MEP 2006, 7-11 November 2006, Guanajuato, Guanajuato, México.
COMPARATIVE ANALYSIS OF SYNCHRONOUS “HICCUP” AND NONSYNCHRONOUS SAW RESONATORS ON QUARTZ Yuri V. Gulyaev1, Valeri I. Grigorievski1, Victor P. Plessky2 1
Institute of Radio-Engineering and Electronics of Russian Academy of Sciences (IRE RAS) 103907, Mohovaya 11, Moscow, Russia Phone: 7-095-5269048, Fax: 7-095-7029572, e-mail:
[email protected] 2 GVR Trade SA CH2022, Bevaix, Switzerland Phone/Fax: 41-32-8463039, e-mail:
[email protected]
Abstract-Designed and measured characteristics of “hiccup” and non-synchronous surface acoustic wave resonators on quartz are presented and analyzed. In “hiccup” configuration basic periodicity of interdigital transducer and reflectors is the same except for additional distance of a quarter of wavelength in between neighboring electrodes at the center of transducer. The configuration with different periods of transducer and reflectors is called as non-synchronous. The hiccup SAW resonators show significant degradation of quality factor that can be attributed to radiation of bulk waves. Measured characteristics of non-synchronous resonators agree well with calculations. Keywords: surface acoustic wave, resonator, quality factor.
SAW resonators find a large variety of applications in modern electronic systems for frequency stabilization and selection. The simplest configuration of a one-port SAWresonator consists of an interdigital transducer (IDT) and two distributed reflectors at both sides of the IDT. If all electrodes are placed periodically, no special gaps being present, such resonator configuration is called as synchronous. The synchronous resonator structure is widely used for surface transverse wave (STW) resonators [1], [2]. Due to the trapping effect of periodic electrode grating on STW propagation [3], [4] and in the absence of breaks in periodicity there is no scattering of STW into bulk acoustic waves. As a result, STWresonators show quality factors as high as 8000 at frequencies up to 1000MHz [2]. However, the resonance frequency of STW resonators on quartz is very sensitive to variations of aluminum electrode thickness and, especially, the metallization ratio. Consequently, high level technology is used in production of STW resonators. The SAW resonators with Rayleigh SAW should be more stable to technological variations, because Rayleigh wave in any material is a well surface localized mode, and some cavities in the resonator structure or breaks in periodicity should influence less significantly. On the other hand, in the synchronous resonator configuration the resonance takes place near the low frequency end of reflector stopband, where the reflectivity of finite gratings is decreased, and rather long reflectors must be used to achieve high resonant quality-factor. For this reason SAW resonators using Rayleigh SAW often include resonant cavities or variations in periodicity. The so called “hiccup” structure has an additional distance of a quarter of wavelength in between neighboring electrodes at the center of IDT also playing the role of reflector [5]. In such a way the resonant cavity is formed, and the resonant frequency is shifted to the center of reflecting stopband, which makes the resonance frequency independent on the reflectivity of fingers. Such resonators are less sensitive to technological variations. There is another way to shift resonant frequency to the center of reflector stopband, namely to slightly increase the period of reflecting structures and shift the stopband to the maximum of SAW excitation 1-4244-0628-5/06/$20.00 ©2006 IEEE
83
MEP 2006, 7-11 November 2006, Guanajuato, Guanajuato, México.
