338

Journal of Electrical Engineering & Technology, Vol. 1, No. 3, pp. 338~342, 2006

Power Loss Calculation of High Frequency Transformers

Geun-Soo Choi*, Shin-Yong Yoon*, Soo-Hyun Baek** and Kim Yong** Abstract This paper analyzed the power loss of transformers considering the magnetic component.
For this, each winding strategy and the effect of air gap between the ferrite core have been an important variable for optimal parameter calculation. Inductors are very well known design rules to devise, but the performance of the flyback converter as a function of transformer winding strategy has not been fully developed. The transformer analysis tool used was PExpert. The influence of the insulator thickness, effect of the air gap, how the window height and variation of the capacitive value effects the coil and insulator materials are some of parameters that have been analyzed in this work. The parameter analysis is calculated to a high frequency of 48[kHz]. Therefore, the final goal of this paper was to calculate and adjust the parameters according to the method of winding array and air gap minimizing the power loss.

Keywords: Flyback, High Frequency Transformer, PExpert, Power loss, Winding strategy

1. Introduction
SMPS (switch mode power supply) is used in the application of industrial control systems, namely OA, FA, robotics, Information machines, and LCD backlights. How it is designed is very important, particularly in the case of transformer design considering power loss at magnetic components. The transformer in this case is the flyback transformer that is able to store energy using a turn off switch. The purpose of this is to transfer power efficiently and instantaneously from an external electrical source to an external load [1-3]. Magnetic components are one of the important components in power electric circuits. The design stage of these components should be carried out not only taking into account the electrical specification of the circuit, but also the influence of the geometrical parameters as well as the material properties of all the parts that form the magnetic components. The usual procedure to design magnetic components allows us to obtain some of their parameters. However there are other very important parameters that are not included in that procedure due to the difficulty to quantify their influence on the magnetic components. The design method proposed in this paper is based on the use of a model that accounts for geometry and frequency effects using the FEA tool. The performance of the final component can be clearly improved by using the right winding strategy and materials for a given electrical
Dept. of Electrical Eng. Cop, DGI, 517 Deok-dong, Yangju-si, Kyunggi-do Korea. Phone : 82-31-860-1200 Fax : 82-31-865-0068 ( research@dgi-net.com ) ** Professor, Dept. of Elec. Eng, Dongguk University, Phil-dong 3, Joong-gu, Seoul, Korea Received: May 17, 2005 ; Accepted: January 27, 2006 *

specification [4-6]. This paper analyzes the effect of winding strategy and air-gap variation of the flyback transformer. Therefore, from those power losses, parameters for resistance, leakage inductance and temperature can be acquired.

2. Flybak Transformer
Fig. 1 indicates the flyback transformer converter driving the CCFL (Cold Cathode Fluorescent Lamp) by amplification of the piezoelectric transformer.

Fig. 1. Inverter of flyback transformer for CCFL Fig. 2 shows the dimension of the flyback transformer for Fig. 1. Here, this construction is applied to the core transformer of EI 28 type of the TDK corp.

(a) EI28 type (b) Dimension values Fig. 2. Transformer construction and dimension

Geun-Soo Choi, Shin-Yong Yoon, Soo-Hyun Baek and Kim Yong

339

2.1 Power Loss

Transformer losses can be put into three major categories: core hysteresis losses, core eddy current losses, and winding losses. Transformer loss is sometimes limited directly by the need to achieve the required overall power supply efficiency. More often, transformer losses are limited by a maximum hot spot temperature rise at the core surface within the center of the windings. The winding loss worst case is always at low Vin, full load. The analysis procedure is based on the use of the Maxwell Equation, shown below as (1) and (2).

eddy current loss is proportional to Vin. Worst case is at high voltage Vin. Winding loss is always greatest at low Vin. Fig. 3 is for self-capacitance and mutual capacitance of equivalency with analysis models.

(1)

Fig. 3. Self-capacitance and mutual capacitance of EI28 models

(2) Here, J = E , B = H DC resistance and AC resistance are represented as the following Equations (3) and (4).
RDC = l A

3. Simulation Result

(3)

RAC = RDC 1.6

(4)

DC resistance is constant to frequency, and AC resistance is high variable to frequency. The power losses of the winding are shown as follows in Eqs. (5), (6), (7).

