Page 308 - The ROV Manual - A User Guide for Remotely Operated Vehicles 2nd edition
P. 308
300 CHAPTER 12 Sensor Theory
phenomenon (e.g., temperature is the simple measurement of molecular excitation but can be mea- sured via thermistor, fluid pressure, thermocoupler, or resistance temperature detector (RTD)). For example, RTDs are sensors used to measure temperature by correlating the resistance of the RTD element with the temperature. Following this example, various technologies for sensing temperature are sensitive/accurate only within certain ranges. Also, the sensor output has a host of issues involved before a logical and practical use of that output can be employed. An example of accurate but unusable sensor output would be an ambient water temperature sensor mounted onto a high- pressure hydraulic pump aboard an ROV—the sensor would accurately measure the water tempera- ture surrounding the pump, but it would indicate a temperature (in all likelihood) higher than the ambient temperature due to the heat generated by the pump. Or if an obstacle avoidance sonar transducer head was embedded in the ROV’s flotation block, the sonar transducer would accurately sense the acoustic reflection from the flotation foam but not the desired echo from the local obstructions. How you incorporate a selected sensor is as critical as the design of the sensor itself.
The perfect sensor has the following characteristics:
1. The sensor senses only the desired physical phenomenon.
2. The sensor is insensitive to other environmental or physical factors.
3. The sensor does not influence the item being measured.
4. The sensor’s electrical output signal is linearly proportional (or can be made as such—e.g.,
plotted linearly on a logarithmic scale).
Have you ever had your eyes “deceive” you? Accurate interpretation of sensor output, as well
as proper sensor placement, is required in order to achieve accurate measurement for either human or machine decision making.
12.1.3 Sensor output
A sensor typically functions as a transducer converting physical phenomena to electrical current for later conversion into machine or human-readable information. With regard to sensors, the term “Garbage In/Garbage Out” is certainly applicable as sensors are prone to a host of calibration errors, sensor performance issues, and user interpretation problems.
Sensor performance varies by device. Sensor performance characteristics are measured with the following qualifiers:
• Transfer function: This defines the relationship between the physical input to the device (i.e., direct/indirect sensing of physical phenomena) and the electrical output from the device. An input/output graph is normally generated for the sensor so that the user of the information may interpret the output of the sensor (e.g., temperature sensor is 420 mA DC or 1050 mA DC) and either display the output (for a visual display) or input the voltage (for machine control) into an intelligent device for commands or other usage.
• Sensitivity: This characteristic (sometimes termed “gain”) is the scaling of the sensor’s input to its output. This parameter is further described as the ratio of a change in output magnitude to a corresponding change in steady-state input (which caused the output). This, of course, is expressed in ratio form. A temperature sensor has a high sensitivity if a small change in
