Sesors & Trasducers 215 by IFSA Publishig, S. L. http://www.sesorsportal.com Uiversal Sesors ad Trasducers Iterface for Mobile Devices: Metrological Characteristics * Sergey Y. YURISH ad Javier CAÑETE Excelera, S. L., Parc UPC-PMT, Edificio RDIT-K2M, c/ Esteve Terradas, 1, 886, Castelldefels, Barceloa, Spai 1 Tel.: +34 93 4137941 E-mail: sergey.yurish@excelera.io, javier.caete@excelera.io Received: 31 March 215 /Accepted: 3 April 215 /Published: 29 May 215 Abstract: A experimetal ivestigatio of metrological characteristics of desiged Uiversal Sesors ad Trasducers Iterface (USTI-MOB) itegrated circuit is described i the article. The USTI-MOB is based o the four ovel pateted methods for frequecy (period), its ratio, phase shift ad duty-cycle measuremets, ad has a o-redudat coversio time. Experimets have cofirmed the high metrological performace at low power cosumptio (.35 ma curret cosumptio at V cc = 1.8 V). So, the relative error of frequecy ad period measuremets is costat i the whole specified measurig rage of frequecies (from.25 Hz to 1.95 MHz) ad with 95 % cofidece does ot exceed ±.89 % i the worst case. Metrological characteristics of measuremets of other frequecy-time parameters of sigals make the USTI-MOB very suitable for various sesor systems desig for mobile devices, IoT, etc. Copyright 215 IFSA Publishig, S. L. Keywords: Uiversal Sesors ad Trasducers Iterface, USTI-MOB, frequecy-to-digital coversio, frequecy measuremet, period measuremet, time measuremet, rotatioal speed measuremet 1. Itroductio The Uiversal Sesors ad Trasducers Iterface (USTI-MOB) Series of itegrated circuits is the ext geeratio of popular Series of ICs UFDC-1, UFDC-1M-16, USTI ad USTI-EXT desiged for accurate measuremets of frequecy-time parameters of sigals [1-6]. Takig ito accout low power cosumptio (<.35 ma) ad a wide fuctioality, the USTI-MOB is suitable for applicatios i various sesor systems i mobile devices (smartphoes ad tablets) ad Iteret-of-Thigs devices (IoT) [7]. I order to be accepted by the moder marker, maufacturers of ay measurig devices must experimetally cofirm its metrological performaces. Today, there are a lot of smartphoes ad tablets with embedded sesors ad meterig applicatios. Ufortuately, a lot of such sesor systems have very poor metrological characteristics. May meterig applicatios ca be cosidered oly as "toys", especially applicatios, based o GPS. For example, barometric ad altimeters applicatios from various maufactures give results of measuremets, which are differ i 1.5 times. Metrological characteristics of ay sesor systems are maily depedet o errors of sesors ad ADC coverters. I order to miimize it at reasoable costs, frequecy output sesors ad appropriate frequecy-to-digital coverters must be used. Such solutio also gives a lot of advatages [7], icludig sigificatly icreased umbers of accurate embedded sesors i the future mobile devices due to a icreased system itegratio level [7]. For these reasos, a curret study ad experimetal ivestigatio aimed at developmet a ovel approach for sesor systems desig usig the USTI-MOB ad frequecy (period)- ad (PWM) output sesors were carry out ad described i this article. http://www.sesorsportal.com/html/digest/p_2655.htm 15
