Design and Development of Low-Loss Transformer for Powering Small Implantable Medical Devices-gWI.pdf

IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS, VOL. 4, NO. 2, APRIL 2010

Design and Development of Low-Loss Transformer

for Powering Small Implantable Medical Devices

Kenji Shiba , Member, IEEE , Akira Morimasa, and Harutoyo Hirano

ABSTRACT— Small implantable medical devices, such as wireless

capsule endoscopes, that can be swallowed have previously been

developed. However, these devices cannot continuously operate

for more than 8 h because of battery limitations; moreover, addi-

tional functionalities cannot be introduced. This paper proposes

a design method for a high-efﬁciency energy transmission trans-

former (ETT) that can transmit energy transcutaneously to small

implantable medical devices using electromagnetic induction.

First, the authors propose an unconventional design method to

develop such a high-efﬁciency ETT. This method can be readily

used to calculate the exact transmission efﬁciency for changes in

the material and design parameters (i.e., the magnetic material,

transmission frequency, load resistance, etc.). Next, the ac-to-ac

energy transmission efﬁciency is calculated and compared with

experimental measurements. Then, suitable conditions for prac-

tical transmission are identiﬁed. A maximum efﬁciency of 33.1%

can be obtained at a transmission frequency of 500 kHz and a

receiving power of 100 mW for a receiving coil size of 5mm 20

mm. Future design optimization is possible by using this method.

INDEX TERMS— Capsule endoscope, energy transmission, im-

plantable medical device, magnetic material, transmission

efﬁciency.

wireless endoscope, and it became possible to transfer energy

up to 50 mW with an efﬁciency of 36% at a distance of 30 mm

(an efﬁciency of less than 10% at a distance of 70 mm). Mori-

masa [23] and Hirano [24] developed the air-core-type trans-

former for implantable medical devices, and their receiving coil

10 mm in diameter could receive 30 mW with an efﬁciency of

16% at a distance of 100 mm. In this research, however, mag-

netic material was not adopted and its loss was not considered.

To obtain more power and to increase energy transmission ef-

ﬁciency, the introduction of magnetic material into small im-

plantable medical devices is expected. Leuerer [25] developed

a planar coil with a magnetic layer for a telemetric system using

ﬁnite-element method (FEM) analysis, and 4–6 mW of energy

could be transferred to the 6-mm diameter receiving coil. How-

ever, FEM analysis is time-consuming and expensive, because

the designer must make and analyze each model with 2-D or

3-D software. Therefore, for a quick analysis of various types of

transformers using various types of magnetic materials at var-

ious frequencies, it is not an adequate method. Incidentally, an

energy transmission transformer (ETT) for an industrial instru-

ment has already been developed [26]–[29]. For example, Kim

et al. [26] transmitted about 40 W using a transmitting coil of

3 000 mm and a receiving coil of 2 500 mm at a transmission

distance of 3 mm. However, this conventional design method

does not consider small magnetic loss and it allows a compar-

atively large power loss because the industrial transformer is

available unless the magnetic material reaches magnetic satu-

ration. Moreover, an industrial transformer is allowed to have a

55–140 C increase in temperature, because there is no limit ex-

cept the Curie temperature of the magnetic core or the melting

temperature of the electrical insulator [30]. For a capsule endo-

scope or implantable medical device, the increase in tempera-

ture of various parts of the transformer must be in the range of

2 –5 at most [31].

In this paper, we propose a design method for the bconstruc-

tion of an ETT for small implantable medical devices. The de-

sign method includes new formulas that accurately express core

loss.

such as a wireless capsule endoscope and a nerve stimu-

lator have been developed [2]–[17]. The M2A [2]–[12] (Given

Imaging Ltd.), commercially released in 2001, is a wireless cap-

sule endoscope that is 26 mm long and 11 mm in diameter. The

Endo Capsule [17] (Olympus Corp.), which is of the same size,

was released in 2005. These capsule endoscopes capture images

at 2 frames/s for 8 h using internal batteries. However, other op-

tional functions for capsule endoscopy are necessary, such as

sampling or spraying medicine [18]. And if these optional func-

tions are included in a capsule endoscope, its working time re-

duces. Thus, some methods for supplying energy to a capsule

endoscope are needed.

