Journal of Science and Tecnology, Vol. 9 No. 3 (017) p. 55-59 Estimation of Dielectric Constant for Various Standard Materials using Microstrip Ring Resonator Pek Jin Low 1, Famiruddin Esa 1*, Kok Yeow You and Zulkifly Abbas 3 1 Department of Pysics and Cemistry, Faculty of Applied Sciences and Tecnology, Universiti Tun Hussein Onn Malaysia, Pago Educational Hub 84600 Pago Joor, Malaysia Faculty of Electrical Engineering, Universiti Teknologi Malaysia, 81310 Skudai Joor, Malaysia 3 Faculty of Science, Universiti Putra Malaysia, 43400 Serdang Selangor, Malaysia Received 30 September 017; accepted 30 November 017; available online 5 December 017 Abstract: Microstrip ring resonator (MRR) is known for dielectric constant determination and many studies used Teflon as a standard sample. However, tere are many oter materials available wic able to perform better or equivalence as te Teflon in calibrating certain dielectric constant measurement. Tis paper presents simulation of te MRR to investigate frequency sift of materials for dielectric constant estimation using te CST STUDIO SUITE 016 software. Te MRR was designed on RT/Duroid 5880 substrate ( =., tanδ = 0.0004) wit 50 Ω matcing impedance were microstrip widt, substrate tickness and ring mean radius were 4.893, 1.575 and 14 mm, respectively to resonate at.65340 GHz. Teflon, Polyimide, Isola, Arlon, Arlon and were alternately selected to be placed on top of te MRR as a standard sample to obtain te frequency sift. Te frequency sifts for te above materials were.5693,.46149,.44680,.53748,.5007 and.48608 GHz, correspondingly. Te differences in frequency sift were used in NetBeans IDE 8.1 algoritm of Java for dielectric constant calculation. Te results indicated tat Polyimide and Arlon ad te lowest and igest mean percentage of 0.83536 and 1.76505 %, respectively. Hence, Polyimide migt as well be te most suitable candidate as a standard sample in MRR tecnique for dielectric constant measurement. Keyword: Microstrip Ring Resonator; Frequency Sift; Dielectric Constant; CST Simulation. 1.0 Introduction Tere are many types of measurement and analysis for determining dielectric properties of materials. Te most common metods are open ended dielectric probe, closed waveguide system and coaxial cavity due to ig sensitivity and accuracy. However, microstrip ring resonator (MRR) also as been widely used in te measurement of te dielectric properties of te low-loss materials due to its simple structure, easy in fabrication, relatively low cost, low profile, and versatility. Information suc as relative dielectric constant, and loss tangent, tanδ of material under test (MUT) can be obtained using te MRR based on resonant frequency sift mecanism [1-3]. Te resonant frequency sift is a measure of frequency deviation between loaded and unloaded MRR. Loaded means wen MUT is placed on top of te MRR wile unloaded is witout MUT. Unknown and unitless parameter, D is required in te calculation of dielectric constant of MUT [4-5]. Tus, it is an essential to investigate te accuracy of te estimation since tis frequency sift metod requires a material to be as standard sample. Tus, in tis work te MRR was designed to resonate at frequency, f r of.6 GHz on RT/Duroid 5880 dielectric substrate wic as uniform of. wit tanδ = 0.0004 and simulated using CST STUDIO SUITE 016 software tested on several different types of dielectric materials tat act as standard sample. Te caracteristic impedance, Z o of te MRR feed line dimension was set at 50 Ω according to microstrip impedance equation so tat transmitted signal lost could be minimised were te substrate tickness was 1.575 mm. Terefore, te effectiveness of measuring te MRR resonant frequency as well as dielectric properties estimation of MUT could be increased [6]. Te minimum return loss (S 11) of microwave signal of te designed MRR could *Corresponding autor: famir@utm.edu.my 017 UTHM Publiser. All rigt reserved. penerbit.utm.edu.my/ojs/index.pp/jst
