UDK 621,3:(53+54+621 +66)(05)(497.1 )=00 ISSN 0352-9045 Strokovno društvo za mikroelektroniko elektronske sestavne dele in materiale 1998 Strokovna revija za mikroelektroniko, elektronske sestavne dele in materiale Journal of Microelectronics, Electronic Components and Materials INFORMACIJE MIDEM, LETNIK 28, ŠT. 3(87), LJUBLJANA, september 1998 Iskra Iskra Avtoelektrika d.d. INFORMACIJE MIDEM 3 01998 INFORMACIJE MIDEM LETNIK 28, ŠT. 3(87), LJUBLJANA, SEPTEMBER 1998 INFORMACIJE MIDEM VOLUME 28, NO. 3(87), LJUBLJANA, SEPTEMBER 1998 Izdaja trimesečno (marec, junij, september, december) Strokovno društvo za mikroelektroniko, elektronske sestavne dele in materiale. Published quarterly (march, june, september, december) by Society for Microelectronics, Electronic Components and Materials - MIDEM. Glavni in odgovorni urednik Dr. Iztok Šorli, dipl.ing., Editor in Chief MIKROIKS d.o.o., Ljubljana Tehnični urednik Executive Editor Uredniški odbor Editorial Board Časopisni svet International Advisory Board Naslov uredništva Headquarters Dr. Iztok Šorli, dipl.ing., Doc. dr. Rudi Babič, dipl.ing., Fakulteta za elektrotehniko, računalništvo in informatiko Maribor Dr.Rudi Ročak, dipl.ing., MIKROIKS d.o.o., Ljubljana mag.Milan Slokan, dipl.ing., MIDEM, Ljubljana Zlatko Bele, dipl.ing., MIKROIKS d.o.o., Ljubljana Dr. Wolfgang Pribyl, SIEMENS EZM, Villach mag. Meta Limpei, dipl.ing., MIDEM, Ljubljana Miloš Kogovšek, dipl.ing., Ljubljana Dr. Marija Kosec, dipl. ing., Inštitut Jožef Stefan, Ljubljana Prof. dr. Slavko Amon, dipl.ing., Fakulteta za elektrotehniko, Ljubljana, PREDSEDNIK - PRESIDENT Prof. dr. CorClaeys, IMEC, Leuven Dr. Jean-Marie Haussonne, EIC-LUSAC, Octeville Dr. Marko Hrovat, dipl.ing., Inštitut Jožef Stefan, Ljubljana Prof. dr. Zvonko Fazarinc, dipl.ing., CIS, Stanford University, Stanford Prof. dr. Drago Kolar, dipl.ing., Inštitut Jožef Stefan, Ljubljana Dr. Giorgio Randone, ITALTEL S.l.T. spa, Milano Prof. dr. Stane Pejovnik, dipl.ing., Kemijski inštitut, Ljubljana Dr. Giovanni Soncini, University of Trento, Trento Prof.dr. Janez Trontelj, dipl.ing., Fakulteta za elektrotehniko, Ljubljana Dr. Anton Zalar, dipl.ing., ITPO, Ljubljana Dr. Peter Weissglas, Swedish Institute of Microelectronics, Stockholm Uredništvo Informacije MIDEM Elektrotehniška zveza Slovenije Dunajska 10, 1000 Ljubljana, Slovenija tel.: +386 (0)61 31 28 98 fax: +386 (0)61 31 91 70 Iztok. Sorli@ guest, arnes. si http://poilux.fer.uni-lj.si/midem/journal.htm Letna naročnina znaša 12.000,00 SIT, cena posamezne številke je 3000,00 SIT. Člani in sponzorji MIDEM prejemajo Informacije MIDEM brezplačno. Annual subscription rate is DEM 200, separate issue is DEM 50. MIDEM members and Society sponsors receive Informacije MIDEM for free. Znanstveni svet za tehnične vede I je podal pozitivno mnenje o reviji kot znanstveno strokovni reviji za mikroelektroniko, elektronske sestavne dele in materiale, izdajo revije sofinanci rajo Ministrstvo za znanost in tehnologijo in sponzorji društva. Scientific Council for Technical Sciences of Slovene Ministry of Science and Technology has recognized Informacije MIDEM as scientific Journal for microelectronics, electronic components and materials. Publishing of the Journal is financed by Slovene Ministry of Science and Technology and by Society sponsors. Znanstveno strokovne prispevke objavljene v Informacijah MIDEM zajemamo v podatkovne baze COBISS in INSPEC. Prispevke iz revije zajema ISI ® v naslednje svoje produkte: Sei Search® , Research Alert® in Materials Science Citation Index™ Scientific and professional papers published in Informacije MIDEM are assessed into COBISS and INSPEC databases. The Journal is indexed by iSI® for Sei Search®, Research Alert® and Material Science Citation Index™ Po mnenju Ministrstva za informiranje št.23/300-92 šteje glasilo Informacije MIDEM med proizvode informativnega značaja, za katere se plačuje davek od prometa proizvodov po stopnji 5 %. Grafična priprava in tisk BIRO M, Ljubljana Printed by Naklada 1000 izvodov Circulation 1000 issues Poštnina plačana pri pošti 1102 Ljubljana Slovenia Taxe Perçue UDK621,3:(53 + 54+621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3,Ljubljana ZNANSTVENO STROKOVNI PRISPEVKI PROFESSIONAL SCIENTIFIC PAPERS P.J. Mach, P.M. Svasta: Študij povezave med diferencialno nelineamostjo, nelinearnostjo in šumom debeloplastnih uporov 149 P.J. Mach, P.M. Svasta: Study of Correlation Among Differential Nonlinearity, Nonlinearity and Noise of Thick Film Resistors D. Raič: Merjenje porabe moči v pomnilnih strukturah CMOS 154 D. Raič: Measuring the Weighted Power of CMOS Latching Circuits V. Jenuš, B. Horvat: Audio vmesnik za računalnik 162 V. Jenuš, B. Horvat: Audio Interface for Workstation W. Pribyi, T. Scheiter, G. Hribemig: CMOS integrirano vezje za senzor prstnih odtisov z ločljivostjo 500 dpi 167 W. Pribyi, T. Scheiter, G. Hribemig: A 500 dpi Fingerprint Sensor IC in CMOS Technology M. Mozetič, B. Praček: Interakcija vodikove plazme s korodirano površino srebra 171 M. Mozetič, B. Praček: Interaction of Hydrogen Plasma With Corroded Silver Surface M. Mozetič: Reakcije na površini katalitične sonde med plazemsko obdelavo polieter sulfona 175 M. Mozetič: Reactions on Catalytic Probe Surface During Oxygen Plasma Treatment of Polyether Sulphone APLIKACIJSKI PRISPEVKI APPLICATION ARTICLES V. Murko: Novitete iz razvoja podjetja Iskra Žarnice Elvelux 180 V. Murko: New Products from the company Iskra Žarnice Elvelux PREDSTAVLJAMO PODJETJE Z NASLOVNICE REPRESENT OF COMPANY FROM FRONT PAGE Iskra Avtoelektrika d.d. 184 Iskra Avtoelektrika d.d. POROČILA REPORTS D. Belavič, R. Ročak: XXII IMAPS - konferenca Poljske sekcije 185 D. Belavič, R. Ročak: XXII IMAPS - Poland Chapter Conference and Exhibition D. Belavič: ISHM'98 - konferenca Nemške sekcije 186 D. Belavič: Deutsche ISHM Konferenze 1998 VESTI 187 NEWS KOLEDAR PRIREDITEV 195 CALENDAR OF EVENTS MIDEM prijavnica 197 MIDEM Registration Form Slika na naslovnici: Brushless DC motor z ločenim (levo) in prigrajenim (desno) elektronskim krmiljem, nov izdelek Iskre Avtoelektrike d.d. Front page: Brushless DC motor with separate (left) and built-in (right) electronic control, new product of Iskra Avtoelektrika d.d. DRUŠTVO MIDEM IN KONFERENCA MIDEM NA INTERNETU Dragi člani društva in bralci revije ! Predstavitev društva MIDEM in predstavitev konferenc MIDEM' lahko poiščete na INTERNETU in sicer : 1. Predstavitev društva MIDEM in revije " Informacije MIDEM " na naslovu http://pollux.fer.uni-lj.si/midem/society.htm http://pollux.fer.uni-lj.si/midem/journal.htm 2. Predstavitev konference MIDEM na naslovu http://polIux.fer.uni-lj.si/midem/conf98.htm 3. Elektronsko pošto lahko pošiljate na naslov : Iztok.SorIi@guest.arnes.si Pri vpisu naslovov pazite na velike in majhne črke !! Vse člane vljudno prosimo, da poravnajo članarino za leto 1998. MIDEM SOCIETY AND MIDEM CONFERENCE ON INTERNET Dear readers and Society members ! Presentation of MIDEM Society and the information on the MIDEM Conference can be found on INTERNET as follows : 1. Presentation of MIDEM Society and Journal "Informacije MIDEM", address http://pollux.fer.uni-lj.si/midem/society.htm http://pollux.fer.uni-lj.si/midem/journal.htm 2. Presentation of the MIDEM'97 Conference, address http://pollux.fer.uni-lj.si/midem/conf98.htm 3. Email can be sent to : Iztok.Sorli@guest.arnes.si Please, use exact lower and upper case letters as indicated. We kindly ask all our members to pay the membership fee for 1998. UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana STUDY OF CORRELATION AMONG DIFFERENTIAL NONLINEARITY, NONLINEARITY AND NOISE OF THICK FILM RESISTORS P. J. Mach, Czech Technical University in Prague, Czech Republic P. M. Svasta, „Politehnica" University of Bucharest, Romania Keywords; electrical thick film resistors, nonlinearities of electrical thick film resistors, noise of electrical thick film resistors, thick resistive films, electrical conductivity of thick resistive films, l-V characteristics, current-voltage characteristics, V-l characteristics, voltage-current characteristics, linear l-V characteristics, linear current-voltage characteristics, nonlinear l-V characteristics, nonlinear voltage-current characteristics, theory of nonlinearlty, differential nonllnearlty of current-voltage characteristics, noise index, second harmonics index, third harmonics index correlation, laser trimming Abstract: Typical parameters which make assessment of reliability of thick film resistors possible are noise and nonllnearity of a current vs. voltage characteristic. Both the parameters are strongly influenced by mechanisms of conductivity of the film and by mechanisms of conductivity which appear Inside a transient area between the resistive film and its conductive contact. A study of correlation among nonlinearity, differential nonlinearlty and noise of the thick film resistors was carried out. It was investigated an influence of the form and trimming of the resistors on these parameters. It was found that the form of the resistors influences these parameters by the same way. Trimming of the resistors influences their noise substantially but Its Influence on the nonlinearity is very low. Študij povezave med diferencialno nelinearnostjo, nelinearnostjo in šumom debeloplastnih uporov Ključne besede: upori električni debeloplastnl, nellnearnosti uporov električnih debeloplastnih, šum uporov električnih debeloplastnih, plasti uporovne debele, prevodnost električna plasti debelih uporovnih, l-U karakteristike tok-napetost, U-l karakteristike napetost-tok, l-U karakteristike tok-napetost linearne, l-U karakteristike tok-napetost nelinearne, teorija nelinearnosti, nelinearnost diferencialna l-U karakteristike tok-napetost, indeks šumni, indeks harmonskih drugih, korelacija Indeksov harmonskih tretjih, doravnavanje lasersko Povzetek: Tipična parametra, s pomočjo katerih lahko ovrednotimo zanesljivost debeloplastnih uporov sta šum in nelinearnost tokovno-napetostne karakteristike. Na oba parametra močno vplivata mehanizem prevajanja znotraj filma, kakor tudi mehanizem prevajanja na prehodnem področju med uporovnim filmom in njegovim prevodnim kontaktom. Študirali smo povezavo med nelinearnostjo, diferencialno nelinearnostjo in šumom debeloplastnih uporov, kakor tudi vpliv oblike in laserskega doravnavanja na omenjene parametre. Ugotovili smo, da oblika uporov vpliva enako na vse tri parametre, med tem ko lasersko doravnavanje močno vpliva na šum In dosti manj na nelinearnost uporov. 