Page 26 - EE Times Europe Magazine | February 2020

24 EE|Times EUROPE

INDUSTRY 4.0
Designers Turning to MEMS

for Condition Monitoring of Power Devices

By Maurizio Di Paolo Emilio

s a fundamental tool to improve
production efficiency, condition
monitoring is accelerating the
A transition to Industry .0. It enables
continuous monitoring of electromechanical
equipment and power devices, facilitating
early detection and real-time diagnosis of
possible anomalies or failures. By identify-
ing and isolating potential problems before
they can compromise production, condition
monitoring makes it possible to optimize
the maintenance function and maximize Figure 1: Block diagram of a typical condition-monitoring system (Image: Analog Devices)
equipment uptime by scheduling mainte-
nance-related downtime periods efficiently.
Power devices such as electric motors, automatically, reliably, and without human A third common problem is parallel or
pumps, bearings, and encoders are subjected intervention. To illustrate, let’s consider angular misalignment of the rotating shafts.
to electrical, mechanical, and environmen- a classic example that’s common to most Misalignment induces considerable mechan-
tal stresses that can negatively affect their industrial production plants rolling- ical stress, which can cause components to
operation. In addition to monitoring electrical element bearing failure that results from fail. The problem typically is detected by
parameters such as voltage and current, bearing wear and tear. analyzing the second harmonic speed (i.e.,
it is essential to perform vibration, noise, Bearing wear occurs in virtually all types 2 the fundamental frequency) of the system
and temperature measurements to ensure of rotating machinery, including motors, tur- rotation speed.
bines, pumps, fans, and computer numerical Misalignment measurement is carried out
Microelectromechanical- control (CNC) machines. Typically, a high-fre- using multiple accelerometer sensors capa-
ble of detecting the number of vibrations at
quency noise outside the audio band is the
system (MEMS) sensors first symptom of a worn bearing. different points on the equipment. In this
Low-noise, high-resolution wideband
provide extended vibration sensors are used to measure those way, it is possible to determine the phase
difference between the measurements and
measurement ranges and frequencies, enabling early detection of bear- thereby determine the type of misalign-
ment (parallel, angular, or a combination of
ing wear caused by the detachment of small
high reliability, even under metal fragments. Mathematical techniques the two).
Figure 1 shows a typical condition-mon-
the most severe operating such as envelope detection — usually com- itoring system. The data acquired by each
bined with complete spectral analysis in the
conditions. frequency domain — are deployed to detect, sensor passes through a low-pass filter
diagnose, and understand the implications of (LPF), which eliminates undesired frequency
failures affecting the bearings. components a low-noise amplifier and an
equipment health. A modern condition-mon- Another important factor to monitor is analog digital converter. A digital system
itoring system typically comprises sensors imbalance, which in this context is an uneven (microcontroller or microprocessor) processes
of various types, usually installed directly on distribution of mass that can cause the center information and transmits it to the remote
power devices; digital electronic circuits for of mass to move away from the center of rota- cloud server via a wired or wireless network
collecting and sending data to the cloud; and tion. Common causes include design error, connection.
an appropriate software application, available faulty installation, component wear, and the Most of the sensors used for this appli-
both on desktops and on mobile devices. presence of dirt or other contaminants. cation are now manufactured using
By accessing data on cloud servers, For example, accumulation of dust or microelectromechanical-system (MEMS)
maintenance personnel can get a clear idea grease on a fan used for cooling motors or technology to provide extended measurement
of the health of the devices and take correc- other power devices can unbalance the fan ranges and high reliability even in the most
tive action to forestall problems. Condition and cause a blade to break. severe operating conditions.
monitoring therefore changes the mainte- Imbalance generates a signal with the
nance approach from reactive to predictive, same frequency of rotation as the system and COMMERCIAL-GRADE SENSORS
potentially lowering maintenance costs amplitude proportional to the square of the Analog Devices’ portfolio of MEMS inertial
as it shortens downtime and increases pro- frequency. Therefore, imbalance detection sensors includes accelerometers, gyroscopes,
duction efficiency. is performed through analysis in the fre- and inertial measurement units (IMUs) for
quency domain. addressing complex application needs. The
CONDITION-MONITORING SENSORS Low-noise, wideband sensors are used for company’s ADIS1622 is a highly integrated
Sensors are deployed in condition-mon- imbalance detection, focusing the analysis on vibration-sensing system that combines
itoring systems to send data to the cloud the first-order harmonic components. triaxial acceleration sensing with advanced

FEBRUARY 2020 | www.eetimes.eu