AN EVALUATION SYSTEM FOR MECHANICAL AND ELECTRICAL
CHARACTERIZATION OF MEMS DEVICES
Except where reference is made to the work of others, the work described in this thesis is my own
or was done in collaboration with my advisory committee. This thesis does not include proprietary
or classified information.
Seong Jin Kim
Certificate of Approval:
George T. Flowers
Professor
Mechanical Engineering
Robert N. Dean, Chair
Assistant Professor
Electrical and Computer Engineering
Bogdan M. Wilamowski
Professor
Electrical and Computer Engineering
Thaddeus Roppel
Associate Professor
Electrical and Computer Engineering
George T. Flowers
Dean
Graduate School
AN EVALUATION SYSTEM FOR MECHANICAL AND ELECTRICAL
CHARACTERIZATION OF MEMS DEVICES
Seong Jin Kim
A Thesis
Submitted to
the Graduate Faculty of
Auburn University
in Partial Fulfillment of the
Requirements for the
Degree of
Master of Science
Auburn, Alabama
May 9, 2009
AN EVALUATION SYSTEM FOR MECHANICAL AND ELECTRICAL
CHARACTERIZATION OF MEMS DEVICES
Seong Jin Kim
Permission is granted to Auburn University to make copies of this thesis at its
discretion, upon the request of individuals or institutions and at
their expense. The author reserves all publication rights.
Signature of Author
Date of Graduation
iii
VITA
Seong Jin Kim graduated from Joongdong High School in Feburary of 1987 and Yonsei Uni-
versity in August of 1992 with a Bachelor of Science of Electronic Engineering. After graduation,
he worked for Halla Engineering and Heavy Industries, Ltd. as an aerospace electrical engineer for
five years. He then continued his electrical design career in MilS Engineering Ltd. for six years,
and then worked for EO System Co., Ltd. After several projects, he worked on a MEMS uncooled
focal plane array (UFPA) detector at EO System Co., Ltd, where he became fascinated with MEMS
technology and then decided to pursue graduate work in MEMS fabrication, packaging and testing
at Auburn University.
Seong Jin Kim was born to Si-Hwan Kim and Byoung-Won Suh on the 16th of September,
1967 in Seoul, Korea. He married Eun Jung An on September 28, 1998. Their daughter, Jimin Kim,
was born on April 18, 2000 and their son, Hyoungmin Kim, was born on Feburary 17, 2004.
iv
THESIS ABSTRACT
AN EVALUATION SYSTEM FOR MECHANICAL AND ELECTRICAL
CHARACTERIZATION OF MEMS DEVICES
Seong Jin Kim
Master of Science, May 9, 2009
(B.S. Electronic Engineering, Yonsei University, Seoul, South Korea, 1992)
95 Typed Pages
Directed by Robert N. Dean
The evaluation of MEMS devices during the developmental phase prior to packaging is chal-
lenging. However, evaluating a MEMS device in die form can be useful, as its electrical and me-
chanical test results can provide useful feedback to improve the design, modeling and fabrication
process. It can be worthwhile to evaluate the performance of MEMS devices in regard to various
parameters, such as in various gas chemistries and at different pressures, and with and without me-
chanical and/or electrical excitation. From these tests, the design and manufacturing processes can
be adjusted before packaged devices are available for in-depth testing.
To accomplish this goal, a vacuum chamber based MEMS evaluation system has been devel-
oped for electrically and mechanically evaluating MEMS die. The system consists of a bell jar
vacuum system with a pressure range from 0.050 Torr to ambient pressure. Additionally, the vac-
uum system has been constructed to allow other gases to be injected into the vacuum chamber for
device testing. A machined plastic fixture is used to mount the bare die during testing. It attaches to
a small electromechanical shaker that resides inside the vacuum chamber for dynamic mechanical
testing of the MEMS device. Two laser interferometers are used through the glass bell jar to then
v
measure the relative motion of two locations on the die, thus allowing the transmissibility and the
range of a microstructure?s motion to be measured. The evaluation system also has nine electrical
feedthroughs for external connection to the MEMS die or to other test equipment located inside the
chamber. Additionally, a graphite heating stage has been integrated into the chamber for thermal
testing.
To demonstrate the usefulness of this evaluation system, electrical and mechanical tests have
been performed on several MEMS devices and systems to measure the mechanical frequency re-
sponse, the mechanical quality factor in various gases at different pressures, and the closed loop
operation with integrated electronics. This system is directly compatible with the microfabrication
clean room. Thus, it allows early testing of new devices before packaging has been completed.
vi
ACKNOWLEDGMENTS
My foremost thanks goes to my advisor, Assistant Professor Robert Dean, who has guided and
encouraged me academically and financially. I could not have imagined having a better advisor for
my study. Without his knowledge, wisdom, and having given me the opportunity to work with him,
none of this work would have been possible.
I also would like to thank Professor Flowers for his help and support for my research. His
constructive feedback and support at critical stages have been immensely helpful in completing
this endeavor. I sincerely thank Professor Wilamowski and Professor Roppel for serving on my
committee. Their insight into my research of the analysis and testing of MEMS devices has been
invaluable.
Additional thanks goes to Mr. Palmer for his assistance with the construction of the MEMS
evaluation system.
One of the most precious persons who has been with me in every moment is my wife Eun
Jung An. I would like to thank her not only for the many sacrifices she has made to support and
understand me in undertaking my study, but also for guiding me along the right path all my life. I
would like to thank my children Jimin and Hyoungmin for their ceaseless love and smiles which
have strengthened me. I would also like to give big thanks to my parents, Si-Hwan Kim and Byoung-
Won Suh, for praying for and always being there for me.
Finally, and most vitally, I would like to thank so much the almighty God, not only for giving
me a faith, a wisdom and health to complete a graduate study, but also for guiding me along the
right path all my life.
vii
Style manual or journal used Journal of Approximation Theory (together with the style known
as ?aums?). Bibliograpy follows van Leunen?s A Handbook for Scholars.
Computer software used The document preparation package TEX (specifically LATEX) together
with the departmental style-file aums.sty.
viii
TABLE OF CONTENTS
LIST OF FIGURES xi
LIST OF TABLES xiv
1 INTRODUCTION 1
2 LITERATURE RIVIEW 4
2.1 MEMS Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Analysis Tools for MEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.1 Finite Element Analysis (FEA) . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2.2 System Level Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2.3 Contamination Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Measurement Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3.1 Electrical Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.2 Optical Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3 TESTING OF MEMS DEVICES 12
3.1 System Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1.1 Electrical Stimulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.1.2 Non-electrical Stimulus . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1.3 Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.4 Fixturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2 Commercial MEMS Evaluation Systems . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.1 Open System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.2 Closed System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4 MEMS EVALUATION SYSTEM 24
4.1 System Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.1.1 Stimulus Requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.1.2 Environmental Requirement . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.1.3 Electrical Requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.1.4 Measurement Requirement . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.3 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.4 Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.4.1 Bell-jar Vacuum Chamber . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.2 Small Shaker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
ix
4.4.3 Optical Measurement System . . . . . . . . . . . . . . . . . . . . . . . . 34
4.4.4 Dynamic Signal Analyzer . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5 APPLICATION OF THE MEMS EVALUATION SYSTEM 36
5.1 The MEMS Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.2 Fabrication of the MEMS Device . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.3 Background for a Micromachined Parallel Plate Capacitor Structure [46] . . . . . . 40
5.4 Relative Velocity Sensor in Low Pressure Environment . . . . . . . . . . . . . . . 43
5.5 Closed-loop Damping Control At Low Pressure[45] . . . . . . . . . . . . . . . . . 46
5.6 Electrostatic Enhancement of MEMS Devices . . . . . . . . . . . . . . . . . . . . 51
5.7 Transmissibility vs. Pressure in Various Gases . . . . . . . . . . . . . . . . . . . . 56
6 CONCLUSIONS 64
7 FUTURE WORK 66
BIBLIOGRAPHY 67
APPENDICES 70
A 71
A.1 Micromachined Si MEMS Device Fabrication Procedure . . . . . . . . . . . . . . 71
A.2 How To Use the Evaluation System . . . . . . . . . . . . . . . . . . . . . . . . . 76
x
LIST OF FIGURES
2.1 Two Simulation Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 System Level Model of the ADXL180 MEMS Accelerometer [50] . . . . . . . . . 6
2.3 Photograph of Comb Defect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.4 Laser Doppler Velocimetry Setup [16] . . . . . . . . . . . . . . . . . . . . . . . . 9
2.5 Illustration of Interferometry [18] . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.6 Stroboscopic Interferometer System [19] . . . . . . . . . . . . . . . . . . . . . . . 10
2.7 Computer Microvision System [22] . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1 System Requirements for Testing Integrated Circuits . . . . . . . . . . . . . . . . 12
3.2 System Requirements for Testing MEMS Device . . . . . . . . . . . . . . . . . . 12
3.3 Closed Platform Test Equipment: SUSS PAP200 [24] . . . . . . . . . . . . . . . . 13
3.4 Open Platform Test Equipment: EM Optomechanical 622-A (courtesy Auburn Uni-
versity) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.5 EM Optomechanical 622-A: (a) MEMS device on the Vibration Platform, (b) GUI
Screen of MEMScriptTM [25] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.6 Electrostatic Parallel Plate Actuator . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.7 LDS V830 Electrodynamic Shaker [36] . . . . . . . . . . . . . . . . . . . . . . . 18
3.8 MEMSense High Speed Angular Rate Table (courtesy Auburn University) . . . . . 19
3.9 Laser Doppler Velocimeter: Polytec LSV 6000 [40] . . . . . . . . . . . . . . . . . 20
3.10 Interferometer: Polytec MSA-500 [41] . . . . . . . . . . . . . . . . . . . . . . . . 21
3.11 Dydnamic MEMS Measurement System: Wyko NT1100 [42] . . . . . . . . . . . . 22
xi
3.12 Pressure Sensor Tester: SUSS PAP200 [43] . . . . . . . . . . . . . . . . . . . . . 22
3.13 Micro System Analyzer with Vacuum Chamber: Polytec MSA-400 [44] . . . . . . 23
4.1 System Configuration Schematic . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.2 MEMS Evaluation System: Front View . . . . . . . . . . . . . . . . . . . . . . . 27
4.3 MEMS Evaluation System: Side View . . . . . . . . . . . . . . . . . . . . . . . . 27
4.4 MEMS Evaluation System: Inside View . . . . . . . . . . . . . . . . . . . . . . . 28
4.5 Photograph of the Mechanical Pump . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.6 Photograph of the Optical Measurement System (two laser units, two controllers,
and a signal analyzer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.7 Valves of the MEMS Evaluation System: (a) Inlet Line Valves, (b) Outlet Line Valve 31
4.8 Illustration of Gas Plumbing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.9 (a) SO2 Cylinder and Valves (b) N2 Gas Control Valves . . . . . . . . . . . . . . . 32
4.10 Photograph of the Mini-shaker Type 4810 . . . . . . . . . . . . . . . . . . . . . . 34
4.11 Optical Measurement System: (a) Vibrometer Controller, (b) Laser Unit . . . . . . 34
4.12 Photograph of the HP 35665 A Dynamic Signal Analyzer . . . . . . . . . . . . . . 35
5.1 Si MEMS Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.2 Parallel Electrode Structure[46] . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.3 Schematic Diagram of Capacitor Interface Circuit . . . . . . . . . . . . . . . . . . 44
5.4 Time Traces of the Measured Relative Electrode Displacement and the Sensor Out-
put at Ambient Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.5 Time Traces of the Measured Relative Electrode Displacement and the Sensor Out-
put at Low Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.6 Feedback Controller Circuit Board [45]: (a) Front side, (b) Back side . . . . . . . . 47
xii
5.7 MEMS Device with the Feedback Controller PCB: (a) Assembled MDUT with the
Feedback Controller Circuit PCB, (b) Illustration of the side view of an assembled
MDUT [45] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.8 The Schematic Diagram of the Feedback Controller Circuit[45] . . . . . . . . . . . 49
5.9 Mechanical Frequency Responses Due to Feedback Control Signal at Low Pressure 50
5.10 Bottom Electrode of the Test PCB . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.11 Photograph of GW Instek LCR-821 . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.12 Capacitance Change in Test1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.13 Capacitance Change in Test2 and Test3 . . . . . . . . . . . . . . . . . . . . . . . 53
5.14 Mechanical Frequency Response in Air . . . . . . . . . . . . . . . . . . . . . . . 57
5.15 Mechanical Frequency Response in N2 . . . . . . . . . . . . . . . . . . . . . . . . 58
5.16 Mechanical Frequency Response in SF6 . . . . . . . . . . . . . . . . . . . . . . . 58
5.17 Mechanical Frequency Response in Water Vapor . . . . . . . . . . . . . . . . . . 59
5.18 Mechanical Frequency Response in SO2 . . . . . . . . . . . . . . . . . . . . . . . 59
5.19 Mechanical Frequency Response in Various Gases and Pressures of Test1 . . . . . 60
5.20 Mechanical Frequency Response in Air . . . . . . . . . . . . . . . . . . . . . . . 60
5.21 Mechanical Frequency Response in Nitrogen Gas . . . . . . . . . . . . . . . . . . 61
5.22 Mechanical Frequency Response in Water Vapor . . . . . . . . . . . . . . . . . . 61
A.1 Cable Connection between Equipment . . . . . . . . . . . . . . . . . . . . . . . . 78
xiii
LIST OF TABLES
5.1 Viscosity, Density, and Permittivity for Air, N2, SO2, SF6, and Water Vapor [48-49] 52
5.2 Capacitance under Various Gases at Different Pressures . . . . . . . . . . . . . . . 55
5.3 Quality (Q) Factor under Air and N2 gas at Different Pressures . . . . . . . . . . . 62
5.4 Quality (Q) Factor under SO2, SF6, and Water Vapor at Different Pressures . . . . . 63
xiv
CHAPTER 1
INTRODUCTION
In contrast to the very standardized design process for VLSI, MEMS device design requires
the consideration of not only electrical and mechanical functionality, but also the complex fabrica-
tion process and its influence on the design, if the MEMS device is fabricated using conventional
IC fabrication techniques. Additionally, a time consuming iterative process covering the entire de-
vice realization process, from design through packaging and test, may be required for successful
device development. In spite of an iterative design process, such as analysis and simulation with a
Computer-Aided Design (CAD) tool [1], testing the packaged MEMS device without fully evaluat-
ing the device at die level may result in the discovery of unexpected device characteristics, which
could result in a whole new iteration.
