Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.



No description

Samuel Vandewaeter

on 25 June 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of MEMS

An Untethered, Electrostatic, Globally Controllable MEMS Micro-Robot
Bruce R. Donald, Member, IEEE, Christopher G. Levey, Member, IEEE, Craig D. McGray, Member, IEEE,
Igor Paprotny, and Daniela Rus

Samuel Vandewaeter
Exam MEMS 25th of June 2013
Globally Controllable
"Development of a micron scale, self-contained locomotive platform that can be controlled in a 2D space by the use of capacitive coupling for electrostatic
power delivery"
Exploration of dangerous environments
Manipulation and assembly of hybrid microsystems
Self-assembly -> applied !
Previous solutions
Power Delivery
photo-thermal transduction
inductive coupling
gold bond wiring

Different , micron-scale actuators, components
macro-scale chassis: mounting of other components for steering, communication,power,...
-> walking chips
Two monolithic actuators controlled by capacitive coupling
One, globally available signal for power and control
Working Principle
Scratch Drive Actuators
Terunobu Akiyama and Hiroyuki Fujita
"Controlled stepwise motion in polysilicon mictrostructures" 1993
"A Quantitative Analysis of Scratch Drive Actuator Using Buckling Motion"
Kazuaki Hayakawa,, Akihiro Torii, Member, Akiteru Ueda
"An Analysis of the Elastic Deformation of an Electrostatic Microactuator"
Capacitive Coupling:
Capacitances C1, C2 proportional to area of overlap
-> Actuator >> width of electrode (18um +2um)
-> Plate dimensions even multiples of electrodes
Different electrode shape
Actuator potential = (V1+V2)/2
->Attraction force
-> Actuator deformation
-> Forward step

Steering arm
Capacitive Coupling
Forward Movement -> Scratch Driver actuator
Similar as for SDA
Snap down of steering arm causes friction
SDA turns around this friction point

Two Parts
SDA: Propulsion
Steering arm: turning
Only one array of electrodes !

Cycle of control voltages lies within another cycle
Forward movement independent of steering
When steering, forward movement is needed
-> Steering arm cycle around SDA cycle
FSM for different control actions
Moving forward
Turning left
Shape of drive signal
Drive waveform: SDA
Shouldn't affect steering arm
Maintain non-zero voltage
Change of actuator performance at change of baseline voltage. (CMMR)
Baseline: 39 V, Peak: 112V
Avoid accumulation of static charge, affects functionality:
change of polarity every 250 pulses
small duty cycle: 10 us peak, 30 us base (~frequency: 4% - 32%)
112V is higher than minimum required (60V)
Frequency -> Speed and turning rate

Control Strategies
Steering Signal
Steering arm: tiff enough so peak waveform doesn't pull the steering arm down: High Vsd
Flexible so base voltage of drive waveform keeps steering arm in place: Low Vr
->Snap Ratio= Vsd/Vr : only dependent on gap sizes !
-> Structure in z-direction
Mems technology is planar ! -> Problem
->Deform out of plane by stress gradients of bilayer materials.
V1=0 V, V4=140 V
Looking back at the course...
forward movement
left turns
-> Other movements composed
-> 2D plane
Multi User MEMS processes (MUMS) 1992
Shared wafer program, customers buy small tiles
Fixed set of standard processes
-> Design based on process
-> Faster, cheaper and more efficient
Price is around €3000 for one tile
Remark: Previous papers didn't use the PolyMUMPS process (A Quantitative Analysis of Scratch Drive Actuator Using Buckling Motion)
8 mask levels
7physical layers
3 layers polysilicon
2 sacrificial layers
1 metal layer
minimum size of features: 2um
heavily doped silicon wafer
Silicon nitride layer by LPCVD (600 nm)
Polysilicon Poly-0 by LPCVD (500 nm)

