In an industrial automation site with increasingly complex systems, the temperature and vibration conditions of various bearings, motors, and pump bodies have been closely related to the health and efficiency of the entire automation system. The huge industrial industry has given birth to an extensive demand for sensors. For a long time, the technology and market of these high-end industrial-grade vibration and temperature sensors have been dominated by European and American companies. Yet with the rise of China’s promising industrial manufacturing industry, the domestic sensor manufacturers are encroaching on this field!
Image: An Industrial Site
Driven by the growing demand in the domestic industrial automation market and the rapidly-developing sensor technology, this industry has seen an integration of vibration sensing and temperature measurement over time. I was fortunate to learn about several domestically-made temperature and vibration integrated sensors in an industrial application laboratory, and assisted in selecting application plans for industrial sites. Then I summarized and analyzed the use experience and test performance of these sensors for the reference and benefit of users and readers in this industry. Any comment is welcomed.
Image: FATRI Temperature and Vibration Integrated Sensor
What is a temperature and vibration integrated sensor?
A temperature and vibration integrated sensor refers to a composite sensor transmitter that integrates a vibration sensor chip and a temperature sensor chip and encapsulates and outputs signals according to application requirements.
Key test points:
• Linearity (the linearity output when the sensor is vibrated with different G values, which is the smaller, the better)
• Transverse sensitivity (it determines the degree of interference to the sensor when it is subjected to vibration or impact in the non-spindle direction, and should be <5% according to national standards. The smaller, the better. )
• Resonant frequency (the upper limit frequency of vibration measurement depends on the resonant frequency in the amplitude-frequency curve)
• Frequency-response characteristic (the upper limit of the frequency of vibration measurement is slightly higher than the vibration frequency of the tested structure)
• Temperature measurement accuracy (the smaller, the better)
• Temperature measurement response time (the shorter, the better)
Environment and apparatus:
Apparatus: Acceleration calibration system (including high, medium and low-frequency vibration exciters), constant current source, constant voltage source, high-insulation resistance tester, high and low-temperature cycle furnace, desktop multimeter, and contact high-precision thermometer
Test reference data (the test data is derived from products purchased on the market in this test, and is for reference only)
Vibration measurement mainly refers to linearity and transverse sensitivity as the basis for consideration of sensor accuracy
Based on the single sensitivity input during sensor measurement, linearity is used to tell whether the sensitivity of the sensor is consistent with the actual sensitivity within a certain frequency response range. In contrast, in the low-frequency band, the sensitivity of the sensor will be less than the actual sensitivity, while in the high-frequency band, it will be greater. Only in the middle-frequency band, the sensitivity conforms to the linear relationship. If the sensor is not in the linear range, there will be a large error in the measured amplitude. Generally speaking, the nonlinearity of the sensor is required to be less than 1%. According to data, Niell CAYD275-10 and FATRI PYDV00-100 performed well in the linearity test.
In the measurement of vibration in a certain direction, theoretically the signal should be output in the vibration sensing direction, but in fact there is also a signal output in the direction perpendicular to this direction, which leads to interference called the transverse effect. The lower the sensitivity of the transverse effect, the better the performance of the sensor. But usually, the sensor has such transverse effect more or less, and it needs to be less than 5% as required in the national standard. In this test, FATRI PYDV00-100 registered a transverse effect of 2.68%. I once suspected that I must have done something wrong, because the the average figures of other products were between 4.9~5%. Even after repeated tests, FATRI’s average figure remained around 2.4~2.5%. To be cautious, I took a relatively high value at 2.68%. In harsh industrial sites, reducing the impact of such transverse effects is the basic guarantee for accurate measurement. I have to admit that FATRI did quite well in this respect!
Being a structure in nature, a sensor comes with an inherent frequency. Generally, the first-order inherent frequency of a sensor is called the resonant frequency. The smaller the structural size of a sensor, the higher the resonant frequency. The upper limit frequency of the vibration sensor depends on the resonant frequency in the amplitude-frequency curve, which is generally required to be greater than 20kHz. It can be seen from this data that both Niell CAYD275V-10 and FATRI PYDV00-100 met the requirements in this test, while the other two products still needed some improvements.
The frequency-response characteristic refers to the error in the sensitivity relative to the reference point when the sensor is subjected to different frequencies of vibration (1.0~7000Hz-±5%, 0.5~10000Hz-±10%, and 0.3~15000Hz-±3dB). Theoretically, the smaller the error and the larger the bandwidth, the better. In this test, FATRI PYDV00-100, SINOCERA CA-YD-170, and Ronds RH103 were all within the normal range. The frequency-response sensitivity error of Niell CAYD275-10 at 1.0-7000Hz was ±5.2%, and the figure at 0.5-10000Hz was ±11.2%, slightly higher than the reference standard value.
Temperature measurement accuracy & Temperature measurement response time
These sensors are all equipped with temperature measurement functions. Undoubtedly, the accuracy and response time of temperature measurement are the most critical indicators, and they should be as small as possible. According to the comparison test, the four products are on a par with each other in measurement accuracy, all basically within the temperature redundancy range in practical applications. However, in terms of temperature response time, FATRI PYDV00-100 performed better with its response time of 3.5min.
This experimental test involves 18 items such as electrical parameters and connection performance. Based on the importance of technical indicators and major references for product selection, I chose six items, as I believe representative, from many test items and data, for a simple analysis. According to the test results, FATRI PYDV00-100 and Niell CAYD275V-10 outperform the other two temperature and vibration integrated sensors whose main performance needs to be improved a lot.
The “charge conversion chip”, the core of the ICP sensor, is the underlying component of a temperature and vibration integrated sensor. In my opinion, an excellent sensor cannot do without its manufacturer’s efforts to develop the underlying technology, and independent chip design and manufacturing capabilities serve as the direct contributor to the final performance and applicability of the sensor. Take FATRI PYDV00-100 for example. As far as I am concerned from the company’s public information, it is already equipped with the chip design and manufacturing technology. Sensor manufacturers are bound to face both opportunities and challenges in the market, for the temporary and in the future, and we are glad to see that a growing number of Chinese sensor companies are making great strides in the right direction. With the improvement of the industrial Internet of Things and intelligent manufacturing, more excellent sensor companies and products will emerge in China.
“Some happy talent, and some fortunate opportunity, may form the two sides of the ladder on which some men mount, but the rounds of that ladder must be made of stuff to stand wear and tear; and there is no substitute for thorough-going, ardent, and sincere earnestness.” – Dickens
My opinions—Some suggestions on the selection of vibration sensors
1. Choose vibration sensors according to the type of data acquisition and spectrum analysis instrument. For example, choose an ICP sensor for ICP conditioning equipment and a piezoelectric sensor for charge conditioning equipment.
2. To measure the structure to be tested in the working state, a sensor with “isolation” is preferred. If the sensor does not have isolation, an insulating material can be added at its bottom as an isolation device.
3. To measure vibration at the focus position, it is advised to set the range between 60-80% of the sensor range, so that a high signal-to-noise ratio can be guaranteed without overloading.
4. The working frequency range of the applied sensor should be slightly higher than the bandwidth actually measured.
5. Choose sensors based on industry applications. For example, for heavy industry, electric power, and petrochemicals, sensors with a large vibration range and a wide frequency are preferred. For the construction industry, in contrast, sensors with a small range, high sensitivity, low frequency or good ultra-low frequency performance are more suitable.
Author: Lu You, a former veteran in the sensor industry, and now a freelancer.