,

Medium frequency signal and less noise from the media platter.

Figures 2.3 and 2.4 are Read spacing. The Medium signal amplitude rises by reading

clearance decrease. Media noise will increase by medium amplitude signal, the media

Signal To Noise Ratio is defined as The ratio of Medium amplitude signal and the media

noise.

Touch down detection and electrical performance improvement are necessary to

control magnetic spacing accurately. We need to reduce the clearance and also maintain

reliability performance. To avoid Head Disk Interference problem (HDI) we need to

determine the touch down point. Wallace equation is used to determine the spacing

reduction.

The Touchdown point can determine the actual operating point. We can apply heat

to the element. Touch down point can be calculated and defined when the head is touching

the media (or very close to the media). We expect that as heater power increases and head

touch the disk surface, the signal amplitude will increase and then we can detect it and

Figure 2.4 Media Signal to Noise Ratio and Read Clearance

Figure 2.5 LOG BER vs. Total SNR in dB

Figure 2.3 MF signal increases as Read clearance reduce

Figure 2.6 Media noise and write clearance

7

report as touch down point. The operation require more add in power to exceed to the

maximum limit or Touch Down point. We will analyze the signal amplitude roll off or

saturated. When the magnetic spacing is decreased by optimum point, the slider will start

to vibrate. This vibration can be detected by AE sensor and amplitude signal. Calculation

of amplitude signal, we need to plug in through a bandpass filter, a numeric full wave

rectifier and a digital low pass filter. The signal is compared to an acoustic signal recorded

under the same condition.

Envelope amplitude signal of two setup of heater was analyzed. The heater power is

160mW and 184 mW for the continuous waveforms blue and red line waveforms,

respectively. Both envelopes reveal the existence of modulation. Both conditions have

same max-signal amplitude but the red line show lower minimum amplitude. Signal

amplitude is going to saturated and reduced when touch the maximum point. Mechanical

perspective can explain that when we apply too high power to the device then head will

impact to the media with high impact power and it will react back with harder. Vibration

has two definitions for TD-detection. HDI contact detector and the modulation signal can

apply as a TDS or touch Down Sensor.

Figure 2.7 SE Signal and a numerically demodulated read signal.

Figure 2.8 Envelope signal as the heater power increases from 160m W to 184 mW. Max amplitude is saturated while min amplitude going down. We can define that head can not move closer to the media.

8

The maximum amplitude increases monotonically until it reaches a maximum and stays at

the maximum. The average amplitude reaches a maximum and starts to decrease. The

maximum is reached when the modulation amplitude starts to suddenly increase. The

increase of modulation amplitude indicates that the element is vibrating and the spacing

between head and media changes. It means that the head is closest to the disk. The curve

show saturated, it mean that head cannot go further from this point (Touch Down Point) .

According to the head can meet to maximum or lowest point in short time, the head also

fly over the media almost of time. Amplitude tend to increase as function of heater power,

We can determine that head fly away far from the media by the more vibrations.

In summary, head-to-disk spacing affects BER during both read and write operation

though the effects are different. Two important factors in controlling clearance, touchdown

detection, and Wallace equation were also analyzed. The analysis suggested the use of

mechanical vibration as the touchdown detection method.

Figure 2.9 Heater power and Signal amplitude

9

2.2 Head Protrusion profile by Heater design

Hard Disk Drives write and detect magnetic signals by compensating and optimizing the

magnetic spacing by thermal fly height control, (TFC). TFC also serves as a new technique

for compensating static flying-height (FH) variations due to manufacturing tolerance and

environmental changes, and reducing the risk of harmful head-disk contact, which is

essential for long-term reliability. In such a drive, a resistive heating element (heater

element) is deposited near the read/write elements, and the gap FH is reduced by applying

a current through the heater element to deliberately induce heater pole-tip protrusion (H-

PTP).

Moreover, in order to Increase the areal density (ad) by reducing the magnetic spacing on

reading and writing process, a “touchdown (TD) and pull back (PB)” scheme has recently

been implemented in current hard disk drives. The operation of touchdown and a pullback

is performed by increasing the power applied to a TFC heater element in the magnetic head

until the head-disk contact, i.e., touchdown, is detected, and the protruded bulge is pulled

back by reducing the power. One major advantage of this operation is the head-to-head FH

variation can be optimized and a sub-1-nm FH can be achieved for read/write operation.

However, due to physical separation between the TMR sensor (reader) and the write gap

(writer), the pole-tip recession (PTR) profile created in the lapping process as well as other

considerations, the heater element has to be properly designed to obtain a preferred

protruded bulge for both reading and writing. The nonlinear thermal–structural coupled-

field problem of the Thermal Fly Height sliders has been fabricated and vary test setup

conditions to see the electrical performance or AFM profile. This study experiment an

iterative approach to numerically predict the DFH sliders' flying performance, such as

actuation efficiency and Flying Hieght profiles.

Figure 2.1 is a cross-section view of the slider. The slider material is Al2O3–TiC substrate.

The heater coils have 3 turns and a bottom pole (P1) thickness of 1.2 um. The top shield

and and bottom shield are about 1 micron. The spacing between the Tunneling Magnetic

Recording sensor and the WG, writ- gap is 6.5 um approximately. Three heater design by

heater locations designed to be experiment. The heater A is stay between the substrate-

Figure 2.10 Device structure

Figure 2.11 AFM image of interesting area

10

Shield1; Heater B and heater C are stay between P1 and the bottom coil, the latter is far

away from the ABS. The ABS is the same for all three sliders to make sure that the

flyability of three heads are the same (Mechanical flying Height should be the same among

three design). Pitch angle is 110 micro radian in the flying condition at a linear velocity of

12.5 meter per second(0 skew angle). The contact area of the sliders is difference base on

the Pole tip Recession profile head design head structure. PTR profile was inducing by

many factors, overcoat condition, removal rate between slider fabrication operation,

removal rate and time and also the slider materials. Metal material can also show higher

recess than alumina because alumina has lower removal rate than metal.

