WO2015035120A1 - Internally generated dft stepped hysteresis sweep for electrostatic mems - Google Patents

Internally generated dft stepped hysteresis sweep for electrostatic mems Download PDF

Info

Publication number
WO2015035120A1
WO2015035120A1 PCT/US2014/054213 US2014054213W WO2015035120A1 WO 2015035120 A1 WO2015035120 A1 WO 2015035120A1 US 2014054213 W US2014054213 W US 2014054213W WO 2015035120 A1 WO2015035120 A1 WO 2015035120A1
Authority
WO
WIPO (PCT)
Prior art keywords
pull
mems
voltage
electrodes
coupled
Prior art date
Application number
PCT/US2014/054213
Other languages
French (fr)
Inventor
James Douglas HUFFMAN
Cong Quoc Khieu
Robertus Petrus Van Kampen
Karl F. Smayling
Vikram Joshi
Original Assignee
Cavendish Kinetics, Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Cavendish Kinetics, Inc filed Critical Cavendish Kinetics, Inc
Publication of WO2015035120A1 publication Critical patent/WO2015035120A1/en
Priority to CN201580001969.3A priority Critical patent/CN105593158B/en
Priority to EP15720864.6A priority patent/EP3189005A1/en
Priority to US14/916,884 priority patent/US10029914B2/en
Priority to PCT/US2015/027201 priority patent/WO2016036422A1/en

Links

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B81MICROSTRUCTURAL TECHNOLOGY
    • B81CPROCESSES OR APPARATUS SPECIALLY ADAPTED FOR THE MANUFACTURE OR TREATMENT OF MICROSTRUCTURAL DEVICES OR SYSTEMS
    • B81C99/00Subject matter not provided for in other groups of this subclass
    • B81C99/003Characterising MEMS devices, e.g. measuring and identifying electrical or mechanical constants
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B81MICROSTRUCTURAL TECHNOLOGY
    • B81BMICROSTRUCTURAL DEVICES OR SYSTEMS, e.g. MICROMECHANICAL DEVICES
    • B81B2201/00Specific applications of microelectromechanical systems
    • B81B2201/01Switches
    • B81B2201/012Switches characterised by the shape
    • B81B2201/016Switches characterised by the shape having a bridge fixed on two ends and connected to one or more dimples
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B81MICROSTRUCTURAL TECHNOLOGY
    • B81BMICROSTRUCTURAL DEVICES OR SYSTEMS, e.g. MICROMECHANICAL DEVICES
    • B81B2207/00Microstructural systems or auxiliary parts thereof
    • B81B2207/01Microstructural systems or auxiliary parts thereof comprising a micromechanical device connected to control or processing electronics, i.e. Smart-MEMS
    • B81B2207/015Microstructural systems or auxiliary parts thereof comprising a micromechanical device connected to control or processing electronics, i.e. Smart-MEMS the micromechanical device and the control or processing electronics being integrated on the same substrate
    • HELECTRICITY
    • H01ELECTRIC ELEMENTS
    • H01GCAPACITORS; CAPACITORS, RECTIFIERS, DETECTORS, SWITCHING DEVICES, LIGHT-SENSITIVE OR TEMPERATURE-SENSITIVE DEVICES OF THE ELECTROLYTIC TYPE
    • H01G5/00Capacitors in which the capacitance is varied by mechanical means, e.g. by turning a shaft; Processes of their manufacture
    • H01G5/16Capacitors in which the capacitance is varied by mechanical means, e.g. by turning a shaft; Processes of their manufacture using variation of distance between electrodes

