EXPERIMENTAL INVESTIGATION ON NONLINEAR

CHARACTERISTICS OF A HIGH-TEMPERATURE

SUPERCONDUCTING DUAL-BAND BANDPASS FILTER

Shoichi Narahashi

, Kei Satoh

, Yasunori Suzuki

, and Yuta Takagi

Research Laboratories, NTT DOCOMO, INC., 3-6 Hikari-no-oka, Yokosuka, Kanagawa, Japan

{narahashi, satokei, suzukiyasu, takagiyu}@nttdocomo.co.jp

Keywords: Dual-band, Bandpass Filter, Superconductor, Intermodulation Distortion, Third-order Intercept Point

Abstract: This paper presents an experimental investigation on the nonlinear characteristics of a high-temperature

superconducting (HTS) dual-band bandpass filter (DBPF) that has two passbands and comprises a coplanar-

waveguide quarter-wavelength resonator with odd- and even-mode resonant frequencies. An HTS filter is a

potential candidate for an RF multi-band bandpass filter (MBPF) that has multiple passbands, which will

support broadband transmission in future mobile communication systems. It is, however, indispensable to

elucidate the nonlinear characteristics of the HTS-MBPF because the nonlinearity of HTS materials causes

intermodulation distortion (IMD). This paper presents the IMD characteristics of a 2.0-/3.5-GHz band HTS-

DBPF using a YBCO thin film deposited on a MgO substrate when two tones with a 30-kHz separation are

individually and simultaneously input to each passband. Experimental investigations confirm that the HTS-

DBPF generates different third-order IMD characteristics depending on how the two tones are input to the

HTS-DBPF. This paper also presents a method for estimating the third-order IMD characteristics based on

the measurement results.

1 INTRODUCTION

In recent years, Carrier Aggregation (CA) has

attracted attention as a technology for improving

transmission speeds in future mobile communication

systems (Miki, Iwamura, Kishiyama, Anil & Ishii

2010). CA achieves broadband transmission using

several fundamental frequency blocks (referred to as

component carriers (CCs)) aggregately and

simultaneously. An RF filter with multiple

passbands, i.e., a multiband bandpass filter (MBPF),

is a basic circuit that is expected to achieve

broadband transmission such as CA using CCs in

different frequency bands.

As a key device for supporting CA, the MBPF is

required to have high performance such as sharp-

skirt characteristics at the band-edges and a high

degree of attenuation in the stopbands with low

insertion losses in the passbands, which is

remarkable when a wireless system with a high

transmitter power is assigned to the frequency band

between the two passbands of the MBPF. A high-

temperature superconducting (HTS) filter is a

potential candidate that satisfies these requirements

(Abu-Hudrouss, Jayyousi & Lancaster, 2008).

It is indispensable to investigate the nonlinear

characteristics of the HTS filter because

intermodulation distortion (IMD) is caused by the

nonlinear characteristics of the HTS materials when

applying the HTS filter to mobile communication

systems. To measure the nonlinear characteristics,

there are two common techniques (Wilker, Shen,

Pang, Holstein & Face, 1995): one is the IMD

measurement method using a signal with two tones

and the other is the single tone harmonic generation

measurement method. It is desirable to utilize the

IMD measurement method using a signal with two

closely-spaced tones in evaluating the nonlinear

characteristics of the HTS filter since it might

attenuate the power level of the harmonic

components of the single tone.

The authors have shown the IMD characteristics

of the 2.0-/3.5-GHz band HTS-DBPF with a YBCO

thin film deposited on a MgO substrate when two

signals are individually and simultaneously input to

the center of the lower passband and the center of

Narahashi S., Satoh K., Suzuki Y. and Takagi Y.

EXPERIMENTAL INVESTIGATION ON NONLINEAR CHARACTERISTICS OF A HIGH-TEMPERATURE SUPERCONDUCTING DUAL-BAND BANDPASS FILTER.

DOI: 10.5220/0004785200540057

In Proceedings of the Second International Conference on Telecommunications and Remote Sensing (ICTRS 2013), pages 54-57

ISBN: 978-989-8565-57-0

higher passband as initial research on evaluating the

IMD characteristics of an MBPF (Satoh, Takagi,

Narahashi & Nojima, 2011; Satoh, Takagi &

Narahashi, 2012a; Satoh, Takagi, Narahashi &

Nojima, 2012b).

