Quality improvemet i measuremet chael icludig of ADC uder operatio coditios 1 Romuald MASNICKI, 2 Jausz MINDYKOWSKI 1, 2 Gdyia Maritime iversity, ul. Morska 81-83, 81-225 Gdyia, POLAND, tel. (+48 58) 6109 440, 1 romas@am.gdyia.pl 2 jamid@am.gdyia.pl Abstract-The possibility of usig of the IEEE STD-1057 ad STD-1241 stadards for measuremet chael features estimatio i operatio coditios is discussed. The stadards cocer the test methods ad circuits for the devices cotaiig ADC i laboratory coditios ad caot be simply applied i the istrumet uder operatio coditios. Some remarks cocerig the specificity of such coditios ad rage of applicability of stadards are preseted. The results of examiig of selected test sigal sources are discussed. I. The IEEE stadards for estimatio of ADC features O the preset developmet of techology, the acquisitio, processig ad trasmissio of the digital data are easier tha aalog oe. Thus, the request for the aalog-to-digital coverter (ADC) has appeared ad still growig. I effect, the geeral usig of ADC i umber of devices makes it possible to covert the iformatio amog these separate two i ature groups of quatities, also i applicatios desiged for use i measuremet ad cotrol. ADC presets the iterface eablig the displacemet of iformatio from a world of aalog quatities to digital oe. The evolutio of digital techology is stimulated by processes coected with the wide use of microprocessor devices ad especially persoal computers. It is rearraged o growth of request for the arragemets eablig the processig of iformatio from aalog form i the digital form as additioal edowmet of computer or autoomous devices, from which the iformatio ca be set with stadard digital iterface. I cosequece, the problem of evaluatio of quality of ADC coverters becomes importat, as well as the devices, the ADCs were applied. The stadards elaborated by orgaizatio IEEE: STD-1057 ad STD-1241 [1,2] have a key meaig i estimatio of parameters of cosidered devices. The mai aim of the metioed stadards is to establish the commo surface for the exchage of iformatio betwee producers ad users of metioed devices i the fields: termiology, testig methods, potetial errors sources idetificatio, device usable parameters. I commo practice, the stadards costitute the base for the comparative estimatio of parameters of devices icludig ADC. Both stadards have may commo refereces. They cotai the basic terms i domai of termiology ad symbols as well as the way of desigatio of the parameters used i ADC ad coected with them digital devices features. The methods ad testig circuits for estimatio of recommeded parameters are also proposed. I differet applicatios, the essetiality of several parameters related to ADC is varied. Tab.1 shows the exemplary domais of ADC applicatios set together with the critical parameters typically desigated for respective applicatio [1,2]. I Fig.1 the geeral cofiguratio of ADC parameters testig system is show. Depedig o kid of the omiated parameter ad applied estimatio algorithm, as the test sigal the sie wave, ramp or pulse/step sigals are used. I selected cases, the other sources of test sigals are additioally used, e.g. the combiatio of sie waveforms (multitoe) or oise sigal geerators. As the referece sigal source, the precise DAC coverters cotrolled by test system are used. I acquisitio block (Acquisitio ad data processig) collectig the output data from ADC, the desigatio of required parameters is achieved.
Table 1. The exemplary domais of ADC applicatio ad critical parameters. Exemplary applicatios Critical parameters Exemplary applicatios Critical parameters Audio SINAD, THD, IMD, Crosstalk Video ad televisio INL, DNL, BW, THD, SINAD Data trasmissio SFDR, SINAD, INL, DNL, Radars ad soars SINAD, SFDR, INL, BW BW Digital istrumets ENOB, BW, SINAD, INL, Spectrum aalyzers SINAD, SFDR, BW, DNL, SFDR ENOB Medical istrumets SFDR, SINAD, INL, BW Telecommuicatio SINAD, SFDR, THD, IMD Automatic cotrol Mootoicity, stability Moitorig ENOB, SINAD Supply Referece voltage Test sigal source S&H (optio) ADC Acquisitio ad data processig Cotrol Figure 1. The geeral cofiguratio of ADC parameters estimatio testig system. The ADC iclusive devices test methods ca be classified cosiderig the domai used for respective parameters desigatio. The classificatio cotais the methods fulfilled i domai of amplitude, frequecy ad time [1,2,3,4]. The methods related to amplitude domai usually are coected with usig of sie wave or ramp sigal source as the test sigal. The data processig algorithm, leadig to the required parameter desigatio, is based o the statistical data aalysis ad eable the makig of histogram of registered data as a result. The methods related to frequecy domai are usually based o use the sie wave or step sigal as the test sigal. The ADC output data are processed usig Fourier trasform. The required parameters are calculated o the basis of correspodig mathematical depedecies. The pulse or step sigals are usually used as the test sigals for the test methods related to time domai. They are applied for