Definitions

  • Embodiments of the present invention generally relate to a mechanism for testing a micro-electromechanical system (MEMS) hysteresis.
  • MEMS micro-electromechanical system
  • a plate moves between a first position and a second position.
  • the plate moves by applying a voltage to an actuation electrode. Once the electrode voltage reaches a certain voltage oftentimes referred to as a snap-in voltage, the plate moves towards the electrode. The plate moves back to the original position once the voltage is lowered to a release voltage.
  • the release voltage is typically lower than the snap-in voltage due to the higher electrostatic forces when the plate is close to the actuation electrode and due to stiction between the plate and the surface to which the plate is in contact once moved closer to the electrode.
  • the MEMS DVC has a hysteresis curve.
  • the snap-in voltage and the release voltage should be known for the MEMS DVC to operate efficiently.
  • the present invention generally relates to a mechanism for testing a MEMS hysteresis.
  • a power management circuit may be coupled to the electrodes that cause the movable plate to move between the electrodes in a MEMS device.
  • the power management circuit may utilize a charge pump, a comparator and a resistor ladder.
  • a device comprises a first MEMS device having a first electrode, a second electrode, and a plate movable between a first position spaced a first distance from the first electrode and a second position spaced a second distance from the first electrode; a power source coupled to both the first electrode and the second electrode; an ammeter coupled to the first electrode; a voltmeter coupled to both the first electrode and the second electrode; a first switch coupled to the plate and to ground; and a second switch coupled to the plate and to a power management circuit.
  • a MEMS DVC comprises at least one MEMS device, the MEMS device comprising a movable plate, an RF electrode, one or more pull-down electrodes and one or more pull-up electrodes; a first switch coupled to either the one or more pull-down electrodes or the one or more pull-up electrodes, wherein the first switch is additionally coupled to ground; a second switch coupled to either the one or more pull-down electrodes or the one or more pull-up electrodes; and a power management system coupled to the second switch, wherein the at least one MEMS device, the first switch, the second switch and the power management system are all disposed on a semiconductor chip.
  • DVC including at least one MEMS device, the MEMS device comprising a movable plate, an RF electrode, one or more pull-down electrodes and one or more pull-up electrodes, is disclosed.
  • the method comprises applying a first voltage to either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring a capacitance of the MEMS device; applying a second voltage to either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring the capacitance of the MEMS device; detecting the capacitance of the MEMS device equals a maximum capacitance of the MEMS device; removing the second voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring the capacitance of the MEMS device; removing the first voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring the capacitance of the MEMS device; and detecting the capacitance of the MEMS device is less than the maximum capacitance.
  • Figure 1 is a schematic cross-sectional illustration of a MEMS DVC device in the free standing state.
  • Figure 2 is a schematic cross-sectional illustration of the MEMS DVC device in the C ma x state.
  • Figure 3 is a schematic cross-sectional illustration of the MEMS DVC device in the C m in state.
  • Figure 4 is a schematic illustration of a waveform controller driving the MEMS DVC device.
  • Figure 5 is a graph showing the hysteresis curve for an electrostatically operated MEMS device.
  • Figure 6 is a schematic illustration of a two terminal MEMS device CV configuration.
  • Figure 7 is a schematic illustration of a three terminal MEMS device CV configuration.
  • Figure 8 is a schematic illustration of a DFT implementation to test a MEMS hysteresis according to one embodiment.
  • Figure 9 is a schematic illustration of a power management implementation for MEMS hysteresis testing according to one embodiment.
  • Figure 10 is a schematic illustration of test methodology for a discrete capacitance hysteresis test using an internal DFT according to one embodiment.
  • Figures 1 1A and 1 1 B are schematic illustrations of a DFT implementation to test a MEMS hysteresis according to additional embodiments.
  • Figure 12 is a flow chart illustrating a method of testing a MEMS DVC according to one embodiment.
  • the present invention generally relates to a mechanism for testing a MEMS hysteresis.
  • a power management circuit may be coupled to the electrodes that cause the movable plate that is disposed between the electrodes to move in a MEMS device.
  • the power management circuit may utilize a charge pump, a comparator and a resistor ladder.
  • a MEMS DVC device may operate with electrostatic forces.
  • the mechanism operated by a force acting on the moveable MEMS element when a voltage V is applied between the movable MEMS element [e.g., movable plate) and a control electrode.
  • This electrostatic force scales with (V/gap) 2 .
  • the mechanical counter-balance force comes from a spring suspension system and typically scales linearly with the displacement. The result is that with an increasing voltage V the MEMS device moves a certain distance ⁇ toward the control-electrode. This movement reduces the gap between the movable MEMS element (oftentimes referred to as a moveable plate) and the electrode, which in turn increases the electrostatic force further.
  • the device displacement is such that the electrostatic force rises faster than the mechanical counterbalance force and the device rapidly snaps-in towards the control-electrode until it comes in contact.
  • the MEMS DVC device may have a control-electrode above (PU- electrode) and below (PD-electrode) the moveable MEMS element, as shown schematically in Figure 1 .
  • an RF-electrode may be present below the moveable MEMS element.
  • the PU-electrode, the PD- electrode and the RF electrode are all covered with dielectric material.
  • the MEMS element is either pulled-up or pulled-down in contact with the dielectric material to provide a stable minimum or maximum capacitance to the RF- electrode.
