The present invention claims priority to U.S. Provisional Application Ser. No. 60/363,139 commonly owned with the application.
Among the applications for micromechanical devices are planar actuators comprising one or typically an array of actuators in a two-dimensional matrix in which individual elements of the array need to be individually and rapidly displaced.
In one application of such actuators to the deflection of radiation, an array of actuators includes mirrors activated by micromechanical electrostatic motivators to provide rapid displacement of the mirror positions in the array in order to alter the phase delay of incoming radiation wavefronts and thereby adjust the phase of the reflected light or the angle of reflection.
In modern high-speed systems such as scanners and pattern recognition systems, the demands for rapid adjustment of the phase of reflected light or beam angle continue to increase, placing severe demands upon control circuitry for an array of large dimensions to precisely and individually control each of hundreds or thousands of mirror elements in the array.
An additional problem encountered in controlling such actuators is the nonlinearity between displacement and applied control voltage due to the mathematical relationship between displacement and applied potential in what is essentially a parallel-plate capacitor geometry. To deal with the nonlinearity in order to provide accurate beam reflection, a heavy demand is placed upon processing electronics to accomplish any adjustment to the hundreds or thousands of individual actuators for controlling the position on an ongoing, rapid sequence basis.
The present invention provides for an effective way of linearizing the response between desired position and applied potential that takes advantage of and places structured designs directly into the micromechanical structure. It is operative directly in response to digital signals, avoiding the complexity and delay of looped processing to accomplish the mathematical linearization.
According to the present invention, at least one electrode of each of the parallel plate actuator elements is divided into a plurality of electrode segments of varying area, from a minimum first area to a maximum, nth area of greatest value, through a plurality of areas increasing from one to the other by a factor of two. Each of the electrode segments is individually addressed through voltage gates that are controlled by binary ones and zeroes directly representative of the applied voltage potential. The resulting nonlinear transfer function that relates the applied potential to the effective displacement caused by the applied potential counteracts some or all of the nonlinearity in the relationship or transfer function between applied potential and actuator displacement.
The nonlinearity in the transfer characteristic between applied potential and actuator displacement is completely eliminated by adjusting the applied voltage that is applied to each of the electrode segments through the use of a plurality of current sources each of magnitude varying from a low minimum first magnitude corresponding to and activated with activation of the first electrode segment and varying, one to the other, by factor of two up to the largest or nth current source, corresponding to and activated simultaneously with activation of the largest or nth electrode segment.
The resulting system utilizes a simple structure not requiring time-consuming electronic processing of the applied potential in order to produce a linear relationship between displacement and input value, but nevertheless linearizing its relationship with respect to the displacement value it causes.
The present invention is more fully understood with respect to the drawings of which:
A reflective surface 22 is fastened above the second electrodes 16 via posts 24 in both versions of
The force created by the voltage source 28, shown only for purposes of illustration of operation of such activators, produces along illustrated electrostatic field lines 32 a force between the electrodes 14 and 16. As the voltage increases, the force increases, and the displacement in the direction X of the electrode 16 toward the electrode 32 increases. The relationship for such a structure is illustrated by the graph of
In order to change or steer the input beam 40 onto different trajectories for output beams 42 it is essential that a progressive change in displacement occur over the entire array of mirrors 22. To accomplish that accurately with the transfer function illustrated in
As illustrated in
The effect of the linearization achieved through the use of switches 80-84 is to linearize the transfer function illustrated in
To achieve a linearization corresponding to the curve 92 an adjustment in the reference voltage corresponding to the same digital word is provided by applying a varying load in the form of current sources 100, 102 and 104, connected via switches 101, 103, and 105 to a junction point 106 common with the application of voltage Vo from a source 108 through a resistor 110. The sources 100, 102, 104 are connected to the common junction point 106 by switches 101, 103, and 103 controlled through the same digital lines controlling the switches 80-84. The magnitude of the current of each current source increases according to a series from a first current of lowest value corresponding to the LSB and the smallest electrode segment, increasing by a factor of two from current source to current source to the largest, nth current source corresponding to the MSB and largest area electrode segment. In this manner, a total linearization of the transfer function as illustrated in curve 92 can be achieved.
The mathematics corresponding to this linearization operate as follows:
Mechanical restoring force for a given displacement of the actuator electrode:
FM=−kx (1)
where k is a mechanical spring constant and x the displacement.
Electrostatic force for a given applied voltage:
where g is the spacing between electrodes 14 and 16 at zero applied voltage, ATOT their area of overlap as seen from a view perpendicular to the surface, and ε a physical constant called the permittivity.
Equilibrium occurs when FM+FE=0:
Define a constant C=2k/ε, so that the above becomes:
Cx(g−x)2=ATOTV2 (4)
In one application of Eq. (4), one can keep the voltage V constant and adjust the area ATOT by activating only some subset of the electrode segments An. In such a case, solving Eq. (4) for the required ATOT as a function of desired displacement x results in
This relationship is nonlinear and is undesirable for the reasons described previously. A desirable condition is one in which the displacement x is linearly proportional to the activation area ATOT, i.e., dx/dA, and therefore dA/dx, are constant. Taking the derivative of Eq. (5) with respect to x leads to
Taking this equation's reciprocal results in:
One can then impose the additional constraint that V also be a function of the desired displacement x. Specifically, let V(x)=Vo(g−x)/g, where Vo is the value of V at zero displacement, and where V is reduced as x approaches in value that of the gap spacing, g. The displacement equation, now a function of both area ATOT and applied voltage V(x), becomes:
The displacement x then becomes
The above-described preferred embodiment is intended as exemplary only, the scope of the invention being described and limited only as shown in the following claims.
This invention was made with Government Support under Contract Number W-7405-ENG-48 awarded by the Department of Energy. The Government has certain rights in the invention.
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/US03/07123 | 3/10/2003 | WO | 00 | 9/3/2004 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO03/077286 | 9/18/2003 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
6215579 | Bloom et al. | Apr 2001 | B1 |
6498870 | Wu et al. | Dec 2002 | B1 |
20040058469 | Kowarz | Mar 2004 | A1 |
Number | Date | Country | |
---|---|---|---|
20050152014 A1 | Jul 2005 | US |