Transcript Section 6- Electronic data collection

Electronic Data Collection
Colin S. Campbell Ph.D.
Research Scientist
Decagon Devices, Inc.
Outline
 Data collection for the early
scientist
 Progress toward modern field
techniques
 Converting electronics into
information
 Electrical Engineering meets the
scientist
 Assessing the requirements of a
project
 Making the right system choice
Field research: Quantify physical
environment
 Early pioneers in environmental biophysics
 Howard Penman at Rothamsted Research
Station
 Manual readings or strip chart/disk recordings
 Sleepless nights
 Seminal paper on evapotranspiration
 Champ Tanner at U. Wisconsin
 Travel trailer
Changing world of measurements
 There’s got to be a better way
 All-nighters at research site not
terribly popular
 Miss fast changes or events with
human sampling
 No control of processes
 Goals
 Make unattended measurements
 Store measurements for analysis
later
 Transform data into information
and understanding
Requirements for field research
 Possible needs
 Sensors that generate electrical signals that
can be correlated with environmental
phenomena
 System to read electronic signals and store
them
 Make decisions based on measurements
 Control external systems based on analysis
Modern field research
 Sensors
 No limit to parameters that can be measured
 Passion for instrumentation design
 Only challenge is to find correlations
 Indirect measurements

Measuring one parameter and inferring the one of interest
 Data logger
 Basically glorified multimeter and oscilloscope
 Repository for raw sensor output
 Interprets electronic signals and stores them
Evolution of measurement:
Temperature example
 First automation by strip chart recorder
 Change in temperature of bimetallic strip
 Deflection calibrated to known temperatures
 Temperature variation changed pen position
 Ink recorded changes over time
 Data evaluated by hand
 Widely used
Conversion to an electrical signal:
The thermocouple
 Seebeck effect
 Two dissimilar metals jointed
together produce voltage
potential when differentially
heated
 Potential related to
temperature difference

Correlation (copper-constantan
thermocouple) ~ 40 mV per oC
 Measurement of minute
voltage changes provides
accurate temperature


Assuming know the
temperature of one junction
Electrical measurement is
accurate
Other measurement techniques:
Temperature
 Thermocouple limitations
 voltage accuracy
requirement
 reference temperature
 Alternatives
 Thermistor, platinum
resistance thermometer

Change electrical resistance
with temperature
 Diode

Voltage drop across a PN
junction
Data loggers and sensors
Sensor signal types
 Four general types of electronic sensor output
 Voltage

Probably the most common type

Includes thermocouples, radiation sensors, some anemometers,
etc.
 Current


Often used over long cable distances
Common to some measurement and control industries

Rain gauge, some anemometers, some soil moisture sensors,
etc.
 Pulse or switch closure
 Digital

Typical of sensors measuring more than one parameter

Allows for more than one signal per input location
Data logger types
 Plug and play (P&P)
 Decagon Em50, Em5b
 Onset Hobo
 CrossBow eKo
 Measurement and Control Systems (MCS)
 Campbell Scientific CR1000, 3000, etc.
 DataTaker DT80
 National Instruments LabView
Choosing a data logger: Things to
consider
 What electronic outputs do you need to
measure?
 Voltage, current, pulse, digital
 How many sensors are you putting at each
research site?
 How often will you be storing a
measurement?
Choosing a data logger: Things to
consider
 Will some measurements need to be made
more often than others?
 >10 Hz (i.e. eddy covariance)
 1 minute (i.e. radiation)
 Do you need to control anything with your
system (lights, heater, valve, etc.)?
 Do you have the time or resources to
program and setup the system?
Choosing the right
system
No
No
Yes
Yes
P&P
MCS
P&P
MCS
P&P
MCS
P&P
MCS
Choosing a data logger
Plug-and-play data logger
 Built for specific sensor
measurements or specific
sensor types
 Allow only a minimum of
configuration
 Date/time
 Measurement interval
 Sensor type
 Limited sensor inputs
 Low flexibility for sensor
types
Plug-and-play data loggers
Advantages
 Fast configuration
 Simple deployment
 No/low programming
complexity
 Simple data collection
and analysis
 Straight-forward sensor
integration
 Low power consumption
 Price
Disadvantages
 Limited sensor types
 Limited input ports
 Little or no
configurability
 No event-based
sampling
 No/little external
control
Choosing a data logger
Measurement and control systems (MCS)
 Build for general purpose measurement