Optical microscope, Scanning electron microscopy (SEM) and AFM images were

measurement tool for analyze this experiments and render to be 3 dimensions pictures for

user friendly analysis. Calculation method program to calculate the Fly Height profile and

protrusion profile at touch down point and overpush conditions. The Fly Height ability

defined as the induce of ABS design and media. The appropriate heater power apply to

each design, convert to be 4 nanometers distance of spacing. The three heater designs

shows difference FH profiles, HeaterA can induce reader area to closer to the media than

other design. (Gain electrical performance but high risk to reliability performance. DFH

heater can also evaluate and analyze to improve Soft Error Rate by optimizing reader and

Writer protrusion profile.

In summary, this paper studies a protrusion profiles and wear patterns of a DFH slider with

three different heater designs. Numerical approach issued (with detailed head structure and

PTR profile measured by SEM and AFM) to calculate the 3 dimensions protrusion profiles

and predict the carbon over coat wear pattern created in the DFH over-push condition.

They experiment by use GUSIK tester to investigate wear pattern profile.

Figure 2.12 Pole tip recession profile by AFM

Figure 2.13 Fly Height and Distance from Slider Substrate (um)

11

2.3 Study Defect mapping by Dynamic Fly Height detector.

This study propose to evaluation DFH detector for defect mapping. Dynamic Fly Height

slider prepared as a detector. 2 defect prepared, Bump (5nm height) and Crater (37nm

depth and 50nm Width). The conventional glide test and Dynamic Fly Height glide test

over 65mm diameter disk. Acoustic signal from 2 glide test method to be compared.

65mm Disk sample with by two types of defects. Bump defect fabricated by laser

irradiation, Crater defect. The roughness of disk is 0.4nm.

Result

Figure 2.14 Defect sample prepare for two types Bump and Crater

Figure 2.15 AE signal output over defects on the conventional glide test

Figure 2.16 AE signal output over defects on the DFH glide test

12

The experiment indicated that the Dynamic Fly height method can detect the defect by fe

nanometers spacing. While the conventional can not detect the defect. According to the

Acoustic signal sensor is saturated below 3 nanometers that why conventional method can

not detect the defect. DFH, Dynamic Fly Height is the propose method for future products,

technology because the new technology require lower magnetic spacing and high areal

density design.

In summary, The conventional glide test could not evaluate the flyability of magnetic disks

below a glide height of 3 nm. However, this was possible with the DFH glide test. The

sensitivity of defect detection on the DFH glide test was higher than the conventional glide

test, because the DFH glide test could work at a larger rotational velocity and lower glide

height and the DFH glide test also was able to detect crater defects at a glide height of a

few nanometers.

13

2.4 Defect mapping by Touch Down Sensor

Hard Drive areal density trend is continuously increase every year by 60% CGR

approximately since 2003. Hard drive for new technology, new product require more areal

density and to achieve the extraordinary areal density, ultra low fly height or magnetic

spacing reduction is also necessary. Dynamic Fly Height is the new technology to

accelerating to achieve low fly height by reduce the magnetic spacing. To achieve 2

nanometer of Hard Drive operation, DFH is need. Reliability performance also concern to

the ultra low fly height, Head Disk Interference problem avoid.

Targeting both high Touchdown sensitivity and defect detection, Thermal contact sensor integrated into head developed. Experimental result show that Thermal contact sensor performance is equivalent to conventional detection, Acoustic detection. A Thermal asperity is the defect that often observed in the drive operation. The Thermal asperity signal can detect by Magneto Resistive sensor. The Touch Down Sensor was used to this experiment as seen in Figure 2.17. The cross-section view th edetermive the sensor position. In general slider body always contains Reader and Writer element. The heater is the latest technology that add into the slider body to improve electrical performance by reducing magnetic spacing. The Touch Down Sensor was fabricated to be stay between Reader and Writer. The function of Touch Down Sensor is that detect the heat that generate beteen slider body (Trailing Edge) to the media. When trailing edge is come closer to ther media then heat is generated and then the sensor will detect the heat and Resistance of the sensor change. The threshold of delta change of resistance determined to be to Touch Down point of TA detection. The relationship between temperature and resistance show in below picture. It called Thermal Coefficient of Resistance. TCR.

14

Waveform of Touch Down Sensor indicated the thermal asperity detection . The

fundamental of uses the Thermal coefficient of Resistance of material.

The power is applied to the coil then the temperature increase in the device and when the

head detects or touches the defect point then the touch down power sigma of head

suddenly increases.

Figure 2.18 Waveform of Touch Down Sensor

Figure 2.19 Sigma as function of heater power. a) Inner diameter b) Middle diameter c) Outer diameter

15

The profile indicated that lowest clearance (2 nanometers) shows the highest amplitude. It

means that when the read elements fly lower or closer to the defect than the amplitude

signal is significantly increased.

Defect mapping by decreasing clearance indicated that the lowest clearance (c) 1

nanometer highest count of defects. (b) is 1.5 nanometers (a) is 2.0 nanometers

In summary, Touch Down Sensor that fabricated into the device was studied. The

experiment confirm that TDS show relationship to the conventional detection, AE sensor.

Spin-Stand result confirmed the relationships between the sensitivity of defect detection,

sensor and the magnetic spacing reduction. Defect mapping also confirm the sensitivity of

magnetic spacing to the detection, low magnetic spacing can detect the defect easier than

the high spacing.