1. INTRODUCTION When an electrical current is fed through a component a voltage drop appears across it. If the amplitude of the current will be changed and the voltage drop will change in proportion to the change of the current the component is linear. If the relationship between the current and the voltage drop is of an another type the component is nonlinear. It is evident that the denotation „the linear component" is related to the component with a linear current vs. voltage characteristic (C-V characteristic) and vice versa. There are two groups of electronic components used in electronics - the nonlinear components and the linear ones. However, the basic assumption about the linear component can be accomplished theoretically only. In practice a higher or lower difference between a straight line and the C-V characteristic of the linear component can be found. Therefore the better denomination for the linear component is „the nominally linear component". The level of nonlinearity of these components is substantially lower than the level of nonlinearlty of the nonlinear components. The measurement of nonliner-arity can be very useful for users as well as producers of electronic components because this parameter makes evaluation of production quality and assessment of reliability possible. The typical linear electronic components are resistors, some types of capacitors, air cored inductors, electrical contacts and leads. The typical nonlinear components are diodes, transistors, iron-cored inductors etc. Non-linearity of the nominally linear component is usually caused by defects inside its electrically functional part. The defects originate as a con sequence of errors in a design of the component (e.g. when materials with substantially different coefficients of thermal expansivity are put together) or errors which were made during its fabrication (e.g. when cleanliness of environment is insufficient) /1/. Thick film resistors are the components made of resistive paste applied on an insulating substrate by screen printing. That is to say that the material of the thick film is not the homogenous material because it consists of two components after firing: of a functional conductive component and of an insulating one. A better description of this structure is that one component has the low resistivity and the second one has the high resistivity /2/. Therefore different mechanisms of conductivity take part in the process of conducting of the current through the film. Some of these mechanisms are linear and some mechanisms are nonlinear. The combination of them influences the final shape of the C-V characteristic. When the linear mechanisms dominate (phonon-elec- 149 P.J. Mach, P.M. Svasta: Study of Correlation Among Differential Informacije MIDEM 28(1998)3, str. 149-153_____Nonlinearity, Nonlinearity and Noise of Thick Film Resistors tron interaction, elastic tunneling) the characteristic is slightly nonlinear only, when the nonlinear mechanisms have a more significant influence nonlinearity of the component is higher. Nonlinearity is usually investigated using a modulating technique. The component is powered by the sinusoidal current and the voltage drop of third harmonic component which arises across it is measured /3/. When the dependence between the fundamental voltage and the third harmonic voltage is a cubic one the relation would be said to be of the third degree. When this dependence is quadratic the component is qualified as the component with the second degree of non-linearity. It is a limited number of electronic components only which have this type of nonlinearity, majority of the components have the third degree of nonlinearity. Nonlinearity of the resistors is often correlated with their noise. Noise also reflects inhomogenity and defects inside the electrically functional part of the component but sources of 1/f noise probably differ from sources of nonlinearity. Nevertheless it was found a good correlation between the noise index and the nonlinearity index of thin metallic films. The results which were found for thick films are presented in this paper. electrons then, under thermal equilibrium, the current lj across the barrier in any direction will be given by equation (2) where A is a constant depending on the material of the film. It was derived that the net increase in the total current Id across the Nt barriers can be described by equation (3) where V is the voltage applied across the barriers. It is assumed that all the barriers are of the same type. Expanding of the sinh term in a series a dependence of the total current Id on the harmonic components of the applied voltage can be found. The terms of the series are given by odd powers of the applied voltage only. Therefore the C-V characteristic is symmetrical and its nonlinearity can be evaluated according to the amplitude of third harmonic component. : Aexp -eVj 12-based Thick Film Resistors: Manufacturing Condition and Percolation", Materials Science, Vol XVIII. No. 3-4, pp. 199-204, 1987 D. Stauffer, A. Aharony, „Introduction to Percolation Theory", Taylor and Francis, London, 1994 P. Mach, „Nonlinearity measurement - equipment and problems", Proc. SilTME'98, TU Bucharest, Bucharest, September 24-27, 1998, will be published Assoc. Prof. Dr. P. J. Mach Department of Electrotechnology Czech Technical University in Prague Faculty of Electrical Engineering Technicka 2, 166 27 Prague 6, Czech Republic Phone: ++420 2 2435 2122, Fax: ++420 2 2435 3949 E-mail: mach@fel.cvut.cz WWW: http://K313.feld.cvut.cz Prof. Dr. P. M. Svasta * Dep. of Electronic Technology and Reliability „Politehnica" University of Bucharest Faculty of Electronics and Telecommunication CETTI - Center for Technological Electronics and Interconnection Techniques Splaiul Independentei nr. 313, 77206 Bucharest, Romania Phone: +40 1 411 6674, Fax: +40 1 410 4488 E-mail: svasta@cadtieccp.pub.ro WWW: http://www.cadtieccp.pub.ro Prispelo (Arrived): 10.07.98 Sprejeto (Accepted): 08.09.98 153 Informacije MIDEM 28(1998)3, Ljubljana UDK621.3: (53 + 54+621 + 66), ISSN0352-9045 MEASURING THE WEIGHTED POWER OF CMOS LATCHING CIRCUITS Dusan Raic, Faculty of Electrical Engineering, Ljubljana, Slovenia Key words: microelectronics, IC, integrated circuits, integrated circuits design, weighted powers, power measuring, CMOS digital integrated circuits, latching circuits, CMOS latching circuits, low-power design, circuit optimization, performance evaluation, PDP, Power-Delay Product, EDP, Energy-Delay Product, flip-flop circuits, Hspice simulators, energy metrics Abstract: We propose a method for power consumption evaluation of CMOS latching circuits, based on the weighting of individual energy-related parameters. By assigning appropriate weighting factors for clock and data inputs the circuit evaluation can be carried out In the context of overall system performance. The clock weighting factor is defined as the power ratio of a complete clocking system and the power needed to drive the clock inputs in the circuit. This factor is found to be =1.8 for a representative CMOS technology with optimally designed clock driver. We show how power parameters of a circuit can be measured and weighted in Hspice environment to evaluate the circuit power, PDP or EDP products. Finally we present comparative results for some well-known CMOS latching circuit types. Merjenje porabe moči v pomnilnih strukturah CMOS Ključne besede: mikroelektronika, IC vezja integrirana, IC snovanje vezij integriranih, moči utežene, merjenje moči, CMOS vezja integrirana digitalna, vezja držalna, CMOS vezja držalna, snovanje za moči male, optimiranje vezij, vrednotenje zmogljivosti, PDP produkt odloga moči, EDP produkt odloga energije, vezja prevesna, Hspice simulatorji, metrika energije Povzetek: Predlagamo metodo za ocenjevanje porabe moči v pomnilnih strukturah CMOS na osnovi utežnostnih faktorjev za posamezne kategorije moči. Uporaba utežnostnih faktorjev omogoča ocenjevanje vezja v luči lastnosti celega sistema. Utežnostni faktor za vhod urinega signala definiramo kot razmerje med porabo moči celotnega podsistema za krmiljenje ure in med močjo, ki jo potrebujemo za krmiljenje vseh urinih vhodov. Za ta faktor ugotovimo, da se giblje okoli vrednosti =1.8 za tipično tehnologijo CMOS ob uporabi optimalno načrtovanega ojačevalnika signala ure. Prikazano tehniko za mejenje utežene povprečne porabe moči, faktorjev PDP in EDP ilustriramo s stavki iz simulatorja Hspice. Na koncu prilagamo izmerjene vrednosti za nekatere znane izvedbe pomnilnih celic CMOS. 1. Introduction Memory cells determine to a large degree the speed and power characteristics of digital systems. They represent relatively large cells that are repeated many times, consume considerable power because of the clock activity and underlay the key architectural decisions of the system (clock distribution scheme, static/dynamic operation, pipelining, etc). In order to improve the speed and power consumption, a number of CMOS latching schemes have been analyzed and new design concepts have been proposed in recent years. However, it is frequently very difficult to compare the design efficiency of these solutions because we are faced with many parameters that are hard to match. The ultimate common cost function is therefore the energy that must be spent to complete the desired function within the limits of the design specification. Similar problems as with the circuit comparison are addressed also when we try to formulate the cost function for the circuit optimization. In fact the technique described in this work can be applied in both cases. When authors compare new concepts with known circuit techniques they rely either on hand calculated data /5/,/10/ or on a number of simulation runs with different data statistics /12/,/13/. The details of power measuring technique are frequently not well documented so that it is hard to reproduce the results. In the present work we try to provide a realistic measure of circuit performance based on power consumption that is weighted against the overall system performance. Different circuit techniques can be therefore compared with a single quantitative measurement. The first section presents a short discussion of performance measurement based on power-delay and energy-delay products. We proceed with individual power consumption components of CMOS circuits and show how they can be monitored in the Hspice environment. Further attention is devoted to the power that is needed to drive the clock and data inputs of the latching cells. We calculate the clock weighting factor and show how the data activity can be used as the data weighting factor to reflect signal statistics on the cell performance evaluation. In the last section we show the implementation details and Hspice code fragments to calculate the weighted PDP, EDP or average circuit power. Experimental results are presented for some representative latching circuits. 