Measure most types of voltage, current, pulse, and digital
sensors
 Highly configurable




Many different measurement and control option
Programming allows for multiple measurement intervals
On board data processing and decision making
High speed measurement
 Expandable

Add additional sensor capacity
 Accurate

May utilize high resolution signal processing for accuracy
Measurement and Control Systems
Advantages
 Configurability
 Precision and accuracy
 Programmability
 Speed
 Decision making and
control
 Data processing
Disadvantages
 Programming
 Configuration
 Installation and setup
 Power
Characteristics to evaluate

Required resolution and range
 Thermocouple


0.1o C resolution = 4 mV data logger resolution
50o C range = 2000 mV data logger range
 Water content sensor



0.1% VWC ~ 1 mV data logger resolution
100% VWC ~ 1000 mV data logger range
Excitation
 Many sensors require a voltage be provided to the sensor


Decagon EC-5 – 2.5 or 3V regulated
Gill WindSonic anemometer – 12V unregulated
 Excitation requirements vary mV to 10s of volts
 Many data loggers have limited excitation options
Characteristics to evaluate (cont.)
 Analog to digital converter (ADC)
 Voltage and current measurements are made by an
ADC
 Precision of ADC defines accuracy of the measurement
 Defined by bits

i.e. 12 bit ADC  0 to 4095  2.5 V range  0.61 mV/bit


Obviously not good enough for the thermocouple, but good for
VWC
24 bit ADC  0 to 16777216  2.5 V range  0.15 mV/bit

Good enough for thermocouples
 Noise rejection
 Multiple sources of ambient electrical noise

60 cycle from electricity, radio frequency
Data logger applications
 Making the decision
 Many choices available
 Sometimes confusion trying to decide which
one will work the best
 Discuss some applications from personal
experience
 Caveat: Vast majority of my experience is with
Decagon and Campbell Scientific data logger
 Many other manufacturers that you may consider

Delta-T, Onset, DataTaker, Stevens, Unidata, etc.
Rice net carbon exchange
Conditional sampling
 Stored 77 different outputs





CO2, H2O concentration (voltage output from IRGA)
Pyranometer, quantum sensor, net radiation (mV)
Water content (pulse count)
Rain gauge (pulse count)
3-D sonic anemometer (digital)
 Data downloaded by cell phone (2.5 h away)
 5 Marine batteries charged by 6-12V solar panels
 2 CR10X dataloggers, 2 MUX, Relay driver

Flexibility, control, programmability, storage, communication
Turf grass watering
Turf field with pop-up
sprinklers
 Control based on
distributed water content
sensors
 VWC at several locations
 Threshold values control
solenoid values for
sprinklers
Decision: things to consider
 P&P data logger
Easy to read VWC sensors
Fast installation
Low power requirements
Data easily collected and
graphed over radio or cell
phone
 Often lack control
capability




 MCS
 Required for system control
 Large sensor input capacity
Distributed field analysis of physical
and morphological interactions
Site description
 37 ha research farm
 Large topographical
variation
 Goal
 Investigate water,
temperature, and EC
variation in relation to soil
morphology
 42 distributed profiles
 Measurement at 5 depths
System choice
 Plug-and-play logging
system well suited for
distributed networks
 Small number of sensors at
each site
 Radio or cell phone
communications
 Fast setup
 Low power use
 No requirement for control
or specialized sampling
Fast, simple plot measurements
Description
 Goal
 Compare performance of
drought tolerant cultivars
 Requirements
 Soil moisture, temperature
in plots
 Weather station parameter
in central location
 Simple deployment
Considerations
 P&P systems require no
programming
 No specialized sample
timing or control
 Self contained loggers
require no enclosure
setup or external power
Summary
 Data loggers simply measure and store
electronic signals
 Art of instrumentation is to dream up new
ways or correlating electronics to science
 Data logger choices are numerous
 Carefully determine all experimental needs
 Evaluate system specifications