Smart Sensors Systems Design: New Challenges. Prof. Sergey Y. Yurish, Technical University of Catalonia (CDEI-UPC Barcelona)

Transcription

1 Smart Sensors Systems Design: New Challenges Prof. Sergey Y. Yurish, Technical University of Catalonia (CDEI-UPC Barcelona) Sensors Expo & Conference, Rosemont, Il, USA, 10 June 2008

2 Introduction Contents Smart Sensors Market Trends Smart Sensors Systems Design Methodology Practical Examples Further Perspectives and MEMS Technologies Conclusions 2 International Frequency Sensor Association

3 Introduction Smart sensor systems are of great interest in many fields of industry Classical approach to such systems means that the information is in the amplitude of a voltage or current signal Another approach relies on resonant phenomena and variable oscillators: information is embedded in the frequency or time parameters of signal 3 International Frequency Sensor Association

4 Smart Sensors World Market Trends The forecast for North American Smart Sensors Market is to reach $635.2 million in 2010 (According to Frost & Sullivan) Strong growth expected for sensors based on MEMS-technologies, smart sensors and sensors with bus capabilities 4 International Frequency Sensor Association

5 Smart Sensor Sensing element Signal conditioning µc DSP ADC or FDC Communications MUX 5 International Frequency Sensor Association

6 Smart Sensor Smart sensor Smart sensor (or intelligent sensor) is one chip, without external components, including the sensing, interfacing, signal processing and intelligence (selftesting, self-identification, self-validation or selfadaptation) functions 6 International Frequency Sensor Association

7 Main Smart Sensor Properties Adaptability: exchange accuracy for speed and conversely; moderate power consumption by adjusting a clock crystal oscillator frequency High Accuracy: measuring error should be programmable (statistical algorithms, weight average, etc.) High Reliability: self-diagnostic is used to check the performance of the system and connection of the sensor wires 7 International Frequency Sensor Association

8 Sensors (IFSA study 2007) Quasi-Digital 20% Digital 25% Analog 55% 8 International Frequency Sensor Association

9 Quasi-Digital Sensors Quasi-digital sensors are discrete frequency-time domain sensors with frequency, period, duty-cycle, time interval, pulse number, pulse width modulated (PWM) or phase shift output f x N x D.c. τ φ T x PWM f 1 /f 2 9 International Frequency Sensor Association

10 Quasi-Digital Sensors Duty-cycle 9% Pulse Number 3% Period 1% Phase-shift 1% PWM 16% Frequency 70% 10 International Frequency Sensor Association

11 Frequency Advantages High Noise Immunity High Power Signal Wide Dynamic Range High Reference Accuracy Simple Interfacing Simple Integration and Coding Multiparametricity 11 International Frequency Sensor Association

12 Temperature Sensors 12 International Frequency Sensor Association

13 Pressure Sensors 13 International Frequency Sensor Association

14 Accelerometers 14 International Frequency Sensor Association

15 Rotation Speed Sensors There are many known rotation speed sensing principles Magnetic sensors (Hall-effect and magnetoresistor based sensors) Inductive sensors Passive and active electromagnetic rpm-sensors are from the frequency-time domain n x 60 Z = f, where Z is the number of modulation rotor s x (encoder s) gradations (teeth) 15 International Frequency Sensor Association

16 TAOS Light and Color Sensors For TSL 230RD: f O = f D + (Re) (Ee), where f O is the output frequency; f D is the output frequency for dark condition (Ee = 0); Re is the device responsivity for a given wavelength of light given in khz/(mw/cm2); Ee is the incident irradiance in mw/cm 2 16 International Frequency Sensor Association

17 Humidity Sensors 17 International Frequency Sensor Association

18 Chemical, Gas and Biosensors Sensors arrays (electronic noses and tongues) Square wave with a frequency inversely proportional to the sensor resistance Sensors Array based on chemisorbing polymer films Acoustic gas sensor based on a gas-filled cell Quartz Crystal Microbalance (QCM) sensors SAW and bulk acoustic wave sensors 18 International Frequency Sensor Association