2. The energy metrics The most common quality measure for logic gates is the energy consumed by the circuit per switching event, usually called the Power-Delay Product (PDP). PDP = Esw = |*T P(t)dt (1) Jo where T is the duration of the switching event. The switching event is normally defined as the low to high and high to low signal transition /1/,/6/. In an ideal CMOS gate without the second order effects this is equal to 154 D. Raic: Measuring the Weighted Power of CMOS Latching Circuits_____ Informacije M1DEM 28(1998)3, str. 154-161^ The problem with PDP is that it ignores the actual speed of the circuit. For a circuit designer the name "power delay" can be misleading as the calculated value has no direct relation with the signal delay in the circuit; it is related to the duration of the switching event that sets the limits for the power integration interval. Faster circuits have larger PDP values because of higher currents involved in the switching process. PDP can be used as an adequate circuit quality measure only under the assumption that the duration of the switching event fits the design specifications and needs no modification. A typical example are ring oscillators that are usually used to measure the PDP and speed limits of a logic family. To combine the circuit speed (or the "specification" speed), characterized by the signal delay TSpec with the power consumption it is better to use the Energy-Delay Product (EDP) as proposed in /4/: FDP = energy . signal delay = pr)p _ j switching event switching event spec (3) The EDP definition implies that the quality of a circuit with PDPa and TSa is equal to the quality of another circuit with PDPb = PDPa/k and TSb = k TSa . In general the speed/power tradings are not linear /11/ and allow various circuit techniques with different topologies to compete for the best solution of a design case. As the power supply voltage is constant the switching energy can be calculated from the total charge flow on the power supply: PDP = fp(t)dt = Vdd iTidd(t)dt = Vdd • Qsw (4) Jo Jo The procedure to measure QSw in Hspice simulation is presented in section 3. TSpec is measured with standard .measure statements for delays relevant to the circuit application. The power consumption of a CMOS cell can be considered as the sum of 3 components: • Internal power (node parasitics and DC currents) 8 Power to drive the inputs 8 Power to drive the loads Of course it is possible to measure the sum of all power components from the power supply current Idd, but this would eliminate the possibility to weight individual components. As we deal with the latching circuits it is of particular interest to weight separately the power consumed by the clock and data inputs. This can be realised by monitoring some internal circuit currents. In Hspice the easiest way to do this is to insert dummy voltage sources of 0 DC volts in the circuit branches, as presented on fig 1. All loads associated with core CMOS logic cells are combinations of some capacitive parasitics to Vdd (Cp) and to Vss (Cn). Figure 2 shows currents associated with the switching event on such loads. As we see from table 1, the power to drive the load can be calculated either from the power supply current Idd or from the measurement of the driver current Id. The third possibility is to measure the input current to the load lj. Because of opposite polarization the average value of this current is 0. However in Hspice it is also possible to measure the absolute current value which would in our case provide the double of the load power. As the capacitive load power is reflected on the cell supply current, the measurement of the Ipp on fig. 1 provides the sum of internal cell power and the load Fig. 1: Dummy voltage sources (Vdr,Vpp,Vin,Vld) to determine individual cell power components Table 1: Power consumption on the CMOS load Power monitor 0 1 1 0 £ — Idd ■ Vdd V'L ■ Cn v;td-cP Vd2d-(Cn + Cp) Id ■ Vdd Vd2d ■ (Cn + CP) .0 V'id ■ (Cn + Cp) Ii ■ Vdd Vjd-(Cn+Cp) 0 \h\ ■ vdd "WT^Tcg" V'dd ' (Cn + Cp) 2 ■ V&-(Cn + Cp) Informacije MIDEM 28(1998)3, str. 154-161 D. Raic: Measuring the Weighted Power of CMOS Latching Circuits 1 -> 0 load IP Vi In V Cp Cn 0-> 1 Q Vdd Vdd Ii = -Ip -In Id = 0 Idd = -Ip Fig. 2: Current flows on the capacitive CMOS load Ii = Ip + In Id = Ip + In Idd = -In power. The measurement of lid is therefore not needed. To calculate the last component, the cell input power, we have two possibilities. One is to build an ideal driving stage without parasitics that consumes no internal power and calculate the power from the supply current Idr of that driver. The second possibility is to measure the absolute value of the input current hn and divide the result by 2. This is more convenient as we don't need extra device models and can use real circuit components to simulate the switching event. 3. Weighting the input power components If we look at the circuit performance from the system level we must recognize that the price for various power components is not the same. To reflect the global system features such as the clock distribution scheme and signal statistics, the circuit evaluation function must be formulated with weighted input power components. To drive clock inputs in synchronous systems we normally need special drivers and distribution networks. The system power penalty is therefore much higher than the sum of locally measured powers on clock inputs. To make an assessment of the real power requirement, consider the internal capacitance of an optimally designed N-stage buffer /6/ with tapering factor u and input capacitance Q. The equivalent gate capacitance is Cdiff - S • Cgate (6) The optimal buffer drives the load Cl = uN • Ci and dissipates the power ^driver ~ ('-'gate + 'ciiff + ^L^dd ' f : (7) f uN 1 ^ (1 + 5)-—— + uN 1 ' u-1 j CrVdVf The measured input power in the cells would be that of the driver load Cl: Pceils=CL-Vd2d.f = uN-CrVd2d-f (8) JL ' vdd "dd The ratio between the total power consumed by the driver and the measured power on clock inputs is defined as the clock weighting factor wc: wr driver cells (Un-1)-(1 + 5) O) :1+ UN-(U-1) C gate = C: + U-C: + C:+... + UN""1 O =Ci uN-1 U-1 (5) We get a similar expression for internal diffusion capacitance if we assume that it is equal to 8 times the buffer input gate capacitance: The number of stages N depends on the size of the clock network, while the optimum tapering factor u is known to be close to 3.6 for technologies with 8= 1, For N values greater than 2 we can then assume uN-1 =uN so that the clock weighting factor can be approximated to w„ 1 + 8 u-1 (10) 161 D. Raic: Measuring the Weighted Power of CMOS Latching Circuits ___ Informacije MIDEM 28(1998)3, str. 154-161 With 8=1 and u=3.6the calculated value is 1,77. It must be pointed out that this is still an optimistic value since we have neglected the wiring capacitance of the clock distribution network and have assumed an optimal driver. Another important issue regarding the input power measurements is the signal activity alpha, defined as the number of complete signal transitions per clock cycle /13/. The clock itself has a = 1 while other input signals may have various values depending on the nature of the system. If the signal comes from another latching cell the maximum value for alpha is 0.5, as data can change once per clock cycle at maximum. A random binary signal has a = 0.25. The natural tendency of binary coded signals is that high-order bits have lower activity than low-order bits. The signal activity depends also on the circuit structure: adder outputs for example may have activities much higher than 1 due to the carry propagation transients. If such signals are connected to the data inputs of flip-flops the toggle power can significantly influence the selection of the optimal latching circuit. When we simulate the switching event of a latching circuit we apply one logic-high and one logic-low state to the data input for each clock cycle. The signal activity of the data input in the switching event is therefore equal to 0.5 (fig. 5). To compensate for the difference between the real application and the calculated input power in the switching event we define the data input weighting factor as The total charge flow across Vp during the switching event with the duration Tfin is given by wr ® system 0.5 (11] It is obvious that some statistical properties of system signals must be known to make better cell comparisons. If this is notthe case, wd = 1 can be used as worst case input power weight. Methods to determine alpha are described elsewhere /14/, /15/. 4. The power measurement technique According to (4) the measurement of average power consumption of CMOS cells with constant power supply is reduced to the calculation of the equivalent charge flow across the power supply. Figure 3 shows the necessary setup to automate this procedure in the Hspice simulator /6/. We insert a new node qt, connected to the measuring capacitor Cq and the current-controlled current source Fp. The later is controlled by the cell supply current (measured on Vp), multiplied by the gain factor. Cq and the gain must be set to scale the voltage Vqt on node qt in the range of reasonable values for the given circuit type, otherwise we will experience difficulties with the numeric precision of the simulator. A good choice is to use gain = 1 and Cq = 1fF, which scale Vqt to 1V for 1fC of charge flow. The voltage on node qt is then given by Qsw = J "n !P (t)^ = QP (Tfin) = —~~ Vpt (T, •'U Od II fin > (13) As we see, QSw can be measured as the voltage on node qt at the end of the switching event, multiplied by the scaling factor Cq/gain. One should not forget to set the initial condition on Cq to 0 volts. The whole procedure requires four Hspice statements: Cq qt 0 1f Fp 0 qt Vp 'gain' . ic V(qt) 0 .measure tran Qsw max V(qt) from=0 to='Tfin' Cq .ic v(qt) 0 Cq qt 0 If Fp 0 qt Vp I .measure tran Qsw max v(qt) from=0 to—'Tfin' Fig. 3: Measuring the equivalent charge flow of a cell Similar technique can be used to measure the charge flow into the cell inputs. In this case we use the absolute value of the controlling current and divide the result by 2. This is done inside the 'F' statement by the 'abs = 1' modifier and the gain multiplied by 0.5. Referring to fig. 4, the charge flow for input CI would be modeled by Fc 0 qt Vc 'gain*0.5*wc' abs = 1 Vqi(t) = -lf,gain.