WO2017198666A1

WO2017198666A1 - Mems resonator sensor and sensing method

Info

Publication number: WO2017198666A1
Application number: PCT/EP2017/061732
Authority: WO
Inventors: Alan James Davie; Noor MANSUR; Jake APSEY
Original assignee: Koninklijke Philips N.V.
Priority date: 2016-05-16
Filing date: 2017-05-16
Publication date: 2017-11-23

Abstract

A MEMS resonator sensor uses a signal generator to generate first and second reference signals of the same frequency, wherein the first reference signal is used to drive a MEMS resonator. A digital controller is used for controlling the signal generator frequency thereby to minimize a phase difference signal during a frequency tracking mode. The phase difference signal represents the phase shift caused by the MEMS resonator, which is a minimum at resonance.

Description

MEMS RESONATOR SENSOR AND SENSING METHOD

FIELD OF THE INVENTION

This invention relates to MEMS resonator sensors and sensing methods.

BACKGROUND OF THE INVENTION MEMS sensors are increasingly being used in consumer electronics devices. MEMS sensors offer the possibility to miniaturize a device's form- factor, and to reduce power consumption and cost. In addition, MEMS technology allows in principle monolithic integration with CMOS circuits that are used for control, signal readout and communication.

By way of example, as a result of the advantages of MEMS technology, MEMS pressure sensors are now present in many tire pressure monitoring systems (TPMS) integrated into the tire valve or rim, or in watches and mobile devices for altitude

measurements.

MEMS pressure measurement may be based on measuring capacitance changes resulting from movement of a membrane or resistance changes of a strain gauge (based on the piezoelectric effect). However, it is also known to use a MEMS resonant device to measure pressure. When the membrane of the MEMS device deflects due to pressure changes, the stress of the membrane makes the stiffness of the resonator change, which results in its resonant frequency changing.

The resonant frequency of a resonator is generally described as

in which k is the stiffness and m is the mass of the resonator. Resonant pressure sensors have been shown to exhibit better pressure sensitivity and lower temperature sensitivity than

piezoresistive sensors.

It is also known that MEMS resonator sensors, such as piezoelectric based MEMS cantilevers, when coated with suitable surface chemistry can be used to detect the presence of selected volatile organic compounds (VOCs) from a mixture of different gases. This detection mechanism is based on measuring the very small changes in the resonant frequency of the cantilever as its mass changes, resulting from gases being selectively absorbed by the surface coating. One possible application of this technology is the detection of volatile organic compounds (VOCs) in exhaled breath, which can act as a biomarker for certain types of disease.

Electronic readout solutions to detect frequency changes in piezoelectric MEMS cantilever structures are already known from the literature. For example, one approach is disclosed in Pettine, Julia, et al. "Power-efficient oscillator-based readout circuit for multichannel resonant volatile sensors." Biomedical Circuits and Systems, IEEE Transactions on 6.6 (2012): 542-551. The solutions are typically based on a self-resonating oscillator architecture, where the piezoelectric cantilever itself forms the feedback element used to set the frequency of oscillation. Although effective and compact, these solutions do have a number of potential disadvantages.

When operating in the analogue domain, MEMS feedback oscillators require careful cancelation of the unwanted intrinsic stray capacitance that is inevitably created as part of the MEMS fabrication process. In common with many self-oscillating circuit arrangements, MEMS oscillators are generally limited to operation at one resonant frequency, usually the fundamental frequency. This can place significant limits on the sensitivity and overall flexibility. The final output frequency of a self-resonant MEMS oscillator is typically determined by the use of a frequency counter. However, frequency counters, by their nature, almost invariably determine frequency values by counting oscillator cycles over an accurately pre-defined period of time. As a result, the reported output frequency values from such counters are never instantaneously available.

There is therefore a need for a MEMS resonator design which addresses one or more of these drawbacks.

SUMMARY OF THE INVENTION

The invention is defined by the claims.

According to examples in accordance with an aspect of the invention, there is provided a MEMS resonator sensor, comprising:

a MEMS resonator comprising a resonator mass which has at least one resonant frequency;

a signal generator arrangement for generating first and second reference signals of the same frequency, wherein the first reference signal is used to drive the MEMS resonator; a digital controller for controlling the signal generator arrangement;

a multiplier for multiplying an output signal from the MEMS resonator with the second reference signal thereby to derive a phase difference signal; a filter for filtering the phase difference signal;

an amplifier for amplifying the filtered phase difference signal to generate an amplifier output signal; and

a feedback control path of the amplifier output signal to the controller, wherein the controller is adapted to adjust the frequency of the first and second reference signals thereby to minimize the phase difference signal during a frequency tracking mode, wherein the sensor output is based on the frequency of the first and second reference signals.

