Testing and improvement of micro-optical-switch dynamics

https://doi.org/10.1016/S0026-2714(00)00216-XGet rights and content

Abstract

The experimental and theoretical investigations presented in this paper have been performed on a micro-opto-electro-mechanical switch developed for switching and attenuation of light in optical fibers. It is demonstrated that high-speed cine photomicrography together with model-based evaluation of the image sequences is a powerful diagnostic tool for reliability testing of dynamic processes in micro-electro-mechanical systems.

Introduction

Micro-electro-mechanical systems (MEMS) are usually characterized by comparing the system output with a defined system input. In the case of the optical microswitch presented in this paper, the device was investigated previously by the characteristic curve of applied voltage and measured attenuation of the light going through the fiber and through the switch [1]. In addition, it was investigated dynamically by applying a sine function and a step function at the device and measuring the light attenuation versus time. The long length of the beam (10 mm) in contrast to the short width (20 μm) results in dynamics of the optical microswitch, that cannot be characterized by measuring only the system output. The step response of the actuator shows a nonreproducible dynamic behavior of the attenuation. However, this characterization method does not result in knowledge of failure mechanisms inside the microactuator. We demonstrate in this paper that high-speed cinematography is capable to measure such rapid nonreproducible events in MEMS.

Section snippets

Optical microswitch

The optical microswitch has been realized with LIGA technique by integrating a flexible beam anchored at the ends to a piezoelectric monolayer substrate [1], [2]. In order to amplify the d31-contraction of the piezoelectric substrate by more than a factor of 20 the beam has a length of l=10 mm. The beam and the structures to fix the fibers are electroplated with Ni. The height h of the beam is 100–200 μm and the width b is 20 μm. The beam is released from the substrate and is only connected

Characterization of micro-electro-mechanical systems

The lack of advanced diagnostic tools for dynamic processes is a big obstacle during the development process and has caused long development periods for MEMS [3]. This obstacle has been noticed and several research facilities for MEMS have started to develop optical measurement methods for static and dynamic behaviors in microdevices [4], [5], [6], [7], [8]. We have shown high-speed cine photomicrography as a powerful diagnostic tool for dynamic processes in MEMS [9], [10]. This visualization

Setup for pseudo-cinematography

A stroboscope light with a flash duration of 250 ns full width at half maximum is attached to a microscope to illuminate the object. The flash duration determines the time resolution of the data capture. The illuminated object is detected with a commercial CCD camera that generates an RGB image with a 756×581 pixel resolution. After receiving the video image, a delay generator changes the phase of the trigger in order to increase the time between the starting point of the process and the flash.

Ultra-high-speed motion analyzer

In most cases nonreproducible transient phenomena are critical to MEMS reliability and cannot be visualized by using stroboscopy. Therefore, we have developed an experimental setup to study dynamics in MEMS with high-speed visualization [10].

The uncertainty of the position data extracted from the image data is approximately the optical resolution of the microscope. Our test rig has a spatial resolution d ofd=0.61λNA≈500nm,where λ is the wavelength of the light and NA is the numerical aperture

Electrical drive

To apply a step function to the micro-optical-switch, high currents have to be supplied for a short moment because of the capacitance CS of the piezoelectric substrate. A commercial high voltage source, LNC 1200-50 from Heinzinger, and a fast drive circuit are used to apply the step function as shown in Fig. 4. The resistance R=1 kΩ in the circuit limits the current to maximal 1 A. Measurements show that a voltage of 1000 V is applied within 1 μs after the trigger signal is applied to the

Pseudo-cinematography of the shutter movement

To determine if the step response of the microswitch shutter is reproducible we have investigated the movement of the shutter with pseudo-cinematography. The result of this investigation for a voltage step with an amplitude of 1000 V is shown in Fig. 5. The measurement demonstrates that the shutter reaches its final position approximately 1 ms after applying the voltage at the substrate. During switching the shutter shows a strong nonreproducible dynamic behavior. By investigating the shutter

Real high-speed cinematography of the shutter movement

Fig. 6a and b demonstrates two visualization sequences of the actuator step response carried out under identical laboratory conditions.