Figure 2.20 Sensor amplitude and Head clearance

Figure 2.21 Defect mapping by decreasing clearance

16

2.5 Magnetic spacing trend

High Areal density requirement is depend on the Writer and the Reader that embedded into

the slider body. The Magnetic spacing distance id the gap between the device (This case is

Reader) and the top surface of the media disk. We can definition of Magnetic spacing my

Head media separation or HMS. Head media Spacing can be optimized by overcoat

process, Too thick over coat thickness will increase the HMS spacing and too low over

coat can reduce the spacing but the reliability problem can occurs.

Head Media Spacing trend can predicted but nobody can define the exactly HMS point

each year but we can refer to the %CGR (Cumulative Growth Rate) of Areal density which

increase by 60% each year and we can estimate the HMS possibility value.

Figure 2.22 shows the dimension vs. areal density for a mobile product and Desktop

product.. The areal density range was about 3 orders of magnitude from 300 Mb/in2 90’s

to 400 Gb/in2 in 2009 (15 years)

According to the %CGR of areal density is increase as 60%. It shows the magnetic write

width is going to narrower or faster rate than the Bit length.

Figure 2.22 Historical trend of bit dimensions vs. AD

17

Figure 2.23 Bit Aspect Ratio vs. Areal Density. BAR trend to reduce by increase the areal

density for example 30 to be 5 of today status.

Magnetic Spacing

Figure 2.24 shows relationship between Head Media Spacing to Areal Density. Per Areal

Density required to be increase, Magnetic Spacing also required to be lowest as possible.

The exponent number is -0.317. 100 Gigabit per square inch require Magnetic spacing

lower than 10 nanometers.

Figure 2.23 Bit Aspect Ratio vs. Areal Density

Figure 2.24 Real Density vs. Magnetic Spacing green are prediction, Red are actual

18

Figure 2.25 confirm the strong relationship between Bit length and Magnetic Spacing

Discussion

Magnetic Spacing and Bit Length

In general the track width is approximate 10% of half the bit length (B) for 15 years. The

HDD trend was phase in new technology both Writer PMR and Reader TuMR the media

technologies also improve. A linear scaling relationship between HMS and B is easy to

justify, but the factor of ½ is not. We’ll begin by presenting a simple argument for the

linear scaling relationship, and then discuss the factor of ½

We can define the Pulse width at 50% as below picture.

G is the Read-Gap, a is the transition parameter, d is the Head Media Spacing, and δ is the

medium thickness.

Figure 2.25 Magnetic Spacing vs. Bit Length

19

Magnetic Spacing Reduction Rate

Head Media Spacing reduction rate as the relationship to the Areal density Growth Rate

shown in Figure 2.21

The equation was approximate as below.

Conclusion

Historical data of HDD industry indicated that Areal Density of HDD product increase b 3

orders over latest 15 years, Magnetic Write Width (MWW), Bit Length and the Magnetic

Spacing HMS have scaled to the function of areal density Log function. BAR Bit aspect

ratio has slowly decrease from 30 to 5 of current. Discontinuity problem was not observed

during 15 years of new technology introduced, MR head to be GMR head then PMR then

DFH.

Figure 2.21 Areal Density vs. Head Media Spacing Contraction rate

20

CHAPTER 3

METHODOLOGY

3.1 Head Gimbal Assembly (HGA)

Head Gimbal Assembly (HGA) is an important component of the Hard Drive. During

HDD operating, it will bring Reader/Writer pole up to the right height relative to top of

Media surface (Head Media spacing). It will allow information (data bits) to be written and

read on/from the media. HGA’s major components are Suspension and Slider. Slider’s

functions are read and write processes in the drive. HGA assembly process contains major

10 operations. First is Suspension load to pallet. This process is an automated process then

Auto adhesive dispensing then Slider binding operation. Suspension and slider are attached

together with the special glue. Next is to bond the slider bond pad with suspension bond

pad (Solder Jet Bonding).HGA inspection under low power inspection to screen

mechanical failures like bonding failure, contamination and slider alignment. The next

operation is Pitch Static Attitude (PSA) and Roll Static Attitude (RSA) measurement audit

then Hi resolution power inspection and then auto OCR unload. The OCR unload operation

is an operation for reading the slider serial number or serializing. The last operation is

Gram load measurement.

Functions of HGA Processes are

a) Maintain the mechanical and electrical performance of the Suspension assembly

and Slider

b) Place Slider onto the Suspension at right position and strong enough to withstand

mechanical/environmental excitation through its expected product life.

c) Properly connects the circuit of Slider’s terminals and Suspension’s terminals, as

well as maintaining the valid circuit through its expected product life.

d) Outgoing HGA will be free from contaminants.

Source: Western Digital Company

21

The HGA process flow, starting with suspension load operation. 1st, Suspension will be

loaded from suspension tray or suspension pallet. This operation is automation by robot.

Suspension will held by fixture then adhesive dispensing machine dispenses the adhesive

2nd , special adhesive is put on the attachment area. Slider is attached to the suspension by

OSC machine, 3rd figure. The next operation is an IR cure process. This operation is to

cure the adhesive with high temperature 4). The 4th operation attached the slider to the

suspension but HGA still can not operate. We need to bond the slider to the bond pad. We

use a Solder Jet bonding machine to put the bonding material to connect slider bond pad to

suspension flexure. This operation is very critical because electrical testing performance

depend on precision of the bonding location, missed bonding location will create electrical

Figure 3.3 Head Gimbal Assembly (HGA) components

Figure 3.4 Slider size compare to a Dime


Figure 3.5 HGA process flow


22

abort problem, misalignment bonding will create low Dynamic electrical performance or

create contamination.

