• Applied Science and Convergence Technology
• /
• v.25 no.6
• /
• pp.124-127
• /
• 2016

Humidity calibration for a temperature-stable quartz oscillator (TSQO) was investigated to exclude the influences of relative humidity on the TSQO output in order to use the corresponding devices outdoors. The TSQO output is a voltage that is inversely proportional to the electric impedance of the quartz oscillator, which depends on the viscosity and density of the measured gas. The TSQO output was humidity calibrated using its humidity dependence, which was obtained by varying the relative humidity (RH) from 0 to 100 RH% while other conditions were kept constant. The humidity dependencies of the TSQO output were fit by a linear function. Subtracting the change in the TSQO output induced by the change in humidity, calculated with the function from the experimentally measured TSQO output for a range of 0-100RH%, eliminated the influence of humidity on the TSQO output. The humidity calibration succeeded in reducing the fluctuations of the TSQO output from 0.4-3% to 0.1-0.3% of the average values for a range of 0-100RH%, at constant temperatures. The necessary stability of the TSQO output for application in hydrogen sensors was below one-third of the change observed for a hydrogen leakage of 1 vol.% hydrogen concentration, corresponding to 0.33% of the change in each background. Therefore, the results in this study indicate that the present humidity calibration effectively suppresses the influence of humidity, for the TSQO output for use as an outdoor hydrogen sensor.

• Journal of Sensor Science and Technology
• /
• v.27 no.6
• /
• pp.368-373
• /
• 2018

The effects of long-term exposure of high and low humidity on thin-film humidity sensors are investigated. Five commercially available thin-film humidity sensors are initially calibrated in a humidity chamber as a reference before longterm exposure to high and low humidity. Then, the sensors are kept in a high-humidity environment (~95 %rh) for four months. After the exposure, the sensors are calibrated in the same manner as the initial calibration. Consequently, the device reading values from the humidity sensors are elevated up to about 5 %rh. Interestingly, the degree of elevation by the high-humidity exposure shows a negative correlation with the price of the humidity sensors. Humidity sensors are then kept in a low-humidity environment (~10 %rh) for another four months. After the exposure, a calibration similar to the initial calibration is performed. As a result, the device reading from humidity sensors is decreased, indicating a recovery from the effect of high-humidity exposure. The durability test conducted in this study provides experimental evidence for the use of thin-film humidity sensors in high-humidity environments such as greenhouses and food factories for a long period of time.

• Journal of Sensor Science and Technology
• /
• v.32 no.6
• /
• pp.447-454
• /
• 2023

The Korea Laboratory Accreditation Scheme (KOLAS) operates accreditation programs for ensuring measurement traceability with the International System (SI) of Units - the highest calibration standard that measurements can be tested against. As of September 2023, there are 70 KOLAS-accredited laboratories in the Republic of Korea that specialize in humidity calibration. Among them, 32 KOLAS laboratories, along with one laboratory not affiliated with KOLAS, participated in the proficiency test (PM 2023-11) for relative humidity transducers in 2023. This proficiency test was conducted within a relative humidity range of 20-90% at a temperature of approximately 20 ℃, taking into consideration the calibration and measurement capability (CMC) of the participating laboratories. The primary objective of the proficiency test was to establish the measurement equivalence between each participating laboratory and the reference laboratory, by calculating the number of equivalence (En). When |En| was less than 1, the measurements from the participating and reference laboratory were equivalent. Out of the 33 participating laboratories, 32 successfully met this criterion and passed the proficiency test.

• Atmosphere
• /
• v.33 no.5
• /
• pp.549-560
• /
• 2023

This study evaluates three distinct observation methods, CRDS, OA-ICOS, and OF-CEAS, in greenhouse gas monitoring equipment for atmospheric CO₂ and CH₄ concentrations. The assessment encompasses fundamental performance, high-concentration measurement accuracy, calibration methods, and the impact of atmospheric humidity on measurement accuracy. Results indicate that within a range of approximately 500 ppm, all three devices demonstrate high accuracy and linearity. However, beyond 1000 ppm, CO₂ accuracy sharply declines (84%), emphasizing the need for caution when interpreting high-concentration CO₂ data. An analysis of calibration methods reveals that both CO₂ and CH₄ measurements achieve high accuracy and linearity through 1-point calibration, suggesting that multi-point calibration is not imperative for precision. In dynamic atmospheric conditions with significant CO₂ and CH₄ concentration variations, a 1-point calibration suffices for reliable data (99% accuracy). The evaluation of humidity impact demonstrates that humidity removal devices significantly reduce air moisture levels, yet this has a negligible effect on dry CO₂ concentrations (less than 0.5% relative error). All three observation method instruments, which have integrated humidity correction to calculate dry CO₂ concentrations, exhibit minor sensitivity to humidity removal devices, implying that additional removal devices may not be essential. Consequently, this study offers valuable insights for comparing data from different measurement devices and provides crucial information to consider in the operation of monitoring sites.

