線上影音

Home > ANSYS HFSS 教學 > Shielding Effect (S.E.) and Shielding Case/Can Analysis

 

本文始於2014年,介紹如何以HFSS 2014分析不同材質的Shielding Effect,並討論肌膚深度對S.E.的影響,另外還觀察不同的shielding case/can與系統接合方式,對EMI的影響,過程中會引出重要觀念:近場屏蔽與遠場屏蔽是完全不同的行為。 另外,對平面波來說,電場屏蔽主要靠反射損耗,磁場屏蔽主要靠入射損耗,後者需要較厚的材料才能有較佳的屏蔽效果。[2]p.6

This article is intended to introduce all about shielding effect (S.E.), shielding case/can, and how to simulate them with 3D EM solver (HFSS). Some important concepts will be induced : near-field shielding (practical shielding) is totally different to far-field shielding.[1][7] By the way, for plane wave, electric-field shielding is via reflection RdB, and magnetic-field shielding is via absorption AdB.[2]p.6

  1. Introduction

  2. Shielding Effect of Different Materials

  3. EMI with Grounding and Opened Shielding Case

  4. EMI with Non-Grounding and Opened Shielding Case

  5. EMI with Grounding and Closed Shielding Case

  6. EMI with Non-Grounding and Closed Shielding Case

  7. CISCO's paper in DesignCon 2013

  8. Another Case with Apertures

  9. 問題與討論

    9.1 如何在HFSS直接plot S.E.?

    9.2 在step2中,為何低頻以外,Ni的S.E.特性比Cu好?

    9.3 在step2中,為何Ni的S21曲線到高頻時開始變水平? (重要)

    9.4 在step2.1中,為何在低頻(<0.4G),Ni的SE特性比銅差?

    9.5 為何step8中,當頻率大於700MHz,加機殼的EMI反而更高? (重要)

    9.6 為何S.E.模擬 ,求解shielding material object需設"solve inside"? (重要)

    9.7 為何step4,5中,Cu shielding的效果比Ni佳,與step2的趨勢正好相反? (重要)

  10. Reference

  1. Introduction 

1.1 Shielding Effect Theory

For E-field:   , Ei is incident wave,Et is transmitted wave

For H-field:   , Hi is incident wave Ht is transmitted wave

在平面波的前提下,以上兩式的結果 相同。但如果是近場或複合材料, 以上兩式的結果則不同 。[1]p.5,7

S.E.dB = Reflection +Absorption+Multiple reflection = R dB +A dB +M dB,

For far-field(plane-wave) and good conductor, S.E.@RdB+AdB (MdB@0). [1]p.13,[2](17), while t>>d, MdB loss may be important. [1]p.18

吸收損耗 ,與肌膚深度(skin depth, d)倒數的指數次方 導體厚度t 的指數次方 呈正比 [2]p.5

而肌膚深度(電荷在導體表面集中的程度)為 , m =導磁率, s=導電率(S/m)

雖然 銅的導電率(5.8e+7)比鎳(1.45e+7)大約四倍,但鎳的相對導磁率(600)比銅(1)大600倍,所以鎳的肌膚深度遠比銅小(2.09um@1G)

肌膚效應越早發生(肌膚深度越小,越厚)的導體,S.E.也越大( absorption loss AdB越大 ),所以遠場(平面波)高頻的屏蔽效果,鎳比銅好。(as step2.1)

For far-field, S.E.@RdB+AdB ,低頻(<100M)以RdB項為主( 隨著頻率增加而減少) , 高頻以AdB項為主

早期的教科書並沒有上圖紅色那條線 [1], p.32 ,而HFSS跑出的結果正是上圖紅色那條線(低頻水平)。 [2], p.8

遠場的低頻屏蔽主要以反射損耗RdB項為主(上圖淺綠色線),而近場的低頻屏蔽要靠導磁性材料包覆(refer to LE shielding)

遠場的高頻屏蔽主要以入射(吸收)損耗 AdB 為主(上圖紅色線),而近場的高頻屏蔽要靠導電性材料包覆產生感應電流底消原磁場(refer to LE shielding)

對遠場(平面波)來說,S.E.的反射損耗RdB與導體厚度無關,但入射(吸收)損耗AdB則與導體厚度息息相關;前者主導E-field的S.E.,後者主導H-field的S.E. [2]p.6

1.2 How to simulate S.E. with HFSS

在屏蔽導體的一側激發能量,看另一測量到的穿透能量是多少。用HFSS建一個金屬平面,在其上下分別設wave port看S21即可。

S.E.越大者,相當於S21越小

For HFSS R17.x and older version, to calculate S.E., you have to use real thickness metal with "solver inside". From now on, R18.x, both real thickness metal with "solver inside" and zero thickness with "two-side shell element" can be used, and they get very consistent results.