The HGA is now almost ready to function. we need to make sure that mechanical defect

can be screen out, Low power in section 6) with eyeball and microscope is used to inspect

mechanical defects including cleanliness of HGA. The 7th operation is PSA/RSA

measurement. We need to make sure that mechanical fly height is within the specification

by PSA/RSA measurement. This data can be fed back to the process performance. After

PSA/RSA complete than HGA is process to Hi- power inspection, the major area function

is cleanliness.

The HGA is now ready to test (DET). The 8th operation is auto OCR unload, or serialize

the serial number on each slider. We can know HGA’s serial number in the tray’s position.

This is also critical operation because we need serial number before loading HGAs to DET

tester (Mag-Test) The Serial number contains information since wafer fabrication date

(when wafer was produced) DET data, we can know the electrical performance on each

head by serial number and the correlate DET data to Bar Quasi static data and then wafer

fabrication data.

We always use DET data to feedback wafer fabrication performance, Slider fabrication

performance and also HGA process performance. After we ship the HGA to the customer,

Head Stack and drive, we can also trace back drive performance by breaking down to HGA

data. Every operation can be accessed by serial number information.

The 10th operation is Gram load measurement and then the 11th is AQ cleaning to remove

some loose contamination from the HGA. The 12th step is DET testing operation. This

operation simulate basic drive operatiing function at HGA level, Head writes signal on the

magnetic layer and reads the signal back from the media and translates those information

into the critical DET parameters. MR-impedance, amplitude MF, amplitude LF, Reverse

Overwrite, Signal to Noise Ratio SNR, Touch down power , Asymmetry, Squash, T50,

WR offset, fundamental signal, noise and also magnetic write width. Major DET

parameters are Reverse Overwrite, SNR and Magnetic write width. We use these top three

parameter to feedback wafer performance and Slider Fab performance. For wafer

fabrication perfornace improvement, we consider Magnetic write width mean and sigma.

MWW sigma is the most important data to feedback to wafer fab performance, we need the

lowest MWW sigma while Reverse Overwrite and SNR always require higher value, (BER

improvement).

We can also use DET data to feedback Slider fab process. MWW, MRR and amplitude is

the data to be considered. Lapping process is very sensitive to the SH (Stripe Height)

value, too much lapping create shorter SH which contribute to higher asymmetry sigma

and wide MWW and also hig MWW sigma, on the other hand longer SH creates narrow

MWW which lowers electrical performance, Reverse Overwrite, SNR. The new lapping

technology to reduce MWW sigma is called “Tilt lapping”. Tilt lapping is the lapping

process which tilt the lapping bar by the required degree with respect to writer, we can

reduce or increase MWW by maintain the same SH,(which means that reader performance

can be maintain but writer performance can be improved). The 13th operation is the Cut

and Bend process. This step is to cut the test pad before shipping to build the Head stack.

The 14th process is AQ cleaning, then low power inspection (15th )and then packing

before shipping to Head stack area.

23

The Static attitude is defined as the angle of orientation of slider to Air Bearing Surface

(ABS)

Figure 3.6 show the pitch motion and Figure 3.7 show Roll motion.

Figure 3.8 indicates that pitch motion can be relate to Fly height gap and Figure 3.9 show

the direction of bonding strength. Bonding strength is the parameter that indicates whether

the adhesive and bonding performance are acceptable or not.

The Slider/Flex bond strength is the ability of HGA to withstand the shear force happening

at the adhesive bond line. Testing is done after the HGA has passed through all adhesive

curing and without Solder JET bonding.

SJB bond strength is the ability of HGA’s Solder bond joints to withstand the shear force

happening at the Solder bond joints. Testing is done after the HGA has passed through all

adhesive curing (UV and IR) and with SJB bonding.

Figure 3.6 Pitch Motion Figure 3.7 Roll Motion


Figure 3.8 Pitch direction Figure 3.9 Bonding strength


24

3.2 Dynamic Electrical Test (DET)

DET testing is the electrical test at spin-stand tester of DET tester. The major purpose of

DET testing is to see head performance both write ability and read ability, and to screen

poor heads before processing to the next operation, (Heads Stack and Drive). DET testing

can reduce Drive cost by screening poor heads before Drive assembly. Important electrical

parameters at DET are Signal to Noise Ratio, Reverse Overwrite and Write Width. DET

parametric test output parameters are:

a) Amplitude LF (6T), MF (2T) & HF(1T) )

b) Noise (electronic off disk noise and media noise)

c) Amplitude Asymmetry

d) MWW, MTW, MCW, EBW, OTC, MRW…. (Triple track test)

e) Micro-track Reader Width

f) Pulse Width (T50)

g) Resolution

h) Squash/ATI

i) Overwrite (Overwrite and Reverse Overwrite)

j) Nonlinear Transition Shift

k) 2T SNR

l) WSSNR

m) Electronic SNR

Figure 3.10 Schematic diagram of DET Tester


25

Architectures of Spin-Stand Tester

The system block diagram show how tester operate. It simulates opertion of disk drive with

good flexibility, precise and stable platform, disk rotation, holds HGA, Loads the head

onto the disk, positions the head to the disk (Position and skew) and moves the head during

testing (radial direction).

Figure 3.11 DET Tester



26

3.2.1 Amplitude

The amplitude of a recording system is characterized by writing single frequency patterns

and averaging the peak-to-peak amplitude of the read-back sigmalon track. These

measurements are referred to as TAA, or “Track-Averaged Amplitude.” Single-Frequency

patterns are often referred to by the multiples of the smallest period from the random

pattern being characterized. The most common measurements are:“Mid Frequency”

measured from a 6T signal.“Hi Frequency” measured from a 2T signal.

