Recent Advances in Electrical and Electronic Engineering
Optimal design of 400 Hz power filter for aircraft switching power
supply
JU MIN LEE, HEON WOOK SEO, SUNG SU AHN, JIN DAE KIM
Daegu Mechatronics & Materials Institute
32, Seongseogongdan-ro 11-gil, Dalseo-gu, Daegu
REPUBLIC OF KOREA
jmlee@dmi.re.kr, seo7712@dmi.re.kr, zecks@dmi.re.kr, jdkim@dmi.re.kr
Abstract: - This study aimed to develop a AC 115 V ± 10 %, 3-phase, 400 Hz, 4 A power filter with an
attenuation efficiency of up to 70 dB and insulation resistance of below 0.5 Ω for application to the switching
power supply of military aircraft. The 400 Hz power filter was designed and fabricated as an all-in-one
structure with components that were minimized in terms of size and weight and a low pass filter for the
switching noise. The test results showed that it had an excellent attenuation efficiency of up to 73.66 dB in the
100 kHz ~ 30 MHz band. Furthermore, a performance verification test was performed by applying the power
filter to an aircraft switching power supply to check for the inhibition of electromagnetic interference in the
power unit, and the results confirmed its reliability.
Key-Words: - EMI (Electromagnetic Interference) Filter, Insertion loss, Attenuation factor, RLC circuits, Power
filters, Switching noise.
Section 2 explains the characteristics of aircraft
switching power supply, which is EUT to which the
power filter is applied, and its noise. In Section 3,
the actual process of designing the filter for a
switching power supply and the power filter that has
been designed and developed as a result are
explained. The results of the experiment, measuring
the characteristics of the power filter and applying
the developed power filter to an aircraft switching
power supply, and the conclusions of the study are
presented.
1 Introduction
Although electronic equipment for aircraft has
become more lightweight with enhanced
performance, output and precision, the increasing
intensity of electromagnetic noise from electronic
components due to the limited internal space has
arisen as an important issue that must be dealt with.
Thus, it has become fundamental to consider the
aspect of removing electronic magnetic inference
(EMI) noise during the design and fabrication
processes, while maintaining the aircraft system
functions at the same time.
Aircraft contains multiple power generation
devices including the main and auxiliary power
supplies in order to supply power to the essential
equipment even in the event of emergency
situations. The main power is supplied by the AC
generator, which is usually driven by the engine,
and because all AC generators use 400 Hz, the
electrical and electronic components of aircraft are
constantly exposed to electromagnetic waves. By
applying an appropriate technology in filter design
for noise removal that can inhibit the EMI, the
electromagnetic compatibility (EMC) of the cable
and internal interconnection can be improved.
This paper introduces the design method for a
power filter, satisfying the 400 Hz requirement, that
has been optimized to raise the reliability of aircraft
switching power supply.
ISBN: 978-960-474-399-5
2 Switching Power Supply
Switching power supply is equipment that supplies
low-voltage, high-current power to the radar
processor unit of military aircraft. The design of
EUT applied for the verification of the 400 Hz
power filter will not be presented in this paper for
security purposes due to its nature as a military
device.
2.1 EUT description
Fig. 1 is a block diagram of the multi-output
switching power supply [1-3]. The output power of
the switching power supply is turned on and off by
the control signal sent from the external system to
the radar processor unit, and this is synchronized
with the input frequency signal, causing the
73
Recent Advances in Electrical and Electronic Engineering
[ dB(μV)
120
switching frequency to vary. The 115 VAC, 3phase, 400 Hz power of the aircraft passes generates
6 types of output power converted from the bus
power generated through the 3-phase, full wave
rectification and smoothing circuits, as well as 28
VDC and 5 VDC AUX power.
“B” Phase
110
100
90
80
Level
70
60
50
40
30
OUTPUT
Rectifier
115 VAC
3 Ф 400 Hz
Cap.
INRUSH
Auxiliary output
+5V
20
+ 3.3 V
10
+5V
+5V
0
10.000
+ 12 V
+ 12 V
- 12 V
+ 15 V
- 15 V
- 12 V
+ 15 V
- 15 V
+ 3.3 V
100.000
1000.000
[ kHz]
(b)
[ dB(μV)
120
“C” Phase
110
On/ Off
Sync.