This paper presents an experimental

investigation on the IMD of the proposed 2.0-/3.5-

GHz band HTS-DBPF. Measurement results show

that the third-order IMD (IMD3) characteristics are

generated differently depending on how the two-

tone signal with a 30-kHz separation is input to the

HTS-DBPF. The two ways that the signal is input to

the HTS-DBPF are that the two-tone signals are

individually input to either the 2.0-GHz band or 3.5-

GHz band, or the two tone signals are

simultaneously input to both the 2.0-GHz band and

3.5-GHz band. This paper also presents a method for

estimating the IMD3 characteristics based on the

third-order polynomial approximation of the input-

output characteristics of the HTS-DBPF.

2 RESONATOR FOR HTS-DBPF

Figure 1 shows the resonator for the HTS-DBPF that

has odd- and even-mode resonant frequencies (Satoh

et al., 2011). These two resonant modes have

different current flows on the short stub.

This suggests that different current flows allow

not only independent configuration of the coupling

coefficients for the lower and higher passbands but

also suppression of the IMD when a signal is

simultaneously input to each passband.

3 IMD MEASUREMENTS

3.1 Measurement System

Figure 2 shows a block diagram of a two-tone IMD

measurement system with a fundamental signal

cancellation circuit for a single-band bandpass filter

(Blount, Olson, Tshudy, Foote & Trantanella, 2004;

Futatsumori, Furuno, Hikage, Nojima, Akasegawa,

Nakanishi & Yamanaka, 2009). This measurement

system enables measurement of the IMD

characteristics without the nonlinear effect of the

spectrum analyzer because two fundamental signals

can be suppressed by more than 60 dB. By using the

fundamental signal cancellation circuit, the IMD

characteristics could be measured in detail at low

power levels.

Figure 3 shows a new block diagram of two

kinds of two-tone IMD measurement systems with a

fundamental signal cancellation circuit for each

High

Low

Odd

Even

(a) Odd-mode frequency (b) Even-mode frequency

Figure 1: Resonator structure and current density

distribution.

DUT

Power

amplifier

Isolator

Signal

generator

Cryostat

Variable

attenuator

Phase

shifter

Variable

attenuator

CouplerIsolatorIsolator

Coupler

Spectrum

analyzer

Variable

attenuator

Isolator

Filter

Figure 2: Block diagram for measuring IMD

characteristics of single-band bandpass filter.

DUT

Power

amplifier

Isolator

Signal

generator

Cryostat

Variable

attenuator

Phase

shifter

Variable

attenuator

CouplerIsolatorIsolator

Coupler

Spectrum

analyzer

Vari able

attenuator

Isolator

Filter

Vari able

attenuator

Coupler

Figure 3: Block diagram for measuring IMD

characteristics of dual-band bandpass filter.

passband in order to evaluate nonlinear

characteristics when the two kinds of two-tone

fundamental signals are input to each passband of

the HTS-DBPF.

Figure 4 is a photograph of the configuration for

the IMD measurement. This measurement system

also enables measurement of the IMD characteristics

with a wide dynamic range.

3.2 Measured Results

Figure 5 shows the measured IMD characteristics of

the HTS-DBPF together with the ideal IMD3

characteristics. Here, the line with the slope of the

third order fitted to the measured results at each

passband represent the ideal IMD3 characteristics. In

Figure 5(a), blue and red squares represent the third-

order IMD (IMD3) characteristics for the 2.0-GHz

band and 3.5-GHz band of the HTS-DBPF when a

Experimental Investigation on Nonlinear Haracteristics of a High-Temperature Superconducting Dual-Band Bandpass Filter

two-tone fundamental signal with the separation of

30-kHz is input to the 2.0-GHz and 3.5-GHz

passbands, respectively. The input and output power

per tone are represented on the horizontal and

Figure 4: Experimental configuration for IMD

measurement.

vertical axes, respectively. The measured IMD

characteristics nearly coincide with the ideal

distortion. The third-order input intercept point

(IIP3) for the 3.5-GHz passband is +50 dBm at 60

K, while the IIP3 for the 2.0-GHz passband is +43

dBm. If the same IMD is assumed for each

passband, the HTS-DBPF can handle an

approximately 2.9-times higher input signal power

in the 3.5-GHz passband than that in the 2.0-GHz

passband at 60 K.