This sensor provides frequency detection based on a minimum phase shift introduced by the MEMS resonator at resonance. A phase difference signal is derived which represents the phase shift caused by the MEMS resonator, and this phase difference signal is used as a feedback signal to control a digital controller which then adjusts a driving frequency of the MEMS resonator to reduce the phase shift.

By using a digital controller, cancellation of the unwanted intrinsic stray capacitance that is inevitably created as part of the MEMS fabrication process is made possible in a simple manner. In particular, phase offsets caused by these stray capacitances can be compensated for by the controlling software.

The frequency control enables a particular resonant mode to be used, and the sensor is not limited to operation at the fundamental frequency. Other resonance modes may for example be of interest because they have better sensitivity. For this purpose, programmable direct digital synthesizers may be used to implement the signal generator arrangement.

The feedback loop generates a tracking frequency by means of the signal generator arrangement. In this way, the output of interest is the resulting tracking frequency, and this is available as an immediate output, for example without the need for frequency counters.

The filter for example comprises a low pass filter for removing higher frequency harmonics and double frequency components which are not of interest.

The controller is for example controllable to adjust the phase of at least one of the first and second reference signals.

In one example, the sensor has a calibration mode, in which the controller is adapted to adjust the relative phase between the first and second reference signals to produce the lowest phase difference signal in the absence of the MEMS resonator. This phase adjustment is used to remove phase errors in the measurement electronics.

The sensor may also or instead have a frequency sweep mode, in which the controller is adapted to adjust the frequency of the first and second reference signals to find a resonant frequency. This enables a suitable operating frequency to be defined before the sensor is used. Multiple resonant frequencies may be identified in the frequency sweep mode, and the controller is adapted to select one of the resonant frequencies either automatically or by user selection. This gives the system flexibility in the selection of an operating frequency.

The sensor may have a DC offset cancellation mode, in which the controller is adapted to provide a phase offset between the first and second reference signals to compensate for parasitic capacitances of the MEMS resonator. This means that the circuit function is optimized for the particular operating frequency, taking into account the electrical characteristics of the MEMS resonator at that frequency.

During the frequency tracking mode, the controller may be adapted to maintain the amplifier output signal at a fixed level. In this way, a rise or fall in the detected signal can be detected and feedback control is then used to provide automatic adjustment of the first and second reference frequencies. The controller for example comprises an analog to digital converter for converting the amplifier output signal to a digital signal, and the fixed level is then a mid-range value of the analog to digital converter.

The MEMS resonator may comprise a cantilever device, although any resonant structure may be used.

The sensor is for example a volatile organic compound (VOC) sensor for detecting resonant frequency changes in response to a deposited mass of VOCs on the MEMS resonator.

Examples in accordance with another aspect of the invention provide a MEMS resonator sensing method, comprising:

generating first and second reference signals of the same frequency, and using the first reference signal to drive a MEMS resonator;

multiplying an output signal from the MEMS resonator with the second reference signal thereby to derive a phase difference signal;

filtering and amplifying the phase difference signal to generate an amplifier output signal;

feeding back the the amplifier output signal to a controller;

using the controller to adjust the frequency of the first and second reference signals thereby to minimize the phase difference signal during a frequency tracking mode; and

providing a sensor output based on the frequency of the first and second reference signals.

This method implements the automatic and digitally controlled frequency adjustment explained above. The method may further comprise:

adjusting the relative phase between the first and second reference signals during a calibration mode to produce the lowest phase difference signal in the absence of the MEMS resonator; and/or

adjusting the frequency of the first and second reference signals to find a resonant frequency during a frequency sweep mode; and/or

providing a phase offset between the first and second reference signals to compensate for parasitic capacitances of the MEMS resonator during a DC offset cancellation mode.

The amplifier output signal may be maintained at a fixed level during the frequency tracking mode.

The control method may be implemented at least in part by software, and the invention also relates to a computer program for implementing the method.

