|
|
| 2. Primary Measuring Element Selection and Characteristics |
|
When selecting the sensor, there are a number of factors to keep in mind. Knowing these
factors in advance, and the needs of the process, saves a buyer time and money. Before any
of these factors can be considered, the intended use of the sensor must be known. Once
that is established, several factors should be considered:
- Range: What is the normal range over which the controlled variable might vary? Are there
extremes to this?
- Response time: The amount of time required for a sensor to completely respond to a
change in its input.
- Accuracy: How close the sensor comes to indicating
the actual value of the measured variable?
- Precision: How consistent the sensor is in
measuring the same value under the same operating conditions over a period of time?
- Sensitivity: How small a change in the
controlled variable the sensor can measure?
- Dead band: How much of a change to the process
is required before the sensor responds to the change?
- Costs: What are the costs involved - not simply the purchase cost, but also the
installed/operating costs?
- Installation problems: Are there special installation problems, e.g., corrosive fluids,
explosive mixtures, size and shape constraints, remote transmission questions, etc.?
|
| 2.1 Range |
|
Range is the region in which the controlled variable might vary, in both normal and
extreme situations? For instance, if a process normally has a pressure of between 200 and
300 pounds per square inch, a sensor that measures from 100 to 400 pounds would be
desirable; this allows for extreme conditions to be measured as well as normal ones. A
broad range also allows the operator time to respond when measurements occur outside the
norm. Ideally, a process should be 40 to 60, or 30 to 70 percent of the range most of the
time and under normal operating conditions.
|
| 2.2 Response time |
|
Response time is the amount of time required for a sensor to respond completely to a
change in input. The response time of the control loop is the combination of the responses
of all the parts, including the sensor. An important objective of control system design is
to match correctly the time responses of the control system (and its measurement systems)
to that of the process. In general, a system with a quicker response time will be more
expensive.
Sensor Response
Time (Time Constant)
This figure shows a key principle of sensor response time. It illustrates that in a
fixed amount of time, known as the "time constant,"
here shown as 3 seconds, the sensor registers 63.2% of the total change. In the next time
constant, the sensor registers 63.2% of the remaining difference, and so on. In each time
constant, the sensor registers the same percent of the remaining difference.
|
| 2.3 Accuracy vs. Precision |
Precision and Accuracy
These figures contrast the term "precision"
as shown above, with the term "accuracy"
below. The actual value (70) is shown as a solid line in both figures. The measured values
are shown as dots. Accuracy is a measure of how close the sensor comes to indicating the
actual value of the measured variable. Here we see that the measured values are scattered
above, below and on the actual value. Accuracy is always given in terms of inaccuracy such
as +/- 2% or +1%, -3%. Precision is a measure of the consistency of a sensor is in
measuring the same value under the same operating conditions over a period of time. The
figure shows precise measured values, all the same distance below the actual value.
Precision is synonymous with repeatability and may be specified as a range or value
excursions or as a percent.
2.3.1 Accuracy
Most devices are rated on their accuracy, not their precision. The specifications
usually state that the device is accurate to plus or minus some value. Thus, with
accuracy, the deviation is known, but not the direction of the deviation. For example, if
a watch has an accuracy of ten minutes, it means an accuracy of plus or minus ten minutes.
The owner of this watch has an appointment for 2:30 p.m., but must arrive at 2:20 in order
to be on time because of the rated accuracy of the watch. The owner cannot know if the
watch is showing the time ten minutes early or ten minutes late or somewhere in between.
2.3.2 Precision
Precision is always within a given value and is always in the same direction. Thus, a
precise measurement may be wrong, but it is consistent. For example, another watch has a
precision of plus five minutes. The owner of this watch knows that she can arrive at her
2:30 p.m. appointment at 2:35 (on the watch) and still be on time.
|
| 2.4 Sensor sensitivity |
|
The sensitivity of the sensor is a measurement
of how small a change in the controlled variable
it can actually measure. The greater the sensitivity, the greater the sensor's reaction to
an input stimulus.
|
| 2.5 Dead band and dead time |
|
Dead band is the "unresponsiveness"
of the sensor. It describes how much change to the process is required before the sensor
actually responds to it or even detects it. The term sensitivity has frequently been used
to denote dead band, but the terms are not truly interchangeable. Sensitivity refers to
the reaction of the sensor. Dead time applies to the time it takes for the sensor to
react.
|
| 2.6 Cost |
|
In addition to the above factors, cost must also be considered. Not only the cost of
the initial purchase, but also the cost of maintaining the instrument. Very often more
money is spent maintaining an instrument than buying it initially.
|
| 2.7 Installation problems |
|
Installation problems can include special problems in the environment such as humidity,
vibration, temperature, or dust. Installation problems can also be anything that causes a
problem to the devices installed, such as, installing the device in a difficult to reach
location.
|
|