Figure 5(b) shows the measured IMD3

characteristics for both passbands of the HTS-DPBF

when two kinds of two-tone fundamental signals

(1957.795- and 1957.825-MHz signals for the 2.0-

GHz passband, and 3463.515- and 3463.545-MHz

signals for the 3.5-GHz passband) are

simultaneously input to each passband. The total

input power to the HTS-DBPF in Figure 5(b) is 3 dB

higher than that in Figure 5(a). The IIP3 for the 3.5-

GHz passband is higher than that for the 2.0-GHz

passband. Based on Figure 5, the HTS-DBPF has

almost the same IIP3 of approximately +43 dBm in

the 2.0-GHz passband. If the same IMD is assumed

for each passband, the power level of the input

signal is defined by the IIP3 for the 2.0-GHz

passband.

3.3 IMD3 Estimation

As shown in Figure 5, the IMD3 characteristics are

generated differently depending on how the two-

tone signal is input to the HTS-DBPF. This section

estimates the IMD3 characteristics shown in Figure

5(b) based on the polynomial approximation of the

input-output characteristics of the HTS-DBPF.

If a two-tone signal with equal separation is

input to each passband of an MBPF that has n

passbands, the following two combinations represent

IMD3 in 3.5GHz band

when fundamental signals

are input to 3.5-GHz band

IMD3 in 2.0-GHz band

when fundamental signals

are input to 2.0-GHz band

Fundamental components

-100

-80

-60

-40

-20

0 5 10 15 20 25 30 35 40 45 50 55

(

)

Input Power (dBm)

IIP3 (2.0-GHz Band) = 43 dBm

IIP3 (3.5-GHz Band) = 50 dBm

Ideal IMD3

(a) Each passband

IMD3 in 2.0-GHz band

IMD3 in 3.5-GHz band

when fundamental signals

are input to 2.0-and 3.5-

GHz bands

Fund amental components

-100

-80

-60

-40

-20

0 5 10 15 20 25 30 35 40 45 50 55

(

)

Input Power (dBm)

IIP3 (2.0-GHz band) = 43 dBm

IIP3 (3.5-GHz band) = 44 dBm

Ideal IMD3

(b) Both passbands

Figure 5: Measured IMD of HTS-DBPF using 30-kHz

separated signals at 60 K.

the frequency components of the IMD3

characteristics generated in the vicinity of the two-

tone signal at each passband.

()

12 21

112 2 21

(i) 2 , 2

(ii) ,

,1,,,

ii i i

ijj i j j

ij n i j

ωωω

ωω ωωω

−

+− +−

=≠"

(1)

where

and

(

) represent the angular

frequencies of the first and second tones at the i-th

passband, respectively. Here, the input signal, x

(t),

to the i-th passband is assumed to be defined as

(

)

(

)()

1112 22

cos cos

iiiii ii

xt a t a t

ωφ

=++ +

(2)

where a

and a

, and

and

represent the

amplitudes and phases of the two-tone signal,

respectively. Assuming that the input-output

characteristics of the HTS-MBPF can be expressed

by the third-order polynomial approximation, the

output signal, y

(t), of the i-th passband of the MBPF

is given as

(

)

(

)

(

)

13iiiii

yt bxt bxt=+

(3)

Second International Conference on Telecommunications and Remote Sensing

where parameters b

and b

are defined as the

coefficients of the first-order term and the third-

order term of x

(t) derived from the measured IMD3

characteristics when x

(t) is input to the i-th passband,

respectively. Then, the total output signal, y(t), of

the HTS-MBPF can be approximated as

-100

-80

-60

-40

-20

0 5 10 15 20 25 30 35 40 45 50 55

(

)

Input Power (dBm)

Calculated IMD3 (2 GHz)

Calculated IMD3 (3.5 GHz)

Measured IMD3 (2 GHz)

Measured IMD3 (3.5 GHz)

Measured fundamental components

Figure 6: Measured and calculated IMD3 for 2-GHz band

and 3.5-GHz band.

