AN LTCC-BASED CAPACITIVE PRESSURE SENSOR WITH A DIGITAL OUTPUT Marina Santo Zarnik1, Matej Možek2, Srečko Maček3, Darko Belavič1 1HIPOT-RR d.o.o., Otočec, Slovenia 2Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia 3Jožef Stefan Institute, Ljubljana, Slovenia Key words: Pressure sensor, 3-D LTCC structure, capacitive sensing, capacitive-to-digital conversion, low power consumption. Abstract: A capacitive pressure sensor, fabricated using low-temperature cofired ceramic (LTCC) materials and technology was considered for an application in a wireless sensor system. LTCC technology is inherently efficient for 3D structuring and exhibits good dimensional definition and stability, appropriate material flexibility (higher than the commonly used alumina), good chemical resistance, and low moisture absorption, which makes it appropriate for a wide range of sensor applications, even in some extreme conditions and harsh environments. However, very often in such applications, remote control and operation in a low power-consumption mode are required. In order to meet such demands, electronics for signal processing and power managing, based on a capacitance-to-digital conversion, were realised by using an Analog Devices AD7746. A sensor characterization system with the corresponding software for an evaluation of the sensors nonlinearity and temperature sensitivity is presented. The typical characteristics of the capacitive sensing elements were as follows: a sensitivity of 1.7 fF/mbar, a temperature dependence of 9 fF/°C and a temperature dependence of the sensitivity of less than 2 aF/mbar/°C. The digital temperature compensation was performed with a two-dimensional rational polynomial approximation, resulting in a less than 0.4% FS temperature error in the compensation range 10 °C to 75 °C. Kapacitivni senzor tlaka z digitalnim izhodom izdelan v LTCC tehnologiji Kjučne besede: senzor tlaka, 3-D LTCC struktura, kapacitivni senzor, kapacitivno digitalna pretvorba, nizka poraba. Izvleček: V prispevku so prikazani rezultati študije kapacitivnega keramičnega senzorja tlaka za uporabo v brezžičnem senzorskem sistemu. Senzor je izdelan z uporabo keramike z nizko temperaturo žganja (LTCC - low temperature cofired ceramic) in z ustreznimi novimi tehnološkimi postopki. LTCC materiali in tehnologija so primerni za oblikovanje tri-dimenzionalnih keramičnih struktur. Poleg tega imajo LTCC materiali približno trikrat nižji Youngov modul elastičnosti v primerjavi z najpogosteje uporabljano korundno keramiko, kar omogoča doseganje večje tlačne občutljivosti senzorja. Zahvaljujoč še nekaterim drugim lastnostim, kot so kemična stabilnost in nizka absorpcija vlage, je LTCC keramika zelo primerna za uporabo v različnih medijih in celo v nekaterih ekstremnih pogojih. Pogosto je pri tovrstni uporabi potrebno zagotoviti tudi prenos in obdelavo merilnih rezultatov v oddaljeni enoti ob energijsko varčnem delovanju. V ta namen smo tudi izbrali kapacitivni senzorski princip, ki je že v osnovi energijsko varčen. Pretvorba in procesiranje senzorskega signala je izvedena z uporabo kapacitivno-digitalnega pretvornika AD7746. Predstavljen je merilni sitem z ustrezno programsko podporo za evaluacijo temperaturne kompenzacije senzorske karakteristike. Tipične karakteristike izmerjenih prototipov so naslednje: tlačna občutljivost je 1.7 fF/mbar, temperaturna odvisnost ničelne kapacitivnosti je 9 fF/°C in temperaturna odvisnost občutljivosti je manjša od 2 aF/mbar/°C. Z digitalno kompenzacijo temperaturne odvisnosti odziva smo dosegli napako manjšo od 0,4% FS na kalibracijskem temperaturnem področju 10 °C do 75 °C. 1. Introduction Thick-film ceramic pressure sensors have been used for many years in a variety of special applications /1, 2/. One advantage of such ceramic sensors is that they can be employed in severe environments: such as high temperatures, magnetic fields, harsh atmospheres and in some aggressive liquids. Realised with the use of Low Temperature Cofired Ceramic (LTCC) materials and technology and offering advantageous features for 3D structuring /3/ (structures with cavities and channels, and buried electronic components, such as conductors, resistors or other functional layers) they have the potential to be an alternative to micro-machined sensors in silicon /6-9/. The most common application is a diaphragm-type ceramic pressure sensor. Depending on the sensing principle, the pressure-induced deformation of the thin ceramic diaphragm is converted into an electrical signal, which is proportional to the changes of the characteristics of the thick-film sensing structure on it (thick-film piezoresistors, thick-film piezoelectric structures or a capacitive sensing structure). The appropriate electronics for the sensors' signal processing can be realised directly on the ceramic structure, i.e., the 3D ceramic structure can be used as a package substrate at the same time. In order to ensure low power consumption, the capacitive sensing principle is one of the most appropriate methods /4-5/. In addition, the ceramic capacitive sensors have further advantages: a very high sensitivity and, accordingly, a potentially high resolution, robustness, good stability, and drift-free measurement capability. However, they traditionally require more complex interfacing circuits, which represented a major disadvantage in the past. Generally, the capacitive sensing utilises the deformation-induced capacitance change to convert the information of the applied pressure into the electrical signal, such as changes of the oscillation frequency, time, charge, and voltage. The translation from voltage or current to a digital word requires an additional analogue-to-digital converter (ADC). The expected variance