Home | Forum | DAQ Fundamentals | DAQ Hardware | DAQ Software
Input Devices | Data Loggers + Recorders | Books | Links + Resources
The highest quality, most up-to-date, most accurately calibrated and most carefully selected sensor can still give totally erroneous data if it’s not correctly applied. This section will address some of the issues that need to be considered to assure correct application of any sensor.
The following check list is derived from a list originally assembled by Applications Engineering in the late 1970s. It has been sporadically updated as additional issues were encountered. It’s generally applicable to all sensor applications, but many of the items mentioned won’t apply to any given specific application. However, it provides a reminder of questions that need to be asked and answered during selection and application of any sensor.
Often one of the most difficult tasks facing an instrumentation engineer is the selection of the proper measuring system. Economic realities and the pressing need for safe, properly functioning hardware create an ever-increasing demand to obtain accurate, reliable data on each and every measurement.
On the other hand, each application will have different characteristics from the next and will probably be subjected to different environments with different data requirements. As test or measurement programs progress, data are usually subjected to increasing manipulation, analysis and scrutiny. In this environment, the instrumentation engineer can no longer depend on his general-purpose measurement systems and expect to obtain acceptable data. Indeed, he must carefully analyze every aspect of the test to be performed, the test article, the environmental conditions, and, if available, the analytical predictions. In most cases, this process will indicate a clear choice of acceptable system components. In some cases, this analysis will identify unavoidable compromises or trade-offs and alert the instrumentation engineer and his customer to possible deficiencies in the results.
The intent of this section is to assist in the process of selecting an acceptable measuring system. While we hope it will be an aid, we understand it cannot totally address the wide variety of situations likely to arise.
Let's look at a few hypothetical cases where instrument selection was made with care, but where the tests were failures.
1. A test requires that low g, low-frequency information be measured on the axle bearings of railroad cars to assess the state of the roadbed. After considerable evaluation of the range of conditions to be measured, a high-sensitivity, low resonance piezoelectric accelerometer is selected. The shocks generated when the wheels hit the gaps between track sections saturate the amplifier, making it impossible to gather any meaningful data.
2. A test article must be exposed to a combined environment of vibration and a rapidly changing temperature. The engineer selects an accelerometer for its high temperature rating without consulting the manufacturer. Thermal transient output swamps the vibration data.
3. Concern over ground loops prompts the selection of an isolated accelerometer.
The test structure is made partially from lightweight composites, and the cases of some accelerometers are not referenced to ground. Capacitive coupling of radiated interference to the signal line overwhelms the data.
From these examples, we hope to make the point that, for all measurement systems, it’s not adequate to consider only that which we wish to measure. In fact, every physical and electrical phenomenon that is present needs to be considered lest it overwhelm or, perhaps worse, subtly contaminate our data. The user must remember that every measurement system responds to its total environment.
The prospective user is generally forced to make a selection based on the characteristics available on the product data sheet. Many performance characteristics are shown on a typical data sheet. Many manufacturers feel that the data sheet should provide as much information as possible. Unfortunately, this abundance of data may create some confusion for a potential user, particularly the new user. Therefore the instrumentation engineer must be sure he or she understands the pertinent characteristics and how they will affect the measurement. If there is any doubt, the manufacturer should be contacted for clarification.
The sensor and signal conditioners must be selected to work together as a system. Moreover, the system must be selected to perform well in the intended applications.
Overall system accuracy is usually affected most by sensor characteristics such as environmental effects and dynamic characteristics. Amplifier characteristics such as nonlinearity, harmonic distortion and flatness of the frequency response curve are usually negligible when compared to sensor errors.
Selecting a sensor/signal conditioner system for highly accurate measurements requires very skillful and careful measurement engineering. All environmental, mechanical, and measurement conditions must be considered. Installation must be carefully planned and carried out. The following guidelines are offered as an aid to selecting and installing measurement systems for the best possible accuracy.
The most important element in a measurement system is the sensor. If the data is distorted or corrupted by the sensor, there is often little that can be done to correct it.
Will the sensor operate satisfactorily in the measurement environment?
Will the sensor characteristics provide the desired data accuracy?
Calibration | Accuracy
Is the proper mounting being used for this application?
Is Insulating Stud Required?
Is Adhesive Mounting Required?
Cables and connectors are usually the weakest link in the measurement system chain.
Will the cable operate satisfactorily in the measurement environment?
Will the cable characteristics provide the desired data accuracy?
Is Sealed Connection Required?
Will the power supply operate satisfactorily in the measurement environment?
Is this the proper power supply for the application?
Will the power supply characteristics provide the desired data accuracy?
The amplifier must provide gain, impedance matching, output drive current, and other signal processing.
Will the amplifier operate satisfactorily in the measurement environment?
Will the amplifier characteristics provide the desired data accuracy?
Data Acquisition and Readout
Does the remainder of the system, including any additional amplifiers, filters, data acquisition and readout devices, introduce any limitation that will tend to degrade the sensor-amplifier characteristics?
Check: ALL of previous check items, plus Adequate Resolution.
Even the most carefully and thoughtfully selected and calibrated system can produce bad data if carelessly or ignorantly installed.
Is the unit in good condition and ready to use?
Is the mounting hardware in good condition and ready to use?
Is the cable in good condition and ready for use?
Are the units in good condition and ready to use?
When the above questions have been answered to the user's satisfaction, the measurement system has a high probability of providing accurate data.
Updated: Monday, September 16, 2019 16:26 PST