Write a high-frequency signal (2T period) and micro-jog for reading

1) Measure positive and negative (0-p) amplitude samples around the track

2) Repeat step 2 for several (such as 3) write/ read cycles

3) Write a low frequency signal (6T period) and micro-jog for reading

4) Measure positive and negative (0-p) amplitude samples around the track

refers to the highest frequency bit cell. For a recording system operating at 1000 kfci, T = 1

µin = 25 nm

Measurements are performed for amplitude at 6T (LFA) and 2T (MFA). Most of the signal

power in random data is at 2T

A square-wave pattern is recorded at 2T or 6T and passed through a narrow band filter to

determine the peak to peak amplitude

The chart shows an illustrative example for 1000 kfci, 1T

27

3.2.2 Track Profile and Micro-track Profile

The track profile is measurement of the width of the writen track. The mewasurement is to

move reader above the writen track in the cross track direction. MWW can be defined as

the width of cross track position at 50% amplitude. For the micro-track the edges of the

written track are trimmed to leave a very narrow written track typically 10% of the original

track width.

The micro-track is used to characterize the reader width.


28

3.2.3 Reverse Overwrite

Reverse overwrite is the ability of a writer to write or erase data over old data. At drive

level, new data is written directly over old data, without erasing the old data first.

Algorithm of Reverse Overwrite

1) AC Erase

2) Write high frequency

3) Measure and record the hi frequency amplitude of the read back signal (V1)

4) Write low frequency

5) Measure and record the high frequency amplitude (residual, remainant) of the read

back signal (V2)

6) Calculate Overwrite. Overwrite = 20log(V2/V1) or 20log(V2)-20log(V1)

7) The result of Overwrite will be compared to specification. Pass or Fail.

Reverse overwrites ia a function of 5 variables

a) Write Field Gradient which depends on pole saturation

b) Write head field strength and writing bubble size

c) Media properties, Hc, Soft under layer properties and grain size

d) Write Gap

e) Head media spacing

Figure 3.12 Recording system. Longitudinal magnetic recording (left) and Perpendicular magnetic recording


29

3.2.4 Signal to Noise Ratio

Signal to Noise Ratio or SNR is the measurement of amplitude signal and root mean

square, RMS of signal after the track is erased. The SNR formula is calculated as

following.

SNR = -20 log (Vnoise RMS/TAA)

TAA = Track amplitude of high frequency signal

Vnoise RMS = Root Mean Square of signal on erased track

3.2.5 Squash Measurement

Figure 3.13 PMR Differentiated Waveforms from Oscilloscope


30

The squash is defined as the ratio of the signal under squeeze to the original signal. It is a

measure of side writing of the head. Smaller numbers are worse.

The test algorithm is:

a) Perform Band erase

b) Write data track

c) Measure the track profile according to the range and step parameters

d) Get maximum TAA

e) Write two adjacent track at the squeeze position

f) Measure TAA

g) Calculate Squash

SQUASH = TAA Squash/ Max TAA

3.2.6 Track profile

31

The Track Profile is a convolution of the MR Read Sensitivity Profile and the Written

Track Profile.

a) Read width can be calculated from either the track profile or a differentiated

track profile.

b) Standard read width should be obtained from micro-track reader width.

c) The differentiated track profile can be thought of as a linear superposition of

2 read sensitivity profiles spaced one write width apart

Magnetic Track Width (MTW) is an amplitude based write with erase test.

a) Write a track and position the reader at the position along one side of the

track where the amplitude from the track scan is 50 to 85% of the maximum on one

side of the track

b) Repeatedly measure amplitude after incrementally writing AC-erased

adjacent tracks closer and closer to the reference track. The amplitude degrades

linearly with adjacent-track position.

c) Repeat this process on the other side of the track.

d) Fit lines thru both amplitude-vs-adjacent-track-position data sets

e) Report the average of the two intercept positions.

MTW = MWW + EB_T/2

MTW = Magnetic Track Width

MWW = Magnetic Write Width

EB = Erase Band

3.2.7 Dynamic Fly Height (DFH)

Figure 3.14 Measurement of Track profile

Figure 3.14 Track pro5ile and Differentiated track profile

32

Dynamic fly height or DFH is the technique to boost up electrical performance by reduces

magnetic spacing by applying heat into materials or device to make it protrude.

The Temperature Coefficient of Resistance (TCR) is the parameter to be watched.

Interesting area is the Trailing edge.

Dynamic Electrical Test (DET) always starts with Touch down point identification. This

process is very important because magnetic spacing can be defined.

The objective of DFH is to achieve optimal balance of reader spacing, reader heating and

bulge profile for maximum performance while maintaining good Head Disk Interference

(HDI) and electrical reliability

DFH process at HGA level, spin stand level are

a) Apply power to the heater (mW)

b) Head protrudes, (Trailing edge area)

c) Touch down point detection

d) Apply back off power to meet operating point.

e) Parametric test

Figure 3.16 Head protrusion, DFH operation

Figure 3.17 DFH Trade offs Figure 3.19 Bulge profile


33

Back off power identification can be determined by the DFH characterization test. The X-

axis is the applied power (mW) to the device and the Y-axis is the protrusion (nm).

Actuation efficiency is defined by the equation of the curve.

Figure 3.21 Origin of material protrusion Figure 3.20 Actuation curve

34

3.2.8 Touch down detection

Conventional touch down detection is “PM”. The new technique is “TDS”.

PM technique defines touch down point by vibration detection. When heat is applied to the

head then the device is protrudes, Trailing edge area is protruded close to the media and

then the head vibrates.

PM can detect the vibration then reports the onset of vibration as the touch down point.