BIT
Control
10000.000
Frequency
100
90
DC EMI Filter
80
+ 28 V
Auxiliary output
70
+5V
(AUX)
+5V
Level
+ 28 VDC
60
50
Fig.1 Switch Power supply block diagram.
40
30
20
10
2.2 Noise Characteristics of the EUT
0
10.000
Test Fixture
LISN LISN LISN
Electric Load
It is necessary to develop a power filter for noise
removal in order to ensure the flight safety and the
quality of the power supply. Unlike general forms of
noise removal devices, it is essential that the
components are compact and lightweight to keep the
weight minimal for maximum performance in such
a limited space [5].
In the case of aircraft using AC power, the size
of the AC power unit is dependent on the frequency.
Of particular note, magnetic components (various
types of transformers, generators and filters, etc.)
take up more than half of the equipment and the
sizes of these components are dependent on the
frequency. Increasing the frequency can minimize
the equipment size, and this is why 400 Hz
equipment is used. The reason for using 400 Hz,
instead of a higher frequency, is that this is the
design parameter that has been globally recognized,
considering the overall constraints such as the
frequency limitations of the applied components and
the impact of the frequency on external electronic
equipment, as well as the safety issues associated
with long-term use.
LISN LISN
Switch
Power supply
Chamber
Fig. 2 CE102 Set-up.
The result of measuring each phase in the CE102
set-up as shown in Fig. 2 showed that 400 Hz
switching noise was produced in the entire 10 kHz ~
10 MHz band, and that it was over the required
limit, as shown in Fig. 3 [4].
Based on this result, the power filter was
designed to inhibit the 400 Hz switching power
noise and its performance was verified. These
processes are explained in the following section.
[ dB(μV)
120
“A” Phase
110
100
90
80
Level
70
60
50
40
30
20
10
1000.000
Frequency
10000.000
[ kHz]
(a)
ISBN: 978-960-474-399-5
[ kHz]
3 Power Filter Development
Test Table(GND)
100.000
10000.000
Fig. 3 Measuring the power supply noise before the
filter application, (a) Conducted noise of A Phase,
(b) Conducted noise of B Phase, (c) Conducted
noise of C Phase.
DC 28 V
0
10.000
1000.000
(c)
Control Room
AC 115 V, 400 Hz
100.000
Frequency
In order to verify the optimal design of the 400 Hz
AC power filter applied to the developed power
supply, the noise in the power unit prior to the
installation of the filter was measured. Based on the
aircraft air force limit for the applied specification
of MIL-STD-461F, an emission test was conducted.
74
Recent Advances in Electrical and Electronic Engineering
induction between the device frames, deteriorating
the frequency characteristics of the filter. Thus, the
grounding wire of the filter should be as thick and
short as possible.
Based on the volumes of the selected
components, the case was designed and fabricated
as shown in Fig. 6. For the case, metal with high
permeability was used to increase the absorption
loss and it was heat treated for a shielding effect.
Fig. 4 Power filter design stages.
(a)
Fig. 4 shows the design method implemented to
develop the filter. Before designing a filter, it is
important to separate the CM noise and the DM
noise for noise flow and component analysis.
As shown in the graph in Fig. 3, DM noise in
400 Hz multiplication was confirmed in the
frequency band below 150 kHz, and it was
discovered that CM noise generated from DC-DC
converter switching multiplied in the frequency
band over 150 kHz.
Next, the cut off frequency ( F0 ) for CM and
DM was chosen as 10 kHz and Eq. 1 was used to
obtain the component capacity values.
L=
Rd
,
2πF0
C=
1
2πF0 Rd
(b)
Fig. 6 Shape and Dimensions, (a) Side view, (b) Top
view.
4 Experiment results
A filter may satisfy the requirements based on a
characteristics measurement test (CISPR17, MILSTD-220B, etc.), but its characteristics tend to vary
when they are used as a single unit and when they
are applied to an actual device. Thus, an experiment
was carried out to examine the attenuation
characteristics when used alone and the conductivity
noise characteristics when it is applied to an actual
device.
(1)
4.1 Filter Characteristics Measurement
Fig. 5 Circuit diagram of the power filter.
Fig. 7 shows the power filter that was actually
developed in this study.
