[go: up one dir, main page]

EE05813B1 - Method and device for frequency response measurement - Google Patents

Method and device for frequency response measurement

Info

Publication number
EE05813B1
EE05813B1 EEP201700005A EEP201700005A EE05813B1 EE 05813 B1 EE05813 B1 EE 05813B1 EE P201700005 A EEP201700005 A EE P201700005A EE P201700005 A EEP201700005 A EE P201700005A EE 05813 B1 EE05813 B1 EE 05813B1
Authority
EE
Estonia
Prior art keywords
signal
reference waveform
response
window
response signal
Prior art date
Application number
EEP201700005A
Other languages
Estonian (et)
Inventor
Olev Märtens
Mart Min
Jaan Ojarand
Raul Land
Marek Rist
Original Assignee
Tallinna Tehnikaülikool
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tallinna Tehnikaülikool filed Critical Tallinna Tehnikaülikool
Priority to EEP201700005A priority Critical patent/EE05813B1/en
Publication of EE201700005A publication Critical patent/EE201700005A/en
Publication of EE05813B1 publication Critical patent/EE05813B1/en

Links

Landscapes

  • Measuring Frequencies, Analyzing Spectra (AREA)

Abstract

The present invention relates to a method and device for frequency response measurement, including generation of an excitation signal (like sinewave) with time-depending frequency (e.g. linear chirp) for the measured object, and in a relatively short time-domain sliding analyzing window of the response signal, wherein the time instants of the sampling of the response signal are generated according to the pre-defined phase increments of the excitation signal and the mentioned sliding analysis window includes the pre-defined number of sequential response samples. This analysis in the mentioned short-time window can be implemented as the multiplication-and-accumulation of the response signal in parallel to first and second reference waveforms, one of which could be the sine-chirp waveform, similar to the excitation signal, and the other reference waveform could be the corresponding cosine-chirp, giving accordingly real and imaginary parts of the frequency response.

Description

Tehnikavaldkond Technical field

Leiutis kuulub mõõtetehnika, täpsemalt sageduskarakteristikute mõõtmise valdkonda. Antud valdkonda kuuluvad ka elektriahelate analüsaatorid ning vektorimpedantsi mõõtjad, mis samuti määravad elektriliste ülekandekarakteristikute sagedussõltuvusi. Leiutise olulisteks kasutusaladeks on elektrilised bioimpedantsi mõõtjad (rakendustega meditsiinilises diagnostikas), elektri- ja elektroonikaskeemide testrid, elektrokeemiliste elementide analüsaatorid (nt patareide seisukorra hindamiseks) ning lahendused materjalide omaduste uurimiseks ja tuvastuseks nende elektriliste omaduste kaudu. The invention belongs to the field of measurement technology, more specifically to the field of frequency response measurement. This field also includes electrical circuit analyzers and vector impedance meters, which also determine the frequency dependence of electrical transfer characteristics. Important areas of application of the invention are electrical bioimpedance meters (with applications in medical diagnostics), electrical and electronic circuit testers, electrochemical cell analyzers (e.g. for assessing the condition of batteries) and solutions for studying and identifying the properties of materials through their electrical properties.

Tehnika tase State of the art

On tuntud lahendused, kus skeemi või ahela ülekannet mõõdetakse (nt siinusekujulise) vahelduvsignaaliga ergutuse tulemusena tekkinud vastussignaali korrutamise ja akumuleerimise teel mainitud ergutussignaaliga, st ergutus- ja vastussignaalide konvolutsiooniga. Samuti võidakse analoogiliselt kasutada (ergutussignaali suhtes 90° faasis nihutatud) teist tugisignaali teise konvolutsiooni leidmiseks (nn kvadratuur-komponent), vt US7428683B2. Taolise lahenduse puuduseks on asjaolu, et tulemus tekib kahe signaali (tugija mõõdetava vastussignaali) konvolutsiooni (korrutamise ja akumuleerimise) tulemusel ühe integraalse tulemusena üle terve mõõtetsükli, mis esiteks ei näita mõõdetava ahela või objekti ülekannet eri sagedustel eraldi ("sageduskäiku") ja teiseks ei saa taoline integraalne mõõtmine kajastada õigesti ka ajas muutuvat (dünaamilist) ahelat või objekti. There are known solutions where the transmission of a circuit or circuit is measured by multiplying and accumulating the response signal generated as a result of excitation with an alternating signal (e.g. a sinusoidal signal) with the said excitation signal, i.e. by convolution of the excitation and response signals. Similarly, a second reference signal (shifted in phase by 90° with respect to the excitation signal) may be used to find the second convolution (so-called quadrature component), see US7428683B2. The disadvantage of such a solution is that the result is generated as a result of the convolution (multiplication and accumulation) of two signals (the measured response signal of the reference) into one integral result over the entire measurement cycle, which, firstly, does not show the transmission of the circuit or object to be measured at different frequencies separately ("frequency response") and, secondly, such an integral measurement cannot correctly reflect a time-varying (dynamic) circuit or object.