BRIEF DESCRIPTION OF THE DRAWINGS

Examples of the invention will now be described in detail with reference to the accompanying drawings, in which:

Figure 1 shows an overall sensor sensor system;

Figure 2 shows a phase calibration procedure;

Figure 3 shows a DC offset removal procedure;

Figurer 4 shows a typical oscillation response from a single cantilever sensor;

Figure 5 shows an example of a response from two different cantilever sensors;

Figure 6 shows a resonance frequency selection procedure;

Figure 7 shows a DC offset compensation procedure;

Figure 8 shows a frequency tracking procedure;

Figure 9 shows the system locking onto a particular resonator peak;

Figure 10 shows the resonator peak moving towards an increased frequency;

Figure 11 shows the synthesizers increasing in frequency to follow the increasing frequency resonator peak; and

Figure 12 illustrates an example of a computer for implementing the controller.

DETAILED DESCRIPTION OF THE EMBODIMENTS

The invention provides a MEMS resonator sensor which uses a signal generator to generate first and second reference signals of the same frequency, wherein the first reference signal is used to drive a MEMS resonator. A digital controller is used for controlling the signal generator frequency thereby to minimize a phase difference signal during a frequency tracking mode. The phase difference signal represents the phase shift caused by the MEMS resonator, which is a minimum at resonance.

The signal processing employed is based on the well-known "lock-in" amplifier concept, which acts as a narrow band phase detector. The standard arrangement is modified to include an adaptive feedback path, under microprocessor control, to facilitate real-time frequency tracking of the MEMS resonator sensor signal in real time.

Figure 1 shows the overall sensor system. It comprises a MEMS resonator 10, functioning as a resonant sensor (and termed a sensor below), for example in the form of a MEMS cantilever structure. The resonant frequency is a function of the mass of the resonant body (i.e. the cantilever), and this for example depends on target species which deposit on the resonator mass for the example of a chemical sensor. In the circuit of Figure 1 , this resonant frequency change effects a phase shift.

A controller 12 controls two synthesizers 14, 16, so that they each generate a reference frequency fo. The first has a phase of φο and the sensor causes a phase shift of the second to a resulting phase φι. The two signals are multiplied by multiplier 18, then filtered by filter 20 and then amplified by amplifier 22.

The output of the amplifier is a DC voltage which is proportional to cos((|)o- φι). It is dependent on the phase shift created by the sensor 10.

The phase shift modulation being detected is that created in the sensor itself, due to the physical deformation (vibration) of the sensor resonant mass when exposed to a controlled gas flow. The gas flow for example is human breath and the aim is to detect VOCs.

The exact amount of signal phase shift created through the sensor 10 is essentially a function of how far away the resonant mass is from its resonant frequency. However, unlike frequency which can be measured in absolute terms, a phase measurement is always a relative quantity and as such always needs to be measured against either a second reference signal or a time reference. Because of this, in a conventional lock-in amplifier arrangement, at least two distinct signal paths are provided. The first path goes through the phase modulated sensor (the MEMS sensor) while the second path acts as a reference and as such must have a constant phase characteristic.

To determine the phase shift caused by the MEMS sensor 10, a common input signal frequency, or alternatively two identical input signal frequencies, are generated and applied to the sensor and a reference path simultaneously. These two signal paths are then multiplied together and low pass filtered to remove higher order products leaving a voltage which is a function of the phase shift caused by the MEMS sensor as mentioned above.

This principle can be best explained by looking at the standard mathematical trig function shown below:

AB , AB

A cos 0_O. B sin 0₁ =— sin(0_o - +— sin(0_o + _X)

In this equation ACos0o and BSin0i, on the left hand side, represent the sensor and the reference signal paths respectively. The output from the multiplier is shown on the right hand side of the equation and contains the "sum" and "difference" products. The double frequency product is not required and is therefore removed by the low pass filter 20 directly following the multiplier.

The difference product, which in this case is the wanted component, is a DC voltage that varies in value in response to the phase shift caused by MEMS cantilever as it moves towards or away from resonance. When the sensor reaches the point of resonance, the phase shift through the MEMS cantilever is zero. By looking at the above equation, it can be seen that when this condition is met, the difference product also has a value that is zero.

Unlike the standard lock-in amplifier configuration, Figure 1 makes use of an additional feedback path 24 that connects the output of the amplifier 22 to a control input of the controller 12. This control input couples to an analog to digital converter of the controller 12 which is used to set the output frequency of the synthesizers 14, 16 that drive the MEMS sensor.

