Data Acquisition Systems
What do you need to do?
As a first step to deciding what type or class of data acquisition system to use, consider the following:
- What physical parameters do you need to measure to answer the research questions?
- How many separate locations need to be observed?
- How frequently do measurements need to be made?
- Do you want to process individual measurements onboard the data acquisition system (e.g., process 1-second temperature observations to create and record 5-minute averages)?
- Is important to have data transmitted back to the researcher's office?
What type of data acquisition makes sense for this project?
There are several basic data acquisition system types that offer different characteristics:
Other aspects of system selection:
Multiplexing refers to using switches so several different sensors can be connected in turn to a single input on a data acquisition system, increasing the capacity of the system to monitor multiple sensors. Multiplexers make use of either electromechanical relays or solid state switches, and are often available through the manufacturers of data acquisition systems.
Communication between a remote research site and the researcher's office is generally desirable, and sometimes critical, in allowing evaluation of the performance of both the system under study and the data acquisition equipment without travel to the site. Data communications from remote sites can use any of several methods:
- Conventional ("landline") telephone service using modems
- Cellular phone service with data plans and communications
- Internet-based communications (TCP/IP)
Considerations in selecting sensors:
Can the sensor provide output across the expected range of experimental conditions to be studied? Can it tolerate exposure to the maximum possible conditions?
Some sensors require external power for operation. Some may allow for a wide range of supply voltage (e.g., 8 to 30 VDC) while others require a well-regulated power supply (e.g., 12 +/- 0.1 VDC). The current draw of sensors should be checked against the output capacity of the power supply.
Sensor output should be compatible with the data acquisition system. Analog-output sensors typically provide output voltages of 0 to 1V, or up to 0 to 12V. Output voltages that exceed the measurement capability of the data acquisition system can readily be dropped to a lower level by adding a voltage divider. In some cases, sensor output is proportional to power supply voltage. Check the capability of the data acquisition system before selecting sensors with low-range output, such as resistive bridge devices and thermocouples.
The digital inputs on a data acquisition system interpret the input as representing either a low value (low, zero, and off are all used to describe this state), or a high value (usually called high, one, or on). Digital output (i.e., an on/off signal) may be implemented using several types of hardware:
- A voltage that switches between 0 and a positive value, such as +5 V.
- A switch; with a voltage applied to one side, the output becomes either an on or off ("high" or "low") value.
- An open collector. This refers to the use of a transistor as a solid state switch.
Data acquisition system digital inputs may also have the ability to count digital pulses. The system manufacturer should specify a maximum input voltage that will always be interpreted as low, a range where the interpretation is ambiguous, and a minimum voltage, above which will always be interpreted as high. Make sure the sensors produce output that will always be interpreted correctly.
Open collector and switch outputs are usually connected to a voltage source through a "pullup resistor" that provides a stable positive signal when the transistor is inactive (or switch is open). When the transistor is energized (or switch closed), the voltage is drawn to near zero. With proper selection of the voltage source and pullup resistor, the input of the data acquisition system sees these two conditions as OFF and ON.
Accuracy and resolution
The accuracy, or maximum expected error, of sensors is usually reported in manufacturers specifications. The importance of sensor accuracy depends on the way the sensor is being used. Measurement of a water temperature difference of 5ºF across a heat exchanger may call for higher accuracy than measuring outdoor temperature for use in studying heating loads. Resolution, the smallest step or change that can be observed, is a function of both the sensor and the device that converts the output to a digital value (usually the data acquisition system). Better resolution becomes more important as smaller changes or differences are to be observed.
Other aspects of sensor selection and operation that may be important in some projects but are beyond our current scope include:
- Interchangeability of sensors on replacement
- Response time of sensors and data acquisition systems (e.g., warmup time, response time constant, switching time)
- Continuous excitation and self-heating of sensors
- Temperature: Thermocouples, thermistors, RTDs, Radiant temperature
- Moisture: Relative humidity, surface wetness, moisture content of soil, concrete, other materials
- Current: Current switches, transducers
- Liquid flow
- Liquid level
- Air flow, wind speed
- Heat flux
- Sound levels, vibration
- Stress & strain
- Gasses (CO, CO2, O2, organics, etc.)
In most cases, the manufacturers of packaged data acquisition equipment provide software for setting operating parameters and for full-featured programming of more sophisticated systems. Manufacturers' software may also provide utilities for saving data files in standard formats, for simple analysis (e.g., finding maximum and minimum values), and for graphing.