|p(t)dt = ^Qp(t) (12) We can get the sum of all charge flows needed for the cell power calculation if we connect the relevant current-controlled current sources to the same node qt. 161 D. Raič: Measuring the Weighted Power of Informacije MIDEM 28(1998)3, str. 154-161_________CMOS Latching Circuits When weighting is required, the gain in the F statement is multiplied by the corresponding weighting factor. In this case the voltage Vqt(T) represents the total weighted switching charge of the circuit. Once QSw is known, one can calculate the average power, PDP or EDP from (3) and (4). The appropriate 'measure' statements would be .measure PWR param = 'Vdd*Qsw*Sq/Tfin' .measure PDP param = 'Vdd*Qsw*Sq' .measure EDP param = 'Vdd*Qsw*Sq*Tspec' The power measuring setup for a CMOS latching cell is presented on fig. 4. Typical waveforms for signals, power supply current and the simulated switching charge are presented on figure 5. Table 2 presents illustrative results for four different M/S static flip-flop structures: the switched inverter or C2MOS type /6/ (fig. 6 a), the pass-transistor type /17/ (fig. 6 b), a variant of the RAM cell type /18/ (fig. 6 c) and the SSTC type/10/ (fig. 6 d). All circuits were simulated with minimum size transistors. To make meaningful evaluations some optimization should be done in the given system environment. This is especially true for the flip-flop types c and d. With gain = 1 the scaling factor Sq is equal to Cq.Tfin is the duration of the simulated switching event. If EDP is needed, Tspec is measured within the same simulation as some combination of the delay times of interest. /r\ r <.\ rrv V'y vy vV V- ic v(c|i) 0 Cc| ql o II Fp 0 C]I Vp I Fd 0 ¡11 V(l v.il ■(! V Fc 0 C]I Vc 'wu*0.S' abs=l .measure [ran Qsw max v(qt) from=0 to="ITm' Table 2: Simulation results for some minimum-size static flip-flops technology w/l Vdd T Ci Wc Wei Conditions: 0.6u, typical mean 0.8u/0.6u, all transistors 3V 25 deg. C 8fF 1.77 measured value type A type B type C type [ avg. power 12.46E-6 W 10.62E-6 W 1 j 4öb bW 12 1UE 6 W PDP 1.12E-12 J 0.95E-12 J 1.39E-12J 3.79E-1 EDP 1.40E-21 Js 0.93E-21 Js 1.85E-21 Js 3.50E-21 Js Fig. 4: Measuring the weighted charge flow of a CMOS latch To characterize flip-flops and latches the switching experiment must assure both high -h> low and low -> high transitions of the outputs which means that input signals must exhibit certain waveforms. We recommend the following rules for the switching experiment: • The circuit must be loaded with realistic loads. 6 The initial states of all nodes must be equal to the final states so that an even number of state transitions is involved. • Inputs must be driven by realistic drivers. In spite of the fact that input currents are measured and contribute to the total power consumption the rise and fall times of ideal voltage sources do not change with loading. Measurement of circuit speed based on heavy loading of ideal input signals would give unrealistic EDP values because of the nonlinear speed-power trading in real drivers. 5. Conclusion The proposed evaluation technique for CMOS latching circuits is based on weighting the energy-related parameters. Special attention has been paid to the power required to drive the clock and data inputs. We show how this power can be weighted to reflect the overall system performance on the circuit under investigation. Fragments of Hspice code illustrate the calculation of average power, PDP and EDP products. All these parameters are measured by the simulation of a simple switching event along with the delay times. The technique can be used to simplify the optimization process and to Improve the comparison of different latching schemes. 158 D. Raic: Measuring the Weighted Power of CMOS Latching Circuits _ Informacije MIDEM 28(1998)3, str. 154-161 Wave Symbol! D0:A0:v(cp) o > 3 2.5 2 1.5 1 500m 0 IBiwiil^M 10n 20n 30n 40n 50n 60n Time (lin) (TIME) 70n 80n 90n _Wave.........Symboij DO:AO:v(d) mmmzmmmmm Wave Symbol DO AO v(q) Wave Symbol: D0:A0:i(vp) O 2.5 2 1.5 1 500m 0 100u 50u ssgssGBsss&assseSi 30n 40n 50n 60n Time (lin) (TIME) Panel 3 70n 80n 90n 10n 20n 30n 40n 50n Time (lin) (TIME) 60n 70n 80n 90n 30n 40n 50n Time (lin) (TIME) Wave______symbol! D0:A0:v(qtj a> o> iS o > 300 250 200 150 100 —i--------------, | 10n 20n 30n 40n 50n 60n 70n 80n 90n Time (lin) (TIME) Fig. 5: Simulation of the weighted charge flow in a CMOS latch 161 Informacije MIDEM 28(1998)3, str. 154-161 D. Raic: Measuring the Weighted Power of CMOS Latching Circuits elk vijdd Fig. 6: Examples of CMOS static MS flip-flops 161 D. Raic: Measuring the Weighted Power of CMOS Latching Circuits _ _ Informacije MIDEM 28(1998)3, str. 154-161 6. References /1/ M. Shoji, CMOS Digital Circuit Technology, Prentice Hall, Englewood Cliffs, NJ, 1988, ISBN 0-13-138850-9 025. /2/ J. Yuan and C.Svensson, "High-speed CMOS circuit technique," IEEE Journal of Solid State Circuits, vol. 24,1989, pp. 62-70. /31 N. Westeand K. Eshraghian, Principles of CMOS VLSI design: a systems perspective, Addison-Wesley, 1993, ISBN 0-201-53376-6 /4/ M.Horowitz, T.lndermaur, and R.Gonzales, "Low-Power Digital Design," in IEEE Symp. Low Power Electr., Oct. 1994, pp. 8-11. /51 P.Day and J.V.Woods, "Investigation into Micropipeline Latch Design Styles," IEEE Transactions on VLSI Systems, vol. 3, 1995, pp. 264-272. /6/ N. M. Rabaey, Digital Integrated Circuits: a design perspective. Prentice-Hall, Englewood Cliffs, NJ, 1996, ISBN 0-201-53376-6 /7/ CTretz and C.Zukowski, "CMOS Transistor Sizing for Minimization of Energy-Delay Product," Proceedings, Sixth Great Lakes Symposium on VLSI, 1996, pp. 168-173. /8/ R.Gonzales and M.Horowitz, "Energy Dissipation In General Purpose Microprocessors," IEEE Journal of Solid State Circuits, vol. 31, 1996, pp. 1277-1283. 191 M.Afghahi, "A Robust Single Phase Clocking for Low Power, High-Speed VLSI Applications," IEEE Journal of Solid State Circuits, vol. 31, 1996, pp. 247-253. /10/ J. Yuan and C.Svensson, "New single-clock CMOS latches and fiipfiops with improved speed and power savings," IEEE Journal of Solid State Circuits, vol. 32, 1997, pp. 62-69. /11/ R.Gonzales, B.Gordon, and M.Horowitz, "Supply and Threshold Voltage Scaling for Low Power CMOS," IEEE Journal of Solid State Circuits, vol. 32, 1997, pp. 1210-1216. /12/ R. Zimmermann and W.Fichtner, "Low-power logic styles: CMOS versus pass-transistor logic," IEEE Journal of Solid State Circuits, vol. 32, 1997, pp. 1079-1090. /13/ C.Svensson and J.Yuan, "Latches and Flip-flops for Low Power Systems," in Low-Power CMOS Design, A.Chan-drakasan and R.Brodersen, editors, IEEE Press, 1998, ISBN 0-7803-3429-9. /14/ /15/ /16/ /17/ /18/ C.Tretz and C.Zukowski, "Conservative Modeling of the Contribution of Spurious Transitions to Power Dissipation in Digital CMOS VLSI Circuits," Proceedings, IEEE 39.th Midwest Symposium on Circuits and Systems, 1997, pp. 317-320. D.Singh, J.Rabaey, M.Pedram, F.Catthoor, S.Rajgopal, N.Sehgal, T.Mozdzen, "Power Conscious CAD Tools and Methodologies: A Perspective," Proceedings of the IEEE, vol. 83, 1995, pp. 570-592. Meta-Software, HSPICE User's Manual, Meta-Software, Inc., Campbell, CA, 1992 G.Gerosa, et. al., "A 2.2 W, 80 MHz Superscalar RISC Microprocessor," IEEE Journal of Solid State Circuits, vol. 29, no. 12, December 1994, pp. 1440-1452. L.Wai, et. al., "A 1-V Programmable DSP for Wireless Communications," IEEE Journal of Solid State Circuits, vol. 32, no. 11, November 1997, pp. 1766-1774. Dr. Dušan Raič, dipl. ing. University Ljubljana, Faculty of Electrical Engineering 1000 Ljubljana, Tržaška 25 Slovenia email: dusan. raic @fe. uni-lj. si Prispelo (Arrived): 06.10.1998 Sprejeto (Accepted): 12.10.1998 161 Informacije MIDEM 28(1998)3, Ljubljana UDK621,3:(53 + 54+621 +66), ISSN0352-9045 AUDIO VMESNIK ZA RAČUNALNIK Vladimir Jenuš, Bogomir Horvat Univerza v Mariboru, Fakulteta za elektrotehniko, računalništvo in informatiko Ključne besede: računalništvo, vmesniki avdio, vmesniki avdio računalniški, mikroprocesorji, SCSI vmesniki sistemski inteligentni računalnikov, SCSI vodila, A-D pretvorniki analogno-digitalni, D-A pretvorniki, zajemanja podatkov, sheme blokovne Povzetek: Področji računalništva in telekomunikacij se vse bolj povezujeta. Telekomunikacijsko omrežje je vse bolj "inteligentno" in omogoča vedno večje hitrosti prenosa informacij. Tako je omogočen prenos jezikovnih, slikovnih in podatkovnih informacij. Računalniki sami pa postajajo vse bolj "naprave za izmenjavo informacij". Na pohodu je večpredstavnost. Za potrebe študija večpredstavnosti in jezikovnih komunikacij je bil razvit audio vmesnik za delovno postajo. V pričujočem članku so predstavljene glavne značilnosti tega audio vmesnika. Audio Interface for Workstation Key words: computer science, audio interfaces, audio interfaces for computers, microprocessors, SCSI interfaces, Smart Computer System Interfaces, SCSI busses, A/D converters, analog-digital converters, D/A converters, digital/analog converters, data acquisition, block diagrams Extended abstract: Telecommunications and computing are more and more tied together. Telecommunication network offers higher and higher bit rates and behaves more and more intelligent. It is possible to transmit speech, movies and data over this network. Computers are more and more machines for exchange of information. The era of multimedia is begun. For studying multimedia and speech communication we develop an audio interface for workstation. It can be used for acquisition, conversion and reproduction of speech information, as a measurement system, etc. In the process of designing we had to decide which material is the most suitable for audio interface. The most important decision was the decision which microprocessor to use. After detailed analysis we chose a 32 bit processor MC68020. After that we choose type and quantity of Memory. For storing basic program we chose to use 256kB of EPROM memory and for storing data 1-3MB of fast (35ns) SRAM. Next vital components were A/D and D/A converters. Because we wanted to design the audio interface, which can offer acquisition and distribution of audio signal in CD quality, we chose AD676 for A/D and AD669 for D/A converter. In the process of designing we divided the audio interface into some basic blocks: 8 Microprocessor with bus interface and clock generation 8 SCSI bus interface 0 RAM - EPROM 8 Frequency generator • A/D converter 8 D/A converter 8 Analog input interface 8 Analog output interface and 8 Power supply unit Functionality of these blocks can be described through four basic phases: 8 Power up sequence 8 Communication with initiator trough SCSI bus 8 Acquisition of analog signals and conversion into digital data 8 Conversion from digital data into analog form and distribution of analog signal After extensive testing, we realize that audio interface fulfill all our expectations. We use it very well in our research of speech recognition. 