[PPT]PowerPoint Presentation - IFSA Web site. Start Page · Web viewSelf-Adaptive Intelligent Sensors and Systems: From Theory to Practical Design Prof. Sergey Y. Yurish, Technical

Prof. Sergey Y. Yurish,Technical University of Catalonia

(CDEI-UPC Barcelona)

Self-Adaptive Intelligent Self-Adaptive Intelligent Sensors and Systems: From Sensors and Systems: From Theory to Practical DesignTheory to Practical Design

IEEE International Workshop on RObotic and Sensors Environments (ROSE’ 2008), Ottawa, Ontario, Canada, 17 October 2008

22 International Frequency Sensor Association ● www.sensorsportal.com

OutlineOutlineIntroductionModern SensorsSensor InterfacingSelf-Adaptive Intelligent Sensors Systems Design MethodologyPractical ExamplesFurther Perspectives and TechnologiesConclusions


IntroductionIntroduction

Intelligent (or smart) sensor systems are of great interest in many fields of industryClassical approach to such systems means that the information is in the amplitude of a voltage or current signalAnother approach relies on resonant phenomena and variable oscillators: information is embedded in the frequency or time parameters of signal


Smart Sensors World Smart Sensors World Market TrendsMarket Trends

According to new sensors market studies (Freedonia Group), the US sensor demand will grow 4.3 percent annually through 2012 and will reach $12.1 billion in 2010According to Frost & Sullivan the forecast for North American Smart Sensors Market is to reach $635.2 million in 2010 The past few decades have witnessed an explosive growth in sensors and sensor-based applications which has led to a greater demand for sensor interfacing integrated circuits (ICs) Strong growth expected for sensors based on MEMS-technologies, intelligent sensors and sensors with bus capabilities


Smart SensorSmart Sensor

Smart sensor Smart sensor (or intelligent sensor) is an electronic device including sensing element, interfacing, signal processing and one- or several intelligent functions as self-testing, self-identification, self-validation or self-adaptation


Smart SensorSmart Sensor

MUXMUX…

Signal Signal conditioningconditioning

Sensing Sensing elementelement

CC

• DSPDSP

• ADC or FDCADC or FDC

• CommunicationsCommunications


Sensors (IFSA study 2008)Sensors (IFSA study 2008)

Analog55%

Digital25%

Quasi-Digital20%


Quasi-Digital SensorsQuasi-Digital SensorsQuasi-digital sensorsQuasi-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

ff11/f/f22

NNxx

D.c.D.c.

TTxxffxx

PWPWMM


Quasi-Digital SensorsQuasi-Digital Sensors


Frequency AdvantagesFrequency Advantages

High Noise ImmunityHigh Power SignalWide Dynamic RangeHigh Reference AccuracySimple InterfacingSimple Integration and CodingMultiparametricity


Temperature SensorsTemperature Sensors


Pressure SensorsPressure Sensors


AccelerometersAccelerometers


Rotation Speed SensorsRotation Speed SensorsThere are many known rotation speed sensing principlesMagnetic sensors (Hall-effect and magnetoresistor based sensors)Inductive sensorsPassive and active electromagnetic rpm-sensors are from the frequency-time domain

Zfn xx

60 , where , where ZZ is the number of modulation rotor’s is the number of modulation rotor’s

(encoder’s) gradations(encoder’s) gradations (teeth) (teeth)


TAOS Light and Color TAOS Light and Color SensorsSensors

For TSL 230RD: fO = fD + (Re) (Ee),

where fO is the output frequency; fD 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/cm2


Humidity SensorsHumidity Sensors


Chemical, Gas andChemical, Gas and BiosensorsBiosensors

Sensors arrays (electronic noses and tongues)Square wave with a frequency inversely proportional to the sensor resistanceSensors Array based on chemisorbing polymer filmsAcoustic gas sensor based on a gas-filled cellQuartz Crystal Microbalance (QCM) sensorsSAW and bulk acoustic wave sensors


Magnetic SensorsMagnetic SensorsHAL810, 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.241-245


Other SensorsOther SensorsTilt and inclination sensors with PWM outputsTorque transducers with frequency outputLevel sensors with frequency outputConductivity sensor SBE4 with frequency outputFlow sensors with frequency output


Multiparameters SensorsMultiparameters SensorsColor sensor (TU Delft, The Netherlands): frequency is proportional to optical intensity (luminance) and duty-cycle is proportional to colour (chrominance)Pressure and temperature sensorsHumidity and temperature sensors (transmitters) from E+E Elektronik, Bitron, etc.