Some researchers have studied the use of wireless energy

transmission systems for implantable medical devices. For ex-

ample, Neagu [19] designed a planar microcoil (air core type)

for implantable microsystems, and conﬁrmed that a receiving

coil with a diameter of 4.5 mm can receive a few milliwatts.

Puers [20]–[22] designed an inductive power-link system for a

II. E NERGY T RANSMISSION S YSTEM

Fig. 1 shows the energy transmission system for implantable

medical devices. The power is transmitted transcutaneously by

electromagnetic induction between two coils, one placed inside

and the other outside the body. For a wireless capsule endo-

scope, the receiving coil must have the dimensions mm

mm [4]. The power consumption of a commercial capsule

endoscope is estimated at about 10–30 mW [16] and for op-

tional functions, such as self-propulsion and drug administration

Manuscript received May 18, 2009; revised August 10, 2009. Current version

published March 24, 2010. This paper was recommended by Associate Editor

R. Sarpeshkar.

The authors are with Graduate school of Engineering, Hiroshima Univer-

sity, Kagamiyama, Hiroshima, 739-8527, Japan (e-mail: kenjishiba@nifty.com;

morimasa@bsys.hiroshima-u.ac.jp; harutoyo@bsys.hiroshima-u.ac.jp).

Digital Object Identiﬁer 10.1109/TBCAS.2009.2034364

AUTHORIZED LICENSED USE LIMITED TO: IEEE XPLORE. DOWNLOADED ON MAY 13,2010 AT 11:55:14 UTC FROM IEEE XPLORE. RESTRICTIONS APPLY.

I. I NTRODUCTION

R ECENTLY, various types of implantable medical devices,

IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS, VOL. 4, NO. 2, APRIL 2010

Fig. 2. Equivalent circuits of the energy transmission system. (a) Series reso-

nant circuit. (b) Parallel resonant circuit.

Fig. 1. Parts of an energy transmission transformer.

TABLE I

M ATERIAL AND D ESIGN P ARAMETERS OF THE T RANSFORMER

which are included, even greater power (20–900 mW) would be

necessary.

III. D ESIGN T HEORY FOR T RANSMITTING

AND R ECEIVING C OILS

To build a small, high-efﬁciency receiving coil, magnetic

material is desirable on the secondary-coil side. An accurate

equivalent circuit that can model the characteristics of a mag-

netic material is needed in cases where the shape, transmission

frequency, magnetic ﬂux density, number of turns, etc., are

changed.

vice, and core-loss resistances. The circuit equation of Fig. 2(a)

is expressed by (1), shown at the bottom of the page. Then, the

resonant condition is expressed by

A. Equivalent Circuit of Transformer With Core Loss

To transmit a large quantity of power to implantable med-

ical devices, the transformer requires a resonance capacitance

with the inductance of the transmitting and receiving coils. Gen-

erally, the transmitting side (primary side) is a series resonant

circuit, and the receiving side (the secondary side) is a series

resonant circuit or a parallel resonant circuit. Fig. 2(a) shows

the equivalent series resonant circuits for the ETT, and Fig. 2(b)

shows a parallel resonant circuit. In these circuits, , and ,

and , and , and , and , , ,

and and represent, respectively, the angular frequency,

input and output voltages, transmitting and receiving coil resis-

tances, transmitting and receiving coil inductances, resonant ca-

pacitances, equivalent series resistances of the resonant capaci-

tances, mutual inductance, load for the implantable medical de-

(2)

B. Elicitation Process of Core Loss

Core loss

is equal to the sum of hysteresis loss

and

eddy-current loss

(3)

and

are expressed as

(4)

(5)

(1)

AUTHORIZED LICENSED USE LIMITED TO: IEEE XPLORE. DOWNLOADED ON MAY 13,2010 AT 11:55:14 UTC FROM IEEE XPLORE. RESTRICTIONS APPLY.