• Journal of the Korean Society for Precision Engineering
• /
• v.32 no.2
• /
• pp.185-190
• /
• 2015

This paper presents a new method to determine sensitivity coefficients of temperature and humidity of torque transducers by using a natural and seasonal variation of ambient conditions at the laboratory. We had measured the sensitivities of the torque transducers over almost one year using the KRISS 2 kN m torque standard machine. The sensitivity data acquired at various ambient conditions were processed using our measurement model to extract the sensitivity coefficients of temperature and humidity simultaneously with high precision. A comparison with a previous method using an environmental control chamber was carried out to test the feasibility of using our new method. Two results agreed within the uncertainty. We revealed that the torque measuring errors could be 8 times higher than the measurement and calibration capability of KRISS torque standard machine if the sensitivity changes due to the temperature and humidity are not properly corrected during a calibration.

• Journal of Sensor Science and Technology
• /
• v.14 no.6
• /
• pp.381-386
• /
• 2005

A Co-C eutectic cell for thermocouple calibration was manufactured and tested to investigate its phase transition characteristics using Type B thermocouples. It was observed that the freezing plateaus were flatter than those of melting, but the melting points were closer to the true transition temperature than the freezing points. The expanded uncertainty of melting temperature was calculated not to exceed $0.2^{\circ}C$ (k = 2). Based on the observed results, the melting process is recommended for the calibration of thermocouples.

• Journal of Sensor Science and Technology
• /
• v.21 no.6
• /
• pp.420-424
• /
• 2012

We have investigated temperature dependence and long-term change of humidity measurement from 32 relative humidity sensors. The readings of the humidity sensors depended not only the reference humidity, but also temperature of the chamber. Approximately, the temperature dependence of the humidity sensor in average was 0.05 %R.H./$^{\circ}C$ in the temperature range from $5^{\circ}C$ to $55^{\circ}C$. For humidity sensors that have an internal temperature compensation circuit, the resulting temperature dependence was weaker by 20%. It should be also noted that for the humidity sensors used in this work underwent ${\pm}3$ %R.H. change per year for level of confidence of 95%. The users of relative humidity sensors may refer this value as a minimum change when they set the calibration interval of the humidity sensors.

• Journal of Sensor Science and Technology
• /
• v.30 no.4
• /
• pp.243-249
• /
• 2021

Humidity is an important physical quantity that is closely related with the quality of everyday life as well as the quality control of products in various industries. Here, we have developed a divided-flow type humidity generator of which humidity generation is faster than the saturator-based humidity generator in ppm level. The operation principle of the divided-flow humidity generator is first introduced. Then, the performance of the divided-flow humidity generator is verified by testing the radiosonde humidity sensor at low temperature. As a result, the humidity generated from the divided-flow humidity generator is consistent with the saturator-based precision humidity generator within 1.6% relative humidity in the range from 10% to 40% at -45 ℃. It is also found that the radiosonde humidity sensor shows measurement errors by 3% - 5% at -45 ℃ when it is only calibrated at room temperature. The response times of radiosonde humidity sensor using the divided-flow humidity generator are between about 2 and 9 minutes, whereas those by the saturator-based humidity generator are about 20 minutes. In this regard, the divided-flow humidity generator has a merit in terms of fast humidity changes for the calibration of radiosonde humidity sensors at low-temperatures.

• Particle and aerosol research
• /
• v.18 no.4
• /
• pp.109-127
• /
• 2022

Recently, low-cost particulate matter (PM) sensors have been widely used in monitoring mass concentration. Maintaining the accuracy of the sensors is important and requires rigorous performance evaluation and calibration. In this study, two commercial low-cost PM sensors(LCS), Plantower PMS3003 and Plantower PMS7003, were evaluated in the laboratory and field with a reference-grade PM monitor (GRIMM 11-D). Laboratory evaluation was conducted with single/mixed particles of PSL (Poly Styrene Latex) in an acrylic chamber at 20℃ and relative humidity of 20%. Field evaluation was conducted inside a building of Yonsei University (Shinchon) from February 12 to March 31, 2022. In both evaluations, LCS measured values became different from reference measured values when the relative humidity was high or the outdoor air PM10/PM2.5 ratio was high. Based on the field evaluation, the LCS measured values were corrected through four different regression analysis models. As a result, the multivariate polynomial regression analysis model showed highest matching with the reference PM monitor (PM2.5 >0.9, PM10 >0.85). In this model, the PM10/PM2.5 ratio and relative humidity were chosen as independent variables.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.814-822
• /
• 2022

The deployment of sensors for Structural Health Monitoring requires a complicated network arrangement, ground truthing, and calibration for validating sensor performance periodically. Any conventional sensor on a structural element is also subjected to static and dynamic vertical loadings in conjunction with other environmental factors, such as brightness, noise, temperature, and humidity. A structural model with strain gauges was built and tested to get realistic sensory information. This paper investigates different deep learning architectures and algorithms, including unsupervised, autoencoder, and supervised methods, to benchmark blind drift calibration methods using deep learning. It involves a fully connected neural network (FCNN), a long short-term memory (LSTM), and a gated recurrent unit (GRU) to address the blind drift calibration problem (i.e., performing calibrations of installed sensors when ground truth is not available). The results show that the supervised methods perform much better than unsupervised methods, such as an autoencoder, when ground truths are available. Furthermore, taking advantage of time-series information, the GRU model generates the most precise predictions to remove the drift overall.