1. Uvod Računalniška in telekomunikacijska tehnologija in tehnika sta doživeli izreden razvoj. Prvim 4 in 8 bitnim enoprocesorskim računalnikom so sledili 16 in 32 bitni računalniki z enim ali več mikroprocesorji, vse do 64 in 128 bitnih večopravilnih računalniških sistemov z velikimi procesnimi močmi. Raziskovalci pa poizkušajo računalnikom še bolj povečati procesno moč in zmanjšati dimenzije, ter porabo energije. Vzporedno z računalniki se hitro razvijajo tudi telekomunikacije. Namesto analognih visokofrekvenčnih sistemov s frekvenčnim multipleksom se vse bolj uporabljajo digitalni sistemi s časovnim multipleksom in skoraj neomejenimi prenosnimi zmogljivostmi. Viden je napredek tudi pri načrtovanju in uporabi prenosnih poti, saj kable z bakrenimi vodniki kot prenosne medije polagoma nadomeščajo kabli s steklenimi optičnimi vlakni. Vse to pogojuje uporabo novih linijskih sistemov in razvoj novih tehnologij ter storitev. Obe področji se med seboj vedno bolj prepletata. Računalniki niso več samo naprave za računanje, ampak tudi naprave za izmenjavo informacij. Z večanjem prenosnih zmogljivosti računalniških omrežij se spreminjajo tudi informacije, ki se prenašajo med računalniki. Najprej so se prenašale samo tekstovne informacije, sedaj pa se prenašata tudi zvok in slika. Pri prenašanju in računalniški obdelavi teh informacij pa je potrebno te informacije pretvoriti v primerno obliko. 162 V. Jenuš, B. Hqrvat: Audio vmesnik za računalnik Informacije MIDEM 28(1998)3, str. 162-166 Zato potrebujemo dodatne vmesnike za zajemanje in pretvorbo podatkov v ustrezno obliko. V pričujočem članku predstavljava značilnosti in osnovno zgradbo audio vmesnika, ki omogoča zajemanje in distribucijo audio signala v večpredstavnost-nih sistemih. 2. Značilnosti audio vmesnika Audio vmesnik omogoča zajemanje analognih signalov in pretvorbo v digitalne ter pretvorbo digitalnih signalov v analogne. Zaradi tega ga lahko uporabljamo za zajemanje, pretvorbo in reprodukcijo govornih informacij, kot merilni sistem, kot vmesnik za krmiljenje naprav z analognim signalom itn. Vmesnik je načrtovan tako, da so uporabljeni elementi večinoma dostopni v prosti prodaji. Priključimo ga na vodilo SCSI, Zasnova audio vmesnika omogoča: • 16 bitno A/D pretvorbo na dveh ločenih kanalih 9 16 bitno D/A pretvorbo na dveh ločenih kanalih • izbiro vzorčevalne frekvence 8, 14.7, 16, 25, 29.4, 32, 44.1 in 48 kHz 9 krmiljenje preko vodila SCSI - II (kabel A) 9 dva dvokanalna audio vhoda: - mikrofon občutljivosti 3mV / 68 kiž - kasetofon občutljivosti 300mV / 10kO 0 dva dvokanalna audio izhoda: - zvočnik izhodnega nivoja 1 Vpp /10 kO - kasetofon izhodnega nivoja 300mV /10kfi 0 prenos podatkov po vodilu SCSI zlog za zlogom ali blok naenkrat 8 programsko izbiro vhodov in izhodov 8 nastavitev naslova audio vmesnika z mostički 8 priključitev 1 - 3 MB pomnilnika RAM Vmesnik je skupaj z napajalnikom vgrajen v ohišje dimenzij 500 x 250 x 90 mm. 2.1 Splošen opis delovanja Audio vmesnik je relativno kompleksna naprava. Njegovo delovanje lahko najboljše ponazorimo z opisom po fazah: 2.1.1 Vklop Ob vklopu audio vmesnika drži vezje za reset linijo RESET* na nivoju logične ničle okoli 0.5s. V tem času se stabilizirajo napajalne napetosti, pri A/D pretvorniku pa se izvede cikel interne kalibracije. Ko linija RESET* preide na logično enico, mikroprocesor naloži v kazalec sklada vrednost na naslovu $0, v programski števec pa vrednost na naslovu $4. Fizično stati vrednosti zapisani v pomnilniku EPROM. S tem začne mikroprocesor izvajati program. Najprej ugotovi koliko pomnilnika RAM je prisotnega in ga preveri. Nato prepiše program iz pomnilnika EPROM v pomnilnik in nadaljuje z izvajanjem programa v pomnilniku RAM. 2.1.2 Komunikacija preko vodila SCSI Po zagonu mikroprocesor preverja kontrolne linije vodila SCSI vse dokler iniciator ne izbere audio vmes- nika. Ko je audio vmesnik izbran, mikroprocesor zaseže vodilo SCSI, poskrbi za izmenjavo sporočil in prebere ukaze. Ukazi, ki jih lahko izvrši audio vmesnik so: • prenos podatkov z audio vmesnika byte za bytom (A/D pretvorba) • prenos podatkov z audio vmesnika blok za blokom (A/D pretvorba) • prenos podatkov v audio vmesnik byte za bytom (D/A pretvorba) • prenos podatkov v audio vmesnik blok za blokom (D/A pretvorba) Mikroprocesor na audio vmesniku kontrolira podatkovni del in kontrolni del vodila SCSI. Po fazi selekcije in prenosu sporočil ter ukazov se navadno začne faza prenosa podatkov preko vodila SCSI, Cikel prenosa podatkov preko vodila SCSI je za mikroprocesor audio vmesnika praktično enak kot pomnilniški cikel. Po prenosu vseh podatkov preko vodila SCSI, mikroprocesor pošlje sporočilo iniciatorju, daje delo končal, in sprosti vodilo. Mikroprocesor nato spet ciklično preverja kontrolne linije vodila SCSI. 2.1.3 Zajemanje podatkov in pretvorba v digitalno obliko Po sprejemu ukaza za začetek zajemanja analognih podatkov mikroprocesor najprej vpiše ustrezne vrednosti v statusni register A/D in D/A pretvornika. Z biti DO - D3 je določena frekvenca vzorčenja, z bitom D5 se izbere zajemanje podatkov z vhoda za mikrofon ali vhoda za kasetofon, bit D7 pa omogoči A/D pretvorbo. Generator frekvenc začne v skladu z vrednostmi v DO -D3 generirati pravokotni signal želene frekvence. Ta signal vodimo na A/D pretvornik. Ob postavljenem bitu D7 za začetek pretvorbe in signalu iz generatorja frekvenc se A/D pretvorba začne. Po končani pretvorbi A/D pretvornik postavi izhod BUSY* na nizek nivo. Ko je končana pretvorba na obeh kanalih, kontrolno vezje A/D pretvornika s postavitvijo linije IPL1* na nizek nivo prekine mikroprocesor. Le-ta izvede avtovektorizirani cikel potrditve prekinitve in izvede prekinitveno rutino. V njej mikroprocesor iz A/D sklopa prebere digitalno vrednost signala z audio vhoda in jo zapiše v pomnilnik ali prenese preko vodila SCSI do iniciatorja. Po končani prekinitveni rutini mikroprocesor nadaljuje z "opazovanjem" kontrolnih linij vodila SCSI. Pri novem impulzu iz generatorja frekvenc se postopek ponovi. Ko se prebere želeno število digitalnih vrednosti z A/D pretvornika, se postavi bit D7 v statusnem registru A/D in D/A pretvornika na 0, s čemer se A/D pretvorba konča. 2.1.4 Pretvorba digitalnih vrednosti v analogne Mikroprocesor na začetku D/A pretvorbe v statusnem registru A/D in D/A pretvornika izbere frekvenco D/A pretvorbe (biti DO - D3), izhod iz D/A pretvornika (bit D4 omogoča izbiro med izhodom na močnostni ojačevalnik ali kasetofon) in omogoči D/A pretvorbo (bit D6). Signal iz generatorja frekvenc aktivira linijo IPL1* in s tem prekine mikroprocesor. Le-ta izvede avtovektorizi- 163 Informacije MIDEM 28(1998)3, str. 162-166 V. Jenuš, B. Horvat: Audio vmesnik za računalnik rani cikel potrditve prekinitve in izvede prekinitveno rutino. V njej najprej vpiše v A/D pretvornik digitalno vrednost, ki naj bi se pretvorila v analogno, nato pa omogoči D/A pretvorbo. Po izvedeni prekinitveni rutini mikroprocesor "servisira" vodilo SCSI in čaka na novo prekinitev. 2.1.5 A/D in D/A pretvorba hkrati Vezje audio vmesnika je zasnovano tako, daje mogoča "hkratna" pretvorba analognega signala v digitalni in digitalnega signala v analognega. V tem primeru je pomnilnik, rezerviran za podatke, razdeljen na del za vhodne in del za izhodne podatke. Iniciator preko vodila SCSI pošlje izhodne podatke v izhodni del pomnilnika, bere pa jih iz vhodnega dela pomnilnika. Mikroprocesor na audio vmesniku v skladu z zahtevo iniciatorja v statusnem registru A/D in D/A pretvornika izbere frekvenco pretvorbe (biti DO - D3), ki je enaka tako za A/D kot za D/A pretvorbo in s postavitvijo bitov D6 in D7 omogoči obe pretvorbi. Signal iz generatorja frekvenc začne A/D pretvorbo. Po izvedbi te pretvorbe vezje A/D pretvornika prekine mikroprocesor s postavitvijo linije IPL1 * na nizek nivo. Mikroprocesor na enak način kot v točki 4.1.3 izvede avtovektorizirani cikel potrditve prekinitve in izvede prekinitveno rutino. V prekinitveni rutini prebere vrednost iz A/D pretvornika in jo shrani v vhodni del pomnilnika. Iz izhodnega dela pomnilnika nato prenese digitalno vrednost v D/A pretvornik in omogoči D/A pretvorbo. Po končani prekinitveni rutini mikroprocesor nadaljuje s pošiljanjem podatkov po vodilu SCSI. Ob prihodu vsakega impulza iz signalnega generatorja se postopek ponovi. A/D in D/A pretvorba se končata, ko mikroprocesor postavi bita D6 in D7 na nizek nivo. 3. Izbira elementov Pri načrtovanju audio vmesnika je bilo potrebno izbrati elemente mikroračunalniškega sistema tako, da je delovanje vmesnika optimalno. Poglejmo, po katerem kriteriju so izbrane posamezne komponente. 3.1 Mikroprocesor Mikroprocesor je srce audio vmesnika. Skrbeti mora za pravočasno A/D pretvorbo, pravočasno zajemanje podatkov z A/D pretvornika, pravočasno pisanje podatkov v D/A pretvornik in komunikacijo po vodilu SCSI. Hitrost zajemanja podatkov z A/D pretvornika kot tudi hitrost pisanja podatkov v D/A pretvornik mora biti po specifikaciji najmanj 48 kS/s. Ker je pri audio vmesniku pretvorba 16 bitna, je potrebno vsakih 20.8 ¡as prenesti do D/A ali A/D pretvornika 4 zloge ali eno dolgo besedo. Po odločitvi, da se v audio vmesniku uporabi Motorolin mikroprocesor, se je potrebno odločiti samo še glede tipa mikroprocesorja. Na razpolago so 16 bitni (MC68000) in 32 bitni mikroprocesorji (MC68020, MC68030). Da bi lahko v oba A/D ali D/A pretvornika hkrati vpisali ali prebrali informacijo, se je potrebno odločiti za 32 bitni mikroprocesor. V poštev prideta torej mikroprocesorja MC68020 in MC68030. Mikroprocesor komunicira z drugimi napravami preko zunanjega vodila. Hitrost komunikacije se lahko meri v urinih ciklih. Oba 32 bitna mikroprocesorja potrebujeta za en cikel na vodilu 3 urine cikle. Za izvršitev enega ukaza branja ali pisanja v periferno napravo pri indi-rektnem naslavljanju pa je potrebnih najmanj 6 in največ 8 urinih ciklov, kar je odvisno od zadetka v interni cache pomnilnik mikroprocesorja oz. od internega načina izvajanja ukazov. V času A/D ali D/A pretvorbe naj bi bil mikroprocesor sposoben izvesti čim več ukazov, saj mora poleg kontrole pretvorbe komunicirati še preko vodila SCSI. Zaradi tega smo se odločili za uporabo mikroprocesorja MC68020, ki deluje z največjo frekvenco ure 25 MHz, medtem ko je ta pri MC68030 "le" 20 MHz. Izvršitev enega ukaza branja ali pisanja na zunanjo vodilo pri MC68020 traja lahko pri indirektnem naslavljanju največ 320 ns. V času ene pretvorbe lahko mikroprocesor izvrši 65 takih instrukcij. Ker pa ne gre samo za izvršitev instrukcij READ ali WRITE z indi-rektnim načinom naslavljanja, ampak tudi drugih, je število instrukcij med dvema pretvorbama še manjše. Izkaže se, da zadošča, če se lahko med dvema A/D in D/A pretvorbama izvrši najmanj 50 instrukcij. Torej MC68020 temu kriteriju "komaj" zadošča. 