PSG (Phosphosilicate glass) by LPCVD (2um)
->1st Oxide and removed at end to release MEMS
Dimple etch: small valley by the use of a mask and RIE (Reactive Ion Etch): 750 nm depth
Removal of the Poly-0 by photolithography
Anchor 1 etch
PSG layer for doping and reduce residual stress
After coating with photoresist, lithographically patterned.
Undoped Polysilicone 1 layer by LPCVD (2 um)
Second oxide layer (0,75 um PSG)
Via etch (RIE) + Anchor 2 etch
Undoped Polysilicon 2 (1,5 um)
200 nm PSG layer: doping+ reduce residual stress
After coating with photoresist, lithographically patterned. (Plasma etching)
Immersing in 49% HF: removes all oxide layers
-> Release of structure
1-cm^2 silicon die
Deposition of chromium layer
(830 A) by thermal evaporation
Lithograpic pattern with Stress Layer
Patterning of contact mask
Phosporus doped silicon wafer
Dry oxidation: high quality, high density thermal silical layer
Wet oxidation: additional thickness
See course
Zirconium dioxide layer
Electron Beam Evaporation
5100 A
insulation of electrodes
Metal layer:resistive boat evaporation
50 A chromium: adhesion with ZrO2
500 A gold: conduction
50A chromium: adhesion with SiO2
Difficult assembly of wheels, propulsion, control,..
Scaling laws
Reinvention of the Wheel?
Scaling down is beneficial:
step ~ L^0.5
L ~ L^1.5
Reference Japanese
Other forces on small scale
Formula only applicable in sub mm range
K. Hayakawa, A. Torii, and A. Ueda, “An analysis of the elastic deformation
of an electrostatic microactuator,” Trans. Inst. Elec. Eng. Jpn.,
Part E, vol. 118-E, no. 3, pp. 205–211, Mar. 1998.
Steering Arm
Simplification to cantilever with load at end
Deflection by force:

Force "F"
Caused by weight of endpoint:
F~L^3 -> yload ~L^2
Maximal possible force F:

F ~ L^2 ->ystress ~L
Less deflection by load, linear scaling of applicable deflection
Snap ratio -> scaling has no effect on ratio
Production Processes
In vacuum chamber
Block of source material is heated till evaporation and condenses on the surfaces
Different evaporation techniques only differ in heat source
Resistive Boat evaporation:
large current through resistive "boat"
in vacuum
high temperatures cause evaporation, material condenses on wafer
boat material: molybdenum, tantalum, tungsten, ceramica materials...
E-beam: electron beam
chemical reaction between gases in reactor.
solid material condenses.
LPCVD (Low Pressure Chemical Vapor Deposition)
sub-atmospheric pressure
reduce unwanted reactions
improved film uniformity
Advantages: good uniformity of thickness and characteristics
Disadvantages: high temperatures, slow deposition
Evaporation techniques
Etching removes non-protected areas
dry etching: with plasma
liquid etching
Electro-Magnetic Control
Electric and magnetic field are perpendicular
Possibility for 3D control ?
Magnetic field controls carrier:
Electric field controls actuator (e.g. tweezers)
Applications: biomedical

Moving backwards
Turning in both directions
Extension to 3 actuators:
All control voltages are nested
Ri<Rj and Dj<Di
Actuator for logic switching
Control signal
Control by FSM
Combination of control+power signal
no internal logic needed
2^n bit saved in states
Interruption of signal: unwanted change in states (surgery)
Tight tolerances for construction: control voltage is representation of physical behaviour. (extensibility)
Micro-scale robots
Older robots: millimeter order -> 1 à 2 orders smaller
Problems when scaling down:
Non-linear scaling of forces
Power delivery
Control: the actuators, movement, ...needs to be controlled

Bruce R. Donald, Christopher G. Levey, Craig D. McGray, Igor Paprotny and Daniela Rus
Power Delivery and Locomotion of
Untethered Microactuators 2003
Deze Paper:
An Untethered, Electrostatic, Globally Controllable
MEMS Micro-Robot
Bruce R Donald: h-index 25
Levey, CG: h-index 8
70 x cited
Average: 8.75/year
Impact? Web of Knowledge
Extra extensions

Robot-> flexible strip
Array of electrodes on both sides
Potential of robot:

When V1=V2=V3=V4
-> Vrobot is constant as long as the robot is parallel with walls
When moving towards a wall
C1,C2 increases
C3,C4 decreases
Vrobot is constant
Upwards force increases
-> Unstable system !
Control voltages of one side nested in between the voltages of other side (means remain equal)
Zone of moving preference becomes bigger
Switching of control voltages between sides:
-> robot moves back and forth around middle
Still not completely stable
External forces may end up pulling it to one side
-> Not completely stable
-> Position feedback to change control voltages (active)
Distance in between electrodes:
Robot "sees" a smaller area of overlap when close to the wall
->smaller attraction force to the wall
Moving forward
Control voltage switching with phase difference between walls:
Vrobot = constant
Different attraction force along robot
->Snape like movement
Different means
->Vrobot changes
->Moment around the robot
The paper was situated in time, technically and academically.
A lot of links with the course were made (scaling,.)
Extra information was provided (PolyMUMPS,..)
Future opportunities, including possible extensions were presented
The cantilever control by capacitive coupling can be used for:
Valve control in micro fluidics (M. De Volder)
As a MEMS actuator, ( Harrie Tilmans):
e.g. micro relay

After process: actuators still attached to substrate by tethers
Automatically : self assembly
Automatic release;
Silicon beam connects two actuators: S: assisting, T: target
Moving at same pace
Target stops, holds down
Assisting continues till beam snaps
Full transcript