SHIBA et al. : DESIGN AND DEVELOPMENT OF LOW-LOSS TRANSFORMER

TABLE II

E XPERIMENTAL C ONDITIONS AND M AXIMUM E FFICIENCIES OF T YPE I–VII

where is the transmission frequency and is the maximum

magnetic ﬂux density. It is known that ,

(there is sometimes a case where at a ferrite core),

, and (there is sometimes a case where at a

ferrite core) [32]. is proportional to the differential voltage

between the terminal voltage of the coil and coil resistance

voltage,

, and inversely proportional to [33]

(6)

Fig. 3. Measurement circuit for core-loss resistance.

Therefore,

is deﬁned as follows:

(7)

to when the voltage drop of the secondary coil resistance is

smaller than

where , , and

are constants determined by the magnetic

material.

is connected in parallel with

as in Fig. 2, by (7),

(11)

is expressed by

where

(there is

(8)

sometimes a case where at a ferrite core).

The material parameters (Table I) , , , , and ( , ,

, , and ) are proposed by the authors and determined by

a one-time measurement. Fig. 3 shows the measurement circuit

for the core-loss resistance. The current is applied to the mag-

netic coil (receiving coil with magnetic material or the trans-

mitting coil with magnetic material); then, the series resistance

and the reactance are measured by using equipment, such as a

power analyzer or a general-purpose oscilloscope. is calcu-

lated from the real part of the impedance by subtracting the coil

resistance. When measuring using Fig. 3, the supply voltage

in Fig. 3 has to be equal to the actual value of the terminal

voltage of the magnetic coil in Fig. 2. For the series resonant

circuit in Fig. 2(a), on the condition that and are con-

stant, the current ﬂowing out of the magnetic coil corresponds

to . Therefore, the supply current in Fig. 3 has to be

equal to .

Likewise, for the parallel resonant circuit in Fig. 2(b), on the

condition that

Using

is expressed by

(9)

C. Core-Loss Resistance

In a series resonant circuit, if it is assumed that and are

constant, (8) is transformed as follows because the differential

voltage between the terminal voltage of the receiving coil and

voltage drop of the coil resistance

is proportional to

(10)

where , . In a parallel resonant

circuit, (8) is transformed as follows because

is nearly equal

and

are constant, the terminal voltage of

AUTHORIZED LICENSED USE LIMITED TO: IEEE XPLORE. DOWNLOADED ON MAY 13,2010 AT 11:55:14 UTC FROM IEEE XPLORE. RESTRICTIONS APPLY.

IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS, VOL. 4, NO. 2, APRIL 2010

the magnetic coil corresponds to . Therefore, the supply

voltage in Fig. 3 has to be equal to .

, , , , and in (10), or , , , , and in (11)

are measured as functions of the transmission frequency.

D. Elicitation Processes for Copper Loss and Capacitor Loss

The transmitting coil’s copper loss

(the receiving coil’s

copper loss

) is deﬁned as the product of

and

are deﬁned as follows because of the skin effect [34]:

(12)

Fig. 4. Arrangement of the two coils.

(13)

where and are the number of wire turns and the radii, re-

spectively. , , and or , , and are the material

parameters (Table I) determined by the wire type. These mate-

rial parameters for a 1-m length of litz-wire are measured with

equipment, such as an LCR meter, as a function of the transmis-

sion frequency.

The losses in the primary-side and secondary-side resonant

capacitors and , respectively, are deﬁned as the prod-

ucts of and , and and . The equivalent series re-

sistances for the resonant capacitors and are the ma-

terial parameters (Table I) determined by the material of the ca-

pacitor. and are measured as functions of the trans-

mission frequency using equipment, such as a general-purpose

LCR meter.

and is the loss in the secondary-side resonance capaci-

tance. In this paper, the ETT is optimally designed by maxi-

mizing the efﬁciency .

