Interferometric Modulator Display Presentation 4: Anatomy of an iMod Influenza del controllo d'imbardata sul rendimento degli aerogeneratori di piccola taglia Interferometric Modulator Display Presentation 4: Anatomy of an iMod ACCOTO Celso ADHAM Mohamed LARRIEU JeanCharles LE GROS Christophe MAQUEDA LÓPEZ Mariazel OTTONELLO BRIANO Floria PIZZATO Daniel SADRINI Jury Team Members: DIPARTIMENTO DI INGEGNERIA DELL’AUTOMAZIONE E DEI SISTEMI 1
Introduction Agenda: New Design Simulations Fabrication Presentation 3: We are designing a new MEMS based Display based on a Fabrey Perot Interferometer (iMod) Previous Problems: - Actuation voltage too high - process: insulating material (PMMA) would have melted… Agenda: New Design Simulations Fabrication
Improving the Design Last time, we planned to: Changed physical design Modify back electrode of device Materials More compliant materials Thicknesses Thinner back plate (technology issue) Lengths Resolution is a constraint Boost V Low power constraint
two beams support the reflective plate New ANSYS Design New single IMOD model: two beams support the reflective plate Former design
MATBLAB Simulations Why did we simulate our model on MATLAB? it’s difficult to set design parameters on ANSYS to have an idea of the values that we will use on ANSYS for the dynamic simulation Hypothesis: Static simulation: just a look at the final state of the system ▪ No viscosity forces ▪ No switching time analysis 2) Clamped beam modeling of the system 3) Everything is made with poly-Silicon
Displacement function of the voltage (3 to 6 V) MATBLAB Results Dimensions: 20 μm plate Displacement function of the voltage (3 to 6 V) Vertical position of beam along its length
MATBLAB Results New Design: Lenght Voltqge 8 μm 31.1 V 10 μm 22.3 V
26 % improvement changing the design! MATBLAB Results: comparison Results: @ 20 μm: from 30 V to 7.8V 26 % improvement changing the design!
Choice of the design Adv: residual stress just twist the structure. Dis: difficult to design. Adv: beams instead of plate
ANSYS Design New design: From plate to beam Less stiff system Plate parallel to substrate
ANSYS Simulation Difficulties… 400 lines code 34 3-D points (102 positions) Problems in the meshing Problems assembling the system
Process Flow & Mask Design New process flow: Previous mistakes New design Mask modeling: AutoCAD Eg. of mask ... Still preliminary
Process Flow Glass substrate Silver deposition: sputtering
Process Flow Silver deposition: sputtering
Process Flow Resist deposition
Process Flow Mask 1
Process Flow Etching
Process Flow Dielectric layer
Process Flow Mask 1 Resist+etch+resist removal
Process Flow SiO2 layer: PECVD
Process Flow Mask 2 Resist: DQN
Process Flow SiO2 etch 80 nm exactly
Process Flow Mask 3 Resist: DQN
Process Flow SiO2 etch 50 nm exactly
Process Flow Remove DQN
Process Flow Deposition of Al: Reflective layer
Process Flow Mask 4 Deposition of protective squares
Process Flow Etch of Al + Removal of protective squares
Process Flow SiO2 deposition
Process Flow Mask 5 Etching of SiO2
Process Flow Mask 6 Etching of SiO2
Process Flow SiN deposition
Process Flow Mask 7 Etching of SiO2
Concept of the system Vhold V+ V- Processor TSP 65131 Battery Column selection Processor TSP 65131 Vhold V+ V- Row selection “Operating Principles of Mirasol Displays: Interferometric Modulation (IMOD) Drive”, QUALCOMM December 2007
PSPICE Design “Operating Principles of Mirasol Displays: Interferometric Modulation (IMOD) Drive”, QUALCOMM December 2007
Conclusions Mask is the 1st approximation of the fabrication process Final masks for the process are under review Ansys/Matlab: Less stiff beams → lower activation voltage: 8V Vibration analysis and switching time to be checked PSPICE Addressing circuit has been defined Power consumption has been evaluated
Thank you for your attention!