Sageduskarakteristikute mõõtmiseks ning ajas dünaamiliste ahelate ja objektide mõõtmiseks on tuntud tehnilised lahendused, kus kasutatakse laiaribalist ergutussignaali, mh ajas muutuva sagedusega (nt chirp) ning (vastus-)signaalide analüüsi osas suhteliselt lühikeses ajas libisevas aknas sageduslikke teisendusi, näiteks lühiajalist Fourier' teisendust (STFT). Selline lahendus on kirjeldatud K. Darowicki, P. Slepski artiklis, influence of the analyzing window on electrode impedance measurement by the continuous frequency scanning method Journal of Electroanalytical Chemistry, vol.533, issue 1-2, 20 Sept 2002, pp.25-31, Elsevier Science. Leiutisele kõige lähedasem ja tehnilise tulemuse poolest kõige samasem lahendus on kirjeldatud patendikirjelduses EE05616B1 (US8854030B2) „Meetod ja seade sageduskarakteristiku mõõtmiseks”. Antud lahenduses kasutatakse ergutussignaalina ajas muutuva sagedusega ergutusignaali (nt chirp) ning mõõtetulemus arvutatakse mõõdetavalt objektilt saadud vastussignaali konvolueerimisega (digitaalsel realiseerituna korrutamise ja For measuring frequency characteristics and measuring time-dynamic circuits and objects, there are known technical solutions that use a broadband excitation signal, including one with a time-varying frequency (e.g. chirp), and for the analysis of (response) signals, frequency transforms in a relatively short time sliding window, for example the short-time Fourier transform (STFT). Such a solution is described in the article by K. Darowicki, P. Slepski, influence of the analyzing window on electrode impedance measurement by the continuous frequency scanning method Journal of Electroanalytical Chemistry, vol.533, issue 1-2, 20 Sept 2002, pp.25-31, Elsevier Science. The solution closest to the invention and most similar in terms of technical result is described in patent specification EE05616B1 (US8854030B2) “Method and device for measuring frequency characteristic”. In this solution, an excitation signal with a time-varying frequency (e.g. chirp) is used as the excitation signal, and the measurement result is calculated by convolution of the response signal received from the measured object (digitally realized by multiplication and

akumuleerimisega) suhteliselt lühikeses ajas libisevas ajaaknas esimese tugisignaaliga, (mis omab ergutussignaaliga ajavallas sama lainekuju) ühes („samas-faasis”) mõõtekanalis ning teise („kvadratuur”) tugisignaaliga (nt koosinus-chirp, kui esimene tugisignaal on siinuschirp) teises mõõtekanalis. accumulation) in a relatively short time sliding time window with a first reference signal (which has the same waveform in the time domain as the excitation signal) in one ("in-phase") measurement channel and a second ("quadrature") reference signal (e.g. cosine chirp, if the first reference signal is a sine chirp) in the other measurement channel.

Seejuures kasutatakse kiijeldatud muutuva (nt chirp ergutusega) signaalisagedusega lahendustes lihtsat (ühtlast) võendamist pidevate (analooog-)signaalide diskreetimisel ajas, analoog-digitaal-muundamisel, vastussignaali hõivel ning ka vastupidi (digitaal-analoogmuundamisel) - ergutussignaali genereerimisel. Samuti on esimene ja teine tugisignaalid (näiteks Fourier’ koefitsendid STFT puhul) antud vastussignaali võendamisega samadel kui diskreetsetel ühtlase ajasammuga hetkedel. In the solutions with variable signal frequency (e.g. chirp excitation), simple (uniform) sampling is used in the sampling of continuous (analog) signals in time, in analog-to-digital conversion, in response signal acquisition, and also vice versa (digital-to-analog conversion) - in the generation of the excitation signal. Also, the first and second reference signals (e.g. Fourier coefficients in the case of STFT) are given by sampling the response signal at the same discrete instants with uniform time steps.

Kõigi kirjeldatud muutuva (nt chirp-ergutusega) signaalisagedusega lahenduste üheks oluliseks puuduseks on mõõtetulemuse viga, mis on seotud signaalide sageduse pideva muutumisega võendamissageduse suhtes, mille tulemusena satub mainitud suhteliselt lühikesse analüüsi (konvolutsiooni) aknasse mittetäisarv võendamisintervalle, mille pidevalt muutuv (ja analüüsist välja jääv) murdarvuline osa tekitab hüpleva veakomponendi. Analoogiline efekt fikseeritud sagedusega signaalide puhul (kui võendamisintervalle on nt Fourier’ teisenduse analüüsiaknas mitte-täisarv) on tuntud „spektri lekkena”. One of the significant drawbacks of all the described solutions with variable signal frequency (e.g. chirp excitation) is the measurement error, which is related to the continuous change of the signal frequency with respect to the sampling frequency, as a result of which a non-integer number of sampling intervals falls into the mentioned relatively short analysis (convolution) window, the constantly changing (and not included in the analysis) fractional part of which creates a hopping error component. An analogous effect in the case of fixed-frequency signals (when the sampling intervals are, for example, non-integer in the Fourier transform analysis window) is known as “spectral leakage”.

