Large industrial plants rely on electronic instrumentation to deliver the accurate, stable measurements for control processes. From fluctuating temperatures and humidity to harsh manufacturing conditions and elapsed time, however, various environmental and application-related factors impact stability.
When sensors or instruments suffer from performance decline, this phenomenon is known as drift. When drift happens, the sensors’ measurement data becomes unreliable, which can, in turn, affect the quality of factory production. Through calibration, we can identify and also resolve drift, but the calibration process can be costly and also contribute to factory downtime.
By gaining a better understanding of drift behaviour and minimizing such behaviour in voltage references, we can tailor an appropriate calibration strategy. This approach reflects the fact that sensors and instruments are only as accurate as their voltage references.
Why It’s Smart to Have More than 1000 Hours of Test Time
The parameter that quantifies drift in voltage references is called long-term stability. Long-term stability is defined as a slow change in output voltage, generally in one direction over months of operation. The value is expressed in ppm/1000 hours. However, given the implications related to voltage reference drift, semiconductor manufacturers are encouraged to exceed 1000 hours of test time to ensure accurate, reliable operation. Typically, these kinds of tests are performed on a selected sample population under set conditions for nominal input voltage, temperature, humidity, and load.
My mini application note, Voltage Reference Long-Term Stability Reduces Industrial Process Control Calibration Costs, provides details on a test case that demonstrates extended long-term stability testing for the MAX6126 ultra-high-precision, ultra-low-noise voltage reference. This particular voltage reference has a long-term stability of 20ppm/1000 hr (typ). Read the application note to learn how increasing long-term stability test time can help instrumentation and process control designers minimize the need for calibration and, therefore, downtime.