Design Task DefinitionDesign Task Definition

There are many quasi-digital sensors for any physical and chemical quantitiesThe 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 intelligent sensor systems ?


ConventionalConventional Methods 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/Tx)

Combined MethodInterpolation method (with digital interpolation)


Disadvantages of Disadvantages of Classical MethodsClassical Methods

High quantization error in low or high frequency range Quantization error dependents on frequencyRedundant conversion time for the standard counting method


Advanced MethodsAdvanced MethodsRatiometric Counting MethodReciprocal Counting MethodM/T Counting MethodConstant Elapsed Time (CET) MethodSingle- and Double Buffered MethodsDMA Transfer MethodMethod of Dependent Count (MDC)Method with Non-redundant Reference Frequency


Benefit and DemeritsBenefit and Demerits

Redundant conversion time due to prescribed conversion timeMeasurement frequency fx < f0

Two references (frequency and gate time)

Disadvantages:Disadvantages:

Advantage:Advantage:

Constant quantization error in all specified conversion range of frequencies


Method of Dependent Method of Dependent Count (MDC)Count (MDC)

Method was proposed in 1980 (modifications: 1993, 2004 and 2006)Most advanced method: constant programmable quantization error in all frequency range; non-redundant conversion time; the possibility to convert frequency fx higher, than the reference frequency f0 (fx >> f0 ); self-adaptation capabilitiesThe method has been developed for conversion of absolute, relative frequencies and periodsSuitable for single channel as well as for multichannel synchronous frequency conversions


Comparison of MethodsComparison of Methods MDC, M/TMDC, M/T


Method SelectionMethod SelectionThe 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 errorSelf-adapting method of dependent count and method with non-redundant reference frequency are very suitable for the usage in frequency-to-digital converters of smart sensors


Integrated FDCsIntegrated FDCsUSP-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 AutoTECTime-to-Digital Converter (TDC) from Acam-messelectronic GmbHSSP1492 - Sensor Signal Processor from Sensor Platforms, Inc. (USA, 2006)


ICs DisadvantagesICs DisadvantagesAll ICs except TDCs are based on conventional methods of measurement, hence, quantization error is dependent on measurand frequency fx , many of ICs have redundant conversion timeThey cannot be used with all existing modern frequency-time domain sensors due to low accuracy or/and narrow frequency rangesThey do not cover all frequency–time informative parameters of electric signals.


Universal Frequency-to-Universal Frequency-to-Digital Converter (UFDC-1)Digital Converter (UFDC-1)

Low cost digital IC with programmable accuracy2 channels, 16 measuring modes for different frequency-time parameters and one generating mode (fosc/2 = 8 MHz)

Based on four patented novel conversion methodsShould be very competitive to ADC and has wide applications


FeaturesFeaturesFrequency range from 0.05 Hz up to 7 MHz without prescaling and 112 MHz with prescalingProgrammable accuracy (relative error) for frequency (period) conversion from 1 up to 0.001 %Relative quantization error is constant in all specified frequency range Non-redundant conversion timeQuartz-accurate automated calibrationRS-232/485, SPI and I2C interfaces


UFDC-1 Block DiagramUFDC-1 Block Diagram


Measuring ModesMeasuring ModesFrequency, fx1 0.05 Hz – 7.5 MHz directly and up to 120 MHz with prescallingPeriod, Tx1 150 ns – 20 sPhase shift, x 0 - 3600 at fx 500 kHzTime interval between start- and stop-pulse, x 2 s – 250 sDuty-cycle, D.C. 0 – 1 at fx 500 kHz Duty-off factor, Q 10-8 – 8.106 at fx 500 kHz Frequency and period difference and ratioRotation speed (rpm) and rotation accelerationPulse width and space interval 2 s – 250 sPulse number (events) counting, Nx 0 – 4.109


Evaluation Board Circuit Evaluation Board Circuit DiagramDiagram


UFDC-1 Evaluation BoardUFDC-1 Evaluation Board


UFDC-1 and Analog UFDC-1 and Analog Signal DomainSignal Domain

Any Voltage-to-Frequency Converter (VFC) can be used to convert an analog signal to quasi-digital (frequency) signal

RS-232/485RS-232/485

SPISPI

II22CC

Voltage (V) Voltage (V)

or or

Current (I)Current (I)

VFCVFCffxx, Hz, Hz

Digital Digital OutputOutput

SSeennssoorrss


Where to use the UFDC-1 ?Where to use the UFDC-1 ?Smart Sensors; Smart Sensors;

Quasi-digital and Quasi-digital and Digital sensors; Digital sensors; Multiparametric Multiparametric

SensorsSensors

Frequency Frequency CountersCounters

Tachometers and Tachometers and Tachometric Tachometric

SystemsSystems

Multimeters for Multimeters for Frequency-time Frequency-time

Parameters Parameters

ABS ABS SystemsSystems

DAQ boards for DAQ boards for Frequency-time Frequency-time Parameters Parameters

Virtual Virtual Instruments Instruments


Color-to-Digital ConverterColor-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-F to 0.1-F capacitor with short leads mounted close to the device package.