  • the capacitance from the moveable element to the RF- electrode (which resides below the moveable element) can be varied from a high capacitance C ma x when pulled to the bottom (See Figure 2) to a low capacitance C m in (See Figure 3) when pulled to the top.
  • the voltages applied to the PD-electrode (Vbottom) and to the top-electrode (Vtop) are typically controlled by a waveform controller (See Figure 4) to ensure a long-life stable performance of the DVC device.
  • the moveable element is typically on DC-ground.
  • the MEMS DVC device may comprise a movable plate disposed in a cavity.
  • the movable plate is coupled to ground and moves between a free standing state shown in Figure 1 to a C ma x state shown in Figure 2 and a C m in state shown in Figure 3.
  • a voltage may be applied to one or more pull-in or pull-down electrodes to pull the plate into close proximity of the RF electrode.
  • the electrodes are covered by a dielectric material.
  • a pull up or pull off electrode may be disposed opposite the pull-in electrodes.
  • Figure 5 shows a typical response of the MEMS DVC device to an applied control voltage to the PD-electrode.
  • the device is in the free-standing state as in Figure 1 and has a capacitance Cf ree -
  • the capacitance slowly increases as the movable plate slowly moves closer to the RF electrode until the snap-in point p1 is reached at a voltage Vpi (pull-in voltage).
  • the device i.e., movable plate
  • the capacitance goes to its maximum value C ma x-
  • the electrostatic force has increased and the voltage has to be reduced down to Vrl (release voltage) in order for the MEMS device to release from the bottom at point p2.
  • the capacitance of the MEMS device is at the maximum value between the MEMS element is in contact with the dielectric material that is disposed on the RF electrode.
  • Vpi and Vrl are important parameters for the MEMS DVC device; both for upward and downward actuation. If the pull-in voltage Vpi is too high then the waveform controller may not be able to pull the MEMS devices into contact intimately, which can impact the obtainable C m in (upward actuation) or C ma x (downward actuation). If the release voltage Vrl is too low this could indicate stiction which impedes proper device operation. Also, if the release voltage Vrl from the bottom is too low then this will impede the device to be released from the RF- electrode in the presence of an RF signal.
  • Vpi and Vrl depend on material parameters (Young's Modulus) as well as geometrical parameters, such as layer thicknesses and CD-control of various layers. Therefore, in production, the MEMS devices will exhibit a certain distribution in both Vpi and Vrl. In order to screen functional devices that meet all required product specs, it is key to test the Vpi and Vrl on every device. As discussed herein, a built-in test methodology can be used facilitate the test.
  • Hysteresis testing because the Vpi and Vrl are separated or the pull-in and the release curves do not overlap as shown in Figure 5.
  • Vpi and Vrl are designed to be in a certain range. Otherwise they can result in non performance as explained in above paragraphs.
  • C max and C min Vpi and Vrl are not product specs, i.e. they are not listed on a product sheet but they are the best gauges for estimating the reliability or robustness of the part. Due to process variations, certain parts on a wafer or across the lot may fall outside the range and, if escape screening, can lead to failures in the field.
  • a typical method for performing a hysteresis test on an electrostatic MEMS device is to perform a CV (capacitance - voltage) sweep.
  • a typical CV sweep can be performed using a CV meter, which uses a combination of a DC source and an AC source to provide the DC bias and the AC signal. The measurements are performed by a combination of an AC voltmeter and an AC ammeter.
  • the basic test configuration to perform a test on a two terminal electrostatic MEMS device is shown in Figure 6.
  • CMEMS iac/2nf * v ac
  • a three or more terminal electrostatic MEMS device does not allow for the same straightforward CV test as shown in Figure 1 .
  • the bias electrodes on the MEMS device provide the actuation bias and are separate from the capacitor electrodes.
  • a CV sweep is performed on this configuration by using the same configuration as shown in Figure 1 , but also including a DC source, Vbias, as shown in Figure 7.
  • the DC source is used in the same manner as the DC source in Figure 6.
  • a semiconductor chip can be composed of one or more MEMS transducers and monolithically integrated CMOS control and power management circuitry. This allows for the Vbias power supply in Figure 7 to be generated internal to the semiconductor chip and not in an external power supply.
  • the Vbias voltage is generated in the integrated power management circuit and passes to the MEMS transducers through switch S2.
  • the actuation voltage for the MEMS transducers is controlled through the power management, where the level of the output voltage is controlled by the digital control bits C ⁇ 0:n>, as shown in Figure 8.
  • the hysteresis sweep can be performed by changing the digital control bits in the power management circuit to the desired actuation voltage.
  • a primary difference between the external power supply version in Figure 7 and the internal DFT mode in Figure 8 is that the digital control bits hold the value at a discrete number of fixed levels (n) instead of a continuous sweep that can be performed by the external version.
  • One representation of a power management circuit that can allow for discrete levels of output voltage is a charge pump with a regulator.
  • the charge pump clock is gated by the output of the comparator.
  • the output voltage of the charge pump is divided by the resistor ladder and compared to the bandgap voltage reference. If the voltage reference at the resistor ladder is lower than the value of the bandgap voltage reference voltage, the charge pump clock is on. This condition will allow the charge pump clock to be toggling and the charge pump voltage will be increasing if the charge generation is greater than the output load current. If the voltage reference at the resistor ladder is higher than the value of the bandgap voltage reference, the charge pump clock is off.
  • the programming of the voltage level is produced by switching in discrete resistors in the resistor ladder, effectively changing the voltage on the compare node to produce a higher or lower output voltage set point.