The weak point of this technique is “Early touch down” problem or ETD. It causes by

many factors, contamination hardware and software limitation. Early touchdown is impact

electrical performance, especially Signal to noise ratio SNR.

The TDS technique can detect the heat generated by contact by using the

Temperature Coefficient of Resistance property of material. When heat is applied to device

and the trailing edge area moves close to the media, then heat between the media and head

is generated. The TDS resistance change by temperature and Touch down point can be

defined.

Temperature Coefficient of Resistance theory

a) The Resistance of materials changes due to atomic activity changing (by

changing Temperature).

b) Conductor’s materials show an increase in resistance as temperature

increases.

c) Insulators materials have decreasing Resistance as temperature increases.

d) The reason for the change of resistivity can be explained as Current flow

through materials.

� Current flow is electron movement from on atom to another atom under

electric field.

� High Resistance is less free electrons movement

� Low Resistance is more free electron movement

e) The heat affect the atomic structure by making atoms vibrate, the more heat

the more vibration.

f) Conductors

� High temperature cause free electrons collision, Resistance is increase.

Figure 3.22 Phase modulation and Touch down sensor

Figure 3.23 Phase modulation and Touch down sensor cartoon

35

g) Insulators

� High temperature allows activate captive electrons to be free electrons

then current flow easier. Resistance is decrease.

dR / Rs = α dT

dR = change in resistance (ohm)

Rs = standard resistance (ohm)

α = temperature coefficient of resistance

dT = change in temperature (K)Insulator

3.2.9 Touchdown Sensor

Figure 3.24 Conductors and Insulators free electron

36

TDS is the thermal probe integrated into the Magnetic head. TDS was designed to be

highly accurate detection sensor. Sensor resistance changes with temperature (+ TCR for

metal, - TCR for semiconductor)

The Sensor detects frictional heating at low magnetic spacing/ touchdown point. Head

bounces at TD so temperature/resistance signal is also modulated, mainly at resonance

frequency.

Signal Amplitude at TD about 2.5 to 10 mV/mA or ~ 3 – 30 mV /V for metal sensor.

a) Sensor located at (or close to) ABS

b) Located above P1

c) POR – metal/ NiFe Sensor

d) Temperature Coefficient of Resistance (TCR) = 0.3%

e) Dimensions: 20nm thick,2.5um long and 0.2um wide

f) Sensor is floating, not grounded

Touchdown Sensor Benefits

Figure 3.25 Touch down sensor cross section view and ABS view


Figure 3.26 Touch down sensor structure

37

a) Function is of TDS is to detect Touch Down point and then identify

operating point.(apply power to the heater)

b) Operate in-field and real time feedback to control the DFH for various

environments, Pressure, Humidity.

c) TDS will be universal Touch Down detector for component (HGA) and

Drive.

d) Expecting to improve Touch down repeatability

e) TDS can increase HDD reliability by detecting DFH overdrive that can

cause drive failure.

CHAPTER 4

RESULTS AND DISCUSSION

4.1 Touch down detection from DET tester

38

According to Touchdown detection at spin stand level is very important operation

for DET process. Touch down detection determines operating touch down point and alow

one to set the magnetic spacing during write or read operation. The optimum operating

point can maximize electrical performance and maintain reliability performance at HGA

level and also drive operating level.

PM data from production indicated that there are some weaknesses in this

approach, Early touchdown (ETD) problem. ETD problem can cause low yield or low

output in high volume and also impact electrical performance, especially Signal to Noise

Ratio which related to Bit Error rate in Drive performance.

Figure 4.1 is the sampling sample from production parts. It contains four quadrants

of touch down data characteristics. The sample size is 532 HGAs.

Quadrant1 is touch down failure (ETD) from PM module and pass at TDS module.

Quadrant2 is touch down passers for both PM and TDS.

Quadrant3 is touch down failure for both PM and TDS.

Quadrant4 is touch down failure for TDS but pass for PM.

Figure 4.2 is Re-Test DET data. This figure is to confirm the electrical

performance, touch down data after Re-Test. The result indicates that there are 3.6% of

heads still perform in quadrant1 which means PM failures repeat. Quadrant 3 and4 from

original data were gone by re-Test. It means these parts are back to normal state for TDS

module.

There is one point (one HGA) that was failure of both TDS and PM, but it has

recovered to normal touch down power in quadrant 2.

Figure 4.3 show the touch down power failure rate or early touch down between two

detection techniques, TDS vs. PM. Graph is clearly indicated that TDS techniques can

eliminate repeat of early touchdown while PM has a remaining 9.45% failures.

Figure 4.1 Touch down detection Phase modulation and Touch down sensor

Figure 4.2 Touch down detection Phase modulation and Touch down sensor after Re Test

39

(19 heads repeat failure from 201 failure sample).

The experiment was designed by sampling HGAs from normal distribution (quadrant2)

and applying mechanical adjustment. Result was considered the touch down power outlier

detection from PM technique and Touch down detection technique. The hypothesis of

experiment is that “TDS detection should perform better detection that PM”.

4.2 Experiment flow and Result

Experiment flow is designed to determine HGA performance by consider whether Touch

down detection between TDS and PM can be changed by mechanical adjustment or not.

Pitch Static Attitude adjustment is the most effective factor to affect fly-ability.

Figure 4.3 % Repeat Early Touch down

40

The experiment flow divides the sample into five groups each group will be threted by

PSA adjustment, -0.5 degree, -0.25 degree, 0 degree, +0.25 degree and +0.5 degrees.

The samples will be tested at DET tester (spin-stand tester) to consider touch down

data.