Light-to-Digital ConvertersLight-to-Digital Converters

(a)

(b)


CommandsCommands Example Example (RS-232 interface)(RS-232 interface)

>M0 ; Frequency measurement initialization

>A0 ; 1 % conversion error set up

>S ; Start a measurement

>R ; Read a result

1000.674946004319 ; Measurement result indication


Multiparameters Sensor Multiparameters Sensor InterfacingInterfacing


Multiparameters Sensor Multiparameters Sensor InterfacingInterfacing (cont.) (cont.)

>M4 ; Duty-cycle measurement initialization


>R ; Read a result

60.9786 ; Duty-cycle measurement result indication

>ME ; Frequency measurement initialization on the 2nd input FX2

>AX ; Appropriate ‘X’ conversion error set up


>R ; Read a result

100.578698673 ; Frequency measurement result indication


II22C Interface to TAOS C Interface to TAOS Opto SensorsOpto Sensors

<06><00> ; Frequency measurement initialization

<02><00> ; 1 % conversion error set up

<09> ; Start a measurement

<07> ; Get measurement result in BCD format


SPISPI Interface to TAOS Interface to TAOS Opto SensorsOpto Sensors


Temperature Sensor Temperature Sensor SystemSystem

>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


MAXIM Temperature MAXIM Temperature Sensors InterfacingSensors Interfacing

MAX6576 period output sensor interfacing (a) and MAX6577 frequency output sensor interfacing (b)

(a) (b)


Accelerometers Based Accelerometers Based SystemsSystems

Інклінометер


Rotation Speed Rotation Speed MeasurementMeasurement System


>MA ;Rotation speed measurement initialization

>Z0C ; Set up Z=12(10)=C(16)

>A9 ;Choose the conversion error 0.001 %

>S ;Start a measurement

>R ;Read a result of measurement in rpm

CommandsCommands Example Example (RS-232 interface)(RS-232 interface)


Digital Humidity Sensors Digital Humidity Sensors and Data Loggersand Data Loggers

PC

RS-2324

6

RH out

7 (+5V)

8 (GND)

EE05

UFDC-1

(a)

(b)


Temperature and Humidity Temperature and Humidity Multisensors SystemMultisensors System

Multisensors systems with the HTF3130 sensor for humidity measurement (the second channel) and temperature sensor

MAX6576 temperature measurement (the first channel)


Pressure Sensors Pressure Sensors InterfacingInterfacing

Connection diagram for 8000 Series of frequency output depth sensors from Paroscientific, Inc.


Commands Example Commands Example (RS-232)(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 0.001 %>S ; Start a measurement>R ; Read a result proportional to pressure


Digital Magnetic Sensors Digital Magnetic Sensors and Systemsand Systems

HAL819 to UFDC-1 interfacing circuit

>M4; Duty-cycle measurement initialization (mode 4)

>S; Start measurement

>R; Read result


UFDC-2 and USTIUFDC-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 0.0005 %, etc.Improved calibration proceduresUSTI – it is the UFDC-2 + resistance, capacitance and resistive bridge measuring modes


Adaptive AlgorithmsAdaptive Algorithms

An adaptation in smart sensors systems can be used for increasing of measurement accuracy, decreasing of measuring time, power consumption reduction, etc.Adaptive measuring algorithms:

where L is the algorithm of measurement; Ts , s and Ps are operations for speed and accuracy increasing; and power consumption decreasing respectively; j (t) is the input action

;* tLtLT jsjsj tLtLP jsjsj *


Parametric AdaptationParametric AdaptationFor the modified MDC:

fxjsj

fxjsj

IFiftL

IFiftLT**

**

,

,

;II f at

For the method with non-redundant reference frequency:

fxjsj

fxjsj

IFiftL

IFiftLP**

**

,

,

at ,II f

where Fx(*) is the characteristic of input action or measuring conditionsIf is the subset of certain area I of possible values of characteristic Fx (*)