  • the value of the output voltage is programmed by the address bits C ⁇ 0:n>.
  • Vactuation will be programmed to be a resistor ratio as compared to the Vbandgap voltage as shown by the following equation for a programmation of c ⁇ 0>:
  • Vactuation ((Rs+R0)/R0) * Vbandgap
  • Vactuation ((Rs+R0+R1 )/(R0+R1 )) * Vbandgap
  • Vactuation ((Rs+R0+R1 +R2...+Rn)/(R0+R1 +R2...+Rn)) * Vbandgap [0039]
  • the test method consists of a voltage programmation using the C address bits, a wait time for settling, and a measurement strobe of the capacitance.
  • the test is implemented in the hardware configuration shown in Figure 10.
  • the device under test, or DUT is preset using the address bits to the regulator to output a voltage level to the MEMS that is lower than the Vpi.
  • a wait time, or Tw for voltage and MEMS settling is implemented in the test sequence before the capacitance level is measured by the CV meter at time Ts.
  • the address bits are incremented to the next voltage level and the measurement is performed using the same timing.
  • the address bits are decremented and the measurements taken until the capacitance meter detects Vrl.
  • Figures 1 1A and 1 1 B are schematic illustrations of a DFT implementation to test a MEMS hysteresis according to additional embodiments.
  • Figure 1 1A shows an embodiment where the test is performed for voltage applied to the pull-down electrode 1 102 while
  • Figure 1 B shows an embodiment where the test is performed for voltage applied to the pull-up electrode 1 104. It is contemplated that the test may be performed on both the pull-down electrode 1 102 and pull-up electrode 1 104.
  • the MEMS device 1 100 includes the pull-down electrodes 1 102, the pull- up electrode 1 104, an RF electrode 1 106 and ground electrodes 1 108.
  • the ground electrodes 1 108 are connected to ground and to the movable plate 1 1 10.
  • a dielectric layer 1 1 12 is disposed over the pull-down electrodes 1 102 and the RF electrode 1 106.
  • Another dielectric layer 1 1 14 is disposed between the pull-up electrode 1 104 and the cavity 1 1 16 within which the movable plate 1 1 10 is disposed.
  • the pull-down electrodes 1 102 are coupled to multiple switches 1 1 18, 1 120.
  • the pull-up electrode 1 104 is connected to ground.
  • the second switch 1 120 in Figure 1 1 A, is connected to a power management device 1 122.
  • the second switch 1 120 is connected to the power management device 1 122.
  • the pull-down electrode 1 102 ( Figure 1 1 A) or the pull-up electrode 1 104 ( Figure 1 1 B) is connected to the power management device 1 122.
  • the power management device 1 122 and the MEMS device 1 100 are both disposed on the semiconductor chip represented by box 1 124.
  • the power management device 1 122 includes a charge pump 1 128 that is coupled to a gate 1 130.
  • the gate 1 130 is coupled to both the Vclock node and the output from a comparator 1 132.
  • the comparator has inputs from the Vbandgap node and the resistive ladder.
  • the resistive ladder is what divides the output of the charge pump 1 128.
  • the resistive ladder includes a plurality of resistors R0... Rn which are coupled together in series.
  • the address bits c ⁇ 0>, c ⁇ 1 >, c ⁇ n> are programmed to incrementally "actuate” or operate such that the next voltage level is achieved and the capacitance of the MEMS device 1 100 is measured.
  • an incremental voltage is applied by operating the address bits c ⁇ 0>, c ⁇ 1 >, c ⁇ n>. Based upon the incremental voltage increase, the Vpi (for Figure 1 1 A) or Vrl (for Figure 1 1 B) is determined. Similarly, by decrementally decreasing the voltage (i.e., operating the address bits c ⁇ 0>, c ⁇ 1 >, c ⁇ n>), the capacitance is again measured the Vrl (for Figure 1 1 A) or Vpi (for Figure 1 1 B) is detected. As such, the hysteresis curve for the particular MEMS device 1 100 is determined. It is to be understood that multiple MEMS devices may be coupled to the power management device 1 122. The multiple MEMS devices may collectively operate as a DVC.
  • Figure 12 is a flow chart 1200 illustrating a method of testing a MEMS
  • a low voltage is applied to either the pull-down electrode or the pull-up electrode in step 1202.
  • the voltage applied to the pull-down electrode is to pull the movable plate closer to the RF electrode.
  • Voltage applied to the pull-up electrode is to push the movable plate closer to the RF electrode.
  • the capacitance of the MEMS device is measured in step 1204. If the capacitance is equal to the maximum capacitance for the MEMS device, then the pull-in (or push-in for voltage applied to the pull-up electrode) has been determined in step 1206. If the measured capacitance is not equal to the maximum capacitance, then the voltage is incrementally increased in step 1208, with the capacitance measured with each incremental voltage increase in step 1204, until the maximum capacitance is reached and the pull-in voltage has been determined in step 1210.
  • the release voltage is determined.
  • the release voltage is determined by reducing the voltage applied to either the pull-down electrode or the pull-up electrode in step 1212.
  • the capacitance is then measured in step 1214. If the capacitance is less than the maximum capacitance, then the release voltage has been determined. If, however, the measured capacitance is equal to the maximum capacitance, then the voltage is decrementally reduced in step 1218. The capacitance is measured for each decremental voltage reduction. If the measured capacitance is less than the maximum capacitance in step 1216, then the release voltage has been determined in step 1220.
  • a first voltage is applied to either the one or more pull-down electrodes or the one or more pull-up electrodes.
  • the capacitance of the MEMS device is measured.
  • a second voltage is then applied to either the one or more pull-down electrodes or the one or more pull-up electrodes.
  • the capacitance is again measured.
  • the capacitance of the MEMS device is detected to equal the maximum capacitance for the MEMS device.
  • the second voltage is then removed from either the one or more pull-down electrodes or the one or more pull-up electrodes.
  • the capacitance of the MEMS device is then measured.
  • the first voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes is removed.
  • the capacitance is again measured and if the measured capacitance is less than the maximum capacitance for the MEMS device, the release voltage has been determined.