Leiutise olemus The essence of the invention

Leiutise eesmärgiks on sageduskarakteristikute mõõtmise lahenduse täpsuse tõstmine. Eesmärk saavutatakse, võrreldes tuntud lahendusega, kus samuti kasutatakse ajas suhteliselt kiiresti muutuva sagedusega ergutussignaali ning vastussignaali analüüsi suhteliselt lühikeses libisevas ajaaknas, sellega, et vastussignaali võendamishetked on genereeritud vastavalt ergutussignaali etteantud faasisammule ning mainitud suhteliselt lühikesse libisevasse ajaaknasse võetakse etteantud arv vastussignaali võendeid. The aim of the invention is to increase the accuracy of the frequency response measurement solution. The aim is achieved, compared to the known solution, which also uses an excitation signal with a frequency that changes relatively quickly in time and an analysis of the response signal in a relatively short sliding time window, by generating the response signal sampling moments according to a given phase step of the excitation signal and taking a given number of response signal samples in said relatively short sliding time window.

Mainitud suhteliselt lühikeses libisevas ajavallaaknas tehtav vastussignaali analüüs on mõistlik teostada korrutamise ja akumuleerimisega mainitud ajaaknas paralleelselt kahe erineva võendatud tugilainekujuga, millest üks on ergutussignaaliga sama kujuga ajavallas ja teine tugilainekuju on 90° faasis nihutatud variant esimesest tugilainekujust ning mainitud korrutamise ja akumuleerimise tulemuseks esimese ja teise tugilainekujuga on vastavalt sageduskarakteristiku reaal- ja imaginaarosad, sageduskäiguna vastavalt akna libisemisele. Mõistlik võib olla, et ergutussignaal ning esimene tugilainekuju on siinus-chirp ning teine tugilainekuju on vastav koosinus-chirp. On otstarbekas, et mainitud chirp-signaalid on ajas lineaarselt muutuva sagedusega (nn „lineaarne“ chirp). It is reasonable to perform the analysis of the response signal in the said relatively short sliding time window by multiplying and accumulating in the said time window in parallel with two different taken reference waveforms, one of which has the same shape as the excitation signal in the time domain and the second reference waveform is a 90° phase shifted variant of the first reference waveform and the result of said multiplication and accumulation with the first and second reference waveforms is the real and imaginary parts of the frequency response, respectively, as a frequency response according to the sliding of the window. It may be reasonable that the excitation signal and the first reference waveform are sine chirps and the second reference waveform is the corresponding cosine chirp. It is expedient that the said chirp signals have a frequency that varies linearly in time (so-called “linear” chirps).

Mõistlik võib olla, et mainitud suhteliselt lühike libiseva ajavalla analüüsiaken sisaldab alati konstantset järjestikuliste võendite arvu (N) ning ergutusignaali faasi muutumise samm, millele vastavalt genereeritakse võendamishetked, on konstantselt 360° (või selle väärtuse täisarvkordse) jagatis (N+1)-ga. It may be reasonable that the aforementioned relatively short sliding time domain analysis window always contains a constant number of consecutive samples (N) and the step of the excitation signal phase change, according to which the sample moments are generated, is constantly a quotient of 360° (or an integer multiple of this value) by (N+1).