Flowchart of Adaptive AlgorithmFlowchart of Adaptive Algorithm


Conversion Times Conversion Times vs. Relative Errorvs. Relative Error

0

0,05

0,1

0,15

0,2

0,25

0,3

0,35

0,0005 0,001 0,0025 0,005 0,01 0,025 0,05 0,1 0,25 0,5 1

Relative error, %

t, s UFDC-1UFDC-1M-16USTI, UFDC-2USTI - 1M-20


Adaptive Rotation Speed Adaptive Rotation Speed MeasurementsMeasurements

>M0A; Rotation speed measurement initialization in the 1st channel>Z30; Set up the modulation rotor teeth number Z=48(10)=30(16)>A09; Choose the relative error of frequency measurement 0.0005 %>S; Start a measurement>R; Read a result of measurement in rpm

; Here microcontroller or computer should check the condition for an algorithm changing and prepare the USTI to measure with highest speed (maximum relative error) if a critical rotation speed has been achieved:

>A00; Choose the relative error of measurement 1 %>S; Start a measurement>R; Read a result of measurement in rpm


Self-adaptive Smart Sensor Self-adaptive Smart Sensor System Based on MSP430System Based on MSP430 KK


Other ApplicationsOther ApplicationsPressure intelligent sensor system for gas pipelineTemperature, humidity, etc. self-adaptive smart sensors systemsRobotics


IEEE 1451 StandardIEEE 1451 StandardThe standard defines the concept of plug-and-play 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 popularSince 2004 more than 3200 different models of sensors were manufactured according to IEEE 1451.4


Standard ExtensionStandard Extension

IEEE 1451.2

IEEE 1451.3

IEEE 1451.5

IEEE 1451.4TEDS

Txdcr

IEEE1451.1

CommonObjectModel

IEEE1451.0

CommonFunctiona -

lity &TEDS

Network-CapableApplicationProcessor

(NCAP)

Network

Frequency+Digital

Wireless

DistributedMultidrop Bus

Digital,Point-to-Point

Mixed -ModeTransducer

WirelessTransducer

Transducer BusInterface Module(TBIM)

Smart TransducerInterface Module(STIM)

Any Network TII - Transducer Independent InterfaceTxdcr - Transducer

Dig

ital

TII

Inte

rfac

e TEDS

FDC Txdcr

Txdc

rB

usIn

terf

ace TEDS

FDC Txdcr

Wire

less

Inte

rfac

e TEDS

FDC Txdcr


Physical Representation Physical Representation of IEEE 1451.2of IEEE 1451.2

Sensor FDC

TEDS

BusInterface

TII busNCAP

IEEE 1451.2

UFDC-1


TEDS ExampleTEDS Example


Mix-Mode Interface for Mix-Mode Interface for Frequency SensorsFrequency Sensors

DigitalSignal I/O

FrequencySignalOutput

TEDS

Sensor Schmitt-Trigger

Frequency

Digital

Data Acquisition SystemIEEE 1451.4 Plug-and-Play Sensor

Class II multiwire interface


TechnologiesTechnologiesIC, ASICHybridMEMSSystem in PackageSystem-on-Chip (SoC)

DigitalOutput

SensingElement 1

SensingElement 2

SensingElement 3 VFC

UFDC or USTI

(Core)

TEDS

Sensors and sensing elements


Next generation oscillator technologySmaller, higher-precision referencesImmune to temperature and vibrationLong-term stability of 0.05 ppm20 ppm frequency variationsCan go in plastic packagesMuch more rugged than quartz crystal oscillators

MEMS OscillatorsMEMS Oscillators


CMOS System-on-Chip

fo

fx1

MEMS Oscillator

Universal Frequency-to-Digital Converter

fx2

MEMS Sensor 1

MEMS Sensor 2

BusOutput

SoC

Sensors system does not require any external time or frequency references UFDC and USTI lets solve problems with the interface circuit design and additional circuitry for MEMS oscillators in order to increase its short frequency stability


ConclusionsConclusionsTask of creation of different smart digital sensors and intelligent sensors systems for various physical and chemical, electric and non electric quantities is one of the most perspective and urgent taskNew smart sensors systems design methodology lets essentially reduce production costs and time-to-market Self-adaptive smart sensors and sensor systems based on novel modified method of the dependent count for frequency (period)-to-digital conversion of sensor’s outputs can be easy realized on UFDC and USTI integrated circuits well-suited for such kind of applications


Questions ?Questions ?

Documents

[PPT]PowerPoint Presentation - IFSA Web site. Start Page · Web viewSelf-Adaptive Intelligent Sensors and Systems: From Theory to Practical Design Prof. Sergey Y. Yurish, Technical