Landscapes

  • Engineering & Computer Science (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Micromachines (AREA)

Abstract

The present invention generally relates to a mechanism for testing a MEMS hysteresis. A power management circuit may be coupled to the electrodes that cause the movable plate that is disposed between the electrodes in a MEMS device to move. The power management circuit may utilize a charge pump, a comparator and a resistor ladder.

Description

INTERNALLY GENERATED DFT STEPPED HYSTERESIS SWEEP FOR
ELECTROSTATIC MEMS
BACKGROUND OF THE INVENTION
Field of the Invention
[0001] Embodiments of the present invention generally relate to a mechanism for testing a micro-electromechanical system (MEMS) hysteresis.
Description of the Related Art
[0002] In operating a MEMS digital variable capacitor (DVC), a plate moves between a first position and a second position. The plate moves by applying a voltage to an actuation electrode. Once the electrode voltage reaches a certain voltage oftentimes referred to as a snap-in voltage, the plate moves towards the electrode. The plate moves back to the original position once the voltage is lowered to a release voltage. The release voltage is typically lower than the snap-in voltage due to the higher electrostatic forces when the plate is close to the actuation electrode and due to stiction between the plate and the surface to which the plate is in contact once moved closer to the electrode.
[0003] Because the plate doesn't release at the same voltage as the snap-in voltage, the MEMS DVC has a hysteresis curve. The snap-in voltage and the release voltage, while different, should be known for the MEMS DVC to operate efficiently.
[0004] Therefore, there is a need in the art for a method and device for effectively measuring the hysteresis curve for a MEMS DVC.
SUMMARY OF THE INVENTION
[0005] The present invention generally relates to a mechanism for testing a MEMS hysteresis. A power management circuit may be coupled to the electrodes that cause the movable plate to move between the electrodes in a MEMS device. The power management circuit may utilize a charge pump, a comparator and a resistor ladder. [0006] In one embodiment, a device comprises a first MEMS device having a first electrode, a second electrode, and a plate movable between a first position spaced a first distance from the first electrode and a second position spaced a second distance from the first electrode; a power source coupled to both the first electrode and the second electrode; an ammeter coupled to the first electrode; a voltmeter coupled to both the first electrode and the second electrode; a first switch coupled to the plate and to ground; and a second switch coupled to the plate and to a power management circuit.
[0007] In another embodiment, a MEMS DVC comprises at least one MEMS device, the MEMS device comprising a movable plate, an RF electrode, one or more pull-down electrodes and one or more pull-up electrodes; a first switch coupled to either the one or more pull-down electrodes or the one or more pull-up electrodes, wherein the first switch is additionally coupled to ground; a second switch coupled to either the one or more pull-down electrodes or the one or more pull-up electrodes; and a power management system coupled to the second switch, wherein the at least one MEMS device, the first switch, the second switch and the power management system are all disposed on a semiconductor chip.
[0008] In another embodiment, a method of testing a MEMS DVC, the MEMS
DVC including at least one MEMS device, the MEMS device comprising a movable plate, an RF electrode, one or more pull-down electrodes and one or more pull-up electrodes, is disclosed. The method comprises applying a first voltage to either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring a capacitance of the MEMS device; applying a second voltage to either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring the capacitance of the MEMS device; detecting the capacitance of the MEMS device equals a maximum capacitance of the MEMS device; removing the second voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring the capacitance of the MEMS device; removing the first voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes; measuring the capacitance of the MEMS device; and detecting the capacitance of the MEMS device is less than the maximum capacitance. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] So that the manner in which the above recited features of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical embodiments of this invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective embodiments.
[0010] Figure 1 is a schematic cross-sectional illustration of a MEMS DVC device in the free standing state.
[0011] Figure 2 is a schematic cross-sectional illustration of the MEMS DVC device in the Cmax state.
[0012] Figure 3 is a schematic cross-sectional illustration of the MEMS DVC device in the Cmin state.
[0013] Figure 4 is a schematic illustration of a waveform controller driving the MEMS DVC device.
[0014] Figure 5 is a graph showing the hysteresis curve for an electrostatically operated MEMS device.
[0015] Figure 6 is a schematic illustration of a two terminal MEMS device CV configuration.
[0016] Figure 7 is a schematic illustration of a three terminal MEMS device CV configuration.
[0017] Figure 8 is a schematic illustration of a DFT implementation to test a MEMS hysteresis according to one embodiment.
[0018] Figure 9 is a schematic illustration of a power management implementation for MEMS hysteresis testing according to one embodiment. [0019] Figure 10 is a schematic illustration of test methodology for a discrete capacitance hysteresis test using an internal DFT according to one embodiment.
[0020] Figures 1 1A and 1 1 B are schematic illustrations of a DFT implementation to test a MEMS hysteresis according to additional embodiments.
[0021] Figure 12 is a flow chart illustrating a method of testing a MEMS DVC according to one embodiment.
[0022] To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements disclosed in one embodiment may be beneficially utilized on other embodiments without specific recitation.
DETAILED DESCRIPTION
[0023] The present invention generally relates to a mechanism for testing a MEMS hysteresis. A power management circuit may be coupled to the electrodes that cause the movable plate that is disposed between the electrodes to move in a MEMS device. The power management circuit may utilize a charge pump, a comparator and a resistor ladder.
[0024] A MEMS DVC device may operate with electrostatic forces. As discussed herein, the mechanism operated by a force acting on the moveable MEMS element when a voltage V is applied between the movable MEMS element [e.g., movable plate) and a control electrode. This electrostatic force scales with (V/gap)2. The mechanical counter-balance force comes from a spring suspension system and typically scales linearly with the displacement. The result is that with an increasing voltage V the MEMS device moves a certain distance δ toward the control-electrode. This movement reduces the gap between the movable MEMS element (oftentimes referred to as a moveable plate) and the electrode, which in turn increases the electrostatic force further. For small voltages, an equilibrium position between the initial position and the electrode is found. However, when the voltage exceeds a certain threshold level (the pull-in voltage), the device displacement is such that the electrostatic force rises faster than the mechanical counterbalance force and the device rapidly snaps-in towards the control-electrode until it comes in contact.