Journal of Science and Tecnology, Vol. 9 No. 3 (017) p. 55-59 be seen at its resonant frequency due to maximum transmission [7]. In tis paper, only was calculated from te simulation result. Te materials used were Teflon, Polyimide, Isola, Arlon, Arlon and..0 Metodology and Zo r 1 r 1 0.11 A 0.3 60 r 1 r 376.73 B Z o r.1 Designation Parameters of te MRR First ting needed for te simulation of te MRR is te dimension parameter of te MRR. For tis paper, te substrate of RT/Duroid 5880 was used wit te tickness, of 1.575 mm. Te resonator as to satisfy te resonance condition: R ng ; for n = 1,, 3,, N (1) were λ g and R are te guided wavelengt and te mean radius (in meter) of te ring circuit. Symbol n is te order of armonic of te resonance. Te λ g can be calculated as: c g () f were c and f are te speed of ligt in vacuum and te resonant frequency. Te effective permittivity, ε eff is expressed as: r 1 r 1 1 0.041 for 1 11 eff r 1 r 1 1 for 1 1 1 eff (3) By giving value of caracteristic impedance, Z o = 50 Ω and dielectric constant, =. of te substrate, te widt, of microstrip ring can be calculated as: Te lengt of te feed line, L can be calculated: g L 8 (5) Te inner radius, R 1 and te outer radius, R can be computed as: R1 R (6) R R (7) All of equations were written using Java algoritm and display calculator was set by NetBeans IDE 8.1. Te calculated MRR parameters were as in Table 1. Table 1: Te parameters of te designed MRR Parameters Value Z o 50 Ω. ε eff 1.87 4.8933 mm 1.575 mm R 14 mm R 1 11.55 mm R 16.45 mm L 11 mm.6 GHz f r. CST STUDIO SUITE 016 8exp A A B exp B1 ln 1 r 1 0.61 ln B 1 0.39 r r were te A and B in (4) are given as: for for (4) CST STUDIO SUITE 016 software was used for te designation and simulation of te MRR to find te dielectric constant of MUT using te resonant frequency sift metod as sown in Fig. 1. Te Planar Device Structure mode and Time Domain Solver were selected as tere is a copper strip on te dielectric substrate. Pysical properties of te surrounding space must be defined after designation of te MRR was completed. Materials suc as 56
Journal of Science and Tecnology, Vol. 9 No. 3 (017) p. 55-59 dielectric substrate and MUT were cosen from model library. Te radius and tickness of MUT were fixed at 17 and 10 mm, respectively. Port one and two were defined at appropriate orientation direction were te incident signal was fed by port one and received by port two. Te simulation frequency was set in te range of 3 GHz. Fig. Variation of resonant frequency for different dielectric materials and unloaded MRR (f unloaded =.65340 GHz). (a) (b) Fig. 1 Te completion model of te (a) unloaded MRR and (b) loaded MRR. 3.0 Result and Discussion Fig. sows S 11 wit a variation of resonant frequency for unloaded and loaded MRR. It was obvious tat te resonant frequency sift mecanism occurred to te loaded MRR gave signatures according to te different properties of MUT. All te loaded MRR sifted to frequencies tat were lower tan te unloaded MRR. Te value of eac resonant frequency of te MUT and unloaded MRR were recorded in Table. All of te recorded values were logged in five decimal points to acieve as iger accuracy as possible. Te difference in frequency cange between te unloaded and loaded MRR for eac MUT was also defined using Δf = f unloaded f loaded. Table Summary of te f loaded and Δf of various dielectric materials Material f loaded Δf (GHz) (GHz) Teflon.5693 0.08408 Polyimide.46149 0.19191 Isola.44680 0.0660 Arlon.53748 0.1159 Arlon.5007 0.13333.48608 0.1673 3.1 Estimation of Dielectric Constant According to te Eq. 8, calculation of for MUT requires frequency of te unloaded and loaded MRR to solve for te value of te parameter, D. Te dielectric materials were used as standard sample for dielectric constant estimation procedure. Teflon was first used as te standard sample and te sift in frequency was observed (Fig. ) and recorded (Table ). Te D Teflon parameter was ten calculated using Eq. 8 given tat te of Teflon was.1. Ten, tis D Teflon parameter was used to estimate te of Polyimide, Isola, Arlon, Arlon and wit te respective resonant frequency as listed in Table 3. All tese mentioned procedures were repeated for different standard sample as sown in Table 4 to Table 8. = 1 + D ( f unloaded f loaded ) (8) f unloaded 57