The article is orgaized as follow. I Sectio I the experimetal set-up ad measuremet techique are described. The experimetal results for frequecy, period, pulse width, pulse space ad rotatioal speed measuremets are provided ad discussed i Sectio II. The article is cocluded i Sectio III. 2. Measuremet Techique ad Experimetal Set-Up The aim of this research was to determie limited metrological characteristics of desiged USTI-MOB such as measurig rages, measurig time ad relative errors of measuremets for various frequecy-time parameters of sigals such as frequecy, period, frequecy/period ratio/differece pulse width ad pulse space. The diagram of experimetal measuremet set-up for the USTI-MOB is show i Fig. 1, ad circuit diagram of USTI-MOB coectio - i Fig. 2. The last circuit diagram is similar to the USTI circuit diagram of coectio [1, 5]. The differece is oly i the voltage of power supply V cc : 1.8 V for USTI-MOB ad 5 V for USTI. Two square waveform pulse sigals whose frequecy-time parameters must be measured, were fed from two chaels of Waveform Geerator Agilet 335B to iputs FX1, ST1 ad FX2, ST2 (the 1 st ad 2 d chael of IC respectively) of the USTI-MOB ruig o a 4 MHz clock. The supply voltage of the evaluatio board was +14 V dc, provided by the Promax FA-851 power supply. The frequecy-time parameters of sigals geerated by the waveform geerator were measured by both: the USTI-MOB ad Uiversal Frequecy Couter/Timer Agilet 5322A with the ultra high ove stability iteral time base. The two-chael digital oscilloscope Promax OD-591 moitored the sigals waveforms. Before measuremets, the USTI-MOB was calibrated i the workig temperature rage: +22.7 + 25.3 o C at 34-38 % RH. The measurads were set to a PC via a RS232 iterface implemeted with the ST22D IC. The user iterface was realized with the help of termial software Termial V1.9b ruig uder Widows XP or Widows 7 operatio systems. As usually, every measuremet were cosisted of 1 values (sample size). The measuremet errors were evaluated from appropriate statistics with the help of NUMERI [8] ad STATISTICA V.12 software. The umber of itervals groupig of experimetal data for histograms was selected accordig to the followig equatio [9]: mmi.55.4 = ad mmax 1.25.4 =, (1) where is the sample size. For =1, the umbers of itervals groupig of experimetal data m are the followig: m mi = 4 ad m max = 8. The Waveform Geerator Agilet 335B has the high-stability OCXO timebase (frequecy referece ±.1 ppm of settig ±15 phz) [1]. The Uiversal Frequecy Couter/Timer Agilet 5322A-1 has the ultra high-stability OCXO timebase (±5 ppb) [11]. The photo of experimetal set-up is show i Fig. 3. Fig. 1. Diagram of experimetal measuremet set-up. 16
USTI-MOB Fig. 2. Circuit diagram of USTI-MOB coectio. Fig. 3. Experimetal measuremet set-up. 3. Experimetal Results 3.1. Frequecy (f x ) ad Period (T x ) Measuremets Frequecy f x ad period T x =1/f x measurig modes are most importat modes of USTI-MOB. Both are based o the pateted advaced modified method of the depedet cout, which guaratees the costat, selectable (programmable) relative quatizatio error i the specified measurig rage of frequecies from ifralow frequecy to high frequecy; o-redudat coversio time, which ca be chaged adaptively accordig to the programmable relative error, ad scalable resolutio. The measurig rage of USTI-MOB, which has bee determiated experimetally is from.25 Hz to 1.95 MHz without prescalig ad 31.2 MHz with prescalig (a digital biary divider to 16). Durig the experimetal ivestigatio, measuremets have bee made for the followig frequecies (1 times for each value):.25 Hz, 1 Hz, 1 Hz, 1 Hz, 1 khz, 1 khz, 1 khz, 1 MHz ad 1.95 MHz. The covetioal true values for these frequecies have bee measured by the Uiversal Frequecy Couter/Timer Agilet 5322A. The miimum possible relative error of USTI-MOB has bee selected (programmed) for all of measuremets (commad "aa" for the RS232 commuicatio iterface [5]). Before the each series 17