Sageduskarakteristiku mõõtmise seadme osas, mis sisaldab ajas muutuva sagedusega võendatud tugilainekujude arvutajat (1) ning esimese ja teise tugilainekuju puhvreid (2 ja 3), mis initsialiseeritakse vastavalt mainitud tugilainekujude arvutaja (1) poolt; võendamise generaatorit (4); digitaal-analoog muundurit (5), mille signaalisisend on ühendatud esimese tugilainekuju puhvri (2) väljundisse ning mille võendamist juhib võendamise generaator (4) ning mille väljund on ühendatud mõõdetava objekti (6) ergutuse sisendiga; analoog-digitaalmuundurit (7), mille signaalisisend on ühendatud mõõdetavalt objektilt (6) saadava vastussignaaliga, ning mille võendamist juhib samuti võendamise generaator (4); vastussignaali puhvrit (8), mille sisend on ühendatud analoog-digitaal-muunduri (7) väljundiga; libisevat kahekanalilist analüüsiakent (9), mis analüüsib etteantud aknas vastussignaali puhvri (8) lainekuju võendite väärtusi paralleelselt esimese ning teise tugilainekuju puhvrite (2, 3) vastavate samas analüüsiaknas (9) olevate lainekujuväärtustega ning mille väljundiks on mõõdetava objekti (6) kompleksne impedantsiväärtuste sageduskäik. Leiutise eesmärk saavutatakse sellega, et seade sisaldab võendushetkede mustri arvutajat (10), mis juhib võendatud tugilainekujude arvutajat (1), ning võendamise generaatorit (4), kusjuures võendushetkede mustri arvutaja genereerib võendamismustri selliselt, et esimese ja teise tugilainekuju puhvrites (2, 3) olev faasisamm kahe järjestikulise võendi vahel on etteantud väärtus ning mainitud libisev analüüsiaken (9) sisaldab alati etteantud arvu võendeid nii esimese kui ka teise tugilainekuju puhvritest (2, 3) kui vastussignaali puhvrist (8). A frequency response measuring device comprising a time-varying frequency reference waveform calculator (1) and first and second reference waveform buffers (2 and 3) initialized by said reference waveform calculator (1), a subtraction generator (4); a digital-to-analog converter (5), the signal input of which is connected to the output of the first reference waveform buffer (2), and the subtraction of which is controlled by the subtraction generator (4), and the output of which is connected to the excitation input of the measured object (6); an analog-to-digital converter (7), the signal input of which is connected to the response signal received from the measured object (6), and the subtraction of which is also controlled by the subtraction generator (4); a response signal buffer (8), the input of which is connected to the output of the analog-to-digital converter (7); a sliding two-channel analysis window (9) which analyses the values of the waveform samples of the response signal buffer (8) in a given window in parallel with the corresponding waveform values of the first and second reference waveform buffers (2, 3) in the same analysis window (9) and which outputs a complex frequency response of the impedance values of the object (6) to be measured. The object of the invention is achieved in that the device includes a sampling instant pattern calculator (10) which controls the sampled reference waveform calculator (1) and a sampling generator (4), wherein the sampling instant pattern calculator generates the sampling pattern in such a way that the phase step between two consecutive samples in the first and second reference waveform buffers (2, 3) is a given value and said sliding analysis window (9) always contains a given number of samples from both the first and second reference waveform buffers (2, 3) as well as from the response signal buffer (8).

Mõistlik on, et mainitud libisev analüüsiaken (9) sisaldab vahendeid vastuvõetud signaali võendite korrutamiseks ja akumuleerimiseks paralleelselt nii esimese kui teise tugisignaali (90° nihutatud esimese suhtes) tugilainekuju puhvrites (2, 3) olevate vastavas analüüsiaknas (9) olevate väärtustega; mainitud akumuleerimise ja korrutamise vahendite väljundid näitavad vastavalt reaal- ja imaginaarosasid mõõdetavast sageduskäigust It is reasonable that said sliding analysis window (9) includes means for multiplying and accumulating the received signal values in parallel with the values of both the first and second reference signals (shifted by 90° with respect to the first) in the reference waveform buffers (2, 3) in the corresponding analysis window (9); the outputs of said accumulation and multiplication means respectively indicate the real and imaginary parts of the frequency response to be measured

Jooniste loetelu List of drawings

Joonisel fig 1 on kujutatud pakutud lahenduse teostamise näite plokkskeem. Figure 1 shows a block diagram of an example of implementing the proposed solution.

Joonisel fig 2 on kujutatud ühtlase sammuga faasinurkasid. Figure 2 shows phase angles with uniform pitch.

Joonisel fig 3 on kujutatud esimest ja teist tugilainekuju ning vastussignaali näiteid ajavallas koos võendamishetkedega. Figure 3 shows examples of the first and second reference waveforms and the response signal in the time domain with sampling times.

Joonisel fig 4 on kujutatud ühe simulatsioonist mõõdetud impedantsi sageduskarakteristiku näide võrrelduna eeldatavaga. Figure 4 shows an example of the impedance frequency response measured from one simulation compared to the expected one.

