Switching power supplies and uninterruptible power supplies (UPS) are widely used. As commercially available products, they were among the first products designated in China to require electromagnetic compatibility (EMC) certification. The assessment focuses on their external interference emissions (including conducted and radiated emissions) during operation, and their impact on the power grid quality. The latter was discussed in detail in the 9th issue of this journal in 2000; this article focuses on the testing of interference emissions. Interference testing of switching power supplies and UPS adopts the methods provided in GB4824-1996, "Methods of Measurement and Limits of Electromagnetic Interference Characteristics of Industrial, Scientific and Medical (ISM) Radio Frequency Equipment" (equivalently converted from CISPR 11-1990). According to the classification of interference sources in GB4824, equipment is divided into two groups: Group 1 refers to all industrial, scientific, and medical equipment that specifically generates and/or uses conducted-coupled radio frequency energy to fulfill its functional requirements; Group 2 refers to all industrial, scientific, and medical equipment that specifically generates and/or uses electromagnetic radiation energy for functions such as material processing and electro-spark corrosion. Based on the above definitions, switching power supplies and uninterruptible power supplies (UPS) undoubtedly belong to the same group of equipment. Furthermore, GB 4824 further classifies equipment into two categories based on the power supply network used: Category A equipment includes industrial, scientific, and medical equipment in facilities not in residential settings and not directly connected to the residential low-voltage power grid; Category B equipment includes industrial, scientific, and medical equipment in residential settings and facilities directly connected to the residential low-voltage power grid. From this classification, both switching power supplies and UPS could potentially belong to either of these categories. Table 1 shows the conducted disturbance voltage limits at the power supply terminals of Group A and Group B equipment in the test field. Table 2 shows the radiated disturbance limits for Group A and Group B equipment in the test field. [/align][img=387,161]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/56-1.jpg[/img][img=387,132]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/56-2.jpg[/img][align=left] 1 Standard Test Methods 1.1 Measuring Instruments and Test Sites (1) The measuring instrument shall be an interference receiver with quasi-peak and average value detectors. Its performance shall meet the requirements of CISPR16-1 or the corresponding national standard GB/T6113.1 "Specification for Radio Interference and Immunity Measurement Equipment". In the frequency range covered by the standard, two interference receivers with different frequency bands shall generally be used, namely 9KHz～30MHz; 30MHz～1000MHz. (2) When conducting conducted interference voltage tests at the power supply end, an artificial power network with an impedance of 50Ω/50μH (V-shaped network) should be used, and its characteristics should meet the requirements of CISPR16-1 and GB/T6113.1. The main function of the artificial power network is to effectively separate the test object from the power supply; at the same time, it provides a stable high-frequency impedance for the test object. (3) In the 9kHz to 30MHz frequency band, a shielded loop antenna should be used. In the 30MHz to 1000MHz frequency band, a balanced dipole antenna should be used. (4) Test site ① Radiation test site in the 9kHz to 1000MHz frequency band The radiation test site should be an open and flat site with no overhead lines within its boundary, no reflective structures (such as reinforced concrete buildings and tall trees) nearby, and a sufficiently large size to allow the antenna, test object and reflective structures to be fully separated. A standard radiation test site should be an ellipse enclosed by an ellipse whose major axis is twice the focal length (F) and whose minor axis is twice the focal length (F), as shown in Figure 1. During the test, the test sample and the measuring antenna will be positioned at the two foci respectively. [Image: http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/57-1.jpg] To obtain stable radio wave transmission characteristics, a fixed, relatively large reflective surface (or grounding plate) is required. The reflective surface is made of metallic materials, such as steel plates (including galvanized steel plates) and metal mesh. The plates must be welded together, without large gaps or holes. The maximum aperture of the metal mesh must be less than 1/10 of the wavelength (for 1000MHz, the aperture should be less than 3cm). In addition, the surface of the site must be flat, and drainage facilities must be considered. Figure 2 shows the minimum dimensions of the metal grounding. ② Conducted disturbance voltage test site The conducted disturbance voltage test can be carried out in the radiation test site or in the shielded room. 1.2 Test method [/align][img=349,184]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/57-2.jpg[/img][align=left] (1) Ambient level The test specimen is connected to the measurement line, but when it is not powered on, the applicability of the test environment should be determined by measuring the ambient noise level. The ambient level should be at least 6dB lower than the specified limit. If the ambient level and the radiation of the test specimen are superimposed and still do not exceed the specified limit, the test specimen is considered to have met the specified limit. When measuring the conducted disturbance voltage at the power supply end, an appropriate radio frequency filter can be connected between the artificial power network and the power grid to reduce the ambient level. However, after connecting the radio frequency filter, the impedance of the artificial power network should still