19 Mass Variation Sensors Typical frequency range: from some khz to some MHz Needs high accuracy (the relative error should be better than %) reduced time of measurement (less than 0.1 s) 19 International Frequency Sensor Association

20 Magnetic Sensors HAL810, HAL819 Hall sensors with duty-cycle output form Micronas; AKL Sensors Series from Rhopoint Component Ltd., High resolution CMOS magnetic field to frequency converter with frequency difference on its output [1] [1]. Shr-Lung Chen, Chien-Hung Kuo, and Shen-Iuan Liu, CMOS Magnetic Field to Frequency Converter, IEEE Sensors Journal, Vol.3, No.2, April 2003, pp International Frequency Sensor Association

21 Other Sensors Tilt and inclination sensors with PWM outputs Torque transducers with frequency output Level sensors with frequency output Conductivity sensor SBE4 with frequency output Flow sensors with frequency output 21 International Frequency Sensor Association

22 Multiparameters Sensors Color sensor (TU Delft, The Netherlands): frequency is proportional to optical intensity (luminance) and duty-cycle is proportional to colour (chrominance) Pressure and temperature sensors Humidity and temperature sensors (transmitters) from E+E Elektronik, Bitron, etc. 22 International Frequency Sensor Association

23 Design Task Definition There are many quasi-digital sensors for any physical and chemical quantities The frequency range of such sensors is very wide (some parts of Hz to some MHz), relative error up to 0.01% and better (0.003 %) How to use them by optimal way and efficiently in a frame of smart sensor systems? 23 International Frequency Sensor Association

24 Conventional Methods Standard counting method (measurement of average frequency for a fixed reference gate time, for example, 1 s) Indirect counting method (measurement of instantaneous frequency 1/T x ) Combined Method Interpolation method (with digital interpolation) 24 International Frequency Sensor Association

25 Disadvantages of Classical Methods High quantization error in low or high frequency range Quantization error dependents on frequency Redundant conversion time for the standard counting method 25 International Frequency Sensor Association

26 Advanced Methods Ratiometric Counting Method Reciprocal Counting Method M/T Counting Method Constant Elapsed Time (CET) Method Single- and Double Buffered Methods DMA Transfer Method Method of Dependent Count (MDC) Method with Non-redundant Reference Frequency 26 International Frequency Sensor Association

27 Benefit and Demerits Advantage: Constant quantization error in all specified conversion range of frequencies Disadvantages: Redundant conversion time due to prescribed conversion time Measurement frequency f x < f 0 Two references (frequency and gate time) 27 International Frequency Sensor Association

28 Method of Dependent Count (MDC) Method was proposed in 1980 (modifications: 1993 and 2006) Most advanced method: constant programmable quantization error in all frequency range; non-redundant conversion time; the possibility to convert frequency f x higher, than the reference frequency f 0 (f x >> f 0 ); selfadaptation capabilities The method has been developed for conversion of absolute, relative frequencies and periods Suitable for single channel as well as for multichannel synchronous frequency conversions 28 International Frequency Sensor Association

29 Comparison of Methods MDC, M/T 29 International Frequency Sensor Association

30 Method Selection The accuracy of conversion is one of the most important quality factor for smart sensor systems. Due to advanced methods it is possible to reach a constant quantization error Self-adapting method of dependent count and method with non-redundant reference frequency are very suitable for the usage in frequency-todigital converters of smart sensors 30 International Frequency Sensor Association

31 Integrated FDCs USP-30 one-chip specialized microprocessor (1980) IC of ALU for time interval measurements (1989) K512PS11 - frequency-to-digital converter (1990) USIC - universal sensor interface chip (1996) Single-chip (FPGA) interpolating time counter ASIC of single channel frequency-to-digital converter (1999) Frequency-to-digital converter from AutoTEC Time-to-Digital Converter (TDC) from Acammesselectronic GmbH SSP Sensor Signal Processor from Sensor Platforms, Inc. (USA, 2006) 31 International Frequency Sensor Association

32 ICs Disadvantages All ICs except TDCs are based on conventional methods of measurement, hence, quantization error is dependent on measurand frequency f x, many of ICs have redundant conversion time They cannot be used with all existing modern frequency-time domain sensors due to low accuracy or/and narrow frequency ranges They do not cover all frequency time informative parameters of electric signals. 32 International Frequency Sensor Association

