Email this Page
Print this Page

PROCESS CONTROL

Important Considerations for Sensor Choice and Placement

Attention to detail in sensor selection, placement, and data acquisition ensures the best possible outcome in any PSV implementation. Here are some factors to take into consideration.
  • Determine what physical properties are important to the process. In this example, force, displacement, and possibly temperature would be important.
  • Determine to what level of accuracy, resolution, and rate of change the process should be measured. Full-scale range should be no more than what is absolutely necessary. The sensor time constant, or frequency response, is one of the most common oversights in the sensor selection process.
  • Choose a sensor that is appropriate. In almost all cases, various types of sensors are available to measure the same physical property. For example, pressure may be measured via strain gauge, piezoelectric deflection, capacitance, etc. Each method of sensing has inherent advantages and disadvantages. Selection of the sensor physics should be carefully considered for a particular process.
  • Place the sensor as close to the device process interaction as possible. Often sensor location is far removed from the physics of interest and, as such, significant system measurement errors are introduced. In the case of this spring test, the load cell should be virtually the last piece of tooling to make contact with the spring.
  • Pay attention to good instrumentation techniques. In almost all cases, sensors produce an electrical output. Instrumentation techniques such as wire shielding and physical routing, galvanic isolation, linear power supplies, sensor environment, and low-impedance, high-bandwidth buffer amplifiers ensure a good signal-to-noise ratio.
  • Optimize the input range to the conversion range. Sampling accuracy of the sensor voltage to a number (A/D conversion) can have a great influence in measurement accuracy if the input range is not optimized to the conversion range.
  • Ensure that the sampling rate is sufficient to prevent aliasing of the signal. Ideally, with current data acquisition systems, 10–100 times oversampling with digital filtering produces systems with excellent gauge reproducibility and repeatability.
Copyright ©2007 Medical Device & Diagnostic Industry