Wide Temperature Range and Low Temperature Drift Eddy Current Displacement Sensor Using Digital Correlation Demodulation

Tianxiang Ma, Yuting Han, Yongsen Xu, Pengzhang Dai, Honghai Shen, Yunqing Liu2023 · DOI

TL;DR

This paper introduces a novel eddy current displacement sensor that significantly reduces temperature drift, a major limitation in traditional sensors. By employing a differential probe design and a digital correlation demodulation technique, the sensor achieves high precision and stability across a wide temperature range, making it suitable for demanding industrial applications.

What Problem Does This Paper Address?

Eddy current sensors are widely used for non-contact measurements due to their high bandwidth and sensitivity. However, their accuracy is severely hampered by temperature drift. This drift arises from changes in the sensor coil's resistance and inductance, as well as the resistivity of the target material, all of which are temperature-dependent. Existing methods to mitigate this issue, such as data post-calibration or incorporating self-compensation circuits, have limitations. Post-calibration is complex and requires individual sensor characterization, while self-compensation circuits can restrict the measurement range. The authors aim to overcome these limitations by developing a sensor system that inherently reduces temperature drift through its design and advanced digital signal processing.

Methods

The core innovation lies in the combination of a differential eddy current sensor probe and a digital correlation demodulation algorithm.

  • Differential Probe: Two identical sensor coils are used in a differential configuration. This setup is designed to cancel out common-mode interference caused by temperature variations affecting both coils, the target material, and the connecting cables.
  • Digital Correlation Demodulation: Instead of traditional analog demodulation, the system digitizes the differential analog carrier signal using a high-speed Analog-to-Digital Converter (ADC). The digitized signal is then processed by a Field-Programmable Gate Array (FPGA) using a double correlation demodulation method. This method involves calculating the cross-correlation function between the measured signal and precisely generated sine and cosine reference signals. This digital approach allows for more robust signal processing and better noise rejection.
  • Error Analysis: The paper meticulously analyzes potential sources of error, including the influence of accumulation cycles and sampling frequency on output accuracy, and the impact of frequency stability. This analysis informs the design choices and experimental setup.

Datasets

Not discussed in the paper. The paper focuses on the design and experimental validation of a new sensor system, rather than training on large datasets.

Key Insights and Ideas

  • Differential Design is Key for Temperature Compensation: The use of a differential probe structure is crucial for effectively canceling out common-mode temperature-induced errors, a significant improvement over single-probe designs.
  • Digital Demodulation Offers Superior Control: Shifting from analog to digital correlation demodulation provides greater flexibility in signal processing, allowing for more precise extraction of amplitude and phase information and better noise immunity.
  • Accumulation Cycles Balance Accuracy and Speed: The number of accumulation cycles in the digital demodulation algorithm directly impacts accuracy but inversely affects data update speed. A trade-off must be made based on application requirements.
  • Frequency Stability is Paramount: Even small deviations in the frequency of the measured signal relative to the reference signal can significantly degrade accuracy, highlighting the need for stable oscillators.

Key Results

The developed differential digital demodulation eddy current sensor demonstrates impressive performance:

  • Nonlinearity: 0.68% within the range of ±2.5 mm.
  • Resolution: 760 nm.
  • Maximum Bandwidth: 25 kHz.
  • Temperature Stability: A significant reduction in temperature drift was achieved. Over a wide temperature range of -40 °C to +50 °C, the system temperature drift rate was ±0.00137% FS/°C. The rate of change in temperature drift for every 10 °C was less than 0.01% at room temperature.
  • Repeatability Error: ±0.62%.

Limitations and Future Work

  • Limitations:
    • While the differential design cancels common-mode interference, the paper acknowledges that nonlinearity in sensor factors can still affect the differential signal.
  • The paper mentions that the sampling frequency of the ADC cannot be increased infinitely due to hardware limitations.
  • Future Work: The authors suggest that this method is flexible, reliable, and convenient for applications with wide temperature variations, implying its suitability for broader adoption. Further research could explore optimizing the FPGA implementation for even higher speeds or integrating more advanced adaptive filtering techniques for dynamic environments.

Summary updated

BibTeX
@misc{ma2023wide,
  author = {Tianxiang Ma and Yuting Han and Yongsen Xu and Pengzhang Dai and Honghai Shen and Yunqing Liu},
  title = {{Wide Temperature Range and Low Temperature Drift Eddy Current Displacement Sensor Using Digital Correlation Demodulation}},
  year = {2023},
  doi = {10.3390/s23104895},
}
Figure