Under normal operating conditions, the system operates in three distinct modes. These are:

(i) Calibration: To optimize the system for most consistent and accurate measurement. (ii) Frequency Sweep : To identify the resonant peak of the cantilever before being exposed to VOCs.

(iii) Tracking: To track the movement in resonant frequency of a cantilever after exposure to VOCs.

Of these three modes, the calibration and sweep modes operate without feedback, while the tracking mode uses feedback to track movements in the sensor resonant frequency. Each of these three modes will now be explained in more detail.

Calibration mode The calibration routines fall into two main categories.

A phase calibration is used to remove phase errors in the measurement electronics before the mixer 18.

A DC offset calibration removes DC offsets in the measurement electronics, post the mixer 18.

The phase calibration tries to remove all sources of phase offset (error) in the measurement electronics that are not due to the sensor itself. Phase errors before the mixer 18 are important and can lead to unwanted DC signals being generated due to the mixer action. These phase errors typically arise from issues such as differences in the phase response of the LC low pass filters in the sensor and reference channels, or differences in the phase response of the various operational amplifiers in the two channels in the design.

One suitable procedure for phase calibration is shown in Figure 2.

In step 30, the system is turned on and the synthesizers 14, 16 are activated.

In step 32, the sensor is short circuited to remove any effect on the signal phase in its channel.

In step 34, the DC gain of the amplifier 22 is set to unity.

In step 36, the DC voltage at the output is measured using the controller 12 and its analog to digital converter.

In step 38 , the phase of the synthesizer 16 in the sensor path is adjusted, either positively or negatively relative to the phase of the synthesizer 14.

This is performed iteratively until the DC value is at a minimum as determined by test 40. When the DC level is at the minimum, the phase value of the synthesizer 16 is recorded in step 42 for use during subsequent measurements.

This completes the phase calibration, and the procedure then proceeds to the DC offset calibration mode in step 44. The purpose is then to minimize the DC offset in the measurement circuits. In general, DC offsets are most significant after the mixing process since they add to the DC generated by the wanted phase shifts caused by the sensor before the mixer.

A typical procedure for removing DC offsets is shown in Figure 3.

In step 50, the synthesizers 14,16 are provided with the phase values calculated from the phase calibration of Figure 2.

In step 52, DC offset control is adjusted in the amplifier 22 to achieve a mid-ADC value, for example of 2.5V. This is performed iteratively until the DC value is at the required level as determined by test 54. When the DC level is at the desired level, the short circuit of the sensor is removed so that the sensor is back in the circuit, in step 56.

This completes the DC offset calibration so that the electronics is calibrated. The procedure then proceeds to the sensor phase calibration mode in step 58.

This is a frequency sweep mode to make a first determination of the resonant frequency of the sensor, typically without VOCs present. The details of this mode are outlined below.

Frequency sweep mode

The sweep mode consists of sweeping the two synthesizers across a range of frequencies that span the likely range of the chosen sensor. This sweep would normally be conducted while the sensor is not exposed to any form of volatile organic compound. The function of the sweep is to identify the starting frequency when entering tracking mode. In general, the more accurately the system can identify the resonant frequency of the sensor cantilever in the sweep mode, the more quickly the system will lock onto the sensor in tracking mode.

Figure 4 shows a typical oscillation response from a single sensor cantilever, and shows the oscillation amplitude (expressed as a value recorded by the analogue to digital converter of the controller 12) versus frequency for a sweep from 92kHz to 98kHz and it shows the resonant frequency in that range. The straight line represents a threshold for detection of resonant oscillation.

The sweep result of Figure 4 only shows one possible harmonic peak. However, if the frequency sweep range is extended, typical MEMS sensors will exhibit multiple resonant modes.

An example of two different cantilever sensors is shown in Figure 5, which shows the output (based on the ADC value in the controller 12) versus frequency for a sweep from 0 to 350kHz.

Up to four resonant peaks are visible. The system can therefore relatively easily choose any one of the available harmonics to lock onto. To select one of these resonance modes a procedure shown in Figure 6 is adopted.

The system is turned on in step 60. The calibration routines described above are performed in step 61.

In step 62, the synthesizers are loaded with the lowest expected resonance mode frequency. In step 63, the synthesizers are swept up in frequency to the highest expected resonance mode of interest.

In step 64, the highest local maximum values as measured at the controller analog to digital converter are recorded. This may instead be based on analyzing the peak gradients of the sweep waveform.

In step 65, a resonance mode is selected, either automatically or by the user through a user interface.

