Ultraviolet detection technology is widely used in military and civilian fields, playing an important role in missile warning, ultraviolet/infrared composite guidance, detection of solar ultraviolet radiation intensity, ozone detection, biomedicine, and other fields. In recent years, with the development of ultraviolet optical remote sensing detection technology, quantitative research on ultraviolet radiation information has become particularly important. As a type of ultraviolet detector, the solar blind phototube can reduce the impact of out-of-band leakage in ultraviolet radiation measurement, thereby improving the detection accuracy of the ultraviolet band. Based on its characteristics, it is often used in various large military equipment missile approaching warning systems and also commonly used in corona detection to effectively and quickly detect fault locations. In Europe and America, it has been used as a standard power detection method and widely applied in equipment. At present, research on solar blind phototube preparation has gradually begun in China, and the study of its radiation quantification is one of the key links for application. Therefore, in the face of the urgent need for high-precision radiation calibration in the ultraviolet band and the exploration of the application of solar blind photocell, it is necessary to study its response nonlinearity.
The research on the response nonlinearity of detectors can be divided into two methods:direct method and indirect method. Based on the indirect method, the responsivity standard is transferred to the detector to be calibrated by the standard detector method. The response nonlinearity of the solar blind phototube detection system is studied, and a standard transfer chain based on the detector is established. In terms of the measurement method, we use an external xenon lamp integrating sphere as the radiation source and control the luminous flux of the external xenon lamp entering the integrating sphere through an adjustable aperture. The adjustment of the radiance of the sphere exit portal is achieved and the spectral consistency during the adjustment process is ensured. In this way, the synchronous measurement of the reference detector and the detector to be tested is easy to achieve under the same radiation conditions. In traditional measurement methods, the optical power level of the radiation source is selected with a set of neutral density filters used either one at a time or in combinations of several filters. It needs to construct a dual optical path to achieve synchronous measurement between the reference detector and the detector to be tested. Otherwise, it needs to establish a standard transfer chain by a motor to continuously exchange the positions of the two detectors. Therefore, we provide a new approach for the study of detector response nonlinearity. Compared to traditional measurement methods, the proposed approach simplifies the complexity of the optical path, reduces the strict requirements for the stability of the light source, and eliminates errors and drifts introduced by non-synchronization during the measurement process. In addition, it reduces sensitivity to environmental and interference factors, improves measurement repeatability and accuracy, and obtains more reliable and accurate measurement results.
We first analyze the stability of the xenon lamp light source spectrum, indicating that there is no drift phenomenon in its relative spectral radiance within 2 h. The peak value (wavelength: 308.558 nm) has a relative standard deviation of 0.254% during the measurement period, with little fluctuation. Then, we propose a method for selecting a reference point, and based on the relative error calculated in Table 3, 0.147 is selected as the reference responsivity. Subsequently, we focus on the influencing factors of indirect synchronous measurement and discuss the interference of light source fluctuations and dark background noise on linear measurement devices from the perspective of the correlation between the measured data of the detector to be tested and the reference detector. Results show that the response nonlinearity of the detector to be tested can be studied within a range of 2.97×10-10-6.61×10-8 A. Finally, based on the principle that the transmittance of the neutral density filter is independent of the radiation output of the light source, we study the nonlinearity of the reference detector responsivity, indicating that its linear error is within the range of 0.69% when the response photocurrent is greater than 1.239×10-9 A, which can be used as a standard detector with excellent linearity in the ultraviolet band.
We introduce an indirect method to establish a transfer chain based on a standard detector and study the application of the standard detector method in laboratory calibration. Compared to the typical light flux superposition method, this method reduces the tedious measurement procedures and the requirements for light source stability. The experiment adopts an external xenon lamp light source integrating sphere as the radiation source, which has good illumination uniformity and reduces the complexity of the optical path compared to the typical dual optical path measurement method. The research results indicate that the response photocurrent of the solar blind phototube detection system is within 2.97×10-10-6.61×10-8 A, and its linear error is within 5.2%. The value range of the nonlinear correction factor is 0.948-1.006. The main factor affecting the nonlinear correction factor is the dark background noise of the detector to be tested, and the uncertainty of the measurement system is 3.59% (k=2).