  1. Shielding Effect of Different Materials 

2.1 Ni(鎳)在低頻(<0.4G)時沒有Cu好,但大致來說具導磁性的Ni (relative permeability=600)的S.E.比Cu好

可以明顯看出由鎳所構成的屏蔽面,其S21隨著頻率的增加快速衰減。這與我們在S.E. theory所得到的結論相符

2.2 S.E.跟導體厚度有絕對的關係 ,越厚的導體屏蔽效果越好(即1.1的式子裡的t越大)

可以明顯看出5um厚度的S21比1um厚度的S21小 ,即越厚的金屬,屏蔽效果越好。

同樣的導體厚度下,鎳較早發生肌膚效應(電荷開始集中於<1/3導體厚度 的表層區域內),銅在較高的頻率才會開始發生肌膚效應,AdB開始快速增加,S21開始快速下降。

為何鎳的S21的下降斜率隨著頻率的增加而漸緩,甚至最後變水平? refer to 9.3

2.3 銅鎳合金的效果

我們可以很容易的預期,1um Cu+1um Ni的S.E.一定比單用Cu或Ni來的好,但如果是0.5um Cu+0.5um Ni呢?

0.5um Cu+0.5um Ni的S21隨頻率衰減的速度較緩

設計shielding case採用合金的原因不只有S.E.的考慮,還有加工時的吃錫效果、抗氧化(鏽蝕)、應力、延展性...等考慮[3]

  1. EMI with Grounding and Opened Shielding Cas

3.1 Uncheck "solve inside"的結果

3.2 Check "solve inside"的結果

  1. EMI with Non-Grounding and Opened Shielding Case 

4.1 Uncheck "solve inside"的結果

4.2 Check "solve inside"的結果

對於只有五個面所組成的屏蔽金屬罩,若下方所處理的EMI source (PCB)沒有完整密合的GND plane,那電磁輻射就會從這開孔/縫隙洩漏出去,而容易洩漏出的頻段一般是這開孔(縫隙)長度的lambda/4~lambda/10以上。

  1. EMI with Grounding and Closed Shielding Case 

5.1 沒有check "solve inside"時,不管是近場或遠場,shielding case外的E值都是0 (錯誤)

5.2 有check "solve inside"時,可以看出屏蔽效果比step4.2好很多,且Cu shielding case效果比Ni好

試著觀察shielding case對1GHz,在外表面上的感應電流,在Cu shielding case上滲出的感應電流小很多,所以銅做的shielding case對EMI source做close loop包覆,所得到的EMI far-field屏蔽效果較鎳好。

為何用closed shielding case包覆EMI source後,反而是Cu得到較佳的屏蔽效果,而不是如S.E. theory所示Ni比較好? refer to 9.7

  1. EMI with Non-Grounding and Closed Shielding Case 

6.1 沒有check "solve inside"時,不管是近場或遠場,shielding case外的值都是0 (錯誤)

6.2 有check "solve inside"時,Cu shielding case效果仍然較好

  1. CISCO's paper in DesignCon 2013 [4] 

待測物

S.E. test量測環境示意

模擬與量測對比

在單位開孔數相同的前提下,開孔大、透氣度佳的,S.E.會比較差。而直接把屏蔽層加厚會改善S.E.

  1. Another Case with Apertures 

3M Far-field emission for a normalize EMI source (the red line) with a copper enclosure (the green line as below)

8.1 Radiation emission at 1.35GHz

8.2 Radiation emission at 2.95GHz

  • Step4,5,6與S.E. theory公式差異之二是,該式子 無法反映真實系統中屏蔽機殼空腔接地與否的效應。

  1. Reference 

[1] Lesson9 -- Shielding, 台大吳宗霖教授2012 EMC課程 (推薦)

[2] Frank Leferink, "Fort Lauderdale, Tutorial on EMC Fundamentals: Shielding", IEEE EMC Symposium 2010. (推薦)

[3] 屏蔽框設計與生產注意事項

[4] X. Zhou, "Validating EMC Simulation by Measurement in Reverberation Chamber", DesignCon 2013. (推薦)

[5] Electromagnetic shielding \ Materials used -- WiKi

[6] Electromagnetic shielding \ How electromagnetic shielding works -- WiKi

[7] 密西西比大學老師的EMC教材 -- 詳細介紹near-field shielding與far-field shielding差異 (推薦)

[8] B. Archambeault, A. Radchenko, "Low-Frequency Magnetic Field Shielding Physics and Discovery for Fabric Enclosures Using Numerical Modeling", IEEE EMC Symposium 2014.

[9] Effective Nuclear Charge and the Shielding Effect -- 以電荷庫倫力的角度來解釋shielding effect

[10] 介紹屏蔽夾取代屏蔽框

[11] EMI Shielding Design - Darcoid Rubber Company