The circuit diagram of the developed power filter
shown in Fig. 5 shows that MPP core was applied to
the L2 component and film capacitor was applied to
the C1 component for reduction of DM noise. In
order to reduce the Cm noise, ferrite core was
applied to the L1 component and ceramic capacitor
was applied to the C2 component as a means to
minimize the distance with the filter case, which
was the grounding point [6-9]. When the grounding
wire of the filter is long, there is a coefficient of
ISBN: 978-960-474-399-5
Fig. 7 External appearance (left) and internal
appearance (right).
75
Recent Advances in Electrical and Electronic Engineering
Table 1 Performance criteria.
No.
Performance
checklist
1
Rated
voltage/current
2
Insulation
resistance
Attenuation(dB) = 20 Log10
Criteria
e2
e1
(2)
e1 : The level reached with the noise filter
e2 : The level reached without the noise filter
AC 115 V, 400 Hz, 4 A
Over 10 ㏁ when applying 250 VDC between
the terminal and the case
Perform conduction test according to the circuit
diagram (below 0.5 Ω )
Equivalent or higher attenuation factor (Method:
MIL-STD-220B)
Attenuation
factor
3
Frequency
(MHz)
0.1
0.5
1
5
10
Normal
20
30
30
20
10
5
5
Common
5
10
10
30
20
20
20
20
30
(a)
(b)
(c)
(d)
(e)
(f)
DC
Source
+
RF- DC
Adapter
Coaxial
switc h
Signal
Generator
-
Buffer
Network
Assembly
Buffer
Network
Assembly
Filter
Under
Test
RF cable
adapter
RG- 214/ U
Isolation
Attenuation
RF- DC
Adapter
Coaxial
switc h
Isolation
Attenuation
Receiver
Output
Meter
Fig. 8 MIL-STD-220B : Concept for measuring the
attenuation characteristics of the filter.
Table 1 shows the performance criteria for the
developed filter, and the attenuation characteristics
were measured using the MIL-STD-220B method,
shown in Fig. 8 [10-11].
A test was conducted using a filter
analyzer(Model FA-2100) and related measurement
equipment(Model LSA 265). The measurement of
the attenuation factor in each phase of the 400 Hz
power filter from 100 kHz to 30 MHz showed that it
was an excellent filter with attenuation factor of up
to 70 dB, which was above the required level, as
shown in Table 2 and Fig. 9.
Fig. 9 Attenuation characteristics of the frequencies
with respect to each line filter, (a) Attenuation of A,
B Normal, (b) Attenuation of A, C Normal, (c)
Attenuation of B, C Normal, (d) Attenuation of A
Common, (e) Attenuation of B Common, (f)
Attenuation of C Common.
4.2 Noise Characteristics of the EUT with
power filter
Table 2 Input loss of EMI filter.
Normal Mode
Mode
Frequency
(MHz)
Attenuation
(dB)
0.1
0.5
1
5
10
20
30
35
47
43
26
32
27
22
Common Mode
Mode
Frequency
(MHz)
Attenuation
(dB)
The switching power supply supplies power to each
of the internal elements through the EMI filter when
it receives an input of 115 VAC from an external
power source. The EMI filter blocks the noise from
coming inside through the 115 VAC input and the
internal noise from leaking to the outside.
Maximum value of Fig. 9 (a)~(c)
Maximum value of Fig 9. (d)~(f)
0.1
11
0.5
32
1
41
5
74
10
51
20
55
AC EMI Filter
30
115 VAC
3 Ф 400 Hz
56
INRUSH
Cap.
Auxiliary output
+5V
On/ Off
Sync.
An attenuation factor of 20 dB means that the
noise level has been reduced by 1/10, while 40 dB
means 1/100 reduction and 60 dB means 1/1,000
reduction. This relationship is shown in Eq. 2.
ISBN: 978-960-474-399-5
OUTPUT
Rectifier
+ 3.3 V
+ 3.3 V
+5V
+5V
+ 12 V
+ 12 V
- 12 V
+ 15 V
- 15 V
- 12 V
+ 15 V
- 15 V
Control
BIT
Fig. 10 Block diagram of the switch power supply
with EMI filter.
76
Recent Advances in Electrical and Electronic Engineering
Fig. 10 shows an input EMI filter in the power input
unit for the reduction of noise generated outside or
inside the power supply unit. The filter should be
mounted as closely as possible to the input/output
ports of the equipment and the input/output cables
of the filter should not be overlapped to allow
maximum attenuation performance of the filter.