Various on-wafer testing systems aid in finding an optimal process with minimized production
costs through identifying possible errors and by maximizing yield. Similarly, a general purpose
MEMS evaluation system, that can be used prior to device packaging, can be useful in minimizing
the iterative MEMS development process. Once the device is fabricated according to the initial
design, it can then be tested and characterized by the MEMS evaluation system.
The MEMS evaluation system in this research endeavor was constructed with off-the-shelf
components, consisting of a vacuum bell-jar chamber with substrate quartz lamp heaters and a
heater controller, a small electromechanical shaker, two laser displacement interferometers and their
controllers, a pressure transducer, a digital pressure meter, and a signal analyzer, a vacuum pumping
system, and various vacuum system fittings and plumbing. The vacuum system supported a pressure
inside the bell-jar between 0.05 Torr and ambient pressure. The internal graphite heater could
1
heat the attached device and its surroundings to 200 C. The chamber possessed a gas inlet for
the supply of various gases. The glass bell-jar could be used with two laser interferometers to
measure useful mechanical properties of the MEMS device, such as the range of motion, the natural
frequency, the transmissibility and the quality factor. The small shaker provided the MEMS device
with a mechanical stimulus in a frequency range of 400Hz to 3.6 kHz. Additionally, the evaluation
system had nine electrical feedthroughs that provided the MEMS device and the small shaker with
electrical connections. The pressure transducer and digital readout meter had a resolution of 0.001
Torr. The dynamic signal analyzer had two input channels used to receive the two interferometer
outputs, and one source output to control the small shaker. The evaluation system utilized the
Microfabrication Laboratory mechanical vacuum pump, which was connected to the outlet port of
the vacuum chamber using flexible plastic pipe.
A literature review of relevant background information is presented in Chapter 2. This includes
MEMS features, analysis tools in design, and measurement methods.
System requirements for evaluating MEMS devices, consisting of stimulus and measurement,
are presented in Chapter 3. A discussion on currently available commercial MEMS evaluation
systems is also presented.
A detailed discussion on the MEMS evaluation system developed through this endeavor is
presented in Chapter 4. It describes mechanical, environmental, electrical, and optoelectrical re-
quirements of the system for various MEMS device evaluation procedures, design philosophy, and
how the system is constructed. Lastly, the capability of the system is presented.
The results from the evaluation of various MEMS devices and systems, performed with this
evaluation system, is presented in Chapter 5. This includes closed-loop damping control at low
2
pressure, relative velocity sensor performance at low pressure, capacitance vs. pressure and trans-
missibility vs. pressure in various gas environments.
3
CHAPTER 2
LITERATURE RIVIEW
2.1 MEMS Features
Microfabricated mechanical elements, structures, electrodes, traces, capacitors, piezoresistors
and sometimes even semiconductor devices, are integrated to realize complex Microelectromechan-
ical Systems (MEMS) devices such as sensors [2], actuators [3], RF switches [4] and energy con-
version devices [5]. Such diverse functionality in MEMS devices has required a new approach to
modeling and testing to ensure proper functioning of the system. MEMS devices are often fab-
ricated with processes borrowed for integrated circuit (IC) manufacturing techniques. The two
primary manufacturing techniques for Si based MEMS devices are Bulk micromachining [6] and
Surface micromachining [7]. These micro-fabrication processes have some limitations, which can
result in defects such as broken structures, under/over etching and stiction. Moreover, due to the
miniature size of the MEMS devices, mechanical characteristics are sensitive to variation in material
properties and to microfabrication limitations and tolerances. Therefore, evaluation and recognition
of mechanical behaviors as well as electrical performance are important in MEMS device develop-
ment.
2.2 Analysis Tools for MEMS
MEMS analysis tools allow for predicting the MEMS device?s behavior and optimizing the
various design parameters prior to device fabrication. While the analysis tools for electronics are
well established, the tools for MEMS devices are not yet well established due to the relative youth
of MEMS technology and the complexity involved.
4
The design flow in the MEMS CAD design environment includes iterative mechanical schematic
generation and MEMS simulation in the early design stage. Most analyses and simulations possess
two levels, as shown in Figure 2.1: the physics level for the detailed behavior in 3-D and the system
level for the behavior of the entire system. Tools for the physics level typically use finite element
(FE), finite volume (FV), or boundary element techniques.[8-9]
Figure 2.1: Two Simulation Levels
2.2.1 Finite Element Analysis (FEA)
FEA divides the device into very tiny elements through a meshing process, which is the first
activity at the physics level. This enables the analysis of complex geometries of MEMS devices in
points of mechanical and thermal interest. Examples of commonly used FEA tools include Coven-
torWare [10], IntelliSuite [11], and MEMS Pro/MEMS Xplorer [12].
2.2.2 System Level Analysis
In the system level analysis, the performance of a MEMS device, which was modeled with a
lumped element model, can be evaluated using tools such as Simulink or SPICE, where the MEMS
device is described by ordinary differential equations and nonlinear functions. An accelerometer,
5
for example, can be modeled using a proof mass, a spring, and a damper. An example of the system
level analysis for a commercial accelerometer, the Analog Devices ADXL180, is shown in Figure
2.2.
Figure 2.2: System Level Model of the ADXL180 MEMS Accelerometer [50]
 Simulink: Simulink enables the implementation of a nodal model [13] for investigating the
dynamics of a MEMS device. Simulated result data can easily be post processed using Matlab [14].
 SPICE: SPICE is one of the ordinary differential equation (ODE) solvers that is used at the
system level. Although initially developed for simulating electrical circuits, it is a powerful tool that
can be used to model any linear system. A PC version, PSPICE [15], is commonly used.
2.2.3 Contamination Analysis
Contamination analysis is a process to analyze and predict feasible damages of physical ge-
ometries of a microstructure occuring during a fabrication process, which affect the structure or
material properties. The contamination leads to faulty MEMS behavior. Therefore, predicting the
effect of the contamination in the design stage is important to improve the quality of MEMS prod-
ucts. On another hand, contamination analysis [16] using software simulation tools can also be a
useful MEMS evaluation method. As the first step, Contamination-Defect-Fault (CODEF) simula-
tion tools perform simulations using microfabrication process recipe data, contamination properties,
6
and microstructure layout, which generates fault macro models of the various defective microstruc-
tures. For example, a CODEF simulation for a microsensor might reveal defects such as shuttle
defect, comb defect, and flexure defect. The photo of a comb defect is shown in the Figure 2.3.
Based on the locations and feasible defects from CODEF, a mechanical simulation with ABAQUS
can be performed to acquire some defects which can cause complete failure in the device. So, sim-
ulating the MEMS device in the design phase can be considered to be an efficient means for MEMS
device development.
Figure 2.3: Photograph of Comb Defect
2.3 Measurement Methods
Many useful test, evaluation, or monitoring methods have been proposed to verify that the
MEMS device mechanical characteristics and motion match the design requirements. There are
two basic categories of testing methods for measuring MEMS mechanical characteristics [16]. One
is a contact method using a microprobe or mass loading to measure a deflection, which is a kind
of static test for measuring mechanical properties of a MEMS structure or material. The other is a
non-contact method which includes electrical or optical technology. Examples include microtension
test, microbeam bend test, axisymmetric plate bend test, resonant frequency test, and M-test to
7
measure Young?s modulus or the fracture strength [17]. Those methods need excitation of the
MEMS structure to measure the changes either electrically or optically.
2.3.1 Electrical Method
This method makes use of a piezoresistive, a piezoelectric or a capacitive property.
 Piezoresistivity: This property is a material effect where the material experiences a change
in resistance due to an applied strain. Semiconductor materials experience the piezoresistive effect
while metal films and foils experience a change in resistance in response to an applied strain due
to the geometric effect. Therefore, a metal foil strain gauge uses the geometric effect and not the
piezoresistive effect.
 Piezoelectricity: Certain classes of crystals produce an electric charge when they experience
a strain, and they also deform in response to an applied electric field. Quartz, zinc oxide, and lithium
niobate are examples of piezoelectric materials.
 Capacitive Technique: For example, one fixed electrode and one or more movable electrodes,
or two fixed electrodes with a movable dielectric material, experience a change in capacitance in
response to a change in the electrode configuration or the dielectric material.
2.3.2 Optical Method
Optical sensing methods primarily utilize the modulating properties of an optical frequency
electromagnetic wave. The non-contact sensing nature of these methods is suitable for measuring
the MEMS mechanical characteristics without a mass loading effect.
 Laser Doppler Velocimetry: The laser light reflected from the moving MEMS surface will
have a shift in frequency by an amount proportional to the velocity of the surface compared to that
8
of the radiated laser light source. The optical setup of laser Doppler velocimetry is presented in
Figure 2.4.
Figure 2.4: Laser Doppler Velocimetry Setup [16]
 Interferometry: The principle of interferometry is that two coherent optical beams will form
a interference or fringing pattern that is dependent on the phase difference between the two beams.
Typically, a single incident source beam of light is split into two identical beams, using a beam
splitter and mirror, and travel along two paths of either continuous or stroboscopic illumination.
The illustration of Figure 2.5 presents the principles of interferometry. The He-Ne laser beam
is split by a first beam splitter (BS1) into a reference beam and a measurement beam. After passing
through a second beam splitter (BS2), the measurement beam is focused onto the object under
investigation and then reflected from the object. This reflected beam is deflected by the second
beam splitter (BS2), then merged with the reference beam by a third beam splitter (BS3). Finally,
they reach the detector. As the length of the reference beam is constant over time, a movement of the
object under investigation generates a fringe pattern on the detector. One complete dark-bright cycle
on the detector corresponds to an object displacement of exactly half of the wavelength of the laser
9
Figure 2.5: Illustration of Interferometry [18]
used. As an example, an illustration of a stroboscopic interferometer system is presented in Figure
2.6. Electronic Speckle Pattern Interferometry (ESPI) is also used for measuring the vibration of
Figure 2.6: Stroboscopic Interferometer System [19]
microstructures possessing a rough surface [20].
 Optical Microscope: This method is a non-interferometry method. The Computer Microvi-
sion System (CMS) developed by MIT is composed of a light microscope with a movable stage, a
10
CCD camera, a signal generator, and a PC [21]. A MEMS device placed on the stage is exposed
to a periodic LED light source and imaged by the microscope with a CCD camera. The LED is
strobed once per stimulus period at a chosen phase to produce a snapshot of the device position.
The repeated snapshots at several stimulus phases and at several locations are processed and result
in the capturing of the motion of the device through one period. An illustration of the Computer
Microvision System is presented in the Figure 2.7.