In the design process, ﬁrst, the material parameters shown in

Table I are measured or speciﬁed, and then, the transmission

efﬁciency is optimized in terms of the design parameters deﬁned

by the user for the given constraints by using (14). Table I shows

a list of candidate design parameters from which the user can

choose the necessary ones. The ETT can be designed on the

basis of the proposed design theory.

IV. D ESIGN E XAMPLE OF THE T RANSMITTING

AND R ECEIVING C OILS

As an example, transmitting and receiving coils that are

wound with the same circumference were designed for the

case where , , and change. In this calculation, the output

voltage was set at 3 V, and and in Fig. 4.

The transmitting coils of types I–V had a diameter of 20

cm and a length of 3 cm [Figs. 1 and 5(a)]. The receiving

coils of types I–V had a diameter of 5 mm and a length

of 2 cm. In addition, the length of the coils

E. Self Inductance and Mutual Inductance

The analytical solution of the self-inductance is used in (1)

and (2). The self-inductance of the transmitting and receiving

coils can be calculated by a simpliﬁed numerical analysis [35].

When the calculation is very complex, depending on the layout

of the magnetic material, the self-inductance is calculated based

on a measured result. The self-inductance of a coil with mag-

netic material is deﬁned as multiplication between the ap-

parent relative magnetic permeability and the self-inductance

of the coil without the magnetic material . The mutual induc-

tance between the transmitting and receiving coils can be

calculated by using Neumann’s law when considering the shape

or relative positions of the coils [35]. In order to calculate the

mutual inductance with magnetic material

and

change due

and

. The number of wire turns per unit length

and

, the apparent rel-

were set at 75 turns/cm and 7.5 turns/cm, respectively.

The transmitting coils of types VI–VII comprised series-con-

nected double solenoidal coils [Fig. 1 and Fig. 5(b)]. Each

solenoidal coil had a diameter of 30–40 cm (ellipse) and a

length of 6 cm. The distance between the two coils was 20

cm. The receiving coils of types VI–VII had a diameter of

10 mm and a length of 2 cm. and for types VI–VII

were set at 4 turns/cm and 47.5 turns/cm, respectively.

ative magnetic permeability

has to be derived by using the

onetime measured

divided by the calculated

(from Neu-

A. Measurement of the Material Parameters

mann’s law) or the onetime measured

First, the material parameters in Table I were measured. The

prototype receiving coils of types I–V (Fig. 5) had 150 turns

of litz-wire (UEW wire, bundles of 15 wires of 0.03 mm)

wound around Material A (EPCOS AG, K1) or Material B (Hi-

tachi Metals Ltd., FINEMET). Material A is a cylindrical ferrite

core. Material B is a ﬁvefold-thinner amorphous magnetic ma-

terial sheet (0.3-mm (thickness) 5) wound around a wooden

stick with a diameter of 2 mm. The prototype receiving coils of

types VI–VII (Fig. 5) had 95 turns of single-wire (UEW wire,

0.2 mm) wound around Material C (a thin amorphous mag-

netic sheet, 0.03 mm in thickness) wound around a wooden stick

F. Energy Transmission Efﬁciency

The energy transmission efﬁciency is expressed by

(14)

where

is the output power, which is deﬁned as

divided by

is the core loss of the receiving coil,

is the trans-

mitting coil’s copper loss,

is the receiving coil’s copper

loss,

is the loss in the primary-side resonance capacitance,

AUTHORIZED LICENSED USE LIMITED TO: IEEE XPLORE. DOWNLOADED ON MAY 13,2010 AT 11:55:14 UTC FROM IEEE XPLORE. RESTRICTIONS APPLY.

SHIBA et al. : DESIGN AND DEVELOPMENT OF LOW-LOSS TRANSFORMER

coil and single wires of 0.2 mm for the receiving coil) was

measured by using an LCR meter (HIOKI E. E. Corp., 3532-50).

The material parameters of the winding wire were

, and

1.7 (types I–V) and

Fig. 5. Transmitting coil and receiving coil of (a) types I–V and (b) types

VI–VII.