meet the specified requirements at the measurement frequency. When measuring radiated interference, if the ambient level cannot meet the requirements, the measuring antenna can be moved closer to the test specimen before measurement, but the limit remains unchanged. This actually makes the requirements for the test specimen more stringent. (2) General requirements for test specimen arrangement The interference level of the test specimen refers to the maximum value of the interference value under various typical usage conditions and different configurations and test arrangements. The test report should describe in detail the configuration and test arrangement of the test specimen during the test. When the test specimen consists of several interconnected devices, the type and length of the interconnecting cable should be consistent with the technical requirements of the test specimen. If the cable length can be changed, the length that can produce the maximum radiation in the radiation test should be taken. (3) Radiation test of 9kHz to 1MHz When the test specimen is placed on the test turntable, the radiation center of the equipment should be as close as possible to the rotation center of the turntable. The distance between the test specimen and the measuring antenna refers to the horizontal distance between the rotation axis of the turntable and the measuring antenna. Regarding the test turntable, if it is elevated above the grounding plate, it should generally not exceed 0.5m above that plane; if it is on the same plane as the grounding plate, the turntable plane must be a metal plane with good electrical connection to the grounding plate. Regardless of the turntable type, for non-floor-mounted test specimens, the height above the grounding plate should be 0.8m. When the test specimen is not placed on the turntable, the distance between the test specimen and the measuring antenna refers to the shortest horizontal distance between the test specimen boundary and the measuring antenna. When the test specimen is placed on the turntable, with the measuring antenna in both horizontal and vertical polarization states, the turntable should rotate at all angles, and the highest level of radiated interference should be recorded at each measurement frequency. When the test specimen is not placed on the turntable, the measuring antenna should be selected at different measurement positions on the ground plane in both horizontal and vertical polarization states. Measurements should be performed in the direction of maximum radiation, and the highest level of radiated interference should be recorded at each measurement frequency. The antenna requirements for measurements are as follows: within the 30MHz–80MHz frequency band, the antenna length should be equal to the resonant length at 80MHz; within the 80MHz–1000MHz frequency band, the antenna length should be equal to the resonant length at the measurement frequency. Additionally, a suitable conversion device should be used to match the antenna to the feed line. A balun should also be configured to connect to the measurement receiver. The antenna should be able to be arbitrarily oriented to measure its vertically and horizontally polarized wave components separately. The antenna center height should be adjustable within 1m–4m. The nearest point of the antenna above the ground should not be less than 0.2m to measure its maximum value. Incidentally, if the measurement results using other types of antennas differ from those using a balanced dipole antenna by within ±2dB, then other types of antennas can also be used. Commonly used broadband antennas in practice are biconical antennas (30MHz–300MHz) and log-periodic antennas (300MHz–1000MHz). [/align][img=333,170]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/58-1.jpg[/img][align=left] (4) Measurement of conducted interference voltage at the power supply end ① When measuring at the radiation test site, the test specimen should be in the same state as the radiation measurement, and the test specimen should be on a metal grounding plate that extends at least 0.5m beyond its boundary or has a minimum size of 2m×2m. ② When measuring in a shielded room, the ground shielding layer or the shielding layer of any wall can be used as the grounding plate. During the test, non-grounded test specimens should be placed on an insulating support or platform 0.4m above the grounding plate. Grounded test specimens should be placed on the grounding plate, and their contact points should be mutually insulated or consistent with normal use. All test specimens should be at a distance greater than 0.8m from other metal objects. The closest distance between the outer surface of the artificial power network and the boundary of the test specimen should be less than 0.8m. The network's reference grounding terminal should be connected to the grounding plate using the shortest possible conductor. The routing of power and signal cables to the grounding plate should be equivalent to actual usage conditions, and cables should be laid out with extreme care to avoid spurious response effects. When the test specimen has a special grounding terminal, it should be grounded using the shortest possible conductor. Test specimens without a special grounding terminal should be tested in their normal connection mode, i.e., grounded from the power grid. For equipment supplied with flexible power cords by the manufacturer, the power cord length should be 1m. If the actual length exceeds 1m, the excess portion should be folded back and forth into a 0.3m to 0.4m bundle. If the test specimen consists of several units, and each unit has a power cord, the following provisions apply when connecting it to an artificial power network: ● Each power cable connected to a standard power plug should be measured separately; ● Power cords or terminals that, unless specified by the manufacturer, require power from another unit in the system, should be measured separately; ● Power cords or terminals that, as specified by the manufacturer, require