desigatio such parameters as the time respose for step, delay, settle time etc. The potetial istrumet user is usually iterested i limited umber of parameters defied i stadards ad they are the subject of comparative aalysis of the differet devices features. The selectio of the proper ADC test method, amog recommeded i stadards, is doe takig ito cosideratio the ecessary scope of iformatio about the features of respective devices (Tab.1). The testig procedures are fulfilled i the especially arraged measuremet statio ad they are very closely related to the certificatio (legalizatio) procedures commoly carried out for examiatio of all measuremet istrumets. The obtaied values of ivestigated parameters testify to a quality of the examied device. Accordig to measuremet istrumet, the quality of device meas usually the accuracy of measuremet results obtaied with use of the istrumet. The certificatio procedures are carried out i laboratory coditios, but additioally the iflueces of exteral referece coditio chages o the selected features of istrumet are examied. The metioed procedures are fulfilled usually selectively ad i such form the effects of various iflueces are preseted for user iformatio. II. The assessmet of istrumet features o-lie The test procedures accordigly to stadards recommedatios eed to apply the test sigal source, properly to the give ADC. The recommedatios icluded i stadards cocer the procedures fulfilled i the laboratory coditios. The result of examiatio is desigatio of the value of coefficiets characterizig the features of the give ADC set. Because of variety of desigated coefficiets, the miscellaeous test sigal are proposed. The istrumet user for adequate accuracy estimatio should to perform the multilateral aalysis of the ifluecig factors ad to evaluate o this base the ucertaity of results obtaied durig measuremet process. The above procedure has the purely hypothetical meaig because the practical measuremets are fulfilled
accordigly to assumptio of the sufficiet accuracy of used istrumet, based o the actuality of its certificates [5]. The operatio coditios usually differ from the laboratory oes. These coditios ifluece as a rule o the coversio characteristics of measuremet chael, especially o the stage of the data acquisitio (the aalog part of the measuremet chael together with ADC). cotrolled chages of the various iflueces result i modificatio of may features of the chael uder cosideratio. I measuremet chael, the iformatio is trasmitted usually o the voltage level, so the iflueces related to its static coversio characteristics course seem to be most importat. The dyamic parameters of measuremet chael (icludig ADC) are established while the istrumet is desiged, usually takig ito accout the idispesable dyamic reserve of its respective characteristics. The istrumet desiger does it maily o the way of selectio of the uits ad elemets havig the adequate dyamic features. Acceptig such assumptio for measuremet istrumet ad takig ito accout the accuracy of measuremet result, the presece of various systematic disturbaces with fluet itesity effects i the course of the coversio characteristics of all data acquisitio uit (the aalog part of measuremet chael icludig ADC). Additioally, the presece of the radom disturbaces i operatio coditio, trasmitted by various ways to the chael, iflueces the accuracy of obtaied measuremet results. The research iteded to apply of the stadards recommedatios for the estimatio of selected features of the measuremet chael, the measuremet result correctio ad its ucertaity estimatio are udertake [6,7]. I Fig.2, the locatio of additioal test procedures (graded blocks) locatio i relatio to the stadard cofiguratio of the measuremet chael is show. The test sigal S t o the iput of measuremet chael eables to complete the digital file of data D t, used for the estimatio of actual characteristics of the chael. The digital data D m gaied durig the coversio of measuremet sigal S m are corrected takig ito accout the actual characteristics of the chael (D k ) ad are processed accordig to the measuremet fuctio applied i the chael. O the base of the data acquired durig test procedure ad data processig algorithms, the ucertaity of measuremet result is estimated. The graded blocks icluded i the measuremet chael, exceptig the Test sigal block, are the software procedures. The Test sigal block represets the test sigal source. S t S m Aalog chael with ADC D m D t Data correctio D k Data processig Result Test sigal Actual characteristics certaity estimatio Result ucertaity Figure 2. The locatio of test procedures i measuremet chael. The mai problem to perform preseted idea is how to arrage the adequate quality test sigal source. A few cofiguratios of sigal sources were examied i test arragemet show i Fig.3. The coectios to the ports of PCI-730 board are preseted. Virtual istrumet u(t) Test sigal t start clear low level high level u w (t) PA0 PA1 PB0 PB1 CH9 DAQ PCI-730 Istrumet iput CH1 Figure 3. The circuitry cofiguratio of virtual istrumet with test sigal source [5].