[0025] The MEMS DVC device may have a control-electrode above (PU- electrode) and below (PD-electrode) the moveable MEMS element, as shown schematically in Figure 1 . In addition an RF-electrode may be present below the moveable MEMS element. As shown in Figures 1 -3, the PU-electrode, the PD- electrode and the RF electrode are all covered with dielectric material. During operation the MEMS element is either pulled-up or pulled-down in contact with the dielectric material to provide a stable minimum or maximum capacitance to the RF- electrode. In this way the capacitance from the moveable element to the RF- electrode (which resides below the moveable element) can be varied from a high capacitance Cmax when pulled to the bottom (See Figure 2) to a low capacitance Cmin (See Figure 3) when pulled to the top. The voltages applied to the PD-electrode (Vbottom) and to the top-electrode (Vtop) are typically controlled by a waveform controller (See Figure 4) to ensure a long-life stable performance of the DVC device. The moveable element is typically on DC-ground.
[0026] As shown in Figure 1 -3, the MEMS DVC device may comprise a movable plate disposed in a cavity. The movable plate is coupled to ground and moves between a free standing state shown in Figure 1 to a Cmax state shown in Figure 2 and a Cmin state shown in Figure 3. A voltage may be applied to one or more pull-in or pull-down electrodes to pull the plate into close proximity of the RF electrode. The electrodes are covered by a dielectric material. A pull up or pull off electrode may be disposed opposite the pull-in electrodes.
[0027] Figure 5 shows a typical response of the MEMS DVC device to an applied control voltage to the PD-electrode. Initially, the device is in the free-standing state as in Figure 1 and has a capacitance Cfree- As the voltage on the bottom control electrode is ramped up, the capacitance slowly increases as the movable plate slowly moves closer to the RF electrode until the snap-in point p1 is reached at a voltage Vpi (pull-in voltage). At this point the device (i.e., movable plate) quickly snaps in and the capacitance goes to its maximum value Cmax- Because the gap between the MEMS element and the PD-electrode is now much smaller, the electrostatic force has increased and the voltage has to be reduced down to Vrl (release voltage) in order for the MEMS device to release from the bottom at point p2. The capacitance of the MEMS device is at the maximum value between the MEMS element is in contact with the dielectric material that is disposed on the RF electrode.
[0028] The Vpi and Vrl are important parameters for the MEMS DVC device; both for upward and downward actuation. If the pull-in voltage Vpi is too high then the waveform controller may not be able to pull the MEMS devices into contact intimately, which can impact the obtainable Cmin (upward actuation) or Cmax (downward actuation). If the release voltage Vrl is too low this could indicate stiction which impedes proper device operation. Also, if the release voltage Vrl from the bottom is too low then this will impede the device to be released from the RF- electrode in the presence of an RF signal.
[0029] Both Vpi and Vrl depend on material parameters (Young's Modulus) as well as geometrical parameters, such as layer thicknesses and CD-control of various layers. Therefore, in production, the MEMS devices will exhibit a certain distribution in both Vpi and Vrl. In order to screen functional devices that meet all required product specs, it is key to test the Vpi and Vrl on every device. As discussed herein, a built-in test methodology can be used facilitate the test.
[0030] The built-in test methodology is termed as "hysteresis testing." Hysteresis, because the Vpi and Vrl are separated or the pull-in and the release curves do not overlap as shown in Figure 5. For a reliable part (MEMS DVC) Vpi and Vrl are designed to be in a certain range. Otherwise they can result in non performance as explained in above paragraphs. Unlike, Cmax and Cmin, Vpi and Vrl are not product specs, i.e. they are not listed on a product sheet but they are the best gauges for estimating the reliability or robustness of the part. Due to process variations, certain parts on a wafer or across the lot may fall outside the range and, if escape screening, can lead to failures in the field. So hysteresis testing, a measure of Vpi and Vrl, allows for screening the bad parts from good. [0031] A typical method for performing a hysteresis test on an electrostatic MEMS device is to perform a CV (capacitance - voltage) sweep. A typical CV sweep can be performed using a CV meter, which uses a combination of a DC source and an AC source to provide the DC bias and the AC signal. The measurements are performed by a combination of an AC voltmeter and an AC ammeter. The basic test configuration to perform a test on a two terminal electrostatic MEMS device is shown in Figure 6.
[0032] The MEMS device capacitance is given by the equation, where f is the frequency of the AC voltage source:
CMEMS = iac/2nf*vac
[0033] A three or more terminal electrostatic MEMS device does not allow for the same straightforward CV test as shown in Figure 1 . The bias electrodes on the MEMS device provide the actuation bias and are separate from the capacitor electrodes. A CV sweep is performed on this configuration by using the same configuration as shown in Figure 1 , but also including a DC source, Vbias, as shown in Figure 7. The DC source is used in the same manner as the DC source in Figure 6.
[0034] A semiconductor chip can be composed of one or more MEMS transducers and monolithically integrated CMOS control and power management circuitry. This allows for the Vbias power supply in Figure 7 to be generated internal to the semiconductor chip and not in an external power supply. This is shown in Figure 8 as a plurality of MEMS transducers each with separate switches to the power management. The Vbias voltage is generated in the integrated power management circuit and passes to the MEMS transducers through switch S2. The actuation voltage for the MEMS transducers is controlled through the power management, where the level of the output voltage is controlled by the digital control bits C<0:n>, as shown in Figure 8. The hysteresis sweep can be performed by changing the digital control bits in the power management circuit to the desired actuation voltage. A primary difference between the external power supply version in Figure 7 and the internal DFT mode in Figure 8 is that the digital control bits hold the value at a discrete number of fixed levels (n) instead of a continuous sweep that can be performed by the external version.
[0035] One representation of a power management circuit that can allow for discrete levels of output voltage is a charge pump with a regulator. In the simple case in Figure 9, the charge pump clock is gated by the output of the comparator. The output voltage of the charge pump is divided by the resistor ladder and compared to the bandgap voltage reference. If the voltage reference at the resistor ladder is lower than the value of the bandgap voltage reference voltage, the charge pump clock is on. This condition will allow the charge pump clock to be toggling and the charge pump voltage will be increasing if the charge generation is greater than the output load current. If the voltage reference at the resistor ladder is higher than the value of the bandgap voltage reference, the charge pump clock is off. As shown in Figure 9, the programming of the voltage level is produced by switching in discrete resistors in the resistor ladder, effectively changing the voltage on the compare node to produce a higher or lower output voltage set point. In this manner, the value of the output voltage is programmed by the address bits C<0:n>.