Touch down power distribution (of two techniques) comparison indicated that touch down

power distribution of five groups are in normal distribution range and “Not significantly

difference”. G1 is -0.5, G2 is -0.25, G3 is 0, G4 is +0.25 and G5 is +0.50 degree of PSA

adjustment. The expectation of the experiment is to see whether the TDS detection can

detect the outlier sample from the mechanical adjustment. The result is not support the

original hypothesis. Figure 4.5 (PM) and Figure 4.6 (TDS) are comparable and the

distribution of each group are in normal specification of touchdown power (mW).

G 1 through G5 distribution indicated that PSA adjustment range can not make the

touchdown power to be the outlier.

Figure 4.4 Experimental Flow

41

PM touch down power data (Figure 4.5)

Figure 4.5 Phase modulation

Figure 4.6 Touch down sensor

42

G1 touch down power 83.67mW and standard deviation 3.77. Sample size is 119.





TDS touch down power data (Figure 4.6)






Figure 4.7 and 4.8 are touch down power data between TDS and PM of before PSA

adjustment and after PSA adjustment respectively.

R-square of TDS and PM of before PSA adjustment is 39% while after PSA

adjustment show R-square 98% which has significantly improved.

Anyway, the range of touch down power data is within normal range, 60 to 120 mW. PSA

maximum adjustment change touchdown power but can not push them to be the outlier of

touchdown power.

43

Since PSA adjustment can not push touch down power to outlier, the next step to confirm

the difference between TDS and PM is to collect outlier DET HGA’s data from production

mode PM module and TDS test module open in the same test period.

Failure or outlier electrical data from PM was considered, (Touch down power less

than 52 mW).

Figure 4.7 Touch down detection of sample before adjust PSA

Figure 4.8 Touch down detection of sample after adjust PSA

44

Dynamic electrical test data between two types of detection was considered. The DET

parameters examined are Signal to Noise Ratio (SNR) and Reverse Overwrite (RevOW).

Touch down power of TDS data mapped to the PM touchdown failure was also considered.

Figure 4.9 Phase modulation’s touch down power distribution

45

Touch down power data between PM and TDS is shown in Figure 4.10. The PM’s touch

down failures pass under the TDS and are also within the normal distribution compared to

production data, 95 mW.

Figure 4.10 indicated that all PM’s touch down’s failure can be recover with TDS

‘s techniques.

The next things to consider are the critical DET parameters for TDS

implementation. Focus parameters are Signal to Noise Ratio or SNR and Reverse

Overwrite or RevOW. SNR is important parameter at spin-stand level because it show

good correlation to Bit Error Rate in Drive level testing. Reverse Overwrite or write ability

is also important at both spin-stand level and drive level.

Figure 4.10 Touch down power mapping between TDS and PM

46

Signal to Noise Ratio (SNR) is the important Dynamic electrical test

parameter. Figure 4.14 is the relationship of touch down power to SNR. The

procedure used in this plot is called “Binning”. SNR is related to write width, SNR

is higher when write width is wide. Binning sorts Write width (Target + 0.1 micro

inch) and plot SNR by touch down power.

The reason why we use Binning technique is because SNR and Reverse Overwrite are

driven by MWW (magnetic Write Width). Wider MWW induce higher SNR and higher

Reverse Overwrite. It can be a bias comparison if we use different MWW data for both

techniques.

Figure 4.11 and Figure 4.12 use data which show relationship between Magnetic Write

Width and SNR and Reverse Overwrite.

Reverse Overwrite has a nearly linear trend with MWW while SNR shows a roll off in the

narrow MWW range. When we would like to compare DET performance, Binning

technique is often use as a tool to analysis.

Figure 4.11 OverWrite vs Magnetic Write Width

Figure 4.12 SNR vs Magnetic Write Width

47

Binning technique considers DET data with a small MWW range, normally referenced to

MWW product target + 0.1.

We consider 2 groups of samples (TDS vs. PM) from 2.7 micro inch target. This analysis

can eliminate MWW population contribution to SNR and Reverse Overwrite.

Figure4.14 confirms that SNR is related to touch down power, SNR is low when

touchdown power is low. Early touch down failure is “Too low Touch down power” which

means SNR should be low.

Both PM and TDS techniques show similar trend of SNR and Overwrite (RevOW)

to Touch down power.

Figure 4.13 Magnetic Write Width distribution

Figure 4.14 Signal to Noise Ratio and Touch down power

48

Reverse Overwrite (RevOW) is a critical DET parameter. The DET touchdown power

binning plot also confirms that too low touch down power head also has low RevOW.

There is RevOW a roll off profile from PM module while TDS trend show NO roll off.

RevOW from TDS is more sensitive to touchdown power than PM. It means that TDS

gives better RevOW sensitivity to touchdown power than PM. (especially high touch down

range from 100mW onward).

Figure 4.15 Overwrite (RevOW) and Touch down power

Figure 4.16 Touch down binning to average of SNR and RevOW table

49

Touch down power binning of SNR sigma also indicate that low Touch down power will

give high SNR sigma (both PM and TDS). Figure 4.17.

The optimum touch down power to SNR sigma is 95 mW The SNR sigma profile

of TDS is better than the profile obtained with the PM, We can see SNR sigma from PM

increases from 100mW onwards while the SNR sigma from the TDS looks saturated.

Touch down power binning also relate to Reverse Overwrite (RevOW) sigma. High

touchdown power gives high RevOW sigma. TDS technique show better RevOW sigma

than PM. We can see RevOW sigma profile to touch down power in figure 4.18. Low

touch down range and high touch down range, TDS RevOW sigma profile is more flat than

PM. (Figure 4.18

Figure 4.17 SNR Sigma and touch down power

50

Based on the experiment result and the analysis of SNR and RevOW binning to touch

down power indicated that TDS performance is better than PM performance.