33 Universal Frequency-to to- Digital Converter (UFDC-1) Low cost digital IC with programmable accuracy 2 channels, 16 measuring modes for different frequency-time parameters and one generating mode (f osc /2 = 8 MHz) Based on four patented novel conversion methods Should be very competitive to ADC and has wide applications 33 International Frequency Sensor Association

34 Features Frequency range from 0.05 Hz up to 7 MHz without prescaling and 112 MHz with prescaling Programmable accuracy (relative error) for frequency (period) conversion from 1 up to % Relative quantization error is constant in all specified frequency range Non-redundant conversion time Quartz-accurate automated calibration RS-232/485, SPI and I 2 C interfaces 34 International Frequency Sensor Association

35 UFDC-1 1 Block Diagram 35 International Frequency Sensor Association

36 Measuring Modes Frequency, f x Hz 7.5 MHz directly and up to 120 MHz with prescalling Period, T x1 150 ns 20 s Phase shift, ϕ x at f x 500 khz Time interval between start- and stop-pulse, τ x 2 µs 250 s Duty-cycle, D.C. 0 1 at f x 500 khz Duty-off factor, Q at f x 500 khz Frequency and period difference and ratio Rotation speed (rpm) and rotation acceleration Pulse width and space interval 2 µs 250 s Pulse number (events) counting, N x International Frequency Sensor Association

37 Evaluation Board Circuit Diagram 37 International Frequency Sensor Association

38 UFDC-1 1 Evaluation Board 38 International Frequency Sensor Association

39 UFDC-1 1 and Analog Signal Domain S e n s o r s Voltage (V) or Current (I) VFC f x, Hz Digital Output RS-232/485 SPI I 2 C Any Voltage-to-Frequency Converter (VFC) can be used to convert an analog signal to quasi-digital (frequency) signal 39 International Frequency Sensor Association

40 Where to use the UFDC-1 1? Smart Sensors; Quasi-digital and Digital sensors; Multiparametric Sensors ABS Systems Frequency Counters Virtual Instruments Tachometers and Tachometric Systems DAQ boards for Frequency-time Parameters Multimeters for Frequency-time Parameters 40 International Frequency Sensor Association

41 Smart Sensors Systems Design Methodology The best modern approach is to use both modern technologies and advanced methods for signal processing and conversion Adaptive measuring algorithms: λ * j ( ) ( * = T Lγ t δ Lγ t); λ = P Lγ ( t) δ Lγ ( t) s j s j j s j s j where L is the algorithm of measurement; T s, δ s and P s are operations for speed and accuracy increasing; and power consumption decreasing respectively; γ j (t) is the input action 41 International Frequency Sensor Association

42 Color-to to-digital Converter Design notes: 100 % scaling mode for TCS230 (S0, S1 =1) and clear photodiode type (no filter, S2=1, S3=0). Power-supply lines must be decoupled by a 0.01-µF to 0.1-µF capacitor with short leads mounted close to the device package. 42 International Frequency Sensor Association

43 Light-to to-digital Converters (a) (b) 43 International Frequency Sensor Association

44 Commands Example (RS-232 interface) >M0 ; Frequency measurement initialization >A0 ; 1 % conversion error set up >S ; Start a measurement >R ; Read a result ; Measurement result indication 44 International Frequency Sensor Association

45 Multiparameters Sensor Interfacing 45 International Frequency Sensor Association

46 Multiparameters Sensor Interfacing (cont.) >M4 ; Duty-cycle measurement initialization >S ; Start a measurement >R ; Read a result ; Duty-cycle measurement result indication >ME >AX ; Frequency measurement initialization on the 2 nd input FX2 ; Appropriate X conversion error set up >S ; Start a measurement >R ; Read a result ; Frequency measurement result indication 46 International Frequency Sensor Association

47 I 2 C Interface to TAOS Opto Sensors <06><00> ; Frequency measurement initialization <02><00> ; 1 % conversion error set up <09> ; Start a measurement <07> ; Get measurement result in BCD format 47 International Frequency Sensor Association