in capacitance is generally of the order of several pF or less. In many cases the signal capacitance is much smaller than the parasitic capacitances present in the measuring circuit, which represents a difficult interfacing task. However, a modification of the conventional sigma delta ADC architecture has been identified as a suitable basis for a monolithic Capacitance to Digital Converter (CDC) /12/. The circuit is parasitic insensitive, and can be configured to work with both a floating and a grounded configuration /16/. Precision capacitive-sensor interface products are based on a well-established sigma-delta (ZA) conversion technology. Converters utilizing the ZA principle offer excellent linearity and very high resolution, and are ideal for most sensor-interfacing applications. In a conventional voltage input ZA converter, the unknown charge is derived from charging a fixed capacitor to an unknown input voltage, while in the CDC realization, the voltage is fixed and the capacitor is variable. Such an arrangement maintains the high precision and accuracy that is typical for ZA ADCs /13/. Modern implementations enable the measurement of capacitances in the atto Farad (aF) range /15, 16/, with an effective noise resolution of 21 bits and a corresponding resolution down to 4aF. They offer measurements of common-mode capacitance up to 17pF on the 4pF range with a 4fF measurement accuracy. These implementations offer complete sensor solutions; however, their application is limited to an indication of temperature and by the humidity dependence problem /17, 18/ of capacitive sensors, while not offering an effective implementation for the compensation of these unwanted quantities. This work reports on the results of a case study of a thick-film capacitive pressure-sensor module made with LTCC technology. The electronics for the signal processing and power managing, based on a capacitance-to-digital conversion, were realised with the use of an Analog Devices AD7746, which was placed directly on the ceramic structure, close to the sensing capacitor electrodes. The sensor is connected via an interface module to the I2C - USB converter, which is used to interface the sensor to the host computer (PC). The prototype sensors were characterised in the pressure range 0-1 bar. The stability and the temperature dependence of the sensors' characteristics were discussed and an effective method for temperature compensation is presented. The sensor nonlinearity and temperature sensitivity were analyzed. In order to achieve the compensation and linearization, an effective method of temperature compensation based on a two-dimensional rational polynomial description of the sensor characteristic was employed. 2 Ceramic capacitive pressure sensors 2.1 Theory of capacitive sensing The construction of the LTCC-based capacitive pressure sensor consists of a circular, edge-clamped deformable diaphragm that is bonded to a rigid ring and the base substrate. One electrode deposited on the rigid substrate and one electrode deposited on the deformable diaphragm form a parallel-plate air-gap capacitor. The operation of the capacitive pressure sensor is based on changing the air gap between the electrodes as a result of the applied pressure (Figure 1). Fig. 1: The construction of the capacitive pressure sensor (above). The deflection of the diaphragm under the measuring pressure applied to the exterior surface of the diaphragm (below). The capacitance between the electrodes on the deflected diaphragm and the rigid substrate can be expressed with equation (1) 2-n-r-dr (1) where e0 and er are the dielectric permittivity of free space and the relative dielectric permittivity of the dielectric, Re is the radius of the electrode, r is the distance from the centre of the electrode/diaphragm, D0 is the distance between the electrodes at zero pressure, p is the measuring pressure and y(r,p) is the deflection of the diaphragm when the pressure is applied. For a clamped circular diaphragm y(r,p) can be calculated from (2) 16 Ef (2) where t and Rd are the thickness and the radius of the diaphragm, E is the Young's modulus and n is the Pois-son's ratio of the diaphragm. Substituting equation (2) into (1) allows preliminary calculations of the sensitivity in the sensor's design phase. Notice that depending on how the pressure is applied, i.e., on the exterior surface of the diaphragm or in the interior of the cavity (or equivalently, if the pressure or an under-pressure is applied), y(r,p) with a positive or negative sign should be considered. However, the sensor's sensitivity to the applied pressure depends on the area of the electrodes and the initial distance between them. Accordingly, the realization of a very small distance between the electrodes is essential if we are to achieve a high resolution. 2.2 LTCC-based capacitive sensing structure In a typical capacitive pressure-sensor construction the thick-film ceramic pressure sensor consists of a ceramic capsule, i.e., a cylindrical cavity (air gap), closed with a thin flexible ceramic diaphragm parallel to a rigid reference substrate. The thick-film electrodes of the sensing capacitor were made on the diaphragm and the substrate plane inside the cavity in co-processing with the LTCC structure. The cross-section of the LTCC-based capacitive pressure sensor considered in this case study is schematically presented in Figure 2. The important dimensions of the prototype sensors were as follows: a diaphragm diameter of 9 mm and a thickness of 200 mm, an electrode diameter of 8.5 mm, and air gap between the electrodes of 50-70 ^m. Such a sensor provides a total capacitance change of approximately 1.5 pF for an applied pressure in the range of 0-1 bar. A