It can be seen clearly that the movements differ dramatically. In Fig. 6a the shutter needs 750 μs to reach its final position and oscillates smoothly after switching, whereas in Fig. 6b the shutter hits the structure, which fixes the fiber. After hitting the structure, the shutter moves backwards. From the pseudo-cinematographic measurements we had concluded

Dynamic behavior of the substrate

To understand the coupling of the beam dynamics with the dynamics of the piezoelectric substrate we have studied the motion of the substrate with and without the switch. We determined that the step response of the substrate is reproducible. Therefore, pseudo-cinematography was used to study the dynamic behavior of the substrate. Cinematographic investigations of the movement of the piezoelectric substrate demonstrate that the contact pads of the beam have been moved together after a few

Finite element analysis

We have calculated the transient behavior of the beam with the commercial finite element software ansys. The contraction at the contact pads has been applied within 6 μs and only one contact pad has been moved in the simulation. The simulation should clarify if the asymmetric deformation of the beam in Fig. 8 results from the inertia of the beam, because we were not able to define the fixed point of the substrate during operation. In the simulation this fixed point is one end of the beam, which

Theoretical limit of actuator operation

The triangle shape given by beam design is almost a straight line because the angle between the two arms is only 0.002 rad. Therefore, the triangle shape of the beam is neglected in the following discussion. In addition we consider only the dynamic cases where the beam stays in its equilibrium transverse displacements and linear beam theory is applicable. Then, the total displacement w(x,t) of the beam along x is given by [13]w(x,t)=u(x,t)+v(x)=(u(t)+v0)121−cos2πxl,where the harmonic function

Improvement of the beam dynamics

The actuator cannot be driven faster to guarantee reliable behavior. For faster operation of the micro-optical-switch the design has to be changed; especially the stiffness of the beam should be improved by increasing width b of the beam. For example, a width b=100 μm leads to a maximum frequency of the driving signal of f=1095 Hz. The resulting higher mass and stiffness of the beam should not cause a problem because of the insignificant influence of the beam to the movement of the

Conclusions and outlook

But even for a 1 mm wide beam, the frequency limit is defined by the beam and not by the piezoelectric substrate. In addition, it is not ensured that the influence of the beam to the substrate dynamics can be neglected for this beam width. Due to the mask layout the beam consists of two unbowed parts (triangle shape) with a short third unbowed beam that connects both long parts. Therefore, an idea to improve the operation is to design the beam at the two connections of the three unbowed parts

References (15)

  • H. Debeda et al.

    Development of miniaturized piezoelectric actuators for optical applications using LIGA technology

    J Microelectromech Syst

    (1999)
  • Debeda H, Wallrabe U, Schulz J, Mohr J, Rembe C. Fully batch fabricated LIGA actuators integrated on piezoelectric...
  • P. Egelhaaf et al.

    Mikrosystemtechnik: Von der Forschung zu innovativen Produkten

    Physikalische Blätter

    (1999)
  • P. Krehl et al.

    High-speed visualization – a useful diagnostic tool for the development and testing of microactuators – retrospect and prospect

    Microsyst Technol

    (1999)
  • Q. Davis et al.

    Using a light microscope with nanometer accuracy

    Opt Eng

    (1998)
  • Freeman D, Aranyosi A, Gordon M, Hong S. Multidimensional motion analysis of MEMS using computer microvision....
  • M. Hart et al.

    Time-resolved measurement of optical MEMS using stroboscopic interferometry

    Transducers

    (1999)
There are more references available in the full text version of this article.

Cited by (4)

1

Tel.: +49-731-50-26300; fax: +49-731-50-26301.

2

Tel.: +49-7247-82-5522; fax: +49 7247-82-3928.

View full text
Copyright © 2001 Elsevier Science Ltd. All rights reserved.