Leiutise teostamise näide Example of carrying out the invention

Pakutud lahendus (fig 1) sisaldab võendushetkede mustri arvutajat (10) ning võendatud tugilainekujude arvutajat (1), mis arvutab esimese ja teise tugilainekuju (fig 2, sinChirp(ti). cosChirp(ti)), võendushetkede mustri arvutaja (10) poolt etteantud ajahetkedel. Fig 2-4 kujutatud näites on 11 võendamispunkti (fig 2, kujutatud 30° faasisammuga) ühes konvolutsiooni libisevas ajaaknas (fig 1 (9), fig 3). Tugilainekuju arvutaja (1) poolt leitud esimese ja teise tugilainekuju võenditega (fig 3, sinChirp(ti) cosChirp(ti)), algväärtustatakse esimese ja teise tugilainekuju puhvrid (2 ja 3). Esimese lainekuju puhvri väljundist saadetakse signaal (fig 3, sinChirp(ti) läbi digitaal-analoog-muunduri (DAC) (5) analoogsignaalina mõõdetavale objektile (6), ergutusena. Vastussignaal mõõdetavalt objektilt (6) digitaliseeritakse analoog-digitaal-muunduris (ADC) (7) ning salvestatakse vastussignaali puhvris (8) (fig 3, BsinChirp(ti)). ADC (7) ning DAC (5) võendamist juhib võendamise generaator (4), vastavalt võendushetkede mustri arvutaja (10) poolt viimasele ette antud ajahetkedele (fig 3 näites 11 võendit BsinChirp(ti) antud konvolutsiooni libisevas ajaaknas). DAC (5) võendamishetked genereeritakse võendushetkede mustri arvutaja (10) ja võendamise generaatori (4) poolt. Nimetatud hetked võivad olla samad, mis ADC (7) osas või genereeritud (nt interpoleeritud vaheväärtustega) ülevõendatuna. The proposed solution (Fig 1) includes a sampling instant pattern calculator (10) and a sampled reference waveform calculator (1), which calculates the first and second reference waveforms (Fig 2, sinChirp(ti). cosChirp(ti)) at the time instants specified by the sampling instant pattern calculator (10). In the example shown in Fig 2-4, there are 11 sampling points (Fig 2, shown with a 30° phase step) in one convolution sliding time window (Fig 1 (9), Fig 3). With the samples of the first and second reference waveforms found by the reference waveform calculator (1) (Fig 3, sinChirp(ti) cosChirp(ti)), the buffers (2 and 3) of the first and second reference waveforms are initialized. From the output of the first waveform buffer, the signal (Fig 3, sinChirp(ti) is sent as an analog signal through the digital-to-analog converter (DAC) (5) to the object to be measured (6), as an excitation. The response signal from the object to be measured (6) is digitized in the analog-to-digital converter (ADC) (7) and stored in the response signal buffer (8) (Fig 3, BsinChirp(ti)). The sampling of the ADC (7) and DAC (5) is controlled by the sampling generator (4), according to the time instants given to the latter by the sampling instant pattern calculator (10) (in the example of Fig 3, 11 sampling instants BsinChirp(ti) in the sliding time window of the given convolution). The sampling instants of the DAC (5) are generated by the sampling instant pattern calculator (10) and the sampling generator (4). Said instants may be the same as those of the ADC (7) or generated (e.g. with interpolated intermediate values) overconfident.

Libisevas kahe kanaliga analüüsiaknas (9) analüüsitakse vastussignaali puhvri (8) lainekuju võendite väärtusi paralleelselt esimese ning teise tugilainekuju puhvrite (2, 3) vastavate samas analüüsiaknas (9) olevate lainekujuväärtustega, mille tulemuseks on mõõdetava objekti (6) impedantsiväärtuste kompleksne sageduskäik. In a sliding two-channel analysis window (9), the values of the waveform samples of the response signal buffer (8) are analyzed in parallel with the corresponding waveform values of the first and second reference waveform buffers (2, 3) in the same analysis window (9), resulting in a complex frequency response of the impedance values of the object being measured (6).

Mainitud analüüs (libisevas analüüsiaknas (9)) on mõistlik teostada korrutamise ja akumuleerimisega üle mainitud analüüsiaknas (9) olevate võendite väärtuste; vastussignaali puhvrist (8) väärtused (11 võendit antud näites, fig 3) korrutatakse ja akumuleeritakse esimese tugilainekuju vastava 11 võendi väärtustega jooksva tulemuse reaalosa („samas faasis”) saamiseks ning teise tugilainekuju vastava 11 võendi väärtustega jooksva tulemuse imaginaarosa („kvadratuurosa“) saamiseks. The said analysis (in the sliding analysis window (9)) is reasonably performed by multiplying and accumulating over the values of the samples in the said analysis window (9); the values from the response signal buffer (8) (11 samples in the given example, Fig. 3) are multiplied and accumulated with the values of the corresponding 11 samples of the first reference waveform to obtain the real part ("in phase") of the current result, and with the values of the corresponding 11 samples of the second reference waveform to obtain the imaginary part ("quadrature part") of the current result.

Sellisel viisil, vastavalt analüüsi akna (9) libisemisele ajas ning vastavalt muutuvale sagedusele, saadakse mõõdetava objekti (6) ülekandefunktsioon, sageduskäiguna. Kompleksset ülekandefunktsiooni teades saab omakorda arvutada mõõdetavate kompleksahelate impedantsi, kui need on mõõteskeemi osaks antud objektil (6). In this way, according to the sliding of the analysis window (9) in time and according to the changing frequency, the transfer function of the measured object (6) is obtained as a frequency response. Knowing the complex transfer function, the impedance of the measured complex circuits can be calculated, if they are part of the measurement scheme for a given object (6).