of measuremets, the USTI-MOB has bee calibrated i order to compesate timebase's errors due to the quartz crystal's o-perfect techology of maufacturig, agig ad temperature chages durig the experimetal ivestigatio. Experimetal results for 1 Hz, 1 khz ad 1.95 MHz frequecies measuremets are show i Fig. 4-6 respectively. The χ 2 test for goodess of fit test was applied to ivestigate the sigificace of the differeces betwee observed data i the histograms ad the theoretical frequecy distributio for data from a ormal, uiform or expoetial populatio. If S < χ 2 max, where S is the sum of deviatios betwee the dataset ad the assumed distributio, ad χ 2 max is the maximum possible allowable deviatio i the χ 2 distributio, the hypothesis of appropriate distributio ca be accepted [8]. The χ 2 test has bee used at 95 % cofidece ad the umber of itervals groupig of experimetal data for histograms m=6 or 7, see equatio (1). The statistical characteristics are adduced i Table 1. As it is visible from the table, the maximal relative error with 95 % cofidece does ot exceed the relative error δ x < ±.89 % i the whole specified rage of frequecies icludig low ad ifralow frequecies. As usually, moder frequecy/period output sesors ad trasducers have relative errors i 2-3 order more i compariso with the USTI-MOB relative error [1, 12]. Hece, the USTI-MOB error ca be eglected at such sesor systems desig i accordace with the rule of eglect small compoets of the error [9]. a) b) fx, Hz f x = 1 Hz 1,15 1,1 1,5 1 9,99995 9,9999 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 c) d) Relative error, % δ x (f x = 1 Hz), %,14,12,1,8,6,4,2 2 4 6 8 1 12,45,4,35,3,25 h,2,41 1,2 1,8 F,6,15,1,5,4,26,13,2,9,5,2 9,9999948 1,163 1,377 1,591 1,86 1,12 1, 9,9999948 1,163 1,377 1,591 1,86 1,12 1, e) f) Fig. 4. Iput sigal oscillogram (a), covetioal true value (b), results of measuremets (c), relative errors (d), distributio desity (e), ad distributio fuctio (f) for 1 Hz measuremets. 18
Table 1. Statistical characteristics of frequecy measuremet results. Frequecy, fx Parameter 1 Hz 1 Hz 1 khz 1 khz 1 khz 1 MHz 1.95 MHz Miimal fx (mi), Hz 1 99.9993 999.9895 9999.9424 99999.4222 999985.415 1949991.47 Maximal fx (max), Hz 1.1 1.23 1.238 1.2418 16.919 12.4 19522.22 Samplig Rage, fx (max) - fx (mi), Hz.2.29.343.2994 7.4974 34.9914 3.742 Media Arithmetic Mea, Hz 1.1 1.8 1.75 1.96 11.719 14.1 1959.98 Variace, Hz 9.3E-1 3.7E-7 4.2E-5.4 2.6661 68.6926 73.2391 Stadard Deviatio, Hz 3.1E-5.6.65.636 1.6328 8.2881 8.558 Coefficiet of Variatio 327347.961 164865.824 154713.577 157184.679 61245.223 12655.427 227858.612 Number of itervals groupig of experimetal data, m Cofidece Iterval at probability P=95 % Maximal Relative Error, δx % 7 7 7 7 7 6 6 fx [1.1 1.19] fx [1.7 1.1] f x [1.62 1.88] f x [1.781 1.13] f x [11.397 12.4] f x [12.38 15.63] f x [1958.3 19511.66] ±.89 ±.19 ±.13 ±.12 ±.32 ±.16 ±.86 Distributio low: S < > χ 2 - uiform - ormal - expoetial S=85.64 > χ 2 =12 S=14.48 > χ 2 =9.4 S= 33589668.9 > > χ 2 =11 S=37.76 > χ 2 =12 S=12.24 > χ 2 =9.4 S= 1242198.7 > > χ 2 =11 S=73.46 > χ 2 =12 S=44.83 > χ 2 =9.4 S= 1365235.3 > > χ 2 =11 S=63.24 > χ 2 =12 S=7.32 < χ 2 =9.4 (accepted) S= 14763268.9 > > χ 2 =11 S=117.72 > χ 2 =12 S=78.46 > χ 2 =9.4 S= 726686.45 > > χ 2 =11 S=87.8 > χ 2 =11 S=68.18 > χ 2 =7.8 S= 1457357.5 > > χ 2 =9.4 S=46.76 > χ 2 =11 S=32.23 > χ 2 =7.8 S= 254626.8 > > χ 2 =9.4
a) b) fx, Hz f x = 1 khz 18 16 14 12 1 99998 99996 99994 1 4 7 1 13 16 19 22 25 28 31 34 37 4 43 46 49 52 55 58 61 64 67 7 73 76 79 82 85 88 91 94 97 Relative error, %,8,7,6,5,4,3,2,1 δ x (f x = 1 khz), % 2 4 6 8 1 12 c) d),6,5,4 h,3,2,51 1,2 1,8 F,6,1,18,12,6,3,6,4,4,2 f x =1 khz, (Hz) 99999,9576911,28812,99813,17914,2419 15,313 16,3841 f x =1 khz, (Hz) e) f) Fig. 5. Iput sigal oscillogram (a), covetioal true value (b), results of measuremets (c), relative errors (d), distributio desity (e) ad distributio fuctio (f) for 1 khz measuremets. 