detail of the parallel-plate air-gap capacitor (cross-section of the LTCC-based sensor structure) is shown in Figure 3. Figure 4 shows the prototype of the LTCC-based capacitive pressure-sensor module with AD7746/45. In order to minimise the stray capacitances that adversely influence the sensitivity to the pressure loads the CDC was placed on the same ceramic substrate, as close as possible to the capacitor electrodes. In this realisation the top surface of the diaphragm is covered with a 3-^m-thick Au film with the same dimensions as the electrode inside the ceramic capsule. Fig. 2: Schematic representation of the cross-section of the sensor structure (not to scale) 3. Capacitive sensor characterization systems The generalized layout of the capacitive-sensor measurement system is depicted in Figure 5. The sensor is con- Fig. 3: A detail of the cross-section of the LTCC structure with the air gap between the cofired thick-film electrodes on the diaphragm and the rigid base. Fig. 4: Prototype of the capacitive ceramic pressure sensor with the AD7745. nected via the interface module to the I2C - USB converter, which is used to interface the sensor to the host PC. A dedicated electronic interface module was designed. This module enables data transmission and the control of the CDC AD7746. The module itself is based on a CY8C24794 Programmable System on Chip (PSoC) circuit. Fig. 5: Capacitive sensor evaluation module The hardware is used to directly map the CDC to the controlling PC. The corresponding PC software performs the functions of CDC status and data reading. In fact, the controlling software implements all the functions of the AD7746: from the capacitance channel setup to the temperature sensor channel setup, as well as the channel excitation, the common mode capacitance setting, the offset and the gain of the capacitive measurement channel. The measurement-range optimization was performed in order to obtain the maximum span of the CDC measurement range. The measured device, the LTCC capacitive sensor /18/, exhibits a negative slope for the sensor characteristic. Therefore, the measurement-range optimization must be performed at the maximum pressure readout with a minimum pressure applied and vice versa. This also imposes the order of the compensation algorithms, i.e., the offset compensation is performed before the gain compensation. The sensor offset response is compensated by setting the AD7746 registers CAPDACA and CAPOFF-SET. The register value CAPDACA value affects the coarse setting of the offset response and the CAPOFFSET affects the fine setting of the sensor response. The procedure of offset setting is composed of the coarse and fine offset settings. Because of the negative sensor characteristic slope, the fine offset value is initially set at maximum and the coarse value is altered from its initial zero value in such manner that the sensor response is maintained at its maximum value. The setting of the CAPDACA register is performed by a successive approximation approach, starting at the MSB of the CAPDACA register. The subsequent bits are tested against the raw sensor output. If the sensor output exceeds the maximum sensor readout (FFFF16) when the corresponding bit is set to 1, then the bit is set to zero and the algorithm advances towards the lower bits. After the coarse register was set, the CAPOFFSET register is processed in a similar manner. The result of this algorithm is a maximum sensor response value at the applied offset pressure. After a successful optimization of the offset value, the gain parameter is set in a similar manner. The minimum sensor response is set with an alteration of the CAPGAIN register, which actually changes the clock rate of the front-end of the CDC. The procedure starts with the minimum setting of the CAPGAIN register. The bits of the CAPGAIN register are tested according to the described successive approximation algorithm, just that the bit-testing criterion is now the minimum CDC readout. The result of this algorithm is a minimum sensor response at the maximum applied pressure. From the capacitive channel output data, where the 0x000000 code represents the zero scale (0 pF), and the 0xFFFFFF code represents the full scale (+4.096 pF), the capacitance can be calculated using the following expressions: Csens = Cq + Coffset + Cs (3) _ Code (4) (5) Cs = (Code-/0xFFFFFF)*Cref (6) where the Code is the corresponding CDC readout and Cref is 4.096 pF. For the temperature sensor on a chip, the temperature can be calculated from the code (readout of the temperature channel) using the following equation /16/: Temperature (°C) = (Code/2048) 4096. (7) 3.1. Characterisation of the ceramic capacitive-sensing structure Initially, the measurements of the prototype sensor were performed for a determination of the optimal settings of the AD7746 and the tested LTCC capacitive sensing structure. The typical sensor characteristic, obtained in an up and down scan of the pressure range 0-700 mbar, at room temperature (25°C ± 0.5 °C) is presented in Figure 6. In order to assess the repeatability the up and down scans were repeated several times. It is evident from Figure 7 that the repeatability was very good and that the tested sensor exhibits practically no hysteresis. For the pressure loads up to 300 mbar the characteristic is almost linear (R2 = 0.9998), and only for a wider pressure range, over 1 bar, does the deviation from an ideal straight line indicate the necessity for sensor characteristic linearization. From the CDC readouts the capacitance was calculated using formula (3). As a result, the variation of the capacitance over the pressure range was assessed, which showed a typical sensitivity of 1.8 fF/mbar. 17.0 16.5 \