Figure 2.7: Computer Microvision System [22]
 High Speed Cine Photomicrography [23]: This method was developed for capturing the
image of fast moving structures due to optical magnification by combining a ultra high speed camera
and a microscope. This method captures the position of moving parts in microdevices with respect
to time. To analyze extremely fast movements of microdevices , a pulsed light source with high
constant intensity is needed. Therefore, the visualization of the movement of the microdevices
is performed using the stroboscopic principle. During the measurement, picture taking is repeated
several times and then visualized at subsequent points of time, which results in a sequence of images.
11
CHAPTER 3
TESTING OF MEMS DEVICES
3.1 System Requirements
In conventional IC testing, wafer level testing is typically performed using precisely controlled
wafer probes. Figure 3.1 presents an illustration of the the input, output, and environments for
Figure 3.1: System Requirements for Testing Integrated Circuits
conventional IC testing. In MEMS devices, excitation as an input is used under a wider range of
environmental conditions. The output of the MEMS devices can be an optical signal or an electrical
signal, as presented in Figure 3.2. The requirements to accomplish the various types of MEMS
Figure 3.2: System Requirements for Testing MEMS Device
12
testing can be very sophisticated. The MEMS evaluation system needs to be equipped with the
capabilities for analyzing the mechanical and electrical characteristics of the device. These testing
capabilities heavily depend on the type of MEMS Device Under Test (MDUT). The system must
provide electrical and mechanical stimuli to the MDUT and then detect the corresponding response
of the device. Electrical and mechanical stimuli are usually device specific. Therefore the stimuli
must be tunable with adjustable parameters, such as amplitude, frequency, bandwidth, duration, etc.
Additionally, the evaluation system should have the capability of evaluating environmental effects,
such as ambient gas pressure and temperature.
In order to evaluate some MEMS devices such as accelerometers, gyroscopes, pressure sensors,
and infrared (IR) detectors, the evaluation system is designed for operation in a low pressure or a
specific gas environment. This type of evaluation system can be called a ?closed platform?. A pho-
tograph of the SUSS PAP200, manufactured by SUSS MicroTech, is presented in Figure 3.3. This
closed platform evaluation system is used not only for testing pressure sensors from 100mbar up to
Figure 3.3: Closed Platform Test Equipment: SUSS PAP200 [24]
50bar absolute pressure, but also for probing in a controlled gas atmosphere and/or with controlled
13
humidity. The chamber has a top-side view-port for the microscope as well as four additional view-
ports. On the rear side, 192 high pressure electrical feedthroughs are located for the chuck stage
for probe-heads, probe-card, and chamber temperature measurement. Five fluidic feedthroughs can
be used with several connectors to the pressure supply system to supply atmospheric pressure. The
pressure supply system of the evaluation system consists of a compressor, a booster, a vacuum
pump, and a controller. The compressor generates a pressure of up to 30 bar, and the booster ex-
pands the compressor?s working pressure to 80 bar. The microscope mounted on the top window
provides features such as recording images and pattern recognition.
An ?open platform? evaluation system can evaluate differential and absolute pressure sensors,
microphones, and micromirrors in an ambient environment. A photograph of the Model 622-A,
manufactured by EM Optomechanical, is presented in Figure 3.4 as an example of the open plat-
Figure 3.4: Open Platform Test Equipment: EM Optomechanical 622-A (courtesy Auburn Univer-
sity)
14
form. This 3D MEMS profiler is composed of a vibration isolation platform, two left-hand vacuum
base probe micropositioners, two right-hand vacuum base probe micropositioners, a non-coherent
green LED as a standard light source, a 2/3-inch black & white CCD camera, and an image pro-
cessing system.
The image processing system consists of a National Instruments analog output board, a Data
Acquisition (DAQ) PCI board, and device drivers. An application software package, MEMScriptTM,
provides analysis features and the ability to control MEMS devices and to make real time mea-
surements of performance, such as pattern matching, phase shifting interferometry, decision mak-
ing capability, stroboscopic imaging, and arbitrary waveform output. The photographs in Figure
3.5 presents a picture of MEMS device on the vibration platform and application software of
MEMScriptTM.
Figure 3.5: EM Optomechanical 622-A: (a) MEMS device on the Vibration Platform, (b) GUI
Screen of MEMScriptTM [25]
3.1.1 Electrical Stimulus
There are several excitation method for MEMS devices, such as, electrostatic stimulus, acoustic
stimulus, electrothermal stimulis, etc. [20, 26]. For a MEMS device that responds to an electrical
stimulus, the stimuli is an applied voltage or current. If a Built-In Self-Test (BIST) function is
15
implemented in the electronics of the MEMS device, the required physical stimulus can be applied
to the device by applying an electrical signal to the BIST input pin, without externally applying
an excitation force [27-28]. As an example of a MEMS device with a BIST function, the Analog
Devices? ADXRS300 angular rate sensor includes a BIST feature that allows two external signals
to be applied to the device to cause the internal microstructure to experience the same response as it
would to angular rates of +55 /s and 55 /s, respectively. This BIST scheme is designed so that a
microcontoller supplies the self-test signals to two BIST pins and then checks the rate from the rate
pin. However, MEMS devices at die level usually need external excitation to be tested.
 Electrostatic stimulus: The electrostatic stimulus uses the principle that an attractive force
will be produced between two parallel electrodes (or plates) when a voltage is applied between
them. An illustration of an electrostatic parallel plate actuator (PPA) is presented in Figure 3.6. The
energy stored in PPA without fringing effect, which is expressed as [29]:
W = CV
2
2 =
eoerAV2
2L ; (3.1)
may be differentiated with respect to L in order to obtain the force between the two plates. This
results in the following nonlinear relationship [29] between force and voltage:
F = dWdL = eoerAV
2
2L2 : (3.2)
However, this method often requires high voltages in order to achieve sufficiently high electrostatic
forces, which often results in utilizing narrow gap microstructures fabricated using surface micro-
machining process. Comb-drive actuators, gap closing actuators and scratch drive actuators are
other types of electrostatic actuators.
16
Figure 3.6: Electrostatic Parallel Plate Actuator
 Electrothermal Stimulus: The principle of the electrothermal stimulus is that two materials
possessing different Coefficients of Thermal Expansion (CTE) are bonded together to realize a
composite structure that bends in response to a temperature change above or below the temperature
at which materials were bonded together. An example device is a thermal vertical bimorph actuator
[30-31]. Joule heating can be used as the electrothermal stimulus by passing an electrical current
through an resistive heater integrated into the MEMS device [26].
 Piezoelectric Stimulus: An applied voltage across the electrodes of a piezoelectric material,
whose crystal structure is asymmetric, produces a deflection. The most commonly used piezoelec-
tric material is lead zirconate titanate (PZT) [32].
 Electromagnetic stimulus [33]: An electromagnetic force (Lorentz force) occurs in a direc-
tion perpendicular to the current and a magnetic field when a current flows within a conductor in
the magnetic field. The externally applied magnetic field may be produced by a permanent magnet
of by an electromagnet, which is brought into close proximity to the MEMS device.
3.1.2 Non-electrical Stimulus
 Vibration: The vibration stimulus provides sinusoidal or wide bandwidth vibrational motion
or velocity to the MEMS device. An electromechanical shaker is often used to produce a vibra-
tion stimulus. Accelerometers [34] and velocity sensors [35] are examples of MEMS devices that
17
may be evaluated using a vibration stimulus. A photograph of the V830 Electrodynamic Shaker
manufactured by LDS is presented in Figure 3.7.
Figure 3.7: LDS V830 Electrodynamic Shaker [36]
 Rotation: Some MEMS devices, such as MEMS gyroscopes, directly respond to angular rate,
where the applied angular rate may be constant or time varying. Additionally, the centrifugal force,
resulting from an object, which is located a fixed distance from the center of rotation, can be used
as an acceleration stimulus for a MEMS accelerometer. A photograph of a high speed angular rate
table is presented in Figure 3.8.
 Pressure: Applied ambient gas pressure can be a stimulus to MEMS devices designed to
respond to pressure changes, such as MEMS pressure sensors [37]. It can also be a stimulus to other
MEMS devices that are affected by ambient gas pressure, such as resonating structures [38], where
the pressure stimulus here would change very slowly over time.
 Acoustic: Sound is actually high frequency pressure waves that propagate in the ambient
fluid that surrounds the MDUT. Although MEMS microphones are designed to respond to sound,
18
Figure 3.8: MEMSense High Speed Angular Rate Table (courtesy Auburn University)
some MEMS devices, such a gyros, may adversely respond to sound [39]. An acoustic chamber
may be used in evaluating MEMS devices with this stimulus.
 Electromagnetic Radiation: This stimulus could include RF energy, IR, visible light, UV, and
X-rays. Different types of evaluation equipment are required for each of these types of radiation.
Many MEMS devices could benefit from evaluation with radiation stimuli, including RF devices,
thermal sensors, optical devices, and radiation sensors.
3.1.3 Measurement
Sensing the mechanical movement of a MEMS device requires the appropriate measurement
methods. As discussed in Chapter 2.3, electrical measurement method includes piezoresistive,
piezoelectric and capacitive techniques. These measurement methods can be implemented in a
MEMS device to convert mechanical movement into an electrical signal. When using the optical
measurement methods, a light beam interacts with one or more movable microstructures on the
19
MEMS device. As a result, microstructure motion can then be measured by detecting changes in in-
tensity, phase, wavelength, frequency, and/or spatial position. The optical measurement method has
a merit of being a non-contact technique, which usually does not affect the microstructure motion
during device testing.
3.1.4 Fixturing
Some type of fixture is usually required to hold or mount the MEMS device during evalua-
tion. This can be particularly challenging for bare die MEMS devices, as they can be very fragile.
Fixturing is very much application specific.
3.2 Commercial MEMS Evaluation Systems
3.2.1 Open System
 Laser Doppler Velocimetry: The Polytec?s LSV 6000, a laser surface velocimeter, is an
example commercial system that uses the laser Doppler velocimetry technique. This system consists
of a sensor head and a controller. The sensor head is a light source using a visible diode laser, and
the controller is a signal processor that evaluates the Doppler frequency change of the laser signal
by means of FFT analysis. A photograph of a LSV 6000 is presented in Figure 3.9.
Figure 3.9: Laser Doppler Velocimeter: Polytec LSV 6000 [40]
20
 Interferometry:
. Micro System Analyzer: The Polytec MSA-500 uses a white light interferometry method
for high resolution topography measurement. This system can visualize and measure system in-
plane and out-of-plane resonances and transient responses of microstructures of MEMS devices. A
photograph of a MSA-500 analyzer is presented in Figure 3.10.
Figure 3.10: Interferometer: Polytec MSA-500 [41]
. Dynamic MEMS Measurement System: The Wyko NT1100 system with the DMEMS option
is a dynamic and static MEMS metrology system that uses phase-shifting and white light vertical
scanning interferometry. The capabilities for this system include a 3D visualization of moving
MEMS devices as well as the measurement of in-plane and out-of plane dimensions. A photograph
of a Wyko NT1100 manufactured by Veeco is presented in Figure 3.11.
3.2.2 Closed System
 Pressure Sensor Tester (SUSS PAP200): This system is a probe station primarily designed for
evaluating pressure sensors from 100 mbar up to 50 bar absolute pressure. Additionally, this system
21
Figure 3.11: Dydnamic MEMS Measurement System: Wyko NT1100 [42]
can simulate a controlled gas atmosphere with controlled humidity. As shown in Figure 3.12, the
pressure chamber has a top-side view-port for a microscope and four additional view-ports. On
the rear side, 192 multi-pole electrical feedthroughs are located for the chuck stage, measurement
cables, chamber temperature measurement, and operating a temperature controlled chuck. Five
fluidic feedthroughs for a relative pressure sensor are connected to the pressure supply system.
Figure 3.12: Pressure Sensor Tester: SUSS PAP200 [43]
 Micro System Analyzer (Polytec MSA-400) combined with a vacuum chamber: This system
is a combination of a vacuum chamber and a micro-scanning laser Doppler vibrometer for charac-
terizing MEMS devices over a wide range of ambient pressures. Through a window in the vacuum
22
chamber, the micro system analyzer measures the motion of a MEMS device excited by broadband
stimuli, including periodic chirp and white noise. A photograph of a MSA-400 is presented in
Figure 3.13.