In step 66, the synthesizer frequency is reset to a value midway between the chosen peak and the adjacent lower frequency peak. The frequency is thus just below the resonant frequency peak of interest.

In step 67, the synthesizer frequencies are swept, with increasing frequency.

In step 68, the frequency value is stored that corresponds to an analog to digital converter value halfway between its maximum and minimum values. This corresponds to a point some way up the low- frequency side of the resonant peak (see Figure 9). It means an increase in frequency can be distinguished from a decrease in frequency (which would not be the case by selecting a frequency of the peak).

In step 69 the frequency sweep is stopped at a frequency equal to a point midway between the chosen peak and the adjacent higher frequency peak. This means the frequency sweep covers the full resonant peak.

The stored frequency value obtained by this process in step 68 is then passed in step 70 to the tracking routine.

Figure 5 shows that the resonant peaks do not lie in a straight horizontal line. Instead they are superimposed on a curve with a rising slope. This slope is due to the underlying parasitic capacitance in the sensor, caused by effects including capacitance between the various processing layers, capacitance to ground between the layers, capacitance in the bond wires, capacitance in the packaging, and capacitance beyond the sensor packaging including the PCB.

This capacitive effect can be removed by effectively programming a phase offset in one of the synthesizers. This can be understood by looking again at the formula for mixing two sine (or cosine) waves:

AB , AB

A cos 0_O. B sin 0₁ =— sin(0_o - +— sin(0_o + _X) By adjusting the phase of either ACos0o or BSin0i, an appropriate DC offset value can be added to the output to compensate (in the opposite direction) for the DC offset caused by the capacitive slope. In this way, parasitic capacitance can be removed from the response.

In general, the higher the order of the harmonic peak chosen the more DC needs to be removed from the output.

Figure 7 shows an example of how parasitic capacitance induced DC offset can be compensated for.

The system is turned on in step 72. The calibration routines described above are performed in step 73.

In step 74, the harmonic (resonant frequency) selection routine of Figure 6 is carried out.

In step 75, the synthesizers are set to the lower end of the intended sweep range.

In step 76, a reading is taken from the analog to digital converter of the controller.

In step 77, it is determined if value is close to zero volts (i.e. below a threshold). If it is not, the phase in one channel is adjusted by a small amount in step 78. If it is, then the process proceeds to the tracking mode in step 79.

Thus, before frequency tracking, a phase adjustment is made which relates to the circuit behavior when operating at the frequency range of interest.

Frequency tracking mode:

Finally, the tracking mode is carried out. This is the only mode that genuinely uses a closed loop feedback technique around the lock-in amplifier. The method is shown in Figure 8.

The system is turned on in step 80. The calibration routines described above are performed in step 81.

In step 82, the harmonic selection routine of Figure 6 is carried out.

In step 83, the synthesizers are set to the resonant frequency obtained by the harmonic selection routine.

In step 84, a reading from the analog to digital converter of the controller is taken and it is inspected in step 85.

If the value is below the mid-range value (e.g. 512 for a 10 bit converter) then step 86 determines that the resonator has increased in frequency so the frequency of the synthesizers is increased to follow. If the value is above the mid-range value (e.g. 512 for a 10 bit converter) then step 87 determines that the resonator has decreased in frequency so the frequency of the synthesizers is decreased to follow.

If the value is at the mid-range value (e.g. 512 for a 10 bit converter) then step 88 keeps the frequency constant.

Steps 86, 87 and 88 return to the reading step 84.

A more specific example of the circuit operation is given via a series of frequency plots below.

Figure 9 shows the system locking onto a particular resonator peak. The synthesizer output 90 is set to a first value which is at the analog to digital converter mid-range value (512).

Figure 10 shows the resonator peak moving towards an increased frequency.

Figure 11 then shows the synthesizers increasing in frequency to a second value 92 in an attempt to follow.

Since the synthesizers are programmed and controlled directly from the controller, their exact frequency is always known. As a result, unlike known self-resonating solutions, the exact tracking frequency is always exactly and instantaneously known, eliminating the need for a frequency counter. This is a significant advantage over known systems.

The system described above makes use of a controller processing the feedback signal and controlling the frequency synthesizers.

Figure 12 illustrates an example of a computer 120 for implementing the controller described above.