(a)

(b)

P1 w = I 12DC R1 DC + I 12AC R1 AC
P2 w = I 22DC R 2 DC + I 22 AC R 2 AC
Pw = P1 w + P2 w

(5) (6) (7) (c) (d)

The power loss of the core is indicated as follows in Eq. (8).

PC = K 1 f
K 2

BK3

(8)

Here, PC : core loss, Pls : power loss, K1 : 0.32,

K 2 : 1.61, K 3 : 2.68.
Total power losses are shown as follows in Eq. (9).
Pls = PC + Pw

(9)

Fixed frequency operation, volts-seconds and therefore flux swing are constant. Hysteresis loss is therefore constant, regardless of changes in Vin or load current. Core

(e) (f) Fig. 4. Winding strategy of EI type core

340

Power Loss Calculation of High Frequency Transformers

Fig. 4 shows the represented analysis result of the EI28 type transformer by FEA Export. Here, constructions of analysis objects are the same as the following conditions: core size EI28, bobbin EI28, core material PC40, manufacture corp. TDK, window width 5,90[mm], window height 12,25[mm], volume 4145.20[mm]. The common condition for analysis of Fig. 4, circuit simulation conditions of the flyback transformer are as follows: duty cycle 0.5%, switching frequency 48kHz, operation condition: continuous mode, magnetizing inductance: 130.23[ H], Vin : 5[V], Irms : 1.417[A], permeability : 0.23[T], at Fig.3(a) winding1 structure: AWG 25, gap: 0.1637[mm], number of turns 15, parallel Turns 3. Fig. 4(b) winding1 structure: AWG 20, gap: 0.1931[mm], number of turns 16, parallel turns 1. Fig. 4(c) winding1 structure: AWG 25, gap: 0.1637[mm], number of turns 15, parallel turns 1. Fig. 4(d) winding1 structure: AWG 30, gap: 0.6151[mm], number of turns 25, parallel Turns 3. Fig. 4(e) winding1 structure: AWG 35, gap: 2.6416[mm], number of turns 42, parallel Turns 3. Fig. 4(f) winding1 structure: AWG 35, gap: 2.6416[mm], number of turns 42, parallel Turns 3. Here yellow rounding coils signify 1 order winding, blue rounding coils signify 2 order winding. Table 1 reveals the AC resistances, DC resistances, and capacitance analysis by PExprt. Table 1. AC/DC Resistance and Capacitance
Items AC Res. Wind.1 [m] Wind.2 [] Wind.1 [m] Wind.2 [m] Self Cap [pF] Mutual. Cap. [pF]
1 35.5 3.8 35.6 35.6 6.86 31.7 2 37.8 7.07 36.1 36.1 4.61 25.8 3 98.8 11.5 98 98 6.8 31.7 4 184.5 19.17 184 184 10.55 48.8 5 966.3 31.52 966.3 966.3 10.55 48.96 6 966.3 10.04 966.3 966.3 10.55 48.96

4 Pow er loss[w ] 3.5 3 2.5 2 1.5 1 0.5 0

Winding type

Fig. 6. Power loss to winding types Fig. 5 shows winding loss, core loss, and total loss for the winding constructor of Fig. 4. Here number 6 is largest at total loss, because the gap and turns are big, and the diameter is small. Number 1 is lowest at total loss, because gap and turns are small, and diameter is large. Fig. 6 represents the power loss of total loss at Fig. 4. It is equal to the explanation of Fig. 5.

3 2. 5 Gap[mm] 2 1. 5 1 0. 5 0 1 2 3 4 5 6 W indings ty pe

DC Res.

Fig. 7. Gap values to winding types

100 90 80 Temperature[^c

Cap.

4.5 4 3.5 3 2.5 2 1.5 1 0.5 0

Total loss Core Windings

70 60 50 40 30 20 10

3 4 Windings type

0 1 2 3 4 5 6 Windings type

Fig. 5. Winding loss, core loss, and total loss for winding constructors

Fig. 8. Temperature for winding arrays

Geun-Soo Choi, Shin-Yong Yoon, Soo-Hyun Baek and Kim Yong

341

Fig. 7 shows the gap value for winding type. The explanation for this is represented in Fig. 4. Fig. 7 depicts temperature loss, with number 6 indicating the biggest temperature rise. Fig. 9 indicates DC loss and resistance for winding type. Here Numbers 1 and 2 represent the least amount of loss.
2500 2000 Resistance, Loss 1500 1000 500 0 1 2 3 4 5 6 Windings type

Fig. 12 represents the flux density distribution of the above analysis model (EI28) by Maxwell-2D.