Figure 3.13: Micro System Analyzer with Vacuum Chamber: Polytec MSA-400 [44]
23
CHAPTER 4
MEMS EVALUATION SYSTEM
4.1 System Requirements
The requirements for the MEMS evaluation system to test MEMS devices can be categorized
into four factors: stimulus, environmental, electrical, and measurement requirements. Due to the
mechanical and electrical functionalities of the MEMS device, the equipment for the stimulus and
measurement can be varied according to the device application. The closed system can simulate
various packaging environments and should be equipped with an electrical interface to provide the
needed electrical signals and to receive output signals.
4.1.1 Stimulus Requirement
The system is equipped with an electromechanical shaker that can shake the MDUT in a con-
trolled fashion for the measurement of mechanical characteristics such as mechanical transmissibil-
ity and quality factor. As the system is designed for various tests using the MEMS velocity sensor,
other stimulus equipment, such as a rate table, a microphone, or an electromagnetic radiation source,
are not needed.
4.1.2 Environmental Requirement
To simulate various packaging conditions at pressures lower than ambient, the pressure needs
to be controlled, and various gases may need to be injected. When the system is a closed platform
and optical measurement equipment is used, the chamber should have a transparent window for the
light source of the optical equipment to shine onto the device surface.
24
4.1.3 Electrical Requirement
The system should provide an electrical interface, such as an electrical feedthrough, which
can be connected between the external equipment and the MDUT in the chamber. The electrical
feedthrough should also have enough pins for electrical power, control and measurement signals.
4.1.4 Measurement Requirement
Since many unpackaged MEMS devices lack on-chip electronics for converting the applied
mechanical stimulus to an output electrical signal, external optical detection equipment, such as a
laser velocimeter, a laser interferometer, or an optical microscope, may be used. However, if the
MDUT possesses suitable on-chip electronics, then electrical characteristics can be measured using
an oscilloscope, a LCR meter, or a multi-meter.
4.2 Design
The system configuration schematic of the MEMS evaluation system is presented in Figure
4.1. The system was designed as a closed platform type. The bell jar vacuum chamber has a small
shaker, which is located on the heater plate. Since the MDUT is mounted on a small shaker located
on the graphite heat plate, the MDUT cannot be heated by the heater. Therefore, the heater and the
small shaker are not used at the same time. If the heater is to be used, then the shaker is removed
from the bell jar first, and the MDUT is then placed on the plate. A MEMS device is attached
to the small shaker. Outside the bell-jar, two optical laser interferometers and their controllers are
connected to the signal analyzer, which also controls the small shaker in specific frequency and
displacement ranges. A pressure transducer is attached to a mechanical feedthrough, and a readout
meter displays the chamber pressure. The vacuum pump and gas cylinder are connected to the bell
25
Figure 4.1: System Configuration Schematic
jar through the gas lines. With this configuration, the evaluation system can excite MEMS devices
and measure the electrical and mechanical characteristics of them.
4.3 Construction
As presented in the Figure 4.2, the constructed MEMS evaluation system is assembled using
a glass bell jar with a L-gasket, a metal base with electrical feedthroughs and a heater, an analog
pressure gauge, and a mechanical vacuum pump.
 Bell-jar Chamber: A glass bell-jar chamber was chosen for the vacuum chamber. The glass
bell-jar size is 12? in height with a 12.5? diameter. The chamber has a transparent glass bell-jar
which is suitable for laser interferometer usage. The chamber is mounted on the four stiff metal
legs. In the middle of the glass bell-jar and metal legs, five feedthroughs for an electrical connector,
26
Figure 4.2: MEMS Evaluation System: Front View
a heater, a pressure transducer, several valves, and an analog pressure gauge are located around the
metal chamber body, as presented in Figure 4.2 and Figure 4.3. On the bottom of the metal body,
Figure 4.3: MEMS Evaluation System: Side View
three additional feedthroughs are also located that can be used for expansion. An inlet gas line is
connected through a ball valve and a precise control valve with a flow meter for fine control of the
27
gas supply, and an outlet gas line through a manual bellows valve to the vacuum pump. An electrical
feedthrough with a 9-pin D-sub connector is also located in the chamber body and can provide nine
electrical interfaces, as presented in the Figure 4.3.
 Stimulus Equipment: A photograph of an Mini-shaker Type 4810, manufactured by B&K, is
presented in Figure 4.4, which is located on the heater plate inside the chamber. The small shaker is
connected to a power amplifier driven by the dynamic signal analyzer. The power amplifier provides
the small shaker with regulated power to vibrate at a specific amplitude and/or over a bandwidth of
frequencies. A photograph of the Power Amplifier Type 2706, manufactured by B&K, is presented
in Figure 4.2. For securely mounting the MDUT, a 10*32NF threaded hole is located at the center
of the shaker head. Therefore, it is relatively easy to attach a printed circuit board (PCB) to the
shaker hole.
Figure 4.4: MEMS Evaluation System: Inside View
 Vacuum Pumping System: A mechanical vacuum pump is connected to the outlet port of the
vacuum chamber using flexible plastic pipe. As the pumping system is the combination of the vac-
uum pump, the chamber, and its sealing, the vacuum system components also depend on how well
the bell-jar chamber is sealed. There may be many leaking places at junctions between feedthroughs
28
and the chamber. Leaks can occur at many locations in the vacuum system. A photograph of the
Model 5KC36PN4 vacuum pump, manufactured by General Electric, is presented in Figure 4.5.
Figure 4.5: Photograph of the Mechanical Pump
 Optical Measurement System: A laser vibrometric system is composed of a laser unit and
a vibrometer controller. A photograph of the optical measurement system including the signal
analyzer is presented in Figure 4.6. The constructed system has two laser units; the motion of the
moving part of a MEMS device is measured by the first laser interferometric system, and the sec-
ond interferometric system is used to measure the motion of the frame of the MEMS device. The
two output signals are recorded simultaneously, then the transmissibility of the MEMS device can
be determined using the dynamic signal analyzer. The dynamic signal analyzer is an instrument
that processes input signals to yield various information about the signals as opposed to an oscillo-
scope, which simply displays the input signals. The analyzer has an FFT mode for calculating the
frequency spectrum of a time domain signal.
 Electrical Measurement System: The electrical measurement system of the evaluation sys-
tem consists of a LCR meter, a digital oscilloscope, and a digital multimeter (DMM). They are
29
Figure 4.6: Photograph of the Optical Measurement System (two laser units, two controllers, and a
signal analyzer)
located outside the vacuum chamber and are connected to the wires coming out from the electrical
feedthrough of the chamber. A tripple output power supply is often a necessary piece of equipment
to supply power to the MDUT.
 Gas Supply: A gas is injected from the gas cylinder outside the chamber and controlled by
a ball valve and a precise control valve of the vacuum chamber for adjusting the gas flow rate. A
photograph of the valves for supplying gas is presented in Figure 4.7 (a). When a different gas is
required for testing, the new gas cylinder must be added to the system. As presented in Figure 4.1,
the inlet gas line combines with nitrogen from the microlab supply lines at the back in the vacuum
pump room. To pump down the pressure in the chamber or vent a gas from the chamber, a bellows
valve, presented in Figure 4.7 (b), is used. An illustration of a connection of gas supply plumbing
is presented in Figure 4.8. The N2 gas is used every test to pump up the pressure inside the bell-jar
chamber. Especially, when a toxic gas is used, the N2 gas is also used to purge residual toxic gas
from the line and the chamber. As an example, the SO2 cylinder and N2 Gas Control Valves in the
30
Figure 4.7: Valves of the MEMS Evaluation System: (a) Inlet Line Valves, (b) Outlet Line Valve
Figure 4.8: Illustration of Gas Plumbing
31
pump room are presented in Figure 4.9. A method to control SO2 and N2 Gas Valves is presented in
Appendix A.2.
Figure 4.9: (a) SO2 Cylinder and Valves (b) N2 Gas Control Valves
4.4 Capabilities
The major components of the MEMS evaluation system consist of a bell-jar vacuum chamber,
a small shaker, two laser displacement interferometers, and a dynamic analyzer. The capabilities of
the components in the MEMS evaluation system of this study are presented below.
32
4.4.1 Bell-jar Vacuum Chamber
This vacuum chamber allows MDUT characterization over the pressure range of 50 mTorr to
ambient. The pressure can be controlled with two hand valves, which are connected to the vacuum
pump and the opened gas cylinder, respectively. The analog gauge provides a rough pressure reading
and a combination of the pressure transducer and the digital readout allows the user to read the
pressure at high resolution. The electrical feedthrough assembly is compatible with the D-sub 9-
pin connector, which allows easy connection to the vacuum chamber. With the substrate quartz
lamp below the plate, thermal evaluation can also be performed. The detailed specifications for the
components of the bell-jar vacuum chamber are:
- Cover Material: Glass
- Chamber Lowest Pressure: 50 mTorr
- Electrical Feedthrough: 9-pin D-sub
- Heater Max. Temperature: 200 C
- Dynamic Range of Pressure Transducer: 0 to 1 Torr for 10V
- Digital Pressure Readout: 1 mTorr resolution
- Analog Pressure Guage: -30 to 200 psi
- Vacuum Pump: Mechanical Pump
4.4.2 Small Shaker
This shaker uses a electromagnetic field to induce sinusoidal motion with a frequency range of
DC to 18kHz in the attached shaker head onto which the MDUT is mounted. A photograph of the
Mini-shaker Type 4810, which is manufactured by B&K, is presented in Figure 4.10.
- Frequency Range: DC to 18 kHz
33
- Force Rating (peak): 2.25 lbf @65 Hz to 4 kHz, 1.5 lbf @ 65 Hz to 18 kHz
- Max. Bare Table Acceleration: 56 g
- Dimension: Diameter- 3 in, Height- 2.9 in
Figure 4.10: Photograph of the Mini-shaker Type 4810
4.4.3 Optical Measurement System
A photograph of an OFV-353 laser unit and an OFV-2610 vibrometer controller, manufactured
by Polytec, is presented in Figure 4.11. The laser unit emits a single, visible laser to the object,
and focusing can be adjusted with the lens housing. The laser unit also has a mirror allowing the
placement of the laser spot to be manually adjusted. The unit has an adjustable measurement range
of 20?m/V to 5120?m/V.
Figure 4.11: Optical Measurement System: (a) Vibrometer Controller, (b) Laser Unit
34
4.4.4 Dynamic Signal Analyzer
Figure 4.12 presents a photograph of the HP 35665 Dynamic Signal Analyzer. The analyzer
can analyze two input signals with a maximum 800-line resolution and a 51.2kHz frequency band-
width. The fast averaging function, over 12KHz in speed, is useful for viewing statistical graphs.
The analyzer allows analysis in the time and frequency domains. The source output port provides
various types of signals, such as sine, random, pink noise, and chirp signals, which can be used to
control the power amplifier for the small shaker.
Figure 4.12: Photograph of the HP 35665 A Dynamic Signal Analyzer
35
CHAPTER 5
APPLICATION OF THE MEMS EVALUATION SYSTEM
To demonstrate the usefulness of this evaluation system, a multipurpose MEMS device was
fabricated and utilized in a variety of tests involving this evaluation system. This micromachined
device was used in several applications, including as a relative velocity sensor [35], as an active
vibration isolator [45-46], and as a testbed for evaluating the performance of a MEMS device in
different gas environments. In the relative velocity sensor application, the performance of the sensor
was investigated by measuring the relative displacement and the relative velocity between a fixed
electrode and a movable electrode while the device was excited by the electromechanical shaker in
the evaluation system. The MEMS device was also used as a passive vibration isolator to evaluate
the effects of packaging a MEMS device in different gas chemistries at various pressures, through
measuring the effect on the mechanical quality factor of the device when excited by the shaker.
This same device was then used in an active vibration isolator, which was evaluated in the system.
Finally, the MEMS device was used as a capacitor structure for evaluating the relative permittivity
of different gas chemistries at various pressures.
5.1 The MEMS Device
The MEMS device consisted of a micromachined Si square structure with a proof mass pad
attached to a surrounding frame by eight springs. The frame and proof mass pad had a nominal
thickness of 375?m. The spring thickness was set during microfabrication by the length of time
the wafer was DRIE etched. The surrounding frame was 19.6mm long and 3.5mm across, and the
center square proof mass pad was 10mm across. Each of the eight springs was 1300 ?m long and
36
225 ?m wide. 40?m was selected as the designed spring thickness, which resulted in a designed
mechanical natural frequency of 1.39KHz. The design for this MEMS device had previously been
used in other applications [47]. A photograph of the fabricated Si MEMS device is presented in
Figure 5.1.