The computer 120 includes, but is not limited to, PCs, workstations, laptops, PDAs, palm devices, servers, storages, and the like. Generally, in terms of hardware architecture, the computer 120 may include one or more processors 121, memory 122, and one or more I/O devices 123 that are communicatively coupled via a local interface (not shown). The local interface can be, for example but not limited to, one or more buses or other wired or wireless connections, as is known in the art. The local interface may have additional elements, such as controllers, buffers (caches), drivers, repeaters, and receivers, to enable communications. Further, the local interface may include address, control, and/or data connections to enable appropriate communications among the aforementioned components.

The processor 121 is a hardware device for executing software that can be stored in the memory 122. The processor 121 can be virtually any custom made or commercially available processor, a central processing unit (CPU), a digital signal processor (DSP), or an auxiliary processor among several processors associated with the computer 120, and the processor 121 may be a semiconductor based microprocessor (in the form of a microchip) or a microprocessor.

The memory 122 can include any one or combination of volatile memory elements (e.g., random access memory (RAM), such as dynamic random access memory (DRAM), static random access memory (SRAM), etc.) and non-volatile memory elements (e.g., ROM, erasable programmable read only memory (EPROM), electronically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), tape, compact disc read only memory (CD-ROM), disk, diskette, cartridge, cassette or the like, etc.). Moreover, the memory 122 may incorporate electronic, magnetic, optical, and/or other types of storage media. Note that the memory 122 can have a distributed architecture, where various components are situated remote from one another, but can be accessed by the processor 121.

The software in the memory 122 may include one or more separate programs, each of which comprises an ordered listing of executable instructions for implementing logical functions. The software in the memory 122 includes a suitable operating system (O/S) 124, compiler 125, source code 126, and one or more applications 127 in accordance with exemplary embodiments.

The application 127 comprises numerous functional components such as computational units, logic, functional units, processes, operations, virtual entities, and/or modules.

The operating system 124 controls the execution of computer programs, and provides scheduling, input-output control, file and data management, memory management, and communication control and related services.

Application 127 may be a source program, executable program (object code), script, or any other entity comprising a set of instructions to be performed. When a source program, then the program is usually translated via a compiler (such as the compiler 125), assembler, interpreter, or the like, which may or may not be included within the memory 122, so as to operate properly in connection with the operating system 124. Furthermore, the application 127 can be written as an object oriented programming language, which has classes of data and methods, or a procedure programming language, which has routines, subroutines, and/or functions, for example but not limited to, C, C++, C#, Pascal, BASIC, API calls, HTML, XHTML, XML, ASP scripts, JavaScript, FORTRAN, COBOL, Perl, Java, ADA, .NET, and the like.

The I/O devices 123 may include input devices such as, for example but not limited to, a mouse, keyboard, scanner, microphone, camera, etc. Furthermore, the I/O devices 123 may also include output devices, for example but not limited to a printer, display, etc. Finally, the I/O devices 123 may further include devices that communicate both inputs and outputs, for instance but not limited to, a network interface controller (NIC) or modulator/demodulator (for accessing remote devices, other files, devices, systems, or a network), a radio frequency (RF) or other transceiver, a telephonic interface, a bridge, a router, etc. The I/O devices 123 also include components for communicating over various networks, such as the Internet or intranet.

When the computer 120 is in operation, the processor 121 is configured to execute software stored within the memory 122, to communicate data to and from the memory 122, and to generally control operations of the computer 120 pursuant to the software. The application 127 and the operating system 124 are read, in whole or in part, by the processor 121, perhaps buffered within the processor 121, and then executed.

When the application 127 is implemented in software it should be noted that the application 127 can be stored on virtually any computer readable medium for use by or in connection with any computer related system or method. In the context of this document, a computer readable medium may be an electronic, magnetic, optical, or other physical device or means that can contain or store a computer program for use by or in connection with a computer related system or method.

The MEMS resonator has not been described in detail above. There are many kinds of MEMS resonators which may be used. The most popular types of MEMS resonators are capacitive resonators, piezoresistive resonators and piezoelectric resonators. For the MHz range, piezoresistive and piezoelectric resonators have the benefit of lower phase noise. At the kHz frequency range, the capacitive principle is generally used.

In capacitive MEMS resonators, the resonant structure is actuated by an electrostatic force which acts over an actuation gap. The electrostatic actuation can be based on either the parallel-plate or comb-drive principle. To sense the vibration of the structure, a capacitance change between the structure and a fixed electrode, separated by a sensing gap (sometimes also by the same actuation gap) is used as output signal. To sense the capacitance change, a DC voltage is applied between the structure and the sensing electrode. The capacitance change manifests itself in the AC current that flows through the sensing gap, which is called the motional current.