, and (types VI–VII).

The material parameters of the equivalent series resistances

of resonant capacitances and of types I–V were set

at 0.25 from the measurement results for the polypropylene

capacitor used (Evox Rifa, PHE450) at a transmission frequency

of 100 kHz. and for types VI–VII were, respec-

tively, set at 0.30 and 0.25 from the measurement results

for the polypropylene capacitor used ( : Evox Rifa, PHE450;

: NISSEI Electric, MPE1600J) at a transmission frequency

of 100 kHz.

The material parameter of the apparent relative magnetic per-

meability of types I–VII was 1.0 because the transmitting

coil had an air core. for Material A was found to be 18.0

from the measurement results H) obtained by

using an LCR meter (HIOKI E. E. Corp., 3532-50), along with

the calculated result H). was calculated by using

the Nagaoka coefﬁcient [35] due to the solenoidal coil.

for Materials B and C was 20.07 and 6.19, respectively. These

values were calculated by one-time measurements.

for Material A in types I–V was calculated by using the

prototype transmitting and receiving coils, as described in Sec-

tion III-E. , , and were set at 10 cm, 0 cm, and 0 rad,

respectively. From the results— H and 98.0

nH— was calculated to be 18.1. for Materials B and C

in types I–VII was calculated by using a one-time measurement

and

. The measured value of

for Materials B and

C is 20.2 and 10.7, respectively.

Fig. 6. Measured results versus approximation of the core-loss resistance.

B. Calculations of the Transmission Efﬁciency, Core Loss, and

Copper Loss of Receiving Coil

The transmission efﬁciency of the type I (Material A) system

was calculated by using the material parameters in Section IV-A

for a series resonant circuit, 3V, ,

100 mW, and 10 cm. The capsule endoscope must be

of a swallowable size. In addition, the terminal voltage of the

resonant capacitor must have a realistic value that is less than

the maximum voltage [37], [38]. Therefore, the limit is set as

follows:

with a diameter of 2.5 mm. A wooden stick is used only to fasten

the thin amorphous magnetic sheet.

The core-loss resistances for the receiving coils were mea-

sured by using a power analyzer (Yokogawa Electric Corp.,

PZ4000) at transmission frequencies of 10 kHz to 1 MHz. Fig. 6

shows the measured values for Material A and an approxi-

mation ﬁtted to (10) and (11) by the least-squares method (using

Mathematica 6.0). From the measured results, the material pa-

rameters for the core-loss resistance of Material A were

2.3, , , 1.0, 2.0,

0.1, , , 0.6, and

0.7. The material parameters for the core-loss resistances of Ma-

terials B and C are as follows:

(15)

(16)

5.3,

The above equation sets the design conditions for the example

capsule endoscope.

Fig. 7 shows an example of the calculated results when

and change. The maximum efﬁciency was 24.5% at

20.0 mm and 477.1 kHz. The equation for the efﬁciency

is shown in Appendix B. Fig. 8 shows a partial differential of

the efﬁciency at 20 mm and 10 mm (example of a

downsized type I system). For

1.0,

2.0,

1.7,

1.0, and

1.0 (Material B);

and

6.9,

2.0,

1.0,

4.4

1.8,

8.6

0.6,

0.7 (Material C).

Then, the resistance of the litz-wire (UEW wire; types I–V,

bundles of 360 wires of 0.05 mm for the transmitting coil and

bundles of 15 wires of 0.03 mm for the receiving coil; types

VI–VII, bundles of 798 wires of 0.05 mm for the transmitting

10 mm,

is 0 at

627.5 kHz. For

20 mm,

is 0 at

477.1 kHz. The

transmission frequency at which

becomes 0 is the same

AUTHORIZED LICENSED USE LIMITED TO: IEEE XPLORE. DOWNLOADED ON MAY 13,2010 AT 11:55:14 UTC FROM IEEE XPLORE. RESTRICTIONS APPLY.

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: