Evaluation of EMI Filter Performance for Uninterruptible Power Supply

碩士 === 國立中山大學 === 電機工程學系 === 87 === The performance of an EMI filter usually is difficult to predict precisely when uninterruptible power supply (UPS) manufacturers try to fix conducted electromagnetic interference problems. They normally involve a cut-and-trial process that wastes a lot...

Full description

Bibliographic Details
Main Authors: LIAO KANG-YU, 廖康佑
Other Authors: LIN KEN-HUANG
Format: Others
Language:zh-TW
Published: 1999
Online Access:http://ndltd.ncl.edu.tw/handle/12199476437336245843
id ndltd-TW-087NSYSU442016
record_format oai_dc
spelling ndltd-TW-087NSYSU4420162016-07-11T04:13:19Z http://ndltd.ncl.edu.tw/handle/12199476437336245843 Evaluation of EMI Filter Performance for Uninterruptible Power Supply 不斷電電源供應器之EMI濾波器特性評估 LIAO KANG-YU 廖康佑 碩士 國立中山大學 電機工程學系 87 The performance of an EMI filter usually is difficult to predict precisely when uninterruptible power supply (UPS) manufacturers try to fix conducted electromagnetic interference problems. They normally involve a cut-and-trial process that wastes a lot of time and increases the production cost. This paper tries to alleviate this annoyance by taking the impedance of noise source into account. Based on the insertion loss, three methods are proposed for comparison. The first method considers neither the noise source impedance nor the parasitic elements, the second method only considers parasitic elements, and the third method considers both. Two different filter topologies are used to evaluate the performance of the three methods. Because EMI filter performance depends not only on the filter itself but also on the noise source impedance, we find the third method is closest to the measurement data, the second method is next, yet the first method has the maximum difference. In the first filter topology, the third method gives an accurate estimate of filter performance with a mean error of 3.14 dB. For the same topology, the second method is worse than the third method by 20 dB, and the first method is worse than the second method by another 30 dB. In the second filter topology, the third method also provides an accurate estimate of filter performance with a mean error of 2.66 dB. Similarly, the second method is worse than the third method by 20 dB, and the first method is worse than the second method by another 40 dB. Two power combiners and an HP 4194A impedance/gain-phase analyzer are used in our measurement. The former includes a power combiner for measuring CM noise and a power combiner for measuring DM noise. The latter is used to measure the frequency responses of filter elements and the parameters of matrix. Using the methods proposed in this paper, EMI engineers can solve the problems quickly, and manufacturers can shorten development time of new products and reduce the production cost. LIN KEN-HUANG 林根煌 1999 學位論文 ; thesis 72 zh-TW
collection NDLTD
language zh-TW
format Others
sources NDLTD
description 碩士 === 國立中山大學 === 電機工程學系 === 87 === The performance of an EMI filter usually is difficult to predict precisely when uninterruptible power supply (UPS) manufacturers try to fix conducted electromagnetic interference problems. They normally involve a cut-and-trial process that wastes a lot of time and increases the production cost. This paper tries to alleviate this annoyance by taking the impedance of noise source into account. Based on the insertion loss, three methods are proposed for comparison. The first method considers neither the noise source impedance nor the parasitic elements, the second method only considers parasitic elements, and the third method considers both. Two different filter topologies are used to evaluate the performance of the three methods. Because EMI filter performance depends not only on the filter itself but also on the noise source impedance, we find the third method is closest to the measurement data, the second method is next, yet the first method has the maximum difference. In the first filter topology, the third method gives an accurate estimate of filter performance with a mean error of 3.14 dB. For the same topology, the second method is worse than the third method by 20 dB, and the first method is worse than the second method by another 30 dB. In the second filter topology, the third method also provides an accurate estimate of filter performance with a mean error of 2.66 dB. Similarly, the second method is worse than the third method by 20 dB, and the first method is worse than the second method by another 40 dB. Two power combiners and an HP 4194A impedance/gain-phase analyzer are used in our measurement. The former includes a power combiner for measuring CM noise and a power combiner for measuring DM noise. The latter is used to measure the frequency responses of filter elements and the parameters of matrix. Using the methods proposed in this paper, EMI engineers can solve the problems quickly, and manufacturers can shorten development time of new products and reduce the production cost.
author2 LIN KEN-HUANG
author_facet LIN KEN-HUANG
LIAO KANG-YU
廖康佑
author LIAO KANG-YU
廖康佑
spellingShingle LIAO KANG-YU
廖康佑
Evaluation of EMI Filter Performance for Uninterruptible Power Supply
author_sort LIAO KANG-YU
title Evaluation of EMI Filter Performance for Uninterruptible Power Supply
title_short Evaluation of EMI Filter Performance for Uninterruptible Power Supply
title_full Evaluation of EMI Filter Performance for Uninterruptible Power Supply
title_fullStr Evaluation of EMI Filter Performance for Uninterruptible Power Supply
title_full_unstemmed Evaluation of EMI Filter Performance for Uninterruptible Power Supply
title_sort evaluation of emi filter performance for uninterruptible power supply
publishDate 1999
url http://ndltd.ncl.edu.tw/handle/12199476437336245843
work_keys_str_mv AT liaokangyu evaluationofemifilterperformanceforuninterruptiblepowersupply
AT liàokāngyòu evaluationofemifilterperformanceforuninterruptiblepowersupply
AT liaokangyu bùduàndiàndiànyuángōngyīngqìzhīemilǜbōqìtèxìngpínggū
AT liàokāngyòu bùduàndiàndiànyuángōngyīngqìzhīemilǜbōqìtèxìngpínggū
_version_ 1718342460850044928