The data acquisitio procedures are cotrolled by LabVIEW applicatio. Selected parts of program are show i Fig.4. The block diagram cosists of the sequece of procedures eablig the cotrol of work of devices icluded ito source test arragemet. a/ b/. c/ d/ Figure 4. The frot pael (a/) ad sequece of block diagrams (b/, c/, d/) of LabVIEW applicatio cotrollig the test arragemet ad data acquisitio for examiatio of test sources. The ADC board iput rage was established o ±5 [V]. The subject of research were a few circuit cofiguratios of the test sources. The examiatio results of some of them are show i Fig.5. They have ot passed the tests. The output sigals have proved to be ot eough liear. 6 4 2 0 1 47 93 139 185 231 277 323 369 415 461 507 553 599 645 691 737-2 -4 6 4 2 0 1 268 535 802 1069 1336 1603 1870 2137 2404 2671 2938 3205-2 -4-6 -6 Figure 5. The exemplary courses of output sigals of the egatively classified test sigal sources (tred lie of sigal is added).
Fially, as a result of umber of modificatios, the test sigal source cofiguratio, providig desirable output sigal, was elaborated. The sigal provided from this source to virtual istrumet is show i Fig.6.a. The liearity of obtaied course is satisfactory. As a referece, i Fig.6.b the course of acquired ramp sigal from 33220A Agilet geerator is show. a/ b/ 6,00 4,00 2,00 0,00 1 94 187 280 373 466 559 652 745 838 931 1024 1117-2,00-4,00 6,00 4,00 2,00 - -2,00-4,00 1 1791 3581 5371 7161 8951 10741 12531 14321-6,00-6,00 Figure 6. The PCI 730 acquired digital data as a result of aalog-to-digital coversio of iput sigals: a/ tested patter source, b/ ramp sigal provided by 33220A Agilet geerator. The liearity of aalyzed courses was estimated by calculatig of the r - Pearso factor. It is the coefficiet used to assess the liear depedece betwee two sets of data. Its value is calculated as i (1). (1) where: x, y, x, y estimated data tables ad their mea values, respectively. The estimatio was carried out for acquired data obtaied by coversio of ADC iput sigals i accordace to liear approximatio of cosidered courses (the referece data sets). For examied source, the r-factor value is equal to 0,9999762, the r-factor for iput patter ramp sigal provided by 33220A Agilet geerator is equal to 0,9999998. The values of r-factor for rejected sources were o the level of 0,995. The stadard deviatio for the test sigal calculated i relatio to its liear approximatio (tred lie) is equal to 0,047 [V] ad for referece Agilet geerator sigal is equal to 0,072 [V]. The value of stadard deviatio of sigal from patter geerator is higher because of the ature of digitally sythetized output waveform. The root mea squared errors are equal 0,00173 [V] ad 0,00058 [V], respectively. The cosidered ADC quatum value correspodig to 1 LSB is equal approximately to 0,0006 [V]. III. Coclusios The IEEE stadards cocerig the estimatio of ADC ad related devices may be the referece for the ivestigatio focused o the evaluatio of measuremet result i operatio coditios. The recommedatios of stadards cocer the methods of test ad estimatio of features of devices ad chaels cotaiig the ADCs. They are used maily for comparative assessmet of the features of several devices. They have limited usability to o-lie estimatio of actual characteristics of measuremet chael as well as the measuremet result. The examiatio of istrumet i operatio coditios eeds to solve the problems of the test sigal geeratig ad providig it to the iput of measuremet chael as well as the data obtaied o the ADC output further processig. I the paper, the cofiguratio of measuremet chael additioal circuits eablig o-lie chael feature estimatio ad the result correctio accordig to the actual operatio coditios are proposed. Selected results coected with the examiatio of several sources cofiguratios are show. The satisfactory results give the assumptio to further research focused o the o-lie measuremet chael feature estimatio.
Refereces [1] IEEE 1057-1994, Stadard for Digitizig Waveform Recorders, 1994. [2] IEEE Std 1241-2000, IEEE Stadard for Termiology ad Test Methods for Aalog-to-Digital Coverters, IEEE, NY, 2001. [3] Arpaia P., Schumy H, Iteratioal stadardizatio of ADC-base measurig systems state of art, Computer Stadards & Iterfaces 19, Elsevier, 1998. [4] Cruz Serra A., Alegria F., Martis M., Foseca da Silva M., Aalog-to-digital coverter testig ew proposals, Computer Stadards & Iterfaces 26, Elsevier, pp. 3-13, 2003. [5] Masicki R, Midykowski J., A ucertaity as a compoet of the measuremet result of the microprocessor istrumets, IMEKO Proceedigs, Athes, Greece, vol.2, pp. 644-649, 2004. [6] Masicki R., The Adaptatioal Procedures Applied i Microprocessor Measuremet Chael, IMEKO Proceedigs, Jurata, Polad, vol.1, pp.114-117, 2005. [7] Masicki R., Midykowski J., Measuremet Accuracy Assessmet i Ship Applied Istrumetatio, IMEKO Proceedigs, Jurata, Polad, vol.1, pp. 87-90, 2005.