Journal of Science and Tecnology, Vol. 9 No. 3 (017) p. 55-59 Table 3 Calculated and for various MUT, D Teflon = 34.71384 Material Calculated Polyimide 3.5 3.5107 0.3069 Isola 3.75 3.7090 1.5600 Arlon.5.51656 0.6640 Arlon.7.74393 1.6704 3. 3.18901 0.34344 Table 4 Calculated and for various MUT, D Polyimide = 34.56568 Material Calculated Teflon.1.09531 0.333 Isola 3.75 3.69137 1.56347 Arlon.5.51008 0.4030 Arlon.7.73649 1.35148 3. 3.17967 0.63531 Table 5 Calculated and for various MUT, D Isola = 35.31873 Material Calculated Teflon.1.11917 0.9186 Polyimide 3.5 3.55447 1.5569 Arlon.5.5498 1.7190 Arlon.7.7743.7559 3. 3.715 0.84844 Table 7 Calculated and for various MUT, D Arlon = 33.83931 Material Calculated Teflon.1.079 1.3195 Polyimide 3.5 3.44746 1.50114 Isola 3.75 3.63481 3.07173 Arlon.5.47835 0.86600 3. 3.13386.06688 Table 8 Calculated and for various MUT, D = 34.8881 Material Calculated Teflon.1.1055 0.686 Polyimide 3.5 3.533 0.6669 Isola 3.75 3.71647 0.89413 Arlon.5.5417 0.96680 Arlon.7.7569 1.95148 Table 9 sows te total percentage of, mean percentage of and standard deviation for eac tested standard sample. It was found tat wen using Polyimide as te standard sample in te simulation, te total percentage of and mean percentage of were te lowest. However its standard deviation was te second smallest among all materials after Teflon wit a difference of 0.00904. Tis verified tat Polyimide could performed better as te standard sample among te tested samples as sown in te simulation. Table 6 Calculated and for various MUT, D Arlon = 34.33489 Material Calculated Teflon.1.08799 0.57190 Polyimide 3.5 3.48331 0.47686 Isola 3.75 3.67340.0467 Arlon.7.7490 0.9 3. 3.16511 1.09031 58
Journal of Science and Tecnology, Vol. 9 No. 3 (017) p. 55-59 Table 9 Total percentage of, mean percentage of and standard deviation for tested standard samples. Standard Sample Total Mean Standard deviation Teflon 4.19517 0.83903 0.581 Polyimide 4.17879 0.83536 0.59116 Isola 7.78938 1.55788 0.77018 Arlon 5.10396 1.0079 0.6373 Arlon 8.857 1.76505 0.84778 4.74156 0.94831 0.6430 4.0 Conclusion Among all materials used as standard sample, Polyimide could be considered as te best standard sample due to te lowest total percentage of and mean percentage of as well as as te most consistent. Based on tis simulation, accuracy of te estimation of dielectric constant is reliable wit te given pysical settings and boundaries condition. However, te accuracy of te dielectric constant in actual experimental may sligtly vary due to te imperfection of fabricating of te MRR. [4] R.Hopkins. C.Free. (008). Equivalent Circuit for te Microstrip Ring Resonator suitable for broadband materials caracterization. IET Microw. Antennas Propagation, Vol., No. 1, 111-137. [5] S.P.Dave, H.J. Jayanti, R.D. Jani. (015) Design and Simulation of Microstrip Ring Resonator is constructed from one quarterwave coupled line Coupler. Carusat University of India, 1-. [6] R. Hopkins (006) Te Microstrip Ring Resonator for Caracterizing Microwave materials. Advanced Tecnology Institute Scool of Electronics and Pysical Sciences University of Surrey, 4-5. [7] S.M. Seng. (016) Investigation of Microwave Absorption Properties of Epoxy/Iron Pospate Glass Composite Using aveguide Metod. Universiti Tun Hussein Onn Malaysia, 13-17. References [1] I. aldron. (006). Ring Resonator Metod for Dielectric Permittivity Measurement of Foams. Degree of Master of Science in Electrical and Computer Engineering. orcester Polytecnic Institute. [] K. Moit, V.R. Gupta and S.K. Rout. (014). Microwave Dielectric Properties of Ni 0. Cu x Zn 0.8 x Fe O 4 4 for Application in Antenna. Progress in Electromagnetic Researc B, Vol.57, 157-175. [3] J.C.A. Santosa, M.H.C. Diasa, A.P. Aguiarb, I.B. Jrb, L.E.P. Borgesb. (009) Using te Coaxial Probe Metod for Permittivity Measurements of Liquids at Hig Temperatures. Journal of Microwaves, Optoelectronics and Electromagnetic Applications, Vol. 8, No. 1, 14-19. 59