power from a particular unit in the system should be connected to that unit, and the cables or terminals of that unit should be connected to the artificial power network for measurement; ● When the test specimen requires grounding for safety purposes, the grounding wire should be connected to the reference grounding point of the artificial power network. Unless otherwise specified by the manufacturer, the grounding wire should be 1m long and laid parallel to the test specimen's power cord, with a spacing of no more than 0.1m. Other grounding wires (such as those for electromagnetic compatibility purposes) specified or provided by the manufacturer and connected to the same terminal used for safety grounding should also be connected to the reference line of the artificial power network. Figure 4 shows a typical setup for conducted disturbance voltage measurement. [/align][img=347,205]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/58-2.jpg[/img][align=left] [b]2 Alternative Testing Methods[/b] The above-mentioned laboratory final test configuration emphasizes the accuracy of test results and comparability with domestic and foreign testing institutions. Therefore, the configuration price is not cheap. If the test site is also involved, it is even more unacceptable to general manufacturing enterprises. This section will recommend alternative testing schemes to reduce configuration costs as much as possible while ensuring a certain degree of comparability, so as to make it acceptable to as many enterprises as possible. 2.1 Measuring Instruments and Test Sites (1) Measuring Instruments Considering the characteristics of switching power supplies and uninterruptible power supplies, their internal components are all composed of electronic circuits. When the power supply is in steady state, it does not produce sparks, arcs, or gas discharges, nor does it produce the unique crackling noise interference of household appliances. It only produces periodic voltage, current, and their harmonics. Therefore, a spectrum analyzer is recommended. Its price is significantly lower than that of the interference receiver used in the final test. Within the test frequency range specified in the standard, only a spectrum analyzer of 9kHz to 1000MHz or higher is required. (2) The artificial power supply network remains unchanged. (3) Since the GTEM cell is used, the function of its upper and lower base plates and internal partitions is similar to that of a receiving antenna, so the receiving antenna used in the final test configuration is cancelled. (4) Since the size of switching power supplies and general-purpose uninterruptible power supplies is not large, they can be accommodated in the recently developed GTEM cell. The price and operating frequency range of the cell are sufficient to meet the needs of general users. Although this test site has not been recognized in CISPR11 and GB4824 standards, the CISPR25 standard for testing the radio interference characteristics of automotive electronic/electrical components has adopted the TEM cell method as the standard test method for testing the radiated emission characteristics of components/modules. The GTEM cell is a development of the TEM cell, with a larger test space and no contradiction with the frequency range used, and therefore it has been increasingly used. Table 3 shows the main performance of the GTEM chamber and the size of the test specimens it can accommodate, for reference. 2.2 Test Methods (1) Radiation test from 9kHz to 1000MHz The test configuration of the GTEM chamber and spectrum analyzer is shown in Figure 5. [/align][img=300,102]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/59-1.jpg[/img][align=left] To ensure the repeatability and comparability of the test results, the position of the test specimen should be fixed for each test. Regarding the polarization test, since the position between the internal partition and the upper and lower bottom plates of the chamber is fixed, it can only be achieved by rotating the test specimen so that several faces of the test specimen face the partition in turn. 2.3 Other Possible Alternative Testing Methods Internationally, some have suggested using the absorbing clamp method, similar to that provided in CISPR 14-1 (the corresponding Chinese national standard is GB 4343-1 "Methods of Measurement and Permissible Values ​​of Radio Interference Characteristics of Household and Similar Electric and Heating Appliances, Power Tools and Similar Circuits"), to test the radiated power of small electronic devices. This method is suitable when the sample is small, and the energy radiated into space mainly escapes through the power cord. Therefore, this portion of energy can be measured using an absorption device surrounding the power cord. This absorption device is called an absorbing clamp (or ferrite clamp). The advantages of this method are its simplicity, low environmental requirements (it can be performed in a shielded room), and good repeatability and comparability of the test results. Figure 6 shows a simplified test diagram. Due to space limitations, detailed methodological information is provided in the relevant standards. [Image: http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/59-2.jpg] This method has been adopted as a revision of CISPR 22 (corresponding to the Chinese national standard GB9254 "Limits and Methods of Measurement of Radio Interference by Information Technology Equipment") by the CISPR subcommittee's technical committee draft (CD), and is now in the stage of drafting an international standard (FDIS) by the secretariat. For the reasons mentioned above, this method is technically feasible, simple, reproducible, and has a low configuration cost. However, there is still a data comparison issue between the absorbing clamp method and the direct measurement method of radiated interference. Nevertheless, in terms of methodology, it remains a good testing method for general enterprises. [img=475,134]http://zszl.cepee.com/cepee_kjlw_pic/files/wx/dyjsyy/2001-5/59-3.jpg[/img]