To confirm TDS and PM performance high volume production data is needed.

Figure 4.20 is the high volume data of PM technique and Figure 4.21 is the TDS technique.

The distribution plot and CDF plot of TDS technique shows better touch down sigma than

PM technique.

PM distribution shows a tail end of early touch down. Early touch down of PM is 2.70%

while TDS is 0.01%. Touch down power’s sigma of “TDS” performs better than PM:

PM ‘s touch down sigma is 12.64 while TDS sigma is 7.88.

Figure 4.18 Reverse Overwrite Sigma and touch down power

Figure 4.19 Touchdown binning to SNR and RevOW sigma table

51

Touch down power distribution of Phase modulation

Touch down power distribution of Touch down sensor

Figure 4.21 High-volume Touch down data Touch down sensor

Figure 4.20 High-volume Touch down data Phase modulation

52

In production data, “TDS” technique has significantly better detection than “PM”.

The number of data points is about 82K.

The average of touch down power is comparable but the Sigma of TDS is better than PM

sigma. TDS’s touch down sigma is 7.88 while PM sigma is 12.64.

Early touch down failure rate of PM is 2.70% while TDS is 0.01%, significantly improved.

Figure 4.22; % early touch down of Phase modulation and Touch down sensor in high volume mode

53

Figure 4.24 Touch down power average of Phase modulation and Touch down sensor in high volume mode

54

CHAPTER 5

CONCLUSION AND RECOMMENDATIONS

5.1 Conclusion and Recommendations

As the result of TDS and PM technique as illustrated in Chapter4, it is clear that

TDS technique performs better electrical performance (DET) than PM. Touch down

power, Signal to Noise Ratio (SNR), Reverse Overwrite (RevOW) and Yield are critical

DET parameters to be considered.

a) Touch down power: TDS has significantly better touch down sigma and

much lower % Early touch down than PM, Figure 5.1 and Figure 5.2

b) Signal To Noise Ratio (SNR): base on touch down relationship to SNR as

shown in Chapter4, Figures 4.11 and 4.14 indicated that TDS gives better SNR than

PM, both average and sigma. Nominal touch down power of both techniques is

comparable but TDS is better for wide touch down power range.

c) Reverse Overwrite (RevOW): base on touch down relationship to RevOW

shown in Chapter4, Figures 4.12 and 4.15 indicated that TDS gives better RevOW

than PM in the high touch down range (higher than 100 mW). The RevOW sigma

of TDS is smaller than the PM, especially in the high touch down range.

d) Yield, “TDS” has lower touch down power’s failure (ETD) than the PM.

2.70% for PM and 0.01% for TDS.

Implementation of TDS is practical for production ramp because component yield

is better and electrical performance is also better than the PM. Cost wise there is no-impact

because new hardware is not required and Test time is not impact.

Touch down power also better for the customer (Drive level) because of the use of

TDS.

Real time feedback is available in various operating environment (humidity, pressure) and

also can improve Drive reliability by detect DFH over drive.

Figure 5.1 % Early Touch down Figure 5.2 Touch down power sigma

55

REFERENCES

Hao, Zheng. Hui, Li. Kensuke, Amemiya. Frank E, Talke (2011). The effect of write

current on thermal flying height control sliders with dual heater/insulator element.

Retrieved from http://link.springer.com

Jia-Yang, Juang. Joel, Forrest. Fu-Ying, Huang (2011). Magnetic Head Protrusion Profiles

and Wear Pattern of Thermal Flying-Height Control Sliders With Difference Heater

Designs. IEEE Transactions on Magnetics Vol.47.NO 10, OCTOBER 2011. Retrieved

from http://ieeexplore.ieee.org

Zhi-Min, Yuan. Bo, Liu. Chun Lian, Ong. Budi, Santoso. (2011). A new Code Pattern of

Tripple Harmonic Method Fore Precise In-Situ FH Measurement in Perpendicular

Recording. IEEE Transactions on Magnetics Vol.47.NO 10, OCTOBER 2011. Retrieved

from http://ieeexplore.ieee.org

Tani, H. Kanda, M. Kubota, M. Tagawa, N. (2009). Study of slider-defect interation at

ultralow flying height by DFH control. JOURNAL OF APPLIED PHYSICS 105, 07B0703

(2009)

Uwe, Boettcher. Hui, Li. Raymond A de, Callafon. Frank E, Talke. (2011). DFH

Adjustment in Hard Disk Drives Through Feedforwad Control. IEEE Transactions on

Magnetics Vol.47.NO 7, FEBUARY 2007. Retrieved from http://ieeexplore.ieee.org

Tawshing, Tang. Soo-Youl, Hong. Na-Young, Kim. Xiaodong, Che. (2007). Overview of

Fly Height Control Applications in Perpendicular Magnetic Recording. IEEE Transactions

on Magnetics Vol.43.NO 2, FEBUARY 2007. Retrieved from http://ieeexplore.ieee.org

Bernhard, Knigge. Tushar, Suthar. Peter, Baumgart. (2006) Friction, Heat, and Slider

Dynamics During Thermal Protrusion Touchdown. IEEE Transactions on Magnetics

Vol.42.NO 10, OCTOBER 2006. Retrieved from http://ieeexplore.ieee.org

Bruno, Marchon. Terry, Olson. (2009). Hitachi GST, San Jose Research Center. Magnetic

Spacing Trends: From LMR to PMR and Beyond. Intermag-2009 (Sacramento). CC-01.

Bargmann, B. Lecture of Magnetic Spacing/ Head Disk Separation. AIT. (2008)

Thesis Touch Down Detection at Spin-Stand Level Raksak111574

Documents