Antud näites kasutati ergutussignaalina ning esimese tugilaine kujuna lineaarset chirpy (10kHz-1MHz, 1ms vältel) ning teise tugilainekujuna vastavat koosinus-chirpi. In this example, a linear chirpy (10kHz-1MHz, 1ms duration) was used as the excitation signal and the first reference waveform, and the corresponding cosine chirpy was used as the second reference waveform.

Et saavutada leiutise eesmärk, et igas libisevas analüüsiajaaknas (9) oleks vaatamata ergutusjt signaalide suhteliselt kiiresti muutuvate sagedusele rangelt ühesugune arv intervalle (antud näites NN = 12, millele vastab NN -1 = 11 võendit) valitakse võendamishetked ja vastavad intervallid ajas niimoodi, et vastav esimese ja teise tugilainekuju faasisammud oleksid ette antud, antud näites konstantselt 360° / 12= 30° (Fig 2). NN mõistlik valik ning piisav suurus sõltuvad vajalikust mõõtetäpsusest reaalsete mitteideaalsuste (mürad, võendamishetkede värin jmt) määrast. In order to achieve the objective of the invention, that in each sliding analysis time window (9) there would be strictly the same number of intervals despite the relatively rapidly changing frequency of the excitation signals (in this example NN = 12, which corresponds to NN -1 = 11 samples), the sampling moments and the corresponding intervals in time are chosen in such a way that the corresponding phase steps of the first and second reference waveforms are given in advance, in this example constantly 360° / 12 = 30° (Fig 2). A reasonable choice and sufficient size of NN depend on the required measurement accuracy and the degree of real non-idealities (noise, jitter of sampling moments, etc.).

Libisevat analüüsiakent (9) võib nihutada jooksvalt ühe võendi kaupa edasi, saades niimoodi ajas ning vastavalt ka sageduslikult tulemuse väärtusi tihedamalt või suurema sammuga, nt veerandi, poole või terve täisakna laiuse võrra ilma ülekatteta. The sliding analysis window (9) can be continuously moved forward one sample at a time, thus obtaining result values in time and, accordingly, frequency more closely or in larger increments, e.g. by a quarter, half or a full window width without overlap.

Pakutud lahenduse näidet on simuleeritud 1 ms lineaarse chirp-signaaliga (sagedusvahemikus 10kHz-1MHz) libisevas ajaaknas (fig 2, 3) 11 võendiga ja 30° faasisammuga, korrutades ja akumuleerides vastussignaali tugilainekujudega mainitud siinus ja vastav koosinus-chirp, Fig 3 reaal- ja imaginaarosa saamiseks. Mõõdetava impedantsi Zx ergutus on läbi 1kΩ takisti ning saadud ülekandefunktsioonist on leitud Zx sageduskäik (fig 4), võrrelduna eeldatavaga (arvutatuna Zx kui tuntud kompleksse ahela sageduskäiguna). Tulemus näitab pakutud lahenduse tulemuse head kokkulangemist eeldatavaga. An example of the proposed solution has been simulated with a 1 ms linear chirp signal (in the frequency range 10kHz-1MHz) in a sliding time window (Fig 2, 3) with 11 samples and a 30° phase step, multiplying and accumulating the response signal with the reference waveforms to obtain the real and imaginary parts of the mentioned sine and corresponding cosine chirp, Fig 3. The excitation of the measured impedance Zx is through a 1kΩ resistor and the frequency response of Zx has been found from the obtained transfer function (Fig 4), compared with the expected one (calculated as the frequency response of Zx of a known complex circuit). The result shows a good agreement of the result of the proposed solution with the expected one.

Claims (7)