Figure 5.1: Si MEMS Structure
5.2 Fabrication of the MEMS Device
A detailed fabrication process traveler for fabricating the MEMS device is included in Ap-
pendix A.1.
I. Wafer & Mask:
A 375?m thick silicon double-side polished 4-inch wafer, with a resistivity of 1-10 W-cm, was
utilized to fabricate the MEMS device. Since the feature size was relatively large, two laser
photoplot masks were used to pattern the front and bask sides of the wafer. Seven devices
were fabricated per wafer.
37
II. Cleaning:
There were several wafer cleaning steps used to get rid of contaminants and foreign materials
from the surface of the wafer. First, the wafer was cleaned with a 50:1 H2O:HF solution
at room temperature to remove oxides from the wafer surface. To remove alkali metal and
hydroxides, the wafer was soaked in a 6:1:1 solution of H2O:HCL:H2O2 at 80 C for 10 min.
Next, the wafer was rinsed and blow dried.
III. Dehydration & HMDS Deposition:
The wafer was cleaned with oxygen plasma in the Matrix etcher for 3 min. Then, the wafer
was dehydrated in a dehydration chamber at 120 C for 20 min to remove moisture. Right
after dehydration, the Hexamethyldisilazane (HMDS) deposition was performed in a HMDS
chamber at ambient temperature for 5 min, after cleaning the inside of the chamber and then
pouring 2 to 3 drops of HMDS into the chamber, which was used as an adhesion promoter.
IV. Top Side Photolithography:
Before patterning with a mask, Clarient 5214E photoresist was spun on at 1000 rpm for 30s,
which resulted in a thickness of approximately 2.5 ?m, and the wafer was then softbaked
on a hotplate at 108 C for 2 min. Mask exposure was performed in the MABA 6 aligner
using two 10s exposures. Finally, the wafer was developed in a 2:1 solution of water:AZ
400K developer. It took more than 30s to develop the wafer. If there had been residual
photoresist on the pattern even after 30s of developing, additional developing was performed
in 30s intervals, followed by microscopic inspection, until the remaining photoresist was
removed.
V. Top Side DRIE Etching & Photoresist Strip:
38
As the spring thickness was designed to be 40?m, the wafer was etched in the STS ASE
DRIE etcher with frequent inspections to meet the 40?m depth as close as possible. After
etching was completed, the wafer was placed in the Matrix for 5 min to strip the remaining
photoresist.
VI. Bottom Side Photolithography:
The bottom side photolithography was very similar to the top side photolithograpy process,
except for the photoresist, the mask, and related parameters. After a 3 min descum in the
Matrix and a 5 min HMDS deposition, Clarient 4620 photoresist was spun on at 1000rpm for
30s. Then a 108 C 2min softbake was followed by a 60s exposure using the bottom mask
in the MABA 6 aligner. In this step, the pattern on the top side of the wafer was aligned to
the bottom side mask using cross alignment marks on both sides. The exposed wafer was
developed in a developer with the same ratio as used in the top side photolithography, and
then finally hardbaked at 120 C for 2 min on a hotplate.
VII. Bottom Side DRIE Etching:
The bottom side of the wafer was DRIE etched until the depth reached 140?m, which was
enough to cleave the wafer.
VIII. Wafer Cleaving:
Once the DRIE etching was finished, the outer square patterns on the bottom side were ex-
tended by scratching with a diamond cutter. The wafer was then cleaved very carefully along
the extended scratch lines.
IX. DIE Mounting;
39
To complete the etching, each cleaved die were attached to a Si wafer. Clarient 4620 pho-
toresist was used as an adhesive. It was spun on at 1500 rpm for 30s, and then the wafer was
softbaked on the hotplate at 108 C for 2 min.
X. Bottom Side DRIE Etching:
To complete the DRIE etching that removes the remaining 170?m of Si, the wafer is placed
back in the STS ASE and further etched in DRIE machine. Checking often to prevent over
etching, the DRIE etching was continued until the red photoresist was seen using a micro-
scope.
XI. Post DRIE MEMS Die Removal & Cleaning:
The wafer and die were soaked in acetone until they separated. After separation, first, a wiper
tissue was soaked in isopropyl alcohol, and then the die was placed on the tissue for 30s. The
wiper tissue was used to take the die out instead of using tweezers.
XII. E-Beam Metallization:
The top side of the die was then metalized with titanium (Ti) and gold (Au) by E-beam
evaporation, after taping the four corners of the die to a clean wafer using Kapton tape.
5.3 Background for a Micromachined Parallel Plate Capacitor Structure [46]
 Mechanical and Electro Dynamics:
An illustration of a two parallel electrode MEMS structure, where the bottom electrode is fixed
and the top electrode can move up and down relative to the bottom electrode, is presented in Figure
5.2.
40
Figure 5.2: Parallel Electrode Structure[46]
For this type of structure, the equation for transmissibility is:
jT(jw)j=
q
(wnwQ )2 +w4n
q
(w2n w2)2 +(wnwQ )2
(5.1)
WherejT(jw)jis the transmissibility, Q is the mechanical quality factor, wn is the natural frequency
of the system, where:
wn =
r
k
m (5.2)
and
Q =
pkm
c (5.3)
Since most MEMS devices have Q values greater than five, jT(jw)j has a high resonant peak at
approximately wn.
 Current Flow into the MEMS Capacitor:
Let the relative electrode motion by modeled by:
x(t) = xAsin(wt) (5.4)
41
and the resulting capacitance between the parallel electrodes can be modeled as:
Cs(t) = eoerAx
o +xAsin(wt)
(5.5)
where eoer is the permittivity of the dielectric material between the two electrodes, A is the over-
lapping electrode area, xo is the rest distance, xA is the amplitude of the sinusoidal relative motion
with frequency wn. This model for the capacitance ignores fringing, which is usually small if the
electrode area is much larger than the separation distance. If a small voltage, Vs(t), which is less
than the pull-in voltage, is applied across the two electrodes, then charge, q(t), accumulates in the
time varying capacitor, Cs(t), according to:
q(t) =Cs(t)Vs(t) (5.6)
The current flowing into the capacitor, Is(t) can then be obtained by taking the derivative of q(t),
resulting in
Is(t) =Cs(t)dV(t)dt +Vs(t)dCs(t)dt (5.7)
Suppose that Vs(t) is constrained to be a small, constant DC voltage, Vb. Therefore the current
flowing into Cs(t) becomes
Is(t) =VbdCs(t)dt = VbeoerAxAwcos(wt)(x
o +xAsin(wt))2
(5.8)
The negative sign in the numerator indicates that current flows out of the capacitor when the distance
between the two electrodes is increasing, which is correct since the capacitance is decreasing. Like-
wise, when the distance between the electrodes is decreasing, capacitance is increasing, resulting in
42
current flow into the capacitor and a positive value for Is(t). For xo/xA having a large ratio,
Is(t) = VbeoerAxAwcos(wt)x
o
(5.9)
and that the relative velocity between the electrodes is
?x(t) = xAwcos(wt): (5.10)
If xo/xA is small, (5.9) is not a valid approximation for Is(t). However the polarity of Is(t) correctly
identifies the direction of relative velocity.
5.4 Relative Velocity Sensor in Low Pressure Environment
The relative velocity sensor was evaluated in the evaluation system at ambient and low pres-
sure.
 Assembly
The micromachined Si structure was assembled with exactly the same parts as for the closed-
loop damping control testing except for the PCB board, which is discussed in detail in the closed-
loop damping control section. The PCB board used in this test had a micromachined capacitor
interface circuit consisting of a transimpedance amplifier and a two stage voltage amplifier on the
front side and bottom electrodes on the back side.
 Micromachined Capacitor Interface Circuit
A schematic diagram of the capacitor interface circuit is presented in Figure 5.3. The first
stage, a transimpedance amplifier (TIA) with a gain of -100,000W, converted the current flowing
into the sensor capacitor into a voltage signal. Then, a two stage amplifier consisting of an inverting
43
amplifier and an non-inverting amplifier, with a combined gain of -10100V/V, were used to amplify
the TIA output voltage to a reasonable level for recording the data.
Figure 5.3: Schematic Diagram of Capacitor Interface Circuit
 Testing
The velocity sensor assembly with the capacitor interface circuit was mounted on the small
shaker inside the MEMS evaluation system. The resonant frequency of the velocity sensor was
investigated first using a wide bandwidth random vibration excitation. A time trace of the velocity
sensor at the resonant frequency was recorded with the signal analyzer by detecting the relative
displacement with two interferometers. This experiment to test the relative velocity sensor was per-
formed at ambient pressure and at low pressure (approximately 0.2 Torr). Time traces of the relative
displacement between the two electrodes, measured using the interferometers, and the output signal
44
from the sensor board are presented in Figure 5.4 (ambient pressure test) and Figure 5.5 (low pres-
sure test). As would be expected for a sinusoidal displacement, the relative velocity sensor?s output
Figure 5.4: Time Traces of the Measured Relative Electrode Displacement and the Sensor Output
at Ambient Pressure
Figure 5.5: Time Traces of the Measured Relative Electrode Displacement and the Sensor Output
at Low Pressure
45
signal is also sinusoidal and leads the displacement waveform by 90 degrees. Much of the noise
in the relative velocity signal is due to the differentiation of the noise in the relative displacement
signal.
This test was performed primarily to verify the operation of the MEMS evaluation system.
Also, since the relative velocity sensor is a key component of the closed-loop damping controller,
it was important to verify that the small shaker in the system was powerful enough to generate
sufficient relative motion between the device?s electrode to measure the relative velocity.
5.5 Closed-loop Damping Control At Low Pressure[45]
A closed-loop damping control system was evaluated using the MEMS evaluation system.
 Assembly:
The backside of the proof mass was the upper or movable electrode. The micromachined Si
structure was mounted onto the back side of a double layer printed circuit board directly over a
planar electrode, the fixed electrode, using a nonconductive spacer with a nominal thickness of
100?m, which set the rest distance between the two electrodes. The feedback controller PCB, a
photograph of which is presented in Figure 5.6, was integrated with the micromachined Si structure
as the top electrode for several evaluations. A photograph and an illustration of the assembled
MEMS device are presented in Figure 5.7. The parallel plate actuator (PPA) and the sensor bottom
electrode are on the feedback control circuit PCB. The small round center sensor electrode was
used to detect the relative velocity and the donut shaped outer electrode acted as the PPA bottom
electrode and connected to the feedback controller circuit.
46
Figure 5.6: Feedback Controller Circuit Board [45]: (a) Front side, (b) Back side
Figure 5.7: MEMS Device with the Feedback Controller PCB: (a) Assembled MDUT with the
Feedback Controller Circuit PCB, (b) Illustration of the side view of an assembled MDUT [45]
47
 Feedback Controller Circuit:
The feedback controller electronics circuit consisted of a low impedance op amp current to
voltage conversion circuit (transimpedance amplifier), a two stage voltage amplifier and a voltage
comparator. The output of the comparator powered a PPA, which was used to increases the me-
chanical damping of the active isolator. A schematic diagram of the feedback controller circuit is
presented in Figure 5.8. CS1 represented the velocity sensor capacitance and CX1 the PPA capac-
itance. When the upper electrode was moving toward the bottom electrodes, the sense capacitance
was increasing in size, resulting in current flow into the sense capacitor. This resulted in a positive
voltage being applied to the comparator?s noninverting input and +25V on the comparator output
being applied to the bottom PPA electrode. With the top electrode hardwired to +25V, both PPA
electrodes had the same potential and no electrostatic force was generated. Therefore the PPA was
off whenever the top electrode was moving toward the bottom electrodes. When the upper electrode
moved away from the bottom electrodes, the sense capacitor decreased in size, resulting in current
flow out of the sense capacitor. This resulted in a negative voltage being applied to the comparator?s
noninverting input and -25V on the comparator output being applied to the bottom PPA electrode.
Since the top electrode was hardwired to +25V, the PPA electrodes had a 50V potential difference,
which resulted in the generation of an electrostatic force that attempted to retard the separating mo-
tion of the upper and bottom electrodes. As soon as the electrodes quit moving away from each
other, the PPA shut off. The net result was that kinetic energy was removed from the oscillating
mechanical system, resulting in an increase in mechanical damping.