48 SPI Interface to TAOS Opto Sensors 48 International Frequency Sensor Association

49 Temperature Sensor System (I) >MB; Pulse interval T1 measurement >S; Start a measurement >R; Read a result for T1 >MC; Space interval T2 measurement >S; Start a measurement >R; Read a result for T2 49 International Frequency Sensor Association

50 Temperature Sensor System (II) >MF; >A3; >S; >R; >MC; >S; >R; T x =T1+T2 measurement 0.1 % T x conversion error Start a measurement Read a result for Tx Space interval T2 Start a measurement Read a result for T2 50 International Frequency Sensor Association

51 TMP05/TMP06 Sensors Interfacing (a) (b) TMP05/TMP06 interfacing: T1 and T2 time intervals measurement (a), and period (T1+T2) and space interval (T2) measurement (b) 51 International Frequency Sensor Association

52 MAXIM Temperature Sensors Interfacing (I) (a) (b) MAX6576 period output sensor interfacing (a) and MAX6577 frequency output sensor interfacing (b) 52 International Frequency Sensor Association

53 MAXIM Temperature Sensors Interfacing (II) MAX6676 to UFDC-1 interfacing functional diagram 53 International Frequency Sensor Association

54 Accelerometers Based Systems (I) ADXL202 to UFDC-2 interfacing functional diagram. 54 International Frequency Sensor Association

55 Accelerometers Based Systems (II) ADXL210 to UFDC-2 interfacing functional diagram. 55 International Frequency Sensor Association

56 Accelerometers Based Systems (III) ADXL213 to UFDC-2 interfacing functional diagram. 56 International Frequency Sensor Association

57 Rotation Speed Measurement System 57 International Frequency Sensor Association

58 Commands Example (RS-232 interface) >MA ;Rotation speed measurement initialization >Z0C ; Set up Z=12 (10) =C (16) >A9 ;Choose the conversion error % >S ;Start a measurement >R ;Read a result of measurement in rpm 58 International Frequency Sensor Association

59 Rotation Acceleration Measurement ε x = n 1 t 2 n 2, where n 1 and n 2 of rotation speed and time interval for the second measurement t 2 59 International Frequency Sensor Association

60 Digital Humidity Sensors and Data Loggers PC UFDC-1 8 (GND) 7 (+5V) RS RH out EE05 (a) (b) 60 International Frequency Sensor Association

61 Temperature and Humidity Multisensors System Multisensors systems with the HTF3130 sensor for humidity measurement (the second channel) and temperature sensor MAX6576 temperature measurement (the first channel) 61 International Frequency Sensor Association

62 Commands Example (RS-232) >M1; Period measurement, 1st channel, MAX6576 temperature sensor >A2; Choose the conversion error 0.25 % >S; Start a measurement >R; Read a result (period proportional to the temperature) >ME; Frequency measurement, 2nd channel, HTF3130 humidity sensor >A2; Choose the conversion error 0.25 % >S; Start a measurement >R; Read a result (frequency proportional to the humidity) 62 International Frequency Sensor Association

63 Pressure Sensors Interfacing Connection diagram for 8000 Series of frequency output depth sensors from Paroscientific, Inc. 63 International Frequency Sensor Association

64 Commands Example (RS-232) >M0 ; Frequency measurement initialization in the first channel >A0 ; Choose the conversion error 0.1 % >S ; Start a measurement >R ; Read a result proportional to temperature >ME ; Frequency measurement initialization in the second channel >A0 ; Choose the conversion error % >S ; Start a measurement >R ; Read a result proportional to pressure 64 International Frequency Sensor Association

65 Pressure Gauges LCD pressure gauge based on UFDC-1 65 International Frequency Sensor Association

66 Digital Magnetic Sensors and Systems HAL819 to UFDC-1 interfacing circuit >M4; Duty-cycle measurement initialization (mode 4) >S; >R; Start measurement Read result 66 International Frequency Sensor Association