frequency. The latter configuration is called as non-synchronous, because the periodicity is not the same throughout the structure. Resonator characteristics were calculated using P-matrix formalism [6]. The P-matrix relates the electric current and the amplitudes of outgoing acoustic waves to the applied source voltage and the amplitudes of incoming waves. The resonator structure was subdivided into basic building blocks, namely, interdigital transducer, reflector, and the gap of a quarter of wavelength between two sections of IDT in the case of hiccup structure. The building blocks were described by their respective P-matrices, which were then cascaded to obtain the P-matrix of the whole resonator structure. The element P33 of this matrix represents the resonator admittance Y=P33. The elements of P-matrices for all building blocks were expressed in closed form on the basis of coupling of modes (COM) theory [6]. To take into account intrinsic attenuation of SAW the complex valued propagation wavenumber was used. In the basic block of the gap in IDT of hiccup resonators an additional attenuation parameter was introduced to describe scattering of SAW energy into bulk waves. The resonators consisted of IDT with number electrodes 165 and two reflectors with 125 shorted electrodes. In non-synchronous resonators the difference in periods of reflectors and IDT of 0.72% was implemented to shift resonant frequency to the stopband of reflectors. The resonators were fabricated on 34O-YX-quartz. Fig.1 and Fig.2 show measured and calculated frequency responses for hiccup and nonsynchronous resonators respectively connected in series in 50-Ohm measurement system. The center frequency is close to 250 MHz.
Insertion loss, dB
0
Calculated Calculated Measured
-5 -10 -15 -20 -25 0.9975
1.0000
1.0025
Relative frequency
Fig.1-Frequency response of hiccup resonator
84
MEP 2006, 7-11 November 2006, Guanajuato, Guanajuato, México.
Insertion loss, dB
0
Calculated Measured
-5 -10 -15 -20 -25 0.9975
1.0000
1.0025
Realtive frequency
Fig.2-Frequency response of non-synchronous resonator
It is seen in Fig.1 that measured (red curve) and calculated data (green curve) differ considerably. The measured unloaded resonant quality factor was equal to approximately 5700, while calculated one was equal to 12000. If an additional attenuation of 0.05dB in the hiccup distance was inserted, then there was more close agreement between calculated (shown in blue color) and measured responses. In the case of non-synchronous resonator (Fig. 2) near the resonant frequency the calculated and measured characteristics agree well. The measured Q-factor at resonance also coincides well with predicted value of 14000. The transverse mode responses to the right from resonant frequency are more pronounced as compared to the hiccup resonator response, because there no additional attenuation in the IDT. Finally, Fig.3 shows frequency response of the non-synchronous resonator with the IDT that was apodized as cosine function to suppress transverse mode responses. Practically there is no difference between calculated and measured responses. The measured resonant Q-factor of this resonator was equal to 14000. 0
Calculated Measured
Insertion loss, dB
-5 -10 -15 -20 -25 -30 0.9975
1.0000
1.0025
Relative frequency
Fig.3-Frequency response of non-synchronous resonator with apodized IDT
As a conclusion, it has been shown that in quartz SAW-resonators using Rayleigh wave an appreciable break in periodicity of electrodes, as in “hiccup” structure, can cause the value of Q-factor to decrease. A SAW resonator with slightly different periods of reflectors and IDT
85
MEP 2006, 7-11 November 2006, Guanajuato, Guanajuato, México.
works well. In such non-synchronous SAW resonator the difference of periods is determined by condition to shift resonant frequency to the center of reflector stopband.
REFERENCES 1. I. D. Avramov, “1 GHz low loss coupled resonator filter using surface skimming bulk waves and Bleustein-Gulyaev waves,” Electron. Lett., vol. 5, pp. 414-415, 1991. 2. I. D. Avramov, F. L. Walls, T. E. Parker,and G. K. Montress, “Extremely low thermal noise floor, high power oscillators using surface transverse wave devices,” IEEE Trans. on UFFC, vol. 43, pp. 20-29, 1996. 3. B. A. Auld, J. J. Gagnepain, and M. Tan, “Horizontal shear waves on corrugated surfaces,” Electron. Left., vol. 12, pp. 650-651, 1976. 4. Yu. V. Gulyaev and V. P. Plessky, “Slow Acoustic Surface Waves in Solids,” Sov. Tech. Phys. Left., vo1. 3, pp. 87-88, 1977. 5. P. V. Wright, "A review of SAW resonator filter technology", in Proc. 1992 IEEE Ultrason. Symp., pp. 29-38. 6. V. P. Plessky and J. Koskela, “Coupling-of-modes analysis of SAW devices,” Int. Journal of High Speed Electronics and Systems, vol.10, pp.1-80, 2000.
86