48
Figure 5.8: The Schematic Diagram of the Feedback Controller Circuit[45]
49
 Testing:
To evaluate the closed loop system, the assembled MEMS device was placed in the chamber
and evaluated at low pressure. Using the electromagnetic shaker, the device was vibrated with and
without the control loop activated. For the closed loop tests, the comparator trip voltage, VTR,
was tied to ground. The feedback controller circuit could lower the quality factor in low pressure,
and the results of this test are presented in Figures 5.9. The data shows that Q decreased a large
amount with the controller activated, from approximately 138 to 50, when the device was tested
at 50 mTorr pressure. Observe that the plot in Figure 5.9 shows a shift in the resonant frequency
of approximately 100Hz when the controller is activated. It is speculated that the change in the
resonant frequency is due to an increase in the system spring constant due to the nonlinearity in
the double clamped spring system when the biased controller electrodes result in a nonzero average
proofmass displacement.
Figure 5.9: Mechanical Frequency Responses Due to Feedback Control Signal at Low Pressure
50
5.6 Electrostatic Enhancement of MEMS Devices
In this test, the effects of packaging an electrostatic MEMS device in a gas with a relatively
high dielectric constant were investigated. The MEMS test article was the same micromachined Si
device used in the other tests. It was attached to a PCB having a single large electrode, just like
the relative velocity sensor was assembled. The capacitance between the two electrodes was then
measured in different gas chemistries at various pressures. A photograph of the bottom electrode of
the PCB is presented in Figure 5.10.
Figure 5.10: Bottom Electrode of the Test PCB
For this test, the MEMS device?s capacitance was measured in air, N2, SO2, SF6, and water
vapor using a model LCR-821 LCR meter, manufactured by GW Instek. A photograph of the LCR-
821 is presented in Figure 5.11.
Figure 5.11: Photograph of GW Instek LCR-821
51
The purpose of these tests were to investigate if the packaging of an electrostatic MEMS device
in various gas environments could be used to enhance performance either by using a gas with a high
relative permittivity or through effecting the system mechanical quality factor. The relevant physical
properties for air, N2, SO2, SF6, and water vapor are presented in Table 5.1. The density properties
were calculated by
r = PRT (5.11)
where R=8.314472(J/K mol).
The permittivity of water vapor was calculated with
e = 1+ r(3aT +3b)T (5.12)
where 3b=3449 and 3a=0.671 [49].
Table 5.1: Viscosity, Density, and Permittivity for Air, N2, SO2, SF6, and Water Vapor [48-49]
Gas Viscosity(?Pa s) Density(Kg/m3) Dielectric Constant
@27 C, 100kPa @25 C, 1atm @20 C, 1atm
Air 18.6 1.1845 1.0005364
N2 17.9 1.1458 1.0005480
SO2 12.9 2.6195 1.0082500
SF6 15.3 5.9738 1.0020000
Water Vapor 9.7 0.7369 10.181060
The measured capacitance values in various gases at different pressures are presented in Table
5.2. The capacitance values decrease as the pressure decreases for each gas chemistry. It was not
meaningful to compare absolute capacitance values to each other as the calibration of the probes of
the LCR meter had a big discrepancy between tests. When the absolute capacitance values of the
results were compared separately, as presented in Figure 5.12 and Figure 5.13, the magnitude
52
Figure 5.12: Capacitance Change in Test1
Figure 5.13: Capacitance Change in Test2 and Test3
53
of the measured capacitance for each gas compared well with the magnitude of the dielectric con-
stants for each gas, except for SF6 in Test1 and air in Test3. However, when the difference of the
capacitances between the lowest and the highest pressure were compared to eliminate the calibration
errors, instead of comparing absolute capacitance values, the capacitance difference is proportional
to the gas permittivity, except for the air. The reason for this was speculated to be that the air was
not dry air. Water vapor has a high relative permittivity, and it is thought that water vapor in the
air injected into the system, or from water molecules that liberated from surfaces in the chamber,
affected the capacitance measurement. In order to perform this test, the gases except for water va-
por were injected into the chamber, while water vapor was evaporated from water in a plastic bottle
inside the chamber for the test. This method to fill the chamber with water vapor was problematic,
because the liquid water in the plastic container would freeze as the pressure was lowered. There-
fore for water vapor, the capacitance was measured while the pressure was deceasing, unlike with
the other gases. As a result of these tests, it was concluded that hermetically packaging an electro-
static MEMS device in a low pressure high dielectric constant gas would not significantly enhance
the device?s performance.
54
Table
5.2:
Capacitance
under
Various
Gases
at
Dif
ferent
Pressures
Pressure
Air
N2
SO
2
SF
6
w.
vapor
(Torr)
Test1
Test2
Test3
Test1
Test2
Test3
Test1
Test1
Test3
760
10.802
4.6361
4.3115
-
-
-
-
-
-
540
10.764
4.6342
4.3088
10.512
4.6391
4.2938
10.607
10.928
4.5110
320
10.806
4.6338
4.2975
10.526
4.6378
4.2893
10.588
10.926
4.5100
100
10.787
4.6319
4.2918
10.510
4.6370
4.2862
10.573
10.922
4.5088
0.9
10.795
4.6328
4.2923
10.521
4.6364
4.2872
10.568
10.921
4.5104
0.8
10.799
4.6320
4.2932
10.517
4.6389
4.2857
10.566
10.917
4.5101
0.7
10.798
4.6325
4.2891
10.511
4.6385
4.2860
10.568
10.919
4.5099
0.6
10.799
4.6337
4.2925
10.495
4.6374
4.2793
10.564
10.916
4.5123
0.5
10.755
4.6332
4.2918
10.493
4.6385
4.2823
10.570
10.916
4.4409
0.4
10.746
4.6330
4.2897
10.495
4.6362
4.2849
10.566
10.915
4.4514
0.3
10.826
4.6321
4.2883
10.492
4.6373
4.2817
10.567
10.913
4.4495
0.2
10.788
4.6325
4.2880
10.491
4.6365
4.2810
10.566
10.917
4.4269
0.1
-
4.6321
4.2879
-
4.6365
4.2846
10.566
-
4.3775
Cm
ax
p-
Cm
in
p
0.014
0.0040
0.0236
0.021
0.0026
0.0092
0.041
0.011
0.1335
Ca
ve
ra
ge
0.014867
0.01093
0.041
0.011
0.1335
55
5.7 Transmissibility vs. Pressure in Various Gases
These tests were performed to investigate the effect on mechanical damping of a MEMS device
by packaging it in various gas chemistries at pressures equal to or lower than ambient pressure.
The same gases were evaluated here as were evaluated in the electrostatic enhancement of MEMS
devices tests. The various gases were evaluated at these pressures: every 220 Torr from 760 Torr to
100 Torr and every 0.1 Torr from 0.9 Torr to 0.2 Torr. The theoretical mechanical frequency response
indicates that the Q values at each testing pressure should increase as the pressure deceases and as
the gas viscosity decreases. The viscosity at 25 C and the density at 25 C, at 1atm for each gas in
Table 5.1 is as follows; Air- 18.5 ?Pa s and 1.19 kg=m3, SF6- 15.3 ?Pa s and 6.1 kg=m3, N2- 17.9
?Pa s and 1.14 kg=m3, water vapor- 9.7 ?Pa s and 0.7369 kg=m3
Measuring the transmissibility yields information regarding the type of response: lowpass,
bandpass, or highpass frequency response, and its resonant frequency or frequencies and the me-
chanical quality factor, Q. The measurement of the quality factor reveals both the susceptibility of
the device to ringing at its resonant frequency and what the time response will be to a step input.
Detailed mechanical frequency response results for each gases are presented in Figures 5.14
through 5.18, and mechanical frequency response results for each test are presented in Figures 5.19
through 5.22. The measured quality factors for all the tests are presented in the Table 5.3. The
results showed that as the pressure decreased, the Q value increased, and the resonant frequency
was substantially shifting to higher values. It is speculated that the gas adds a term to the system
mass due to the gas having to be displaced by the motion of the vibrating proof mass. As the pressure
decreases, the mass term would then decrease, thus increasing the resonant frequency of the device.
The ambient pressure test with SF6, SO2, N2, and water vapor was not performed, because it was
difficult to keep the vacuum chamber at ambient pressure while supplying a pressurized gas to the
56
bell jar. Even though transmissibility tests in ambient pressure with SF6, SO2, N2, and water vapor
could not been performed, when comparing the average quality factor values at pressures between
760 Torr and 100 Torr in Table 5.3, SF6 gas resulted in the lowest quality factor. The order of
average measured quality factors in this pressure range is exactly the reverse order of each gases?
density factor. This results indicate that gas density is a more critical factor than the viscosity at the
higher end of the pressures evaluated in this study: the lower the density, the higher the Q (order
of density at 25 C, 1atm: SF6 > SO2 > Air > N2 > water vapor). As presented in Figures 5.14
through 5.18, the quality factor commonly increased as the pressure decreased.
Figure 5.14: Mechanical Frequency Response in Air
57
Figure 5.15: Mechanical Frequency Response in N2
Figure 5.16: Mechanical Frequency Response in SF6
58
Figure 5.17: Mechanical Frequency Response in Water Vapor
Figure 5.18: Mechanical Frequency Response in SO2
59
Figure 5.19: Mechanical Frequency Response in Various Gases and Pressures of Test1
Figure 5.20: Mechanical Frequency Response in Air
60
Figure 5.21: Mechanical Frequency Response in Nitrogen Gas
Figure 5.22: Mechanical Frequency Response in Water Vapor
61
Table
5.3:
Quality
(Q)
Factor
under
Air
and
N2
gas
at
Dif
ferent
Pressures
Pressure
Air
N2
(Torr)
Test1
Test2
Test3
Test1
Test2
Test3
760
20.2@1.316
27.7@1.324
20.4@1.296
-
-
-
540
17.6@1.224
26.1@1.308
28.0@1.300
17.1@1.264
27.9@1.316
21.6@1.320
320
18.4@1.216
27.0@1.316
26.4@1.300
23.8@1.288
29.6@1.312
27.0@1.320
100
23.0@1.220
29.7@1.320
34.4@1.304
22.0@1.288
34.0@1.304
29.4@1.328
Avg.
Q
24.9
25.8
0.9
50.4@1.228
72.7@1.320
107.3@1.312
89.5@1.300
124.7@1.312
68.0@1.332
0.8
80.6@1.236
83.6@1.324
104.6@1.316
61.7@1.300
94.2@1.312
78.3@1.336
0.7
61.7@1.244
113.4@1.324
111.9@1.320
61.4@1.300
113.8@1.312
74.7@1.336
0.6
69.1@1.248
111.5@1.320
154.0@1.320
69.0@1.304
108.9@1.312
81.8@1.340
0.5
65.4@1.252
146.5@1.320
139.2@1.328
59.0@1.304
156.0@1.308
90.2@1.340
0.4
66.1@1.276
114.0@1.320
137.5@1.328
64.7@1.308
169.1@1.308
109.0@1.344
0.3
115.0@1.284
149.7@1.324
145.3@1.328
83.8@1.308
133.6@1.312
105.6@1.344
0.2
149.9@1.288
179.4@1.324
150.1@1.332
95.6@1.312
219.5@1.308
118.4@1.348
0.1
-
170.0@1.324
156.3@1.332
-
190.3@1.308
134.1@1.344
62
Table 5.4: Quality (Q) Factor under SO2, SF6, and Water Vapor at Different Pressures
Pressure SO2 SF6 w. vapor
(Torr) Test1 Test1 Test3
760 - - -
540 26.5@1.292 15.7@1.276 28.8@1.348
320 21.2@1.300 18.7@1.296 31.6@1.352
100 21.8@1.300 21.7@1.324 28.8@1.360
Avg. Q 23.2 18.7 29.7
0.9 68.3@1.308 58.3@1.328 68.4@1.356
0.8 54.9@1.304 61.6@1.336 79.2@1.352
0.7 69.5@1.308 59.1@1.336 87.8@1.352
0.6 50.3@1.308 65.8@1.340 80.3@1.348
0.5 70.4@1.308 95.5@1.336 87.9@1.340
0.4 62.0@1.304 104.0@1.344 99.3@1.336
0.3 60.1@1.308 131.9@1.344 125.5@1.328
0.2 76.0@1.304 150.2@1.348 124.0@1.320
0.1 124.1@1.304 - 133.9@1.312
63
CHAPTER 6
CONCLUSIONS
In this thesis, the MEMS evaluation system and its application have been discussed. The eval-
uation system was designed and constructed for the mechanical and electrical characterization of
unpackaged MEMS die and was used to measure the mechanical frequency response, the capac-
itance, and the behavior of closed-loop damping control, at various pressures in different gases.
Current research efforts in MEMS evaluation system development show how such a system can be
useful for testing MEMS die. The evaluation system provides practical mechanical and electrical
characterization data on the MEMS device, which can be useful for design modification during the
device development stage.