3.2. Frequecy Ratio (f x1 /f x2 ) ad Period Ratio (T x1 /T x2 ) Measuremets There are two ways for frequecy/period ratio measuremets i the USTI-MOB. I both cases ukow frequecies f x1, f x2 or periods T x1, T x2 must be fed o two iputs FX1, ST1 ad FX2, ST2 respectively of the IC. Accordig to the first method, the frequecy/ period ratio is measured with the help of a special measurig mode (commads "m8" or "m9" for the RS232 commuicatio iterface) [6]. Accordig to the secod method, both frequecies/periods are measured with the help of frequecy/period mode i each of chael (commads "m" ad "me" respectively [6]), the the ratios f x1 /f x2 or T x1 /T x2 must be calculated with the help of a exteral microcotroller (i the case of digital sesors ad sesor systems) or by PC (i the case of DAQ systems). The secod measurig method is a little bit more accurate i compariso with the first oe, but eeds i twice loger coversio time. It is suitable for measuremet coditios, whe a dyamic error is quite low. Durig the experimetal ivestigatio, the ukow frequecies/periods were geerated by the 2-chael Waveform Geerator Agilet 335B (see oscillograms of appropriate sigals i Fig. 7). I parallel to the USTI-MOB, the measuremets were made by the Uiversal Frequecy Couter/Timer Agilet 5322A i order to get the covetioal true value of ratio for measuremet errors calculatio (Fig. 8). 2
fx, Hz f x max = 1.95 MHz 19525 1952 19515 1951 1955 195 1949995 194999 1949985 194998 1949975 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 Relative error, % δ x (f x = 1.95 MHz), %,12,1,8,6,4,2 2 4 6 8 1 12 a) b),35,7,3,6,25,5,2 h,15,1,5,32,25,14,5 194999, 195, 1951, 1951, f x max (Hz),4 F,3,2,1 1,95E+6 1,95E+6 1,95E+6 1,95E+6 f xmax (Hz) c) d) Fig. 6. Results of measuremets (a), relative errors (b), distributio desity (c), ad distributio fuctio (d) for 1.95 MHz measuremets. The experimetal results of frequecy ratio measuremet for f x1 = 2 khz ad f x2 = 1 khz are show i Figs. 9 ad 1 respectively. The statistical characteristics are adduced i Table 2. As it is visible from the table, the maximal relative errors with 95 % cofidece do ot exceed the error δ x < ±.25 % for the first ad the secod measurig methods. 3.3. Pulse Width t p ad Pulse Space t s Measuremets Fig. 7. Oscillograms of iput sigals at frequecy ratio measuremets: 1 - fx1=2 khz, 2 - fx2= 1 khz Fig. 8. Covetioal true value for frequecy ratio measuremet. Alogside with the period of sigal T x, the pulse width t p ad pulse space ts are two importat time parameters, which are ecessary for duty-cycle (D.C.=t p /T x ), duty-off factor (1/D.C.) ad PWM (t p /t s ) idirect measuremet ad calculatio. Such time parameters of sigals are widely used as output iformative parameters for differet sesors [12]. Accordig to the IFSA Study 213, such sesors have 16 % of share amog all quasi-digital sesors o the moder market. The pulse width t p ad pulse space t s measuremets i the USTI-MOD are performed accordig to the classical method with the maximum possible accuracy for a curret measurad ad i the rage from 7 μs to 1 s. The pulse width t p ad pulse space t s ca be measured i both USTI-MOB's chaels: the mode 'B' ad 'C', ad mode '17' ad '18' accordigly [5]. 21
f x 1/f x 2 f x 1/f x 2 = 2 2,12 2,1 2,8 2,6 2,4 2,2 2 1,99998 1,99996 1,99994 1,99992 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 Relative error. %,6,5,4,3,2,1 δ x (f x1 / f x2 = 2), % 2 4 6 8 1 12 a) b) Fig. 9. Results of measuremets (a) ad relative errors (b) for frequecy ratio measuremets by the 1 st measurig method. f x 1/f x 2 2,8 2,6 2,4 2,2 2 1,99998 1,99996 1,99994 1,99992 7 1 13 16 19 22 25 28 31 34 37 4 43 46 49 52 55 58 61 64 67 7 73 76 79 82 85 88 91 94 97 1 1 4 f x 1/f x 2 = 2 Relative error, %,35,3,25,2,15,1,5 δ x (f x1 / f x2 = 2), % 2 4 6 8 1 12 a) b) Fig. 1. Results of measuremets (a) ad relative errors (b) for frequecy ratio measuremets by the 2 d measurig method. Table 2. Statistical characteristics of frequecy ratio fx1/fx2 measuremet results. Method of Measuremet Parameter 1 st 2 