() () ()

1/3

ii i i

yt bx t b x t

⎛⎞

⎜⎟

⎝⎠

∑∑

(4)

Figure 6 shows the measured and calculated

IMD3 characteristics for each passband as well as

the fundamental components of the HTS-DPBF

using Equation (4) and parameters b

and b

(i = 1,

2). Comparing Figure 5(a) to Figure 5(b) leads to the

fact that the IMD3 characteristics are almost the

same in the 2-GHz band whereas they are 15.1 dB

higher in the 3.5-GHz band at the input power level

of 5 dBm. This difference is calculated as 16.6 dB

using Eq. (4), which indicates good agreement

between the measured and calculated results. The

increase in the IMD3 characteristics in the 3.5-GHz

band is considered to be due to the following reason.

The IMD3 components generated by the

combination of angular frequency

(which depends on parameter b

) appear at the

angular frequency of 2

since parameter b

4.9-times (13.7 dB) greater than parameter b

4 CONCLUSIONS

This paper presented an experimental investigation

on the IMD characteristics of an HTS-DBPF. A new

two-tone IMD measurement system enables the

evaluation of the IMD characteristics of the HTS-

DBPF when the HTS-DBPF simultaneously deals

with two kinds of two-tone fundamental signals.

This paper also presented a method for estimating

the IMD3 characteristics based on the third-order

polynomial approximation of the input-output

characteristics of the HTS-DBPF.

There still remain technical issues such as

clarifying the effective range of the HTS-DBPF

using widely-separated signals or a modulated signal,

investigating the nonlinearity of the HTS-DBPF

when interference signals are input to its passband

and when signals are input to its passband edges or

stopbands, and confirming whether or not the

proposed IMD3 estimation method is available when

the two-tone signal employs a frequency separation

other than 30 kHz.

REFERENCES

Miki, N., Iwamura, M., Kishiyama, Y., Anil, U., and Ishii,

H. 2010. CA for Bandwidth Extension in LTE-

Advanced. NTT DOCOMO Technical Journal, 12(2),

10-19.

Abu-Hudrouss, A. M., Jayyousi A. B., and Lancaster M. J.

2008. Triple-Band HTS Filter Using Dual Spiral

Resonators with Capacitive-Loading. IEEE Trans.

Appl. Supercond., 18(3), 1728-32.

Wilker, C., Shen, Z., Pang, P., Holstein, W. L., and Face,

D. W. 1995. Nonlinear Effects in High Temperature

Superconductors: 3

Order Intercept from Harmonic

Generation. IEEE Trans. Appl. Supercond., 5(2),

1665-70.

Satoh, K., Takagi, Y., Narahashi, S., and Nojima, T. 2011.

Intermodulation Distortion Characteristics of High-

Temperature Superconducting Dual-Band Bandpass

Filter. Proc. of CJMW.

Satoh, K., Takagi, Y., and Narahashi, S. 2012a.

Experimental Investigation of the Intermodulation

Distortion Characteristics of High-Temperature

Superconducting Dual-Band Bandpass Filter. Proc. of

TJMW, FR4-6.

Satoh, K., Takagi, Y., Narahashi, S., and Nojima, T.

2012b. Experimental Investigation of the Relationship

between Operating Temperature and Power Handling

Capability of High-Temperature Superconducting

Dual-Band Bandpass Filter. Physics Procedia, 36, 13-

18.

Blount, P., Olson, M., Tshudy, R., Foote, D., and

Trantanella, C. 2004. An Automated Phase

Cancellation Method for Measuring Ultra-High Third

Order Intercept Points. IEEE MTT-S Int. Microwave

Symp. Dig., 1727-30.

Futatsumori, S., Furuno, M., Hikage, T., Nojima, T.,

Akasegawa, A., Nakanishi, T., and Yamanaka, K.

2009. Precise Measurement of IMD Behavior in 5-

GHz HTS Resonators and Evaluation of Nonlinear

Microwave Characteristics. IEEE Trans. Appl.

Supercond., 19(3), 3595-99.

Experimental Investigation on Nonlinear Haracteristics of a High-Temperature Superconducting Dual-Band Bandpass Filter