The test results for the closed-loop damping control system at low pressure showed that damp-
ing by the feedback control circuit reduced the quality factor by one-third from the quality factor
when the feedback controller was off. This evaluation system enabled the investigation of the per-
formance of the MEMS active vibration isolator in a low pressure environment. This is important
since many applications would require its use in a hermetically sealed package at low pressure.
The MEMS evaluation system was useful for investigating the MEMS relative velocity sensor?s
output response at low pressure. This was important because it is a key part of the active vibration
isolator. Previously, the results showed that the displacement in low pressure was greater than the
displacement at ambient pressure, where the MEMS structure had a lower quality factor.
The damping effect in various gases at different pressures was investigated in the MEMS eval-
uation system. It was observed that the density of the gas was more critical than the viscosity. As
64
the pressure decreased, the damping also decreased, which was detected by observing changes in
the device?s quality factor from the recorded transmissibility plots.
The evaluation system was also used to investigate the effects on capacitance in various gases
at various pressures. Since the force produced by MEMS electrostatic actuators is proportional to
capacitance, this test was performed to see if the performance of these actuators could be enhanced
by optimizing the surrounding gas chemistry and pressure. A comparison with absolute capaci-
tance values was meaningless due to a large discrepancy in probe calibration of the LCR meter.
The measured capacitance decreased as the pressure and permittivity of the gas under investigation
decreased, except for air. It is believed that this was due to moisture in the air used during the test.
The moisture resulted in higher measurements than would be expected with dry air. Furthermore,
it is speculated that as the pressure was lowered, water molecules were liberated from surfaces in
the vacuum chamber, which would keep the measured capacitance values higher than expected, due
to the high dielectric constant of water vapor. It is also speculated that the ?pure? gases evaluated
in the system may not have had the same interaction with water molecules in the vacuum system,
thereby occluding their effect on the overall capacitance measurements.
In summary, the primary accomplishment of this endeavor was the development of an evalua-
tion system for unpackaged MEMS die that allowed both electrical and mechanical characterization.
Furthermore, this system allowed the evaluation to be performed in various gas chemistries at pres-
sures lower than ambient air pressure. This system was then used to evaluate several MEMS devices
and integrated microsystems.
65
CHAPTER 7
FUTURE WORK
The current MEMS evaluation system does not have the capability to control vaporization of
water at very low pressure inside the bell jar chamber for evaluating the effects of high concentra-
tions of water vapor at low pressures. This could be useful due to the high dielectric constant of
water vapor. Attempts to control the water vaporization were not successful with a Thermal Electric
Cooler (TEC). If a TEC with more power capacity is used, and a hygrometer is added, water vapor
control could possibly be achieved at a low pressure for a more accurate evaluation.
The leak rate of the vacuum chamber was observed to be approximately 0.007 Torr/min. This
can affect the gas environment inside the vacuum chamber during the test by adding moisture as well
as air into the chamber. Therefore decreasing the leak rate would improve the evaluation system.
Thermal testing using the built-in heater was not performed under this research effort. Evalu-
ating the viscosity of various gas chemistries over pressure and temperature would be useful. Other
evaluation tools could be added to the system to enhance its capabilities. Examples could include
a rotating stage for MEMS gyroscope testing, a residual gas analyzer (RGA), and an embedded
optical camera with magnification optics. An integrated miniature probe station would enhance the
evaluation of bare die without having to use machined fixtures to hold the die.
66
BIBLIOGRAPHY
[1] Jeff LaFrenz, Giorgio Grattiker, Karan V.I.S. Kaler, Martin P. Mintchev, ?Starting from
Scratch: Creating an Information Technology Infrastructure for MEMS-related Research and
Development,? International Journal ? Information and Communication Technologies and Pro-
gramming,? 2003, pp. 12-21.
[2] Chi-Yuan Lee, Guan-Wei Wu, Wei-jung Hsieh, ?Fabrication of micro sensors on a flexible
substrate,? Sensors and Actuators A 147 (2008) pp. 173-176.
[3] A. Nisar, Nitin Afzulpurkar, Banchong Mahaisavariya, Adisorn Tuantranont, ?MEMS-based
micropump in drug delivery and biomedical applications,? Sensors and Actuators B 130 (2008)
pp. 917-942.
[4] A. K. Poddar, K. N. Pandey, ?Microwave Switch Using MEMS-technology,? Sensors and
Actuators A 147 (2008) pp. 173-176.
[5] S. Hurlebaus, L. Gaul, ?Smart structure dynamics,? Mechanical Systems and Signal Proces-
sign 20 (2006) pp. 255-281.
[6] Gregory T. A. Kovacs, NADIM I. Maluf, and Kurt E. Petersen, ?Bulk Micromachining of
Silicon,? Proceedings of The IEEE, Vol. 86, No. 8, August 1998 pp. 1536-1551.
[7] James M. Bustillo, Roger T. Howe, and Richard S. Muller, ?Surface Micromachining for Mi-
croelectromechanical Systems,? Proceedings of The IEEE, Vol. 86, No. 8, August 1998 pp.
1552-1574.
[8] Yunn-Shiuan Liao, Shiang-Woei Chyuan, and Jeng-Tzong Chen, ?FEM versus BEM,? IEEE
Circuits & Devices Magazine, September/October 2004 pp. 25-34.
[9] M.H. Zaman, S.F. Bart, J.R. Gillbert, N.R. Swart, and M. Mariappan, ?An Environment for
Design and Modeling of Electromechanical Micro-systems,? JMSM, Vol. 1, No. 1, 1999 pp.
65-76.
[10] Software: http://www.coventor.com/mems/analyzer/index.html, CoventorWare
[11] Software: http://intellisensesoftware.com/Solutions/Solutions.html, IntelliSuite.
[12] Software: http://www.softmems.com/products.html, MEMS Pro/MEMS Xplorer
[13] D.J. Combes, K.M. Brunson and R. J. T. Bunyan, ?Dynamic Simulation of MEMS Switches
Using Simulink,? Nanotech, Vol. 2, 2003 pp. 456-459.
67
[14] Software: http://www.mathworks.com/products/simulink/?BB=1, Matlab, Simulation
[15] Abhijeet Kolpekwar, R. D. Blanton, ?Development of a MEMS Testing Methodology,? Inter-
national Test Conference, IEEE, 1997 pp. 923-931.
[16] Gino Rinaldi, Muthukumaran Packirisamy, Ion Stiharu, ?Dynamic Testing of Micromechani-
cal Structures under thermo-electro-mechanical Influences,? ScienceDirect, Measurement 40,
2007, pp. 563-574.
[17] V. T. Srikar, S. M. Spearing, ?A Critical Review of Microscale Mechanical Testing Methods
Used in the Design of Microelectromechanical Systems,? Society for Experimental Mechan-
ics, Vol. 43, No. 3, September 2003 pp. 238-247.
[18] http://www.polytec.com/usa/ files/LM AN INFO 0103 E Vibrometry Basics.pdf, Measure-
ment Solutions
[19] Christian Rembe, Richard S. Muller, ?Measurement System for Full Three-Dimensional Mo-
tion Characterization of MEMS,? IEEE, Journal of Microelectromechanical Systems, Vol. 11,
No. 5, October 2002 pp. 479-488.
[20] R. M. Lin, W. J. Wang, ?Structural dynamics of microsystems - current state of research and
future directions,? ScienceDirect, Mechanical Systems and Signal Processing 20 (2006) pp.
1015-1043.
[21] Dennis M. Freeman, C. Quentin Davis, ?Using Video Microscopy to Characterize microme-
chanical Systems,? IEEE/LEOS Summer Topical Meetings, 1998 pp. 9-10.
[22] http://umech.mit.edu/people/freeman/freeman mems/
[23] C. Rembe, B. Tibken, E. P. Hofer, ?Analysis of the dynamics in microactuators using high-
speed cine photography,? IEEE Journal of Microelectromechanical Systems 10 (2001) pp.
137-145.
[24] http://www.suss.com/products/test systems/jp/test systems, PAP200
[25] http://www.emopto.com/, Model 622-A 3-D MEMS Profiling System
[26] Beno?it Charlot, Salvador Mir, Fabien Parrain, and Bernard Courtois, ?Electrically Induced
Stimuli For MEMS Self-Test,? IEEE 2001 pp. 210-215.
[27] S. Mir, L. Rufer, A. Dhayni, ?Built-in-self-test techniques for MEMS,? Microelectronics Jour-
nal 37 (2006), pp. 1591-1597.
[28] Xingguo Xiong, Yu-Liang Wu, and wen-Ben Jone, ?A Dual-Mode Built-in Self-Test technique
for Capacitive MEMS Devices,? IEEE Transaction on instrumentation and measurement, Vol.
54, No. 5, October 2005 pp. 1739-1750.
68
[29] Stephen Beeby, Graham Ensell, Michael Kraft, Neil White, ? MEMS Mechanical Sensors,?
Artech House, Inc, Chapter 5, pp. 104-105.
[30] Harald Sehr, Alan G R Evans, Arthur Brunnschweiler, Graham J Ensell, and Trevor E G
Niblock, ?Fabrication and test of thermal vertical bimorph actuators for movement in the wafer
plane,? Journal of Micromechanics and Microengineering, 11 (2001), pp. 306-310.
[31] Ankur Jain, Hongwei Qu, Shane Todd, Huikai Xie, ?A thermal bimorph micromirror with
large bi-directional and vertical actuation,? Sensors and Actuators, A 122 (2005), pp. 9-15.
[32] Sergey Edward Lyshevshi, ? MEMS and NEMS Systems, Devices, and Structures,? CRC
Press, Chapter 2, p. 51.
[33] Chia-Yen Lee, Hsien-Tsung Chang, and Chih-Yung Wen, ?A MEMS-based valveless
impedance pump utilizing electromagnetic actuation,? Journal of Micromechanics and Mi-
croengineering, 18 (2008), pp. 35-44.
[34] Sangkyung Sung, Jang Gyu Lee, Byeungleul Lee, and Taesam Kang, ?design and performance
test of an oscillation loop for a MEMS resonant accelerometer,? Journal of Micromechanics
and Microengineering, 13 (2003), pp. 246-253.
[35] Robert Neal Dean, Jr., George T. Flowers, Roland Horvath, Nicole Sanders, A. Scottedward
Hodel, John Y. Hung, Thaddeus Roppel, ?Characterization and Experimental Verification of
the Nonlinear Distortion in a Technique for Measuring the Relative Velocity Between Micro-
machined Structures in Normal Translational Motion,? IEEE Sensors Journal, Vol. 7, No. 4,
April 2007 pp. 496-501.
[36] http://www.lds-group.com/docs/display product.php?product item id=442, Model V830
Electrodynamic Shaker, LDS
[37] Ashwin Mohan, Ajay P. Malshe, Shyan Aravamudhan, and Shekhar Bhansali ?Piezoresistive
MEMS Pressure Sensor and Packaging for Harsh Oceanic Environment,? IEEE Electronic
Components and technology Conference, 2004 pp. 948-950.
[38] Keith B Brown, Walter Allegretto, Fred E Vermeulen and Alexander M Robinson, ?Simple
resonating microstructures for gas pressure measurement.? Journal of Micromechanics and
Microengineering, 12 (2002), pp. 204-210.
[39] R.N. Dean, G.T. Flowers, A.S. Hodel, G. Roth, S. Castro, R. Zhou, A. Moreira, A. Ahmed,
R. Rifki, B.E. Grantham, D. Bittle and J. Brunsch, ?On the Degradation of MEMS Gyroscope
Performance in the Presence of High Power Acoustic Noise,? Proc. of the 2007 IEEE Int.
Symposium on Industrial Electronics (ISIE 2007), Vigo, Spain, June 4-7, 2007, pp. 1435-
1440.
[40] http://www.polytec.com/usa/158 379.asp, LSV 6000 Surface Velocimetry Optical Profiler
69
[41] http://www.polytec.com/usa/158 426.asp, MSA-500 Micro System Analyzer
[42] http://www.veeco.com/pdfs/datasheets/ds525 20dmems 20data 20sheet 396,pdf, Wyko
NT1100 Dynamic MEMS Measurement System
[43] http://www.suss.com/products/test systems/jp/test systems, SUSS PAP200 Pressure Sensor
Tester
[44] http://www.polytec.com/usa/158 426.asp, Polytec MSA-400 Microscope Scanning Vibrome-
ter
[45] S.J. Kim, C. Chen, G. Flowers and R. Dean, ?Active damping control of micromachined de-
vices in a low atmospheric pressure environment,? Proc. of SPIE Smart Structures and Mate-
rials & Nondestructive Evaluation and Health Monitoring 2008, Vol. 6928, San Diego, CA,
March 9-13, 2008 (11pp).