3.2 Pomnilnik Pri izbiri pomnilnika se je bilo potrebno odločiti predvsem za tip, hitrost in kapaciteto. V mikroprocesorskem sistemu sta potrebna dva tipa pomnilnika: pomnilnik v katerem je spravljen program in ostane v njem tudi ob izklopu napajanja in pomnilnik namenjen shranjevanju podatkov med delovanjem. Pri pomnilnikih iz prve skupine je mogoče izbirati med pomnilniki ROM, EPROM, EEPROM, FLASH RAM... Pomnilnika iz druge skupine pa sta statični in dinamični pomnilnik RAM. Zaradi relativne cenenosti in razširjenosti za pomnilnik, v katerem bo shranjena programska koda, izberemo pomnilnik tipa EPROM. Dinamični pomnilniki tipa RAM so enostavnejši, imajo večje kapacitete, vendar potrebujejo osveževanje, kar pomeni dodatna vezja in v določenih pogojih tudi dodatne zakasnitve pri prenosu informacij. Statični pomnilniki so glede izdelave zahtevnejši, porabijo malce več energije, so hitrejši in ne potrebujejo dodatnih vezij za osveževanje. Predvsem zaradi hitrosti in enostavnejše izvedbe drugih vezij smo se odločili za uporabo statičnih vezij RAM. 3.2.1 Pomnilnik EPROM Pri izbiri pomnilnika igra pomembno vlogo hitrost. Dostopni časi za pomnilnike tipa EPROM so relativno veliki. Gibljejo se od 120 ns do 300 ns. To je za naš primer preveč, saj mikroprocesor v najslabšem primeru zahteva dostopni čas 45 ns. Zato se v vezju audio vmesnika ob vklopu napajanja ves program iz pomnilnika EPROM prepiše v hitrejši pomnilnik RAM in se izvaja tam. Torej nam hitrost pomnilnika EPROM določa samo še čas, potreben od vklopa vmesnika do njegove pripravljenosti za delovanje, na samo delovanje pa nima vpliva. Izbrali smo pomnilnik organizacije 64k x8 in oznako 27512. Zaradi 32 bitne širine podatkovnega vodila potrebujemo 4 integrirana vezja 27512. To nam da 164 V. Jenuš, B. Horvat: Audio vmesnik za računalnik Informacije MIDEM 28(1998)3, str. 162-166 velikost pomnilnika 256kB, kar zadošča za kontrolni program audio vmesnika. 3.2.2 Pomnilnik RAM Najvažnejša lastnost, ki jo pri audio vmesniku zahtevamo od pomnilnika, je njegova hitrost. Če naj bi mikroprocesor deloval s polno hitrostjo, morajo imeti integrirana vezja pomnilnika RAM dostopni čas manjši od 45ns. Za pomnilnik RAM smo izbrali pomnilniške čipe proizvajalca HITACHI z oznako 624256. Kapaciteta le-teh je 256kB x 4 z dostopnim časom 45 ali 35 ns, odvisno od verzije. Za izvedbo audio vmesnika potrebujemo hitrejšo verzijo. Zaradi zunanjega podatkovnega vodila širine 4 bitov in podatkovnega vodila mikroprocesorja širine 32 bitov je potrebno uporabiti 8 integriranih vezij. To pomeni 1 MB pomnilnika. Pri delovanju audio vmesnika je pomembna tudi količina pomnilnika. V primeru, če bi se samo program izvajal v pomnilniku RAM, bi zadoščalo za delovanje do 512kB pomnilnika. Če pa pomnilnik RAM uporabljamo tudi kot vmesni pomnilnik za podatke, se potrebna količina pomnilnika poveča. Zato je na audio vmesniku predvideno, da količino lahko spreminjamo v odvisnosti od potreb. Predvidene so tri pomnilniške banke po 1MB. Za samo delovanje je potrebno vstaviti na ploščo audio vmesnika 1 pomnilniški modul - 1MB. Postopek prenosa podatkov preko vodila SCSI v načinu "zlog naenkrat", kjer je za vsak prenesen zlog potreben postopek ponovne izbire iniciatorja, največkrat ni izvedljiv (novi podatek iz A/D pretvornika je na razpolago vsakih 20.8 ¡as, ponovna izbira pa lahko traja tudi nekaj sekund), zato je potrebno prenos podatkov opravljati v blokih. 1MB pomnilnika se pri vzorčevalni frekvenci 48 kHz in dvokanalnem vzorčenju s 16 biti napolni v 5.4 s. Torej je potrebno najmanj vsake 5.4s izvesti preko vodila SCSI prenos bloka podatkov dolžine 1 MB. S povečevanjem pomnilnika RAM se ta interval lahko podaljša in znaša pri pomnilniku velikosti 3 MB okrog 15 s. Torej je priporočljivo napolniti s pomnilnikom vse tri pomnilniške banke. 3.3 A/D pretvornik Od A/D pretvornika se zahteva 16 bitna A/D pretvorba s frekvenco vzorčenja najmanj 48kHz. Zaradi uporabe standardnega mikroprocesorja za krmiljenje A/D pretvornika in za manipulacijo s podatki mora imeti A/D pretvornik paralelno zunanje podatkovno vodilo. Glede na zgornje zahteve je popolnoma ustrezal A/D pretvornik proizvajalca Analog devices z oznako AD676. To je 16 bitni A/D pretvornik z največjo hitrostjo vzorčenja 100kS/s in s paralelnim izhodom. Vhodno območje pretvorbe je enako ±Vref. 3.4 D/A pretvornik Od D/A pretvornika je pričakovati, da 16 bitno besedo pretvori v analogno vrednost. Hitrost D/A pretvorbe mora biti najmanj 48kS/s. Zaradi priključitve na vmesnik mikroprocesorja je zahtevano paralelno podatkovno vodilo. V skladu s temi zahtevami smo izbrali D/A pretvornik z oznako AD669, proizvajalca Analog devices. To je 16 bitni D/A pretvornik s paralelnim vhodom. Največji čas D/A pretvorbe je 13 |as, izhodno analogno območje pa do ±10V. 3.5 Vhodni in izhodni multiplekser Na vhod audio vmesnika lahko priključimo mikrofon in kasetofon. Med tema dvema izbiramo s postavitvijo ustreznega bita v statusnem registru A/D - D/A pretvornika. Postavitev tega bita pomeni preklop vhodnega multipleksorja v želeni položaj. Slika 1. Blokovni diagram audio vmesnika 165 Informacije MIDEM 28(1998)3, str. 162-166 V. Jenuš, B. Horvat: Audio vmesnik za računalnik Pri izbiri vhodnega multipleksorja sta v splošnem pomembna predvsem dva parametra: 9 hitrost preklopa in 0 električne stikalne karakteristike (Ron, Rof) Z uporabo dveh A/D in D/A pretvornikov smo se izognili kritičnim hitrostim preklopa in smo za vhodne in izhodne multipleksorje lahko zaradi idealnih stikalnih karakteristik izbrali stikala, izvedena z miniaturnimi releji. Le-ti so preko ustreznih tranzistorjev krmiljeni s postavitvijo ustreznega bita v statusnem registru A/D -D/A pretvornika. 4. Blokovna shema audio vmesnika Sam audio vmesnik smo pri načrtovanju razdelili na več enot: 8 mikroprocesor z gonilniki za vodilo in generatorjem urinega signala • naslovni dekodirnik 8 vmesnik za vodilo SCSI • pomnilnik RAM • pomnilnik EPROM 8 generator vzorčevalnih frekvenc 0 A/D pretvornik • D/A pretvornik • analogni vhodni del •> analogni izhodni del 0 napajalno enoto Povezavo enot prikazuje blokovna shema na sliki 1. Naloge zgoraj naštetih enot so naslednje: • Mikroprocesor izvaja program, zapisan v pomnilniku. S tem koordinira delovanje vseh enot in izvršuje naloge, ki jih preko vodila SCSI zahteva iniciator. • Naslovni dekodirnik razvrsti enote audio vmesnika v naslovni prostor mikroprocesorja in poskrbi za pravilne časovne poteke krmilnih signalov na vodilu audio vmesnika. • Vmesnik za vodilo SCSI pretvarja signale na vodilu SCSI v interne signale audio vmesnika (mikroprocesor) in obratno. • V pomnilniku RAM se izvaja program in v njem shranjujejo rezultati A/D pretvorbe kot tudi podatki za D/A pretvorbo. • V pomnilniku EPROM je shranjen program audio vmesnika. • Generator vzorčevalnih frekvenc generira vse potrebne frekvence za A/D in D/A pretvorbo. 9 A/D pretvornik je s svojimi dodatnimi vezji namenjen za vzorčenje, kvantiziranje in kodiranje analognega signala na izhodu analogne vhodne enote v digitalno obliko. 8 D/A pretvornik pretvarja digitalne vrednosti v analogne in jih posreduje analogni izhodni enoti. • Analogna vhodna enota amplitudno in impedančno prilagodi signal na vhodu audio vmesnika na vhod A/D pretvornika. Vrši tudi izbiro med signalom iz mikrofona in kasetofona. • Analogna izhodna enota omogoča izbiro močnostnega ojačevalnika ali kasetofona na izhodu in prilagodi signal z izhoda D/A pretvornika na vhoda močnostnega ojačevalnika ali kasetofona. • Napajalna enota zagotavlja vse potrebne napetosti za delovanje audio vmesnika. 5. Sklep Opisani audio vmesnik smo izdelali v laboratoriju za telekomunikacije. Po začetnem testiranju in odpravi otroških bolezni v programski kodi se je pokazal kot zelo dobro orodje pri zajemanju audio signalov in ga uporabljamo tako v raziskovalne namene kot v pedagoškem procesu. V prihodnosti bi mogoče bilo potrebno razmisliti o nadgradnji vmesnika še z DSP procesorjem. 6. Literatura /1./ Multimedia Systems, John F. Koegel Buford, University of Massachusetts Lowell 12.1 SCSI: Understanding The Small Computer System Interface, PTR Prentice Hall, Englewood Cliffs, New Jesey /3./ MC 68020 32-bit Microprocesspr User's Manual Second Edition, Prentice Hall, Englewood Cliffs, New Jesey /4./ Standard X3T9.2, Project 375D, revision 10L - standard za vodilo SCSI-2 /5./ Data Converter Reference Manual Volume I, Analog Devices, 1992 /6./ Data Converter Reference Manual Volume II, Analog Devices, 1992 /7./ Mikroprocesor - Datebuch 1, Elektor Verlag GmbH, 5100 Aachen, 1990 /8./ Mikroprocesor - Datebuch 2, Elektor Verlag GmbH, 5100 Aachen, 1990 /9./ Mikroprocesor - Datebuch 3, Elektor Verlag GmbH, 5100 Aachen, 1990 /10./ Hitachi IC Memory DATA BOOK, Hitachi /11./ Dahms elektronik HANDBUCH, 1989 /12./ Tehniški opis za komunikacijski krmilnik za mikroračunalnik TK 68000, dr. Bogomir Horvat, Nenad Črnko ing., Rado Slat-inek dipl. ing., Univerza v Mariboru, Tehniška fakulteta, VTO Elektrotehnika, računalništvo in informatika, Maribor, junij 1987 /13,/ Diagnostični program za komunikacijski krmilnik za mikroračunalnik TK 68000, dr. Bogomir Horvat, Rado Slatinek dipi. ing., Nenad Črnko ing., Univerza v Mariboru, Tehniška fakulteta, VTO Elektrotehnika, računalništvo in informatika Maribor, junij 1987 /14./ Priročnik za laboratorijske vaje iz mikroračunalniških sistemov, Zmago Brezočnik, Bogomir Horvat, Univerza v Mariboru, Tehniška fakulteta, Elektrotehnika, računalništvo in informatika, Maribor, september 1990 Vladimir Jenuš, Bogomir Horvat Univerza v Mariboru, Fakulteta za elektrotehniko, računalništvo in informatiko Smetanova 17 2000 Maribor tel.: +386 62 25 461 fax: + 386 62 225 013 Prispelo (Arrived): 14.10.1998 Sprejeto (Accepted): 16.10.1998 164 UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana A 500 DPI FINGERPRINT SENSOR IC IN CMOS TECHNOLOGY Wolfgang PRIBYL, Siemens EZM Ges.m.b.H., Graz-Villach, Austria Thomas SCHEITER, Siemens Semiconductor Division, Munich, Germany Gerd HRIBERNIG, Siemens PSE, G raz, Austria Keywords: biometrics, fingerprints, monolithic silicon sensor systems, CMOS, Complementary Metal Oxide Semiconductors, passwords, PIN codes, Personal Identification Number codes, security systems, identification systems, Identification devices, authorization, pattern recognition, chip cards Abstract: A monolithic fingerprint sensor has been implemented in a CMOS technology. This new sensor has a spatial resolution of 50 fjm and translates the fingerprint pattern into a gray scale Image of 256 x 256 pixels with 8 bit resolution. The sensor chip measures about 160 mm' together with a pattern recognition software, which extracts characteristic features, it forms a low cost system, which is well suited for a variety of applications as e.g. access control for buildings, computer networks and many other services. CMOS integrirano vezje za senzor prstnih odtisov z ločljivostjo 500 dpi Ključne besede: biometrlka, odtisi prstni, sistemi senzorjev silicijevih monolitnih, CMOS polprevodniki kovinskooksidnl komplementarni, gesla, PIN kode številčne Identifikacijske osebne, sistemi varnostni, sistemi identifikacijski, naprave Identifikacijske, odobritev, razpoznavanje vzorcev, chip kartice Povzetek: Izdelali smo senzor prstnih odtisov v CMOS tehnologiji. Njegova prostorska ločljivost je 50 j.im pri prevajanju oblike prstnega odtisa v sivo sliko velikosti 256x256 pik z 8 bitno ločljivostjo. Površina senzorskega vezja je 160 mm2. Skupaj s programsko opremo za prepoznavanje oblik sestavlja sistem, katerega nizka cena omogoča različne uporabe, kot so nadzor vstopa v objekte, računalniške mreže In podobno. cess to sensible areas of corporations and government agencies has to be restricted, but also electronic banking and electronic commerce in general need a proper identification and authorization, so that only the legitimate user may initiate transactions. If passwords and/or PIN codes are forgotten, the respective service is no longer available. For the identification of the legitimate users biometric methods are best suited, which are based on constant features of the user's human body. Examples are the human speech, some characteristics of the face, the signature, structures of the human eye (pattern of the retina and iris) and many more. According to (Fig. 1, /1/) most applications use pattern of the hand and fingers (65,7%) followed by voice recognition systems (21,4%). All other systems only have a small share of the overall biometrics market, either because of a high complexity of the systems needed, or because of the somehow invasive character of the measurement (e.g. illumination of the eye needed for the evaluation of the retina patterns). Simple, reliable and cheap systems are needed for the emerging security market for applications in PCs, telecommunication systems and electronic commerce applications. 