[0036] As shown in Figure 9, the value of Vactuation will be programmed to be a resistor ratio as compared to the Vbandgap voltage as shown by the following equation for a programmation of c<0>:
Vactuation = ((Rs+R0)/R0)*Vbandgap
[0037] The value for Vactuation with a c<1 > programmation is:
Vactuation = ((Rs+R0+R1 )/(R0+R1 ))*Vbandgap
[0038] The value for Vactuation with a c<n> programmation is:
Vactuation = ((Rs+R0+R1 +R2...+Rn)/(R0+R1 +R2...+Rn))*Vbandgap [0039] For a discrete hysteresis curve using this DFT method, the test method consists of a voltage programmation using the C address bits, a wait time for settling, and a measurement strobe of the capacitance.
[0040] The test is implemented in the hardware configuration shown in Figure 10. The device under test, or DUT, is preset using the address bits to the regulator to output a voltage level to the MEMS that is lower than the Vpi. After the DUT is powered up, a wait time, or Tw, for voltage and MEMS settling is implemented in the test sequence before the capacitance level is measured by the CV meter at time Ts. After the capacitance is measured, the address bits are incremented to the next voltage level and the measurement is performed using the same timing. Once the Vpi is detected, the address bits are decremented and the measurements taken until the capacitance meter detects Vrl. By utilizing this test sequence, along with the internal DFT, a continuous hysteresis curve can be represented by discretizing the voltage levels as shown in Figure 10.
[0041] Figures 1 1A and 1 1 B are schematic illustrations of a DFT implementation to test a MEMS hysteresis according to additional embodiments. Figure 1 1A shows an embodiment where the test is performed for voltage applied to the pull-down electrode 1 102 while Figure 1 B shows an embodiment where the test is performed for voltage applied to the pull-up electrode 1 104. It is contemplated that the test may be performed on both the pull-down electrode 1 102 and pull-up electrode 1 104.
[0042] The MEMS device 1 100 includes the pull-down electrodes 1 102, the pull- up electrode 1 104, an RF electrode 1 106 and ground electrodes 1 108. The ground electrodes 1 108 are connected to ground and to the movable plate 1 1 10. A dielectric layer 1 1 12 is disposed over the pull-down electrodes 1 102 and the RF electrode 1 106. Another dielectric layer 1 1 14 is disposed between the pull-up electrode 1 104 and the cavity 1 1 16 within which the movable plate 1 1 10 is disposed.
[0043] As shown in Figure 1 1 A, the pull-down electrodes 1 102 are coupled to multiple switches 1 1 18, 1 120. The first switch 1 1 18, when engaged, connects the pull-down electrodes 1 102 to ground. In Figure 1 1 B, when the first switch 1 1 18 is engaged, the pull-up electrode 1 104 is connected to ground. The second switch 1 120, in Figure 1 1 A, is connected to a power management device 1 122. Similarly, in Figure 1 1 B, the second switch 1 120 is connected to the power management device 1 122. Thus, when the second switch 1 120 is engaged, the pull-down electrode 1 102 (Figure 1 1 A) or the pull-up electrode 1 104 (Figure 1 1 B) is connected to the power management device 1 122. The power management device 1 122 and the MEMS device 1 100 are both disposed on the semiconductor chip represented by box 1 124.
[0044] The power management device 1 122 includes a charge pump 1 128 that is coupled to a gate 1 130. The gate 1 130 is coupled to both the Vclock node and the output from a comparator 1 132. The comparator has inputs from the Vbandgap node and the resistive ladder. The resistive ladder is what divides the output of the charge pump 1 128. The resistive ladder includes a plurality of resistors R0... Rn which are coupled together in series. The address bits c<0>, c<1 >, c<n> are programmed to incrementally "actuate" or operate such that the next voltage level is achieved and the capacitance of the MEMS device 1 100 is measured. Hence, an incremental voltage is applied by operating the address bits c<0>, c<1 >, c<n>. Based upon the incremental voltage increase, the Vpi (for Figure 1 1 A) or Vrl (for Figure 1 1 B) is determined. Similarly, by decrementally decreasing the voltage (i.e., operating the address bits c<0>, c<1 >, c<n>), the capacitance is again measured the Vrl (for Figure 1 1 A) or Vpi (for Figure 1 1 B) is detected. As such, the hysteresis curve for the particular MEMS device 1 100 is determined. It is to be understood that multiple MEMS devices may be coupled to the power management device 1 122. The multiple MEMS devices may collectively operate as a DVC.
[0045] Figure 12 is a flow chart 1200 illustrating a method of testing a MEMS
DVC according to one embodiment. Initially, a low voltage is applied to either the pull-down electrode or the pull-up electrode in step 1202. The voltage applied to the pull-down electrode is to pull the movable plate closer to the RF electrode. Voltage applied to the pull-up electrode, on the other hand, is to push the movable plate closer to the RF electrode. After the voltage has been applied, the capacitance of the MEMS device is measured in step 1204. If the capacitance is equal to the maximum capacitance for the MEMS device, then the pull-in (or push-in for voltage applied to the pull-up electrode) has been determined in step 1206. If the measured capacitance is not equal to the maximum capacitance, then the voltage is incrementally increased in step 1208, with the capacitance measured with each incremental voltage increase in step 1204, until the maximum capacitance is reached and the pull-in voltage has been determined in step 1210.
[0046] Once the pull-in voltage (or push-in voltage) has been determined, the release voltage is determined. The release voltage is determined by reducing the voltage applied to either the pull-down electrode or the pull-up electrode in step 1212. The capacitance is then measured in step 1214. If the capacitance is less than the maximum capacitance, then the release voltage has been determined. If, however, the measured capacitance is equal to the maximum capacitance, then the voltage is decrementally reduced in step 1218. The capacitance is measured for each decremental voltage reduction. If the measured capacitance is less than the maximum capacitance in step 1216, then the release voltage has been determined in step 1220.
[0047] In one particular embodiment of determining the pull-in and release voltages, a first voltage is applied to either the one or more pull-down electrodes or the one or more pull-up electrodes. The capacitance of the MEMS device is measured. A second voltage is then applied to either the one or more pull-down electrodes or the one or more pull-up electrodes. The capacitance is again measured. The capacitance of the MEMS device is detected to equal the maximum capacitance for the MEMS device. The second voltage is then removed from either the one or more pull-down electrodes or the one or more pull-up electrodes. The capacitance of the MEMS device is then measured. If the measured capacitance equals the maximum capacitance, then the first voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes is removed. The capacitance is again measured and if the measured capacitance is less than the maximum capacitance for the MEMS device, the release voltage has been determined. [0048] While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.