[46] Seong Jin Kim, Chen Chen, Robert Dean, George Flowers, A. Scottedward Hodel, Stanley
J. Reeves, ?A Packaging Solution to Reduce Electrical Noise in MEMS Capacitive Elements
Resulting from Environmental Mechanical Vibration,? IMAPS 4th International Conference
on Device Packaging, AZ, March 2008.
[47] ROBERT N. DEAN, ?A Technique for the Measurement of Relative Velocity between Par-
allel Plate Electrodes in Micromachined Structures?, Auburn University Graduate School,
5/11/2006.
[48] David R. Lide, ?Handbook of Chemistry and Physics,? 84th Edition, 2003-2004, CRC Press
[49] N. Ernest Dorsey, ?Properties of ordinary water-substance,? No81 ReinHold Publishing Corp.,
1940.
[50] http://www.analog.com/en/mems-and-sensors/imems-accelerometers/adxl180/products/product.html
APENDICES
70
APPENDIX A
A.1 Micromachined Si MEMS Device Fabrication Procedure
I. 350?m DSP Si Wafer Clean
1. 50:1 H2O:HF at ambient for 30s
2. 6:1:1 H20:HCL:H202 at 80 C for 10min
3. DI water rinse and blow dry
II. Dehydration
1. 20min at 120 C in dehydration oven
III. Top Side Photolithography
1. 3min O2 clean in Matrix
2. 15min vapor phase HMDS deposition at ambient
3. PR spin: Clarient 5214E, 1000RPM, 30s
4. Softbake: 108 C for 2min on hotplate
5. Soak glass plate in cooled (after 10min) piranha etchant for 2min
6. DI water rinse and N blow dry glass plate
7. Mount ?Pas Top 11/12/04? photoplot on glass plate with tape
8. MABA6: 14s exposure with photoplot attached to glass plate mask
9. Develop: Clarient AZ 400K, 2:1, 30s
71
10. DI water rinse and blow dry
IV. Top Side DRIE
1. 40?m Si etch in STS ASE
2. Inspect often
3. As close to 40um as possible
V. Top Side Photoresist Strip
1. 5min O2 clean in Matrix (500W, 5Torr)
VI. Bottom Side Photolithography
1. 3min O2 clean in Matrix
2. 15min vapor phase HMDS deposition at ambient
3. PR spin: 1045, 1000RPM, 30s
4. Softbake: 108 C for 2min on hotplate
5. Soak glass plate in cooled (after 10min) piranha for 2min
6. DI water rinse and N blow dry glass plate
7. Mount ?Pas Bot 11/12/04? photoplot on glass plate with tape
8. MABA6: 60s exposure with photoplot attached to glass plate mask
9. Develop: Clarient AZ 400K, 2:1, 30s+ (until fully developed)
10. DI water rinse and blow dry
11. Hardbake: 120 C for 2min on hotplate
12. DI water rinse and blow dry
72
VII. Bottom Side DRIE 1
1. 140?m Si etch in STS ASE
VIII. Cleave Wafer
1. Using a diamond cutter, the etched lines parallel to the flat edge of the wafer are extended
to each end of the wafer. The wafer is then aligned in a clamp and carefully snapped at
each horizontal cut. This snapping is then repeated for the perpendicular etched lines.
IX. Oxidize Test Wafer
1. Push in at 900 C
2. Ramp to 1050 C
3. N2 - 5 minutes
4. O2 - 5 minutes
5. O2, H2 - 1 hour
6. O2 - 5 minutes
7. N2 - 5 minutes
8. Ramp down to 900 C and pull out
X. Die Mount 1
1. PR spin: 4620, 2000 rpm, 700 ramp, 30s
2. Place four die symmetrically on the test wafer with bottom side up and gently apply
pressure onto each die.
3. Softbake: 108 C for 3min on hotplate
73
XI. Bottom Side DRIE 2
1. 170?m Si etch in STS ASE
2. Etch until part released - check if the top and bottom etching side meet each other under
the microscope; if they meet, the color turned to red or black. If not, check it after every
cycle: rotate the wafer by 90 every cycle.
3. Check often to prevent over etching
XII. Post DRIE MEMS Die Removal
1. Prepare two glass dishes and pour the acetone. Cut the wiper tissue and put it into each
dishes; this tissue will be used for moving the die from the dishes not to break them.
2. Soak the wafer for 30 to 60 min.
3. Gently stir the dish and check if each die are away from the wafer. If not, wait more.
Repeat it every 15min.
4. Move the wafer and die using the wiper tissue. Gently remove each die and move them
to another dish to clean.
5. Prepare a glass dish and pour the isopropyl alcohol. Rinse1: Move the die and put them
into the the isopropyl alcohol using the wiper tissue.
6. Rinse2: Move them out of the dish using the wiper tissue. Change the alcohol to a new
one and put them into it again using the wiper tissue.
7. Soak MEMS die in DI water bath using the wiper tissue.
8. Allow MEMS die to air dry in a very soft blow.
9. Rinse test wafer in DI water and N blow dry.
74
XIII. Post DRIE Clean
1. Place MEMS die on holder wafer, bottom side up.
2. 5min O2 clean in Matrix (500W, 5Torr) [until all photoresist removed]
3. Place MEMS die on holder wafer, top side up.
4. 5min O2 clean in Matrix (500W, 5Torr) [until all photoresist removed]
XIV. E-Beam Metallization
1. Mount die onto a holder wafer with top side up. Die are held down with a narrow piece
of Kapton tape on each corner.
2. 2min Ar ion clean
3. 800A Ti - E-beam
4. 2000A Au - E-beam
5. CAREFULLY remove blue tape from corners of MEMS die (The edges of the tape are
carefully pealed up from the wafer, but left on the die. Once all the edges have been
removed, and the die can be lifted from the wafer, the tape can be fully removed from
the die.)
6. Inspect MEMS die
75
A.2 How To Use the Evaluation System
This section explains how to use the MEMS evaluation system to test MEMS devices.
1. Locating the small shaker
- The shaker is located on top of the graphite heating plate. Make sure that the heating
plate is securely mounted over heating lamp and put the shaker on the even heating plate.
Once the shaker is located on the plate, the shaker?s power cable should be connected
through the electrical feedthrough to the shaker. Then, DUT article can then be attached
to the shaker head.
2. Cable Setup Between Equipment
- The Nine Connection wires between Inside and Outside: Inside and outside the bell jar,
there are nine electrical connections passing through the 9-D sub electrical connector
in the electrical feedthrough. Among the nine connection wires, two lines are used for
the shaker power cable, which is a very thin coaxial cable connecting to the shaker. For
the applications under this endeavor, four connections are used not only to provide an
electronic circuit board with +/-25V power and ground but also to receive the output
signal of the capacitor voltage from the board. The rest of the three connection wires
are reserved for additional purpose. For example, if a TEC peltier and a thermistor are
installed inside the chamber to control the water vaporization, four more wires would
be needed. So, three reserved wires and one wire among the used ones need to connect
to them. Nine connection wires are connected to the external cable through the elec-
trical feedthrough. Under this endeavor, the nine electrical wires from the 9-pin D-sub
connector are used to connect to a power supply, a LCR meter and/or an oscilloscope.
76
- Cables between Laser Interferometer Controller and Signal Analyzer: The coaxial cable
is connected between the BNC port of the interferometer controller and one of two
input channel ports of the signal analyzer. The output from the interferometer controller
measuring the displacement of MEMS outer frame should be connected to channel 1
port of the signal analyzer.
- Cables between Shaker Power Amplifier and Small Shaker: The small shaker needs to
receive power from the shaker amplifier, so the power line from the red and black ports
are connected to the small shaker through the 9-pin D-sub connector. A BNC line port
in the small shaker, which receives the control signal, is connected to two line from the
9-pin D-sub connector inside. (Refer to Figure A.1)
- Cables between Shaker Power Amplifier and Signal Analyzer: A BNC output port of
signal analyzer is connected to the input port of the shaker power amplifier by using a
coaxial wire. (Refer to Figure A.1)
3. Dynamic Signal Analyzer Setup
- Measuring the Mechanical Frequency Response: To measure the mechanical frequency
response, the analyzer should be set to FFT mode for two channels. Then, set it to the
dual display mode and Hamming mode. Adjust the input frequency range, source level,
and input wave type. (400Hz to 3600 Hz, 70 to 100 mVrms, and random signal are
the parameters used for transmissibility testing for this endeavor). Once the resonant
frequency is detected, while shaking the MDUT with a sine wave input at the resonant
frequency, time traces of the velocity direction are recorded. Especially, in order to get
77
Figure A.1: Cable Connection between Equipment
78
a clean output result, change unit of ?m/V in the laser controller until noise in the output
wave disappears.
4. Laser Controller Setup
- The laser controller has a range of 20?m/V to 5120?m/V unit. Depending on the output
signal level, a suitable unit needs to be set for a clean output result.
5. Focusing the Laser Spot of the Interferometer
- For a test in ambient pressure, the glass bell jar is not used. Therefore, the laser beam
directly reflects from the surface of MDUT, and the signal strength is over 90%. For a
low pressure test, the chamber is covered with the glass bell jar, which leads to difficul-
ties in focusing and obtaining a high signal strength due to the refraction by the curved
surface of the bell jar. Move the laser beam position and adjust the mirror until over
90% of the signal strength can be received.
6. Gas Handling Valves
- Please refer to the valve pluming configuration diagram in Figure 4.8. The N2 valve and
SO2 gas valve are located in a metal cabinet at the back of the yellow room. The N2 is
used for purging as well as an evaluation gas. To supply the N2 and SO2 (or other gases)
to the chamber, the valve in the cabinet needs to be open or closed. In the case of SF6
gas, the gas cylinder in its cart is connected to the Tee of the flow meter, which has the
main ball valve and precise control valve. These three valves including a valve of the
SF6 gas cylinder are used to open or close for supplying the SF6.
7. Pumping Down the Chamber
79
- To pump down the chamber, the main ball valve and precise valve are closed and the
bellows valve is opened. The analog pressure gauge is used over the range of 760 Torr
to 1 Torr, and for lower pressures (less than 1 Torr), the digital, digital readout displays
accurate pressure reading. The minimum pressure obtainable in the current system is 50
mTorr. It takes around four hours to reach at 50 mTorr and less than 15 minutes to reach
at 1 Torr. If it takes longer, the operator should check for leaks in the system.
8. Pumping Up the Chamber
- The N2 is injected into the chamber while the bellows valve is closed and the precise
valve is open slightly, so that the bell jar cover does not pop up due to the high pressure
of the N2. While watching the pressure gauge and the flow meter, the precise valve
should be used to regulate the gas flow.
9. Maintaining a Certain Pressure
- Maintaining a certain pressure is not easy due to leaks in the system. By adjusting the
bellows valve precisely with the two inlet valves closed, equilibrium status between
pumping down and air injection can usually be accomplished.
10. Controlling the N2 and SO2 Gas Valves in the Pump Room
- As presented in Figure 4.9, the main valve should be open to supply SO2 gas. Then, a
proper pressure can be adjusted with the regulator valve of the SO2 cylinder. N2 gas can
be supplied by opening the N2 main valve and then the N2 control valve. The N2 check
valve allows the operator to determine if there is N2 gas in the tank. In order to stop
supplying a gas, the main valves for each gas are closed.
80
11. Purging a Gas
- After a toxic gas such as SO2 is used for a test, valves of the SO2 and N2 gases are
closed. Then the bellows valve is opened to vent the residual SO2 gas from the plumbing
and the chamber for a few minutes. Then N2 gas valve is opened to inject N2 gas into
the plumbing line and the chamber. The N2 gas valve should not be opened fully when
the bellows valve is open, to avoid overconsumption of N2 gas. The operator can also
adjust the N2 gas flow using the precise valve, even when the N2 valve is fully open.
12. The Order of Turning on Equipment
- Once all the cables are connected, as explained in the cable setup section, turn on the
equipment in the following order. Vibrometer controller!laser unit!(signal analyzer
!set all the parameter of the signal analyzer)!(the small shaker power amplifier!
increase the amplification by rotating the adjustment knob to the designated line), When
turning off the equipment, do this procedure in the reverse order. NOTE: if the signal
analyzer is off with the shaker power amplifier is on, the shaker can be damaged due to
the peak surge from the analyzer.
81