2. The fingerprint as a unique characteristic of human beings A typical fingerprint, covering an area of ca. 100 mm2, shows about 12 characteristic positions, the so called minutia (Fig. 2). These minutia can be used as unique features for the identification of people. If compared to other methods, the fingerprint recognition systems offer several advantages: 1. Introduction For a long time keys, made from metal in a more or less sophisticated way, have been a well accepted measure to protect property and the private sphere of people and organizations. But these keys can be stolen or lost, copying of the keys may reduce the security or privacy level to zero. Besides those well known key systems, which are based on the possession of a key also knowledge based systems are used. These systems substitute the physical keys by agreements of secret codes. Many of today's systems as e.g. mobile phones, credit cards, computers require special passwords or PIN-codes (personal identification number). The ac- N. Uutina 1?o 35,Jli> M ur tof schnitt' ^ Gesicht 3,9°o "" ^ APIS 4,7% t Stimme 21,4% Finget 30.4% Fig. 1: The global biometrics market, share of different systems 159 Informacije MIDEM 28(1998)3m str. 167-170 W. Prybil, T. Scheiter, G. Hribernig: A 500 dpi Fingerprint Sensor IC in CMOS Technology 8 Since about 150 years the criminal science has carried out exhaustive theoretical and practical investigations on the use of fingerprints Op 8 An almost infinite number of variations (>>10 ) enable the reliable and unique identification of all living people and even future generations 8 The characteristics of the fingerprint do not change over the lifetime of a human being, even after most injuries the skin recovers in exactly the same pattern. 8 The use of fingerprint identification systems is easy and pleasant, compared e.g. to systems based on the retina patterns and needing illumination and camera systems. 9 Authentication: The person enters his/her ID, afterwards the system verifies the identity and rejects the access in the case of a mismatch. 8 Identification: The system identifies the person based on the fingerprint presented and rejects all not previously stored persons. Up to now the recording and the evaluation of finger prints was a costly procedure using very sophisticated equipment. With the availability of the monolithic „Fingertip-Sensor-System" it is now possible to record fingerprints very easily and to evaluate and identify the respective persons in a very short time. The whole process takes only a few seconds. Only if the extracted features correspond to the previously stored ones the protected device, system or transaction is released for the desired operation. Fig. 2: Fingerprint record, features marked The fingerprint recognition system described in this paper is based on the well established knowledge about the characteristics of fingerprints. The monolithic silicon sensor delivers a grayscale image and contains the significant characteristics (minutia) for each fingerprint, as e.g. ending lines, branching lines and characteristic turns and eddies. The features and characteristics and their relative position to each other are stored in a feature list. Such the overall memory needed can be reduced from 64 Kbytes for the whole gray scale image to a few hundreds of bytes for the still unique feature list. Due to this reduction, the whole system including a limited archive of legitimate users eventually will be implemented on a chip card. If the system has to learn the features of a new person (enrollment procedure) several fingerprints have to be recorded and their respective features be stored in the feature list. Usually several fingers are stored at a time for each person, so that in case of an injury the access to the system is still possible. The available system can act in two different ways, depending on the planned application: 3. The Fingertip Sensor Chip - Hardware A prototype of the sensor chip, fabricated in a 0,8 ¡jm CMOS technology with 2 layers of metalization has been produced and evaluated; first product samples are shipped to the customers now. This is an important basis for a low cost mass production of fingerprint based identification and authorization systems. The sensor area consists of an array of 256 by 256 sensor electrodes with a grid of 50 ¡jm and such cover an area of about 160 mm2 with a spatial resolution of 500 dpi (dots per inch). The sensor electrodes are covered with a special dielectric layer, which protects the surface of the chip from the direct contact with the outside world. The finger has to touch the chip surface and depending on the local structure of ridges and valleys different capacitance values in the range of a few femtofarads (10~12 F) can be read out from the individual electrodes in the array. Such the information on the whole structure of the finger print is available as 65.536 coded capacitance values. A schematic cross section through the sensor array with a finger in contact can be seen in fig. 3, the block diagram of the whole sensor chip in fig.4. Besides the sensor array, the chip contains a direct analog-to-digital conversion of the capacitance values, a control unit based on a 1 MHz clock generator and an interface for the transmission of the gray scale image, which is coded as 8 bit/pixel. For the prototypes a parallel interface for an easy connection to a PC has been chosen, later on standard interfaces for different microcontrollers or customized versions will be available. Fig. 5 shows a view of finished wafer of sensor chips with a finger applied. Aktive Pixel Silicium Wafer Fig. 3: Schematic cross section of sensor 168 ? W. Prybil, T. Scheiter, G. Hribernig: A 500 dpi Fingerprint Sensor IC in CMOS Technology Informacije MIDEM 28(1998)_3m str. 167-170 Fig. 4: Block diagram of the integrated finger print sensor. Fig. 6: Structure of software network, the archive then stores only the result of these processes. When searching for a match later on, only the characteristic features are used. The compare algorithm is able to compensate for eventual translatorial or rotational changes in the pattern. The recording and identification process for finger prints was a very complex procedure and requeired expensive instruments up to now. The new system can be used for a much broader range of applications, even in the low cost market segments. Almost every man -machine interface may use a fingerprint recognition system for increased security. The access to various types of computers, chipcards, immobilizer systems for vehicles, access control for buildings. At the CeBit exhibition 1998 application examples have been demonstrated: achip card incorporating a sensor, a prototype of a mobile phone and a PC keyboard (fig. 7, 8). Fig. 5: Microphotograph of silicon sensor wafer with a finger applied 4. The Software for the Pattern Fig■ 7: Recognition Process Since 1992 the PSE group in Graz is working on algorithms for the detection and identification of fingerprints. For this task conventional techniques of pattern recognition have been applied, in recent time newtechniques as e.g. neural networks have been successfully applied. The software for this project has been developed within two years by a 5 person team. The processing chain (fig. 6) starts with the preprocessing of the sensor data Literature: and the transformation of these data into a normalized /y image. Then the identification of line structures, the /2/ completion of these lines and the extraction of the pure line structures is carried out. Then the characteristic features are extracted and classified using a neural Prototype of a GSM mobile phone with fingerprint authorization. Markt und Technik, volume 34/98 S.Jung, R.Thewes, T.Scheiter: CMOS Fingerprint Sensor with Automatic Local Contrast Adjustment and Pixel-Parallel Encoding Logic, submitted to the ISSCC'99, San Francisco, 2/99. 169 Informacije MIDEM 28(1998)_3m str. 167-170 W. Prybil, T. Scheiter, G. Hribernig: A 500 dpi Fingerprint Sensor IC in CMOS Technology Dr. Wolfgang Pribyl Siemens Entwicklungszentrum für Mikroeelektronik GmbH Hilmteichstr. 