DC Losses[m W] DC Res.[m ohm ]

Fig. 9. DC loss and resistance for winding types Fig. 12. Flux density distribution of EI28 type core Fig. 10 represents AC resistance at 1 and number 2 shows winding. Number 6 indicates the largest ac resistance.
Winding 1 120 100 AC Resistance 80 60 40 20 0 1 2 3 4 5 6 Windings type Winding 2

3. Conclusions
Transformer design at the SMPS is very important to considering loss. For this, it is required to obtain the parameters for loss considering winding strategy. This paper analyzed power loss according to the winding strategy of the flyback transformer. Here, the simulation tool used was PExprt. From the above analysis result considering parameters, namely winding diameter, gap, turns, resistance, temperature, leakage inductance etc., the EI28 type of number 1 shows the smallest power loss.

Fig. 10. AC resistance of 1, 2 winding Fig. 11 illustrates leakage inductance to winding type. Here number 3 indicates lowest leakage inductance, while numbers 5 and 6 represent the biggest leakage inductance.
2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 1 2 3 4 5 6 Winding types

References
[1] J. P. Vandelac and P. Zioggas, A Novel Approach for Minimizing High Frequency Transformer Copper Losses, Journal of IEEE, pp. 355-367, 1987. [2] R. Petkov, Design Issues of the High Power High Frequency Transformer, IEEE PCIM, pp. 401-410, 1995. [3] R. Prieto, J. A. Cobos, O. Garcia and J. Uceda, Study of the Sensitivity of the Parameters in Magnetic Components, PESC, pp. 355-360, 1997. [4] R. Prieto, J. A. Cobos, O. Garcia and J. Uceda, 1D Magnetic Component Model for Planar Structures, PESC, pp. 574-579, 1999. [5] R. Prieto, J. A. Cobos, O. Garcia and J. Uceda, Using

Leakage inductance[uH

Fig. 11. Leakage inductance to winding types

342

Power Loss Calculation of High Frequency Transformers

Parallel Winding in Planar Magnetic Components, PESC , pp. 2055-2060, 2001. [6] L.M. Escribano, R.Prieto, J.A. Oliver, J.A. Cobos and J. Uceda, New modeling strategy for the fringing energy in magnetic components with air gap, PESC, pp. 144-150, 2002.

Geun-Soo Choi He received his B.S. degree in Electrical Engineering from Wonkwang University in 1983. He received his M.S. degree from Dongguk University, Seoul, Korea in 1995. He has been studying to obtain his Ph.D at Dongguk University since 2002. His main area of interest includes Inkjet printing systems, Digital Textile Printing, Servo motor control, FPCB control systems and Cutting Plotters.

Soo-Hyun Baek He received his B.S., M.S. and Ph.D. degrees in Electrical Engineering from Hanyang University in 1972, 1974 and 1983, respectively. Since 1977, he has been a Professor in the Department of Electrical Engineering at Dongguk University, Seoul, Korea. He has also been a member of the KIEE and ICEE. His major research interests are in the areas of motor control systems, power electronics, ultrasonic motors and the design and analysis of electrical machines.

Shin-Yong Yoon He received his B.S. degree in Electrical Engineering from Seoul National University of Technology in 1992. He received his M.S. and Ph.D degrees in Electrical Engineering from Dongguk University, Seoul, Korea in 1994 and 1999, respectively. His main research interest includes Inkjet printing, Servo motor control, and Linear motor design.

Kim Yong He received his B.S., M.S. and Ph.D. degrees in Electrical Engineering from Dongguk University in 1981, 1984 and 1994, respectively. Since 1977, he has been a Professor in the Department of Electrical Engineering at Dongguk University, Seoul, Korea. His major research interests are in the areas of motor control systems, SMPS and Grounding systems.