Claims

Claims:
1 . A MEMS DVC, comprising:
at least one MEMS device, the MEMS device comprising a movable plate, an RF electrode, one or more pull-down electrodes and one or more pull-up electrodes; a first switch coupled to either the one or more pull-down electrodes or the one or more pull-up electrodes, wherein the first switch is additionally coupled to ground;
a second switch coupled to either the one or more pull-down electrodes or the one or more pull-up electrodes; and
a power management system coupled to the second switch, wherein the at least one MEMS device, the first switch, the second switch and the power management system are all disposed on a semiconductor chip.
2. The MEMS DVC of claim 1 , wherein the power management system comprises:
a charge pump; and
a resistive ladder coupled between the charge pump and the second switch.
3. The MEMS DVC of claim 2, wherein the power management system further comprises a comparator coupled between the charge pump and the resistive ladder.
4. The MEMS DVC of claim 3, wherein the power management system further comprises a gate coupled between the comparator and the charge pump.
5. The MEMS DVC of claim 4, wherein the comparator is coupled to a bandgap voltage node on the semiconductor chip.
6. The MEMS DVC of claim 5, wherein the gate is coupled to a clock voltage node on the semiconductor chip.
7. The MEMS DVC of claim 6, wherein the resistive ladder comprises a plurality of resistors.
8. The MEMS DVC of claim 7, further comprising a plurality of address bits coupled to the resistive ladder.
9. The MEMS DVC of claims 8, wherein the second switch is coupled to the one or more pull-down electrodes.
10. The MEMS DVC of claim 8, wherein the second switch is coupled to the one or more pull-up electrodes.
1 1 . The MEMS DVC of claim 1 , wherein the second switch is coupled to the one or more pull-down electrodes.
12. The MEMS DVC of claim 1 , wherein the second switch is coupled to the one or more pull-up electrodes.
13. The MEMS DVC of claim 1 , wherein the power management system includes a resistive ladder having a plurality of resistors.
14. The MEMS DVC of claim 13, wherein the resistive ladder is coupled to one or more address bits.
15. The MEMS DVC of claim 14, wherein the plurality of resistors are coupled together in series.
16. A method of testing a MEMS DVC, the MEMS DVC including at least one MEMS device, the MEMS device comprising a movable plate, an RF electrode, one or more pull-down electrodes and one or more pull-up electrodes, the method comprising: applying a first voltage to either the one or more pull-down electrodes or the one or more pull-up electrodes;
measuring a capacitance of the MEMS device;
applying a second voltage to either the one or more pull-down electrodes or the one or more pull-up electrodes;
measuring the capacitance of the MEMS device;
detecting the capacitance of the MEMS device equals a maximum capacitance of the MEMS device;
removing the second voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes;
measuring the capacitance of the MEMS device;
removing the first voltage from either the one or more pull-down electrodes or the one or more pull-up electrodes;
measuring the capacitance of the MEMS device; and
detecting the capacitance of the MEMS device is less than the maximum capacitance.
17. The method of claim 16, wherein the first voltage and the second voltage are applied incrementally.
18. The method of claim 17, wherein the second voltage and the first voltage are removed decrementally.
19. The method of claim 16, wherein the first voltage and the second voltage are applied to the one or more pull-down electrodes.
20. The method of claim 16, wherein the first voltage and the second voltage are applied to the one or more pull-up electrodes.
PCT/US2014/054213 2013-09-06 2014-09-05 Internally generated dft stepped hysteresis sweep for electrostatic mems WO2015035120A1 (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
CN201580001969.3A CN105593158B (en) 2014-09-05 2015-04-23 The DFT staged sluggishness for electrostatic MEMS that inside generates scans
EP15720864.6A EP3189005A1 (en) 2014-09-05 2015-04-23 Internally generated dft stepped hysteresis sweep for electrostatic mems
US14/916,884 US10029914B2 (en) 2013-09-06 2015-04-23 Internally generated DFT stepped hysteresis sweep for electrostatic MEMS
PCT/US2015/027201 WO2016036422A1 (en) 2014-09-05 2015-04-23 Internally generated dft stepped hysteresis sweep for electrostatic mems

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US201361874653P 2013-09-06 2013-09-06
US61/874,653 2013-09-06

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US14/916,884 Continuation-In-Part US10029914B2 (en) 2013-09-06 2015-04-23 Internally generated DFT stepped hysteresis sweep for electrostatic MEMS

Publications (1)

Publication Number Publication Date
WO2015035120A1 true WO2015035120A1 (en) 2015-03-12

Family

ID=51541393

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2014/054213 WO2015035120A1 (en) 2013-09-06 2014-09-05 Internally generated dft stepped hysteresis sweep for electrostatic mems

Country Status (1)

Country Link
WO (1) WO2015035120A1 (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090320557A1 (en) * 2008-06-26 2009-12-31 Analog Devices, Inc. MEMS Stiction Testing Apparatus and Method
WO2013033613A2 (en) * 2011-09-02 2013-03-07 Cavendish Kinetics, Inc Rf mems isolation, series and shunt dvc, and small mems

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090320557A1 (en) * 2008-06-26 2009-12-31 Analog Devices, Inc. MEMS Stiction Testing Apparatus and Method
WO2013033613A2 (en) * 2011-09-02 2013-03-07 Cavendish Kinetics, Inc Rf mems isolation, series and shunt dvc, and small mems

Similar Documents

Publication Publication Date Title
US10029914B2 (en) Internally generated DFT stepped hysteresis sweep for electrostatic MEMS
US6744258B2 (en) Capacitive sensor apparatus
EP2756317B1 (en) Linearity enhancement of capacitive transducers by auto-calibration using on-chip neutralization capacitors and linear actuation
JP5823394B2 (en) System and method for measuring capacitance
US8373424B2 (en) Apparatus for detecting fault of flying capacitor of insulated condition detecting unit
TWI383947B (en) Mems stiction testing apparatus and method
KR20050039867A (en) Micro-electromechanical switch performance enhancement
US20110169509A1 (en) Electrostatic mems driver with on-chip capacitance measurement for autofocus applications
WO2011108037A1 (en) Capacitive touch panel
EP3171181B1 (en) Circuitry and method for generating a discrete-time high voltage
US20220397593A1 (en) Detecting capacitive faults and sensivity faults in capacitive sensors
TWI550482B (en) Touch sensing circuit and method thereof
WO2015035120A1 (en) Internally generated dft stepped hysteresis sweep for electrostatic mems
TWI383158B (en) Capacitance measurement circuit and method
US9625520B2 (en) Latch-up test device and method for testing wafer under test
Czarnecki et al. New insights into charging in capacitive RF MEMS switches
JP6086008B2 (en) Resistive touch panel device
KR20170019740A (en) Apparatus for touch sensor test
JP6541456B2 (en) Test equipment
KR101279737B1 (en) The inspection apparatus for the capacitive touch screen panel using switching device and the inspection method using the same
US11927645B2 (en) Detecting sensitivity faults in capacitive sensors by using pull-in functionality
Kouteoureli et al. The discharge current through the dielectric film in MEMS capacitive switches
US20160077144A1 (en) Electronic device
JP2014211360A (en) Semiconductor tester
WO2020002449A1 (en) Mutual capacitance measurement

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14766365

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14766365

Country of ref document: EP

Kind code of ref document: A1