d Miimal fx1/fx2 (mi) 2 2 Maximal fx1/fx2 (max) 2.1 2.1 Samplig Rage, fx1/fx2 (max) - (mi).1.1 Media fx1/fx2 Arithmetic Mea fx1/fx2 2.5 2.5 Variace fx1/fx2 4.3E-1 3.6E-1 Stadard Deviatio fx1/fx2 2.1E-5 1.9E-5 Coefficiet of Variatio fx1/fx2 96219.6783 15727.359 Number of itervals groupig of experimetal data, m 7 7 Cofidece Iterval at probability P=95 % f x1 /f x2 [2 2.1] f x1 /f x2 [2 2.5] Maximal Relative Error, δx % ±.25 ±.25 Distributio low: S < > χ 2 - uiform S=69.96 > χ 2 =12 S=51.34 > χ 2 =12 - ormal S=15.93 > χ 2 =9.4 S= 6.97 < χ 2 =9.4 (accepted) - expoetial S=8513325.69 > χ 2 = 11 S= 999178.78 > χ 2 = 11 22
I this method, the relative error (i %) ca be calculated accordig to the followig equatio: 11 1,5 t p = 1 µs 1 δ tx = 1 4 1 6 t p (2) fx, Hz 1 99,5 99 Before measuremets, the USTI-MOB was calibrated i the workig temperature rage: +23.2 o C at 37-38 % RH. Oscillograms of iput sigal ad the covetioal true value of pulse width for measuremet errors calculatio are show i Fig. 11 ad Fig. 12 respectively for the t p =1 μs measuremets at 1 khz iput frequecy. The lead ad trail edges of iput pulse sigal were from 8.4 to 9 s. Relative error, % 98,5 98 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97,8,7,6,5,4,3,2,1 a) δ x (t p = 1 µs), % 2 4 6 8 1 12 b) Fig. 13. Results of measuremets (a) ad relative errors (b) for pulse width measuremets (t p =1 μs). Table 3. Statistical characteristics of pulse width tp measuremet results. Fig. 11. Oscillograms of iput sigals at pulse width measuremets (t p =1 μs). Parameter Value Miimal tp, (mi), μs 99.2495 Maximal tp, (max), μs 1.7495 Samplig Rage, tp, (max) - (mi), μs 1.5 Media tp μs Arithmetic Mea tp, μs 1.645 Variace tp, μs.131 Stadard Deviatio tp, μs.321 Coefficiet of Variatio tp, μs 311.769 Cofidece Iterval at probability P=95 % tp [99.2495335 1.749526] Maximal Relative Error, δx % ±.69 Fig. 12. Covetioal true value for pulse width measuremet (t p =1 μs). The experimetal results of pulse width measuremets are show i Fig. 13, ad statistical characteristics - i Table 3. As it is visible from the Table 3 the relative error does ot exceed ±.69 % at probability 95 %. Based o the χ 2 test, the hypothesis about uiform, ormal ad expoetial distributios were rejected because of S > χ 2 [8]. The similar metrological characteristics are predictable for measuremets of the same values of pulse space t s ad time iterval T betwee STARTad STOP-pulses. 3.5. Other Frequecy-time Parameters' Measuremets I additio to discussed frequecy-time parameters of sigals, the USTI-MOB ca measure rotatioal speed, frequecy/period differece, relative ad absolute frequecy deviatio, ad ca to cout the pulse umber. The high precisio, accurate rotatioal speed measurig mode makes the USTI-MOB very suitable for the applicatios i had-held tachometers, digital rotatioal speed sesors, appropriate sesor systems i mobile devices, etc. I this measurig mode, the USTI-MOB has the same metrological characteristics, as i the frequecy measurig mode. The rotatioal speed is calculated by the USTI-MOB accordig to the followig equatio: 23
x 6 = f x, (3) Z where Z is the umber of gradatios ("teeth") of modulatig rotor. The USTI-MOB returs a result of measuremet i the rpm uit. The modes for frequecy/period differece, relative ad absolute frequecy deviatio are very suitable for various chemical sesors ad biosesors [1]. The pulse cout mode ca be used i sesor systems based o the pulse cout output sesors, for example, pressure sesors series D8M-D1/D2 ad D8M-D82 from Omro; pulse desity output accelerometers from Silico Desigs, etc. [12]. As it was described i [7], the USTI-MOB ca work also with ay aalog output sesors with the help of itermediate voltage-to-frequecy coverters. I such case may sesors ca be embedded ito mobile devices without ay complex sesor fusio algorithms. 