A conductivity test was conducted in accordance
with the military specifications to check whether the
power filter satisfied the reliability and filter
characteristic criteria set forth for military
equipment during actual application. As shown in
Fig. 11, the test results showed that it conformed to
all the criteria.
supply units of military aircraft. The filter was
designed to minimize the 400 Hz switching power
noise and as an all-in-one structure to enhance the
utility of the limited space. The structural design is
characterized by a through pipe for the withdrawal
of the wiring that is welded to the body, which helps
maintain high shield effectiveness.
In an experiment, the power filter showed
outstanding performance with an attenuation factor
of up to 70 dB higher than the required value. On
the other hand, a power unit EMC reliability test
showed that the switching noise characteristics of
the power supply improved when a model with
enhanced capacity to block the DM noise was
applied to an aircraft switching power supply.
[ dB(μV)
120
“A” Phase
110
100
6 Acknowledgement
90
80
This research was supported by Economic Region
Collaborative Program through the Korea Institute
for the Advancement of Technology (KIAT) funded
by the Ministry of Trade, Industry and Energy
(R0002611)
Level
70
60
50
40
30
20
10
0
10.00
100.00
1000.00
10000.00
Frequency
[ kHz]
(a)
References:
[1] S. J. Robert, Synchronous Rectification in
High-Performance Power Converter Design,
Texas Instruments, www.ti.com.
[2] B. Keith and M. Taylor, Switch mode Power
Supply Handbook 3rd edition, McGraw Hill,
2011.
[3] S. Lopez Arevalo, IEEE Trans. on Industrial
Electronics, Control and Implementation of a
Matrix-Converter-Based AC Ground PowerSupply Unit for Aircraft Servicing, Vol. 57,
Issue.6, 2010, PP. 2076-2084.
[4] Military
Standard
:
MIL-STD-461F,
Requirements
for
the
Control
of
Electromagnetic Interference Characteristics
of Subsystem and Equipment, Department of
Defense, December 2007.
[5] J. Milton and E. Greenwood, IEEE Trans. on
Electromagnetic Compatibility Special Filter,
Improving the Specifications For Power-Line
Filters, Vol.EMC-10, Issue.2, 1968, pp. 264268.
[6] H. M. Hoffart, IEEE Trans. on Electromagnetic
Compatibility, Electromagnetic Interference
Reduction Filters, Vol.EMC-10, Issue.2, 1968,
pp. 225-232.
[7] D. J. Jobe, Electromagnetic Compatibility
Symposium Record, 1969 IEEE, Selection and
Test of Power Line Filters for Use in
Equipment Designed to Meet Government
[ dB(μV)
120
“B” Phase
110
100
90
80
Level
70
60
50
40
30
20
10
0
10.00
100.00
1000.00
10000.00
Frequency
[ kHz]
(b)
[ dB(μV)
120
“C” Phase
110
100
90
80
Level
70
60
50
40
30
20
10
0
10.00
100.00
1000.00
Frequency
10000.00
[ kHz]
(c)
Fig. 11 Measuring the power supply noise after the
filter application, (a) Conducted noise of A Phase,
(b) Conducted noise of B Phase, (c) Conducted
noise of C Phase.
5 Conclusion
In this study, a 400 Hz power filter was developed
for application to 8 multi-output switching power
ISBN: 978-960-474-399-5
77
Recent Advances in Electrical and Electronic Engineering
Electromagnetic Compatibility Specifications,
1969, pp. 283-289.
[8] H. M. Schlicke and H. Weidmann, IEEE
Spectrum, Compatible EMI Filters, Vol.4,
Issue.10, 1967, pp. 59-68.
[9] H. Weidmann and W. J. McMartin, IEEE Trans.
on Electromagnetic Compatibility Special
Filter, Two Worst-Case Insertion Loss Test
Methods for Passive Power-Line Interference
Filters, Vol.EMC-10, Issue.2, 1968, pp. 257263.
[10] Military Standard : MIL-STD-220B, Method of
Insertion Loss Measurement, Department of
Defense, 24 January 2000.
[11] H. M. Schlicke, H. Weidmann and H. S.
Dudley,
Electromagnetic
Compatibility
Symposium Record, IEEE, The Controversial
MIL STD 220A, 1969, pp. 215-226.
ISBN: 978-960-474-399-5
78