1. Meetod sageduskarakteristiku mõõtmiseks, mis sisaldab ajas muutuva sagedusega mõõdetavale objektile rakendatud ergutussignaali genereerimist ning suhteliselt lühikeses libisevas ajaaknas teostatud mõõdetava objekti vastussignaali analüüsi vastussignaali analüsaatoris, mis erineb selle poolest, et vastussignaali võendamishetked genereeritakse vastavalt ergutussignaali ette antud faasisammule ning suhteliselt lühikesse libisevasse ajaaknasse võetakse etteantud arv vastussignaali võendeid.1. A method for measuring the frequency response, which includes generating an excitation signal applied to a measured object with a time-varying frequency and analyzing the response signal of the measured object in a response signal analyzer in a relatively short sliding time window, characterized in that the response signal sampling moments are generated according to a given phase step of the excitation signal and a given number of response signal samples are taken in a relatively short sliding time window. 2. Meetod vastavalt nõudluspunktile 1, mis erineb selle poolest, et vastussignaali analüüs teostatakse suhteliselt lühikeses libisevas ajavallaaknas konvolutsiooni arvutamise seadmes korrutamise ja akumuleerimisega mainitud ajaaknas paralleelselt kahe erineva võendatud tugilainekujuga, millest üks on ergutussignaaliga sama kujuga ajavallas ja teine tugilainekuju on 90° faasis nihutatud variant esimesest tugilainekujust ning vastussignaali korrutamise ja akumuleerimise tulemuseks esimese ja teise tugilainekujuga on vastavalt sageduskarakteristiku reaal- ja imaginaarosad, sageduskäiguna vastavalt akna libisemisele.2. The method according to claim 1, characterized in that the analysis of the response signal is performed in a relatively short sliding time window in the convolution calculation device by multiplying and accumulating in said time window in parallel with two different taken reference waveforms, one of which has the same shape as the excitation signal in the time domain and the second reference waveform is a variant shifted in phase by 90° from the first reference waveform, and the result of multiplying and accumulating the response signal with the first and second reference waveforms is the real and imaginary parts of the frequency response, respectively, in the form of a frequency response according to the sliding of the window. 3. Meetod vastavalt nõudluspunktile 2, mis erineb selle poolest, et ergutussignaal ning esimene tugilainekuju on siinus-chirp ning teine tugilainekuju on vastav koosinus-chirp.3. The method according to claim 2, characterized in that the excitation signal and the first reference waveform are sine chirps and the second reference waveform is a corresponding cosine chirp. 4. Meetod vastavalt nõudluspunktile 3, mis erineb selle poolest, et ergutussignaali ning esimese tugilainekuju siinus-chirp on ajas lineaarselt muutuva sagedusega ning teine tugilainekuju on mainitud siinus-chirbile vastav koosinus-chirp.4. The method according to claim 3, characterized in that the sine chirp of the excitation signal and the first reference waveform has a frequency that varies linearly in time, and the second reference waveform is a cosine chirp corresponding to said sine chirp. 5. Meetod vastavalt nõudluspunktile 1, mis erineb selle poolest, et mainitud suhteliselt lühike libiseva ajavalla analüüsiaken sisaldab alati konstantset arvu (N) järjestikulisi võendeid ning ergutusignaali faasi muutumise samm, millele vastavalt genereeritakse võendamishetked, on konstantselt 360° (või selle täisarvkordse väärtuse) jagatis (N+1)-ga.5. The method according to claim 1, characterized in that said relatively short sliding time domain analysis window always contains a constant number (N) of consecutive samples, and the step of the phase change of the excitation signal, according to which the sample moments are generated, is constantly a quotient of 360° (or an integer multiple thereof) by (N+1). 6. Seade sageduskarakteristiku mõõtmiseks, mis sisaldab ajas muutuva sagedusega võendatud tugilainekujude arvutajat (1) ning esimese ja teise tugilainekuju puhvreid (2 ja 3), mis on initsialiseeritud vastavalt tugilainekujude arvutaja (1) poolt; võendamise generaatorit (4); digitaal-analoog muundurit (5), mille signaalisisend on ühendatud esimese tugilainekuju puhvri (2) väljundisse ning mille võendamine on juhitud võendamise generaatori poolt (4) ning mille väljund on ühendatud mõõdetava objekti (6) ergutuse sisendiga; analoog-digitaal-muundurit (7), mille signaalisisend on ühendatud mõõdetavalt objektilt (6) saadava vastussignaaliga, ning mille võendamine on juhitud samuti võendamise generaatoriga (4); vastussignaali puhvrit (8), mille sisend on ühendatud analoog-digitaal-muunduri (7) väljundiga; libisevat kahekanalilist analüüsiakent (9), mis analüüsib etteantud aknas vastussignaali puhvri (8) lainekuju võendite väärtusi paralleelselt esimese ning teise tugilainekuju puhvrite (2, 3) vastavate samas analüüsiaknas (9) olevate lainekujuväärtustega, ning mille väljundiks on kompleksne mõõdetava objekti (6) impedantsiväärtuste sageduskäik, mis erineb selle poolest, et seade sisaldab võendushetkede mustri arvutajat (10), mis on juhitud võendatud tugilainekujude arvutaja (1) poolt ning võendamise generaatorit (4), kusjuures võendushetkede mustri arvutaja (10) poolt on võendamismuster genereeritud selliselt, et esimese ja teise tugilainekuju puhvrites (2, 3) olev faasisamm kahe jägestikulise võendi vahel on etteantud väärtus ning mainitud libisev analüüsiaken (9) sisaldab alati etteantud arvu võendeid nii esimese ja teise tugilainekuju (2, 3) kui vastussignaali puhvritest (8).6. A device for measuring frequency response, comprising a time-varying frequency reference waveform calculator (1) and first and second reference waveform buffers (2 and 3) initialized by the reference waveform calculator (1), respectively; a subtraction generator (4); a digital-to-analog converter (5), the signal input of which is connected to the output of the first reference waveform buffer (2), and the subtraction of which is controlled by the subtraction generator (4), and the output of which is connected to the excitation input of the measured object (6); an analog-to-digital converter (7), the signal input of which is connected to the response signal received from the measured object (6), and the subtraction of which is also controlled by the subtraction generator (4); a response signal buffer (8), the input of which is connected to the output of the analog-to-digital converter (7); a sliding two-channel analysis window (9) that analyzes the values of the waveform samples of the response signal buffer (8) in a given window in parallel with the corresponding waveform values of the first and second reference waveform buffers (2, 3) in the same analysis window (9), and the output of which is a complex frequency response of the impedance values of the measured object (6), characterized in that the device includes a sampling instant pattern calculator (10), which is controlled by the sampled reference waveform calculator (1), and a sampling generator (4), wherein the sampling instant pattern is generated by the sampling instant pattern calculator (10) in such a way that the phase step in the first and second reference waveform buffers (2, 3) between two consecutive samples is a given value, and said sliding analysis window (9) always contains a given number of samples from both the first and second reference waveforms (2, 3) and the response signal buffers (8). 7. Seade vastavalt nõudluspunktile 6, mis erineb selle poolest, et mainitud libisev analüüsiaken (9) sisaldab vahendeid vastuvõetud signaali võendite korrutamiseks ja akumuleerimiseks (MAC) paralleelselt nii esimese kui teise tugisignaali (90° nihutatud esimese suhtes) tugilainekujude puhvrites (2, 3) olevate vastavas analüüsiaknas (9) olevate väärtustega; mainitud akumuleerimise ja korrutamise vahendite väljundid näitavad vastavalt reaal- ja imaginaarosasid mõõdetavast sageduskäigust.7. The device according to claim 6, characterized in that said sliding analysis window (9) comprises means for multiplying and accumulating (MAC) the received signal values in parallel with the values of both the first and second reference signals (shifted by 90° with respect to the first) in the reference waveform buffers (2, 3) in the respective analysis window (9); the outputs of said accumulation and multiplication means respectively indicate the real and imaginary parts of the frequency response to be measured.
EEP201700005A 2017-02-01 2017-02-01 Method and device for frequency response measurement EE05813B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EEP201700005A EE05813B1 (en) 2017-02-01 2017-02-01 Method and device for frequency response measurement