67 Further Development New method with non-redundant reference frequency New ICs: UFDC-1M-16 (a high speed version of UFDC-1), UFDC-2 and Universal Sensors and Transducers Interface (USTI) IEEE 1451 standard support Modern MEMS-based technologies 67 International Frequency Sensor Association

68 UFDC-2 2 and USTI UFDC-2 it is the UFDC-1 + frequency deviation (absolute and relative) measuring mode. Improved metrological performances: extended frequency range up to 9 MHz (144 MHz with prescaling), programmable relative error up to %, etc. Improved calibration procedures USTI it is the UFDC-2 + resistance, capacitance and resistive bridge measuring modes 68 International Frequency Sensor Association

69 IEEE 1451 Standard The standard defines the concept of plug-andplay sensors with analog outputs, maintaining compatibility with the large existing base of analog instrumentation and interfaces. IEEE 1451 family of standards become more and more popular Since 2004 more than 3200 different models of sensors were manufactured according to IEEE International Frequency Sensor Association

70 Standard Extension Network-Capable Application Processor (NCAP) Digital, Point-to-Point IEEE Distributed Multidrop Bus IEEE Digital TII Interface Txdcr Bus Interface TEDS FDC Txdcr TEDS FDC Txdcr Smart Transducer Interface Module (STIM) Transducer Bus Interface Module (TBIM) Network IEEE Common Object Model IEEE Common Functiona - lity & TEDS Wireless IEEE Frequency+ Digital IEEE Wireless Interface TEDS Txdcr FDC TEDS Txdcr Wireless Transducer Mixed -Mode Transducer Any Network TII - Transducer Independent Interface Txdcr - Transducer 70 International Frequency Sensor Association

71 Physical Representation of IEEE UFDC-1 Sensor FDC TEDS Bus Interface TII bus NCAP IEEE International Frequency Sensor Association

72 TEDS Example 72 International Frequency Sensor Association

73 Mix-Mode Mode Interface for Frequency Sensors IEEE Plug-and-Play Sensor Data Acquisition System Sensor Frequency Schmitt- Trigger Frequency Signal Output TEDS Digital Digital Signal I/O Class II multiwire interface 73 International Frequency Sensor Association

74 Technologies IC, ASIC Hybrid Sensing Element 1 UFDC or USTI (Core) Digital Output MEMS Sensing Element 2 TEDS System in Package System-on-Chip (SoC) Sensing Element 3 VFC Sensors and sensing elements 74 International Frequency Sensor Association

75 MEMS Oscillators Next generation oscillator technology Smaller, higher-precision references Immune to temperature and vibration Long-term stability of 0.05 ppm 20 ppm frequency variations Can go in plastic packages Much more rugged than quartz crystal oscillators 75 International Frequency Sensor Association

76 CMOS System-on-Chip Sensors system does not require any external time or frequency references UFDC lets solve problems with the interface circuit design and additional circuitry for MEMS oscillators in order to increase its short frequency stability MEMS Oscillator f x1 f o Universal Frequency-to- Digital Converter MEMS Sensor 1 f x2 MEMS Sensor 2 Bus Output SoC 76 International Frequency Sensor Association

77 Conclusions Task of creation of different smart digital sensors and systems for various physical and chemical, electric and non electric quantities is one of the most perspective and urgent task New smart sensors systems design methodology lets essentially reduce production costs and time-to-market Integrating all components of sensor system into a single SoC or MEMS with advanced processing and conversion methods in many cases will allow to achieve magnificent technical and metrological performances 77 International Frequency Sensor Association

78 References [1]. Kirianaki N.V., Yurish S.Y., Shpak N.O., Deynega V.P., Data Acquisition and Signal Processing for Smart Sensors, John Wiley & Sons, Chichester, UK, 2002 [2]. Smart Sensors and MEMS, ed. by S.Y. Yurish and M.T. Gomes, Springer Verlag, 2005 [3]. Sensors Web Portal: 78 International Frequency Sensor Association

79 Acknowledgment This research and development were supported in the frame of EU Marie Curie Chairs (EXC) project MEXT-CT CT SMARTSES and partly by the International Frequency Sensor Association (IFSA). 79 International Frequency Sensor Association

80 Questions? 80 International Frequency Sensor Association