4. Measuremet Time Similar to the USTI itegrated circuit the USTI- MOB's measuremet time also has three compoets: coversio rate, calculatios time ad commuicatio time [1, 4]. Nevertheless the coversio rate is o-redudat ad depedet oly o the required relative error of measuremet, it is i 5 times more i compariso with the USTI's coversio rate due to the reduced clock frequecy (4 MHz for USTI-MOB i compariso with 2 MHz for USTI). The calculatio time is also reduced i the same umber. The serial clock frequecy for the I 2 C iterface is 2 khz; ad 28 khz for the SPI iterface. But the maximal possible baud rate for the RS232 iterface has bee icreased from 1/384 for USTI to 1/768 bps for the USTI-MOB. 5. Coclusios The experimetal ivestigatio of the desiged USTI-MOB itegrated circuit cofirms its high metrological characteristics at low power cosumptio (.35 ma curret cosumptio at V cc = 1.8 V). So, the relative error of frequecy ad period measuremets is costat i the whole specified measurig rage of frequecies (from.25 Hz to 1.95 MHz) ad with 95 % cofidece does ot exceed ±.89 % i the worst case. Metrological characteristics of measuremets of other frequecy-time parameters of sigal such as pulse width ad pulse space, time iterval, frequecy/period differece ad ratio, pulse umber) demostrate a possibility to use the USTI-MOB i differet sesor systems (icludig smart sesor systems) based o appropriate quasi-digital sesors with metioed iformative parameters o its outputs. The optimal trade-off betwee accuracy, power cosumptio ad commuicatio speed has achieved. It makes the USTI-MOB suitable ot oly for mobile devices but also for other low power cosumptio applicatios such as IoT sesig devices, digital sesors, measurig istrumets, etc. The future ivestigatio ad testig will iclude a experimetal metrological characteristics determiatio for duty-cycle (duty-off factor), phase shift, capacitace, resistace ad resistive bridge measuremets. The USTI-MOB IC will be itroduced o the moder market i the curret year by Techology Assistace BCNA 21, S.L. (Excelera), Barceloa, Spai (http://www.excelera.io). Refereces [1]. S. Y. Yurish, Digital Sesors ad Sesor Systems: Practical Desig, IFSA Publishig, 211. [2]. Uiversal Frequecy-to-Digital Coverter (UFDC-1), Specificatio ad Applicatio Note, Techology Assistace BCNA 21, S. L. (Excelera), 21. [3]. Uiversal Frequecy-to-Digital Coverter (UFDC-1M-16), Specificatio ad Applicatio Note, Techology Assistace BCNA 21, S. L. (Excelera), 21. [4]. S. Y. Yurish, High-Speed Uiversal Frequecy-to- Digital Coverter for Quasi-Digital Sesors ad Trasducers, Sesors & Trasducers, Vol. 8, Issue 6, Jue 27, pp. 1225-1229. [5]. Uiversal Sesors ad Trasducers Iterface (USTI), Specificatio ad Applicatio Note, Techology Assistace BCNA 21, S. L. (Excelera), 21. [6]. Uiversal Sesors ad Trasducers Iterfacig IC Meets Exteded Temperature Rage Applicatios Specificatio at +15 C, Press Release, Techology Assistace BCNA 21, S. L. (Excelera), 21, April 212. [7]. J. Cañete, S. Y. Yurish, Sesors Systems for Smartphoes, Tablets ad IoT: New Advaced Desig Approach, Sesors & Trasducers, Vol. 187, Issue 4, April 215, pp. 1-9. [8]. E. Schrufer, Sigal Processig: Digital Sigal Processig of Discrete Sigals, Lybid', Kiev, 1992, (i Ukraiia). [9]. P. V. Novitskiy, I. A. Zograf, Errors Estimatio for Measurig Results, Eergoatomizdat, Leigrad, 1991 (i Russia). [1]. 335B Series Waveform Geerators, Data Sheet, Agilet Techologies, Ic., USA, 212. [11]. Agilet 532A Series RF/Uiversal Frequecy Couter/Timers, Data Sheet, Agilet Techologies, Ic., USA, 21. [12]. Sesors Web Portal (http://www.sesorsportal.com). 215 Copyright, Iteratioal Frequecy Sesor Associatio (IFSA) Publishig, S. L. All rights reserved. (http://www.sesorsportal.com) 24
25