Simulated Sensor-Based Vibration Measurement Chain
Koukkula, Sanna (2026)
2026
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202604015464
https://urn.fi/URN:NBN:fi:amk-202604015464
Tiivistelmä
This thesis presents the design of a vibration measurement specified for condition monitoring on industrial machinery. The system was based on a flexible capacitive electromechanical (EMFi) force sensor, which functions as a quasi-piezoelectric transducer, using a permanently charged dielectric film. The sensor was simulated with a substitute circuit, followed with an analog front-end including a fully differential amplifier (FDA) and an analog-to-digital converter (ADC). The focus of the thesis was on designing a configurable measurement chain with adjustable frequency band and gain parameters for various use cases and requirements.
Requirement for the measurement system was to achieve a desired frequency response with carefully implementing the filtering in three different stages. The sensor’s inherent capacitance created a high-pass characteristic, followed by a passive RC circuit which defined the low-frequency cutoff. Anti-aliasing low-pass filter with FDA was then added to the circuit to complete the desired band-pass filter. Different use cases require tuning of both the bandwidth and gain. Therefore, the analog front-end adjustment was presented together with the corresponding design equations. The performance of the measurement chain was evaluated with using LTspice simulations. An amplitude sweep test with two different gain configurations was performed to characterize dynamic range and clipping threshold. Total Harmonic Distortion (THD) percentage was monitored from the ADC output to identify the voltage level where the output signal started to distort. Linearity of the circuit with adjusted gain configuration was tested to verify that the signal remains linear throughout the chain. A frequency sweep test was performed to confirm the filters band-pass behaviour by comparing the results to the theoretical frequency response. The study concluded that the proposed architecture provides a flexible and reusable platform for EMFi-based vibration measurement and can be adapted to different sensor sizes and use cases. LTspice validation framework supports both synthetic test signals and real-world vibration data, enabling evaluation for multiple use cases. This approach ensures an optimized signal chain for each application, streamlining and cost-reducing product development.
Requirement for the measurement system was to achieve a desired frequency response with carefully implementing the filtering in three different stages. The sensor’s inherent capacitance created a high-pass characteristic, followed by a passive RC circuit which defined the low-frequency cutoff. Anti-aliasing low-pass filter with FDA was then added to the circuit to complete the desired band-pass filter. Different use cases require tuning of both the bandwidth and gain. Therefore, the analog front-end adjustment was presented together with the corresponding design equations. The performance of the measurement chain was evaluated with using LTspice simulations. An amplitude sweep test with two different gain configurations was performed to characterize dynamic range and clipping threshold. Total Harmonic Distortion (THD) percentage was monitored from the ADC output to identify the voltage level where the output signal started to distort. Linearity of the circuit with adjusted gain configuration was tested to verify that the signal remains linear throughout the chain. A frequency sweep test was performed to confirm the filters band-pass behaviour by comparing the results to the theoretical frequency response. The study concluded that the proposed architecture provides a flexible and reusable platform for EMFi-based vibration measurement and can be adapted to different sensor sizes and use cases. LTspice validation framework supports both synthetic test signals and real-world vibration data, enabling evaluation for multiple use cases. This approach ensures an optimized signal chain for each application, streamlining and cost-reducing product development.