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
EEP201700005A EE05813B1 (en) 2017-02-01 2017-02-01 Method and device for frequency response measurement

Publications (2)

Publication Number Publication Date
EE201700005A EE201700005A (en) 2018-09-17
EE05813B1 true EE05813B1 (en) 2018-11-15

Family

ID=63517611

Family Applications (1)

Application Number Title Priority Date Filing Date
EEP201700005A EE05813B1 (en) 2017-02-01 2017-02-01 Method and device for frequency response measurement

Country Status (1)

Country Link
EE (1) EE05813B1 (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112713911B (en) * 2020-12-08 2022-02-22 重庆邮电大学 Method for increasing reference signal in Chirp modulation signal

Also Published As

Publication number Publication date
EE201700005A (en) 2018-09-17

Similar Documents

Publication Publication Date Title
US8854030B2 (en) Method and device for frequency response measurement
CN100353169C (en) Method for testing electronic component and its instrument
EE05788B1 (en) Method and device for impedance analysis with binary excitation
Zou et al. Blind timing skew estimation using source spectrum sparsity in time-interleaved ADCs
Vucijak et al. A simple algorithm for the estimation of phase difference between two sinusoidal voltages
Pjevalica et al. Further generalization of the low-frequency true-RMS instrument
CN103983849A (en) Real-time high-accuracy power harmonic analysis method
Toth et al. Power and energy reference system, applying dual-channel sampling
CN104849569B (en) Dielectric loss measuring method
CN102495280B (en) Anti-noise broadband frequency measurement method and phase-locked frequency meter
Ireland et al. A method for using Josephson voltage standards for direct characterization of high performance digitizers to establish AC voltage and current traceability to SI
Sheu et al. High precision, low complexity, and fast calculation based on hybrid recursive DFT and FFT algorithms for electrochemical impedance spectroscopy system
RU2435168C1 (en) Method for harmonic analysis of periodic multifrequency signal
RU2363005C1 (en) Method of spectral analysis of polyharmonic signals and device to this end
D'Apuzzo et al. Modeling DAC output waveforms
Ihlenfeld The ac-ac transfer technique with subsampling
JPH0723901B2 (en) Frequency response function measurement method
Ecker et al. A digital method for phase noise measurement
CN101300497B (en) Data compression to generate spectral traces
EE05813B1 (en) Method and device for frequency response measurement
CN1968161A (en) Filter equalization using magnitude measurement data
Ivanisevic et al. Impedance spectroscopy systems: Review and an all-digital adaptive IIR filtering approach
Ojarand et al. Comparison of excitation signals and methods for a wideband bioimpedance measurement
US9759751B1 (en) Line cycle correlated spectral analysis for power measurement systems
Kitzig et al. Accuracy of power quality measurement based on interpolated sampling

Legal Events

Date Code Title Description
MM4A Lapsed by not paying the annual fees

Effective date: 20210201