In a piezoresistive or piezeoelectric resonator, a current may be driven through the resonator element, and its change in resistance or impedance is measured by reading out a resulting voltage. This is a current-biasing and voltage sensing approach. Current sensing can instead be employed, wherein a fixed voltage is applied to the piezoresistive or piezoelectric element to provide voltage-biasing, and the current is measured. Whether voltage-biasing and current-sensing is used, or current-biasing in combination with voltage-sensing, depends on the requirements of the complete oscillator.

In the example above, the all of the circuitry between the MEMS resonator sensor 10 and the input to the analog to digital converter of the controller 12 is analogue. This includes the mixer 18, amplifiers 22 and filters 20. However, other topologies are possible, for example including digitization immediately after the resonator sensor 10 so that all of the following circuitry is in the digital domain. Thus, other more digitally intensive embodiments are also possible.

The resonator mass of the MEMS resonator may be based on a cantilever which is clamped at one end (clamped-free) or clamped at both ends (clamped-clamped) or indeed a resonant frame shape with two or more anchors connecting the frame to a fixed substrate.

The MEMS device typically has micrometer scale dimensions. For example, a MEMS cantilever typically has a length less than 500μιη and a width less than 50 μιη. However, smaller nanometer scale device are also possible, for example with a length and width of the order of 1 μιη. Various possible resonant MEMS device architectures will be known to those skilled in the art.

The invention is of particular interest for the non-invasive detection of exhaled breath for disease detection. For example, one area of interest is for the detection of ventilator acquired pneumonia (VAP) in intensive care unit (ICU) patients. The application could of course also be extended to the detection of other diseases where suitable breath biomarkers or volatile organic compounds in general need to be identified. The invention may be used in a hospital or, because of the low cost and easy-to-use nature of the system, it could also be suitable for disease identification or monitoring within the home.

Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.

Claims

CLAIMS:

1. A MEMS resonator sensor, comprising:

a signal generator arrangement for generating first and second reference signals of the same frequency, wherein the first reference signal is used to drive the MEMS resonator; a controller for controlling the signal generator arrangement;

a multiplier for multiplying an output signal from the MEMS resonator with the second reference signal thereby to derive a phase difference signal;

a filter for filtering the phase difference signal;

2. A sensor as claimed in claim 1, wherein the filter comprises a low pass filter for removing frequency harmonics.

3. A sensor as claimed in claim 1, wherein the controller is controllable to adjust the phase of at least one of the first and second reference signals.

4. A sensor as claimed in claim 3, wherein the sensor has a calibration mode, in which the controller is adapted to adjust the relative phase between the first and second reference signals to produce the lowest phase difference signal in the absence of the MEMS resonator.

5. A sensor as claimed in claim 1, wherein the sensor has a frequency sweep mode, in which the controller is adapted to adjust the frequency of the first and second reference signals to find a resonant frequency.

6. A sensor as claimed in claim 5, wherein multiple resonant frequencies are identified in the frequency sweep mode, and wherein the controller is adapted to select one of the resonant frequencies either automatically or by user selection.

7. A sensor as claimed in claim 1, wherein the sensor has a DC offset cancellation mode, in which the controller is adapted to provide a phase offset between the first and second reference signals to compensate for parasitic capacitances of the MEMS resonator.

8 A sensor as claimed in claim 1 , wherein during the frequency tracking mode the controller is adapted to maintain the amplifier output signal at a fixed level.

9. A sensor as claimed in claim 8, wherein the controller comprises an analog to digital converter for converting the amplifier output signal to a digital signal, and the fixed level is a mid-range value of the analog to digital converter.

10. A sensor as claimed in claim 1, wherein the MEMS resonator comprises a cantilever device.

11. A sensor as claimed in claim 1 , comprising a VOC sensor for detecting resonant frequency changes in response to a deposited mass of VOCs on the MEMS resonator.

12. A MEMS resonator sensing method, comprising:

feeding back the the amplifier output signal to a controller;

13. A method as claimed in claim 12, comprising:

14. A method as claimed in claim 12, comprising maintaining the amplifier output signal at a fixed level during the frequency tracking mode.

15. A computer program comprising computer program code which is adapted to implement the method of claim 12, 13 or 14 when said program is run on a computer.