113 A-8043 Graz Tel.: ++43 316 321210-10 Fax: ++ 316 32 1210-44 mailto: Wolfgang.Prybil@siemens-scg.com Fig. 8: Sample of a chip card with integrated fingerprint sensor Prispelo (Arrived): 16.9.1998 Sprejeto (Accepted): 2.10.1998 169 UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana INTERACTION OF HYDROGEN PLASMA WITH CORRODED SILVER SURFACE M. Mozetič and B. Praček Institute of Surface Engineering and Optoelectronics, Ljubljana Key words: materials for electrical contacts, cleaning of contacts, surface cleaning, cleaning of contact surfaces, cleaning of metals, plasma cleaning, discharge cleaning, electrical discharges, H-plasma, hydrogen plasma, archeological artifacts, H-plasma interaction with surface, corroded surfaces, material corrosion, surface corrosion Abstract: Experimental investigation on discharge cleaning of corroded silver surface has been performed. Samples of silver with a thin and thick corroded layer were used. In the case of a thin layer, the surface of pure silver was contaminated with fingerprints. AES analysis showed that the layer with the thickness of 20 nm consisting of O, S, CI and C was formed. In the case of a thick layer, a well corroded silver coin from 19th century was used. In this case, the thickness of the corroded layer was of the order of 0.1 mm and the EMPA investigation showed it was an agglomerate consisting of O, S, CI, Si, C, Fe, Ti. All samples were exposed to a low pressure weakly ionized hydrogen plasma at the pressure of 1 mbar. Plasma parameters were measured with a double Langmuir probe and a catalytic probe. The electron temperature was 6eV, plasma density 2.1016m~3, and the degree of dissociation of hydrogen moleculs about 1%. Samples with the thin corroded layer were exposed to plasma for 10 minutes. AES analysis of the treated samples showed that all impurities were completely removed from the surface. The coin was exposed to hydrogen plasma at the same conditions, and the EMPA analysis showed that concentration of oxidizing impurities was lowered but other impurities still persisted.They could not be removed solely by treatments in plasma, but successive treatment in plasma and ultrasound bath. By the combination of both treatment we were able to decrease the concentration of any impurities below the detection limit of the EMPA, Interakcija vodikove plazme s korodirano površino srebra Ključne besede: materiali kontaktov električnih, čiščenje kontaktov, čiščenje površin, čiščenje površin kontaktov, čiščenje kovin, čiščenje plazemsko, čiščenje razelektritveno, razelektritve električne, H-plazma vodikova, artefakti arheološki, interakcija H-plazme vodikove s površino, površine korodirane, korozija materialov, korozija površin Povzetek: Prikazujemo rezultate eksperimentalne preiskave plazemskega čiščenja korodirane površine srebra. Uporabili smo srebrne vzorce s tanko in debelo korodirano plastjo. V primeru tanke korodirane plasti smo površino čistega srebra kontaminirali s prstnimi odtisi. AES preiskava teh vzorcev je pokazala, da je na površini prisotna tanka plast nečistoč debeline 20 nm, ki vsebuje poleg srebra še O, S, Cl in C. V primeru debele plasti smo izbrali močno korodiran srebrni kovanec iz 19. stoletja. Vtem primeru je bila debelina korodirane plasti reda velikosti 0,1 mm. Preiskava vzorca z elektronskim mikroanalizatorjem je pokazala, da površinski aglomerat vsebuje naslednje elemente: 0,S,CI,Si,C,Fe in Ti. Vsi vzorci so bili izpostavljeni nizkotlačni šibkoionizirani vodikovi plazmi pri tlaku 1 mbar. Plazemske parametre smo merili z dvojno Langmuirjevo sondo in katalitično sondo. Elektronska temperatura je bila 6 eV, gostota plazme 2x1016 m'3 in stopnja disociiranosti vodikovih molekul okoli 1%. Vzorci s tanko korodirano plastjo so bili izpostavljeni vodikovi plazmi za 10 minut. AES preiskava tako obdelanih vzorcev je pokazala, da smo odstranili vse nečistoče površin. Kovanec je bil izpostavljen vodikovi plazmi pri enakih pogojih in analiza z elektronskim mimkroanalizatorjem je pokazala, da je po obdelavi koncentracija oksidativnih nečistoč bistveno zmanjšana, medtem ko so ostale nečistoče na površini še vedno prisotne. Slednje nismo uspeli odstraniti samo z plazemskim čiščenjem temveč smo jih uspešno odstranili s kombinacijo plazemskega in ultrazvočnega čiščenja. S kombinacijo obeh metod smo uspeli kovanec tako dobro očistiti, da je koncentracija katerih koli nečistoč na površini manjša kot je meja detekcije elektronskega mikroanalizatorja. 1 Introduction Discharge cleaning has become a widely used method for removing surface impurities. Active particles which are created in plasma, interact with surface impurities forming volatile molecules which are easily desorbed and pumped away. By creating plasma in a mixture of various gases, it is possible to remove different types of impurities. For reduction of surface layers of oxides, hydrogen or a mixture of a noble gas and hydrogen is usually used. In the past decade, this method has been widely investigated as it is of a great scientific and commercial importance. The method has been successfully applied in cleaning of stainless steal surfaces of tokamaks /1,2,3/, silicon in microelectronic devices /4,5,6,7/ and a variety of metals and alloys archaeological artifacts are made from /8,9,10/. Although hydrogen plasma can be created by different DC and high frequency discharges, the radio frequency (RF) and microwave (MW) discharges are the most suitable for this application. This is due to the fact that a plasma with a smal space potential can be created with a high frequency discharge, and it penetrates fairly well in gaps between samples. The degree of ionization in a high frequency discahrge is usually low (between 10~6 and 10~2) /11,12/, except in the case the electron cyclotrone resonance (ECR) conditions are met /13/. In the latter case, the degree of ionization can be more than 10% /14/. In any case, the degree of dissociation of hydrogen molecules is more than 1 % /12/ and can approach unity if a plasma is created in a discharge vessel made of material with a low recombination coefficient for the reaction 2H->H2, i.e. different glasses, alumina, some ceramics /14/. Despite the discharge cleaning of silver and its alloys is of a great importance in electronic industry as it is an excellent method of final treatment of contact materials, little work on this subject has been published. In the present paper, we describe experimental work recently done at our laboratories on discharge cleaning of silver with a thin and a thick layer of impurities. 2 Experimental Experiments were carried out in a vacuum system, which consisted of a discharge vessel, a liquid nitrogen cooled trap and a two stage mechanical rotary pump. 163 M. Mozetič, B. Praček: Interaction of Hydrogen Plasma Informacije MIDEM 28(1998)3, str, 171-174 with Corroded Silver Surface The base pressure in the system was 10~3 mbar, The discharge vessel was a glass cylindrical tube with the lenght of 80 cm and the diameter of 4 cm. Plasma in the discharge vessel was created at the pressure of 0,5 mbar by an inductively coupled RF generator with the frequency of 27.12 MHz and the maximum output power of 700 W. Plasma parameters were measured with a double Langmuir probe /15/, and a catalytic probe /14,16/. The electron temperature in plasma was 6eV, while the plasma density was 2.1016 rrr3. The use of a liquid nitogen cooled trap was found to be very important not only as a trap for oil form the rotary pump, but especially as a trap for aggresive gases forming during the discharge treatment of well corroded samples. Namely, during the treatment of ar-cheological artifacts, a rather large amounts of H2S and HCI were produced, and both of the gases could have been harmful to the pump. Two types of samples were prepared: i) strips made of pure silver, and ii) old silver coins. The strips were first cleaned with freon and than touched well with fingers in order to obtain a thin layer of different impurities on the surface. The coins were discovered recently, and only the layer of sand and soil was removed by an archeologist. We treated them in ultrasound bath in a mixture of water and detergent in order to remove weakly bonded impurities from the samples, mostly hydrocarbonaceus compounds. Samples with a thin layer of surface impurities were analyzed with an Auger electron spectroscopy (AES) depth profiling. We used a scanning Auger microprobe (Physical Electronics Ind. SAM 545 A) with a static primary electron beam with the energy of 5 keV, the beam current of 0.5 p.A, and the beam diameter of about 10 jim. The incidence angle of the electron beam with respect to the normal of the surface plain was 30°. In order to obtain depth profile, samples were sputtered with Ar+ ions with the energy of 1keV, rastered on a surface area larger than 10x10 mm2. Atomic concentrations were calculated by taking into account relative sensitivity factors according to the literature /17/. The concentrations are plotted against ion gun sputter time instead of the depth, and 1 minute sputter time corresponds to the depth of about 2 nm. The thickness of the layer of impurities on silver coins was too large to be analyzed with AES depth profiling, and the method is destructive anyway. The concentration of elements on the surface of these samples was determined with an electron microprobe analysis (EMPA). Ascanning electron microscope JEOLJSM 35 with an energy dispersive X ray microanalyser (EDX-TRACOR TN 2000) was used to determine the concentration of elements in the surface layer. An electron beam with the voltage of 20 kV and current of approx. 0.5 mA was focused at the sample in a spot of about 1 x 1 |im2. The thickness of the layer analyzed by EDX was about 1 |im. 3 Results The composition of the surface layer of industrial cleaned silver strips and those contaminated with fingerprints was analyzed with AES depth profiling and is shown in Figure 1 and 2, respectively. It is noticeable that both samples contain a surface layer of impurities with the thickness of the order 0.1 ¡im . The composition of the impurity film on the samples is only slightly different. Both samples contain mostly carbon (probably an organic compound), oxygen, chlorine, sulphur and potassium. C o n c e 11 t r a i 0 11 at. m Fig. 1. 20 40 60 80 10( Sputter time [min] AES depth profile of the surface layer of the industrial cleaned silver strip. [%] Fig. 2. 100 '200 300 Sputter time ¡miu] AES depth profile of the surface layer of the silver strip cleaned with freon and later contaminated with finger prints. Both samples were mounted in the middle of the discharge vessel and treated with hydrogen plasma for 10 minutes. Due to the recombination of atomic hydrogen on the sample surfaces, the absorption of UV light from plasma, and the bombardment of the surfaces with charged particles, the temperature of the samples raised to about 150°C. After the treatment the samples were exposed to air for a short time and analyzed with AES sputter depth profiling again. The composition of the surface of the samples is shown in Figure 3 and 4. It is noticeable that the surface of both samples is clean except of traces of oxygen and sulphur which were probably adsorbed on the surface during the exposure of the samples to air. 172 M. Mozetič, B. Praček: Interaction of Hydrogen Plasma with Corroded Silver Surface Informacije MIDEM 28(1998)3, str. 171-174 100 Sputter time [minj Fig. 3. AES depth profile of the surface layer of an industrial cleaned silver strip which was exposed to hydrogen plasma for 10 minutes. Sputter time [minj Fig. 4. AES depth profile of the surface layer of a silver strip first cleaned with freon, than contaminated with finger prints and exposed to hydrogen plasma for 10 minutes. The sample was treated with plasma for 10 minutes. After the treatment it was analyzed with the electron microprobe. The only change was a substantial enlargement of iron peak. It was also found that, after the plasma treatment, the layer of impurities on the surface can be rather easily removed mechanically by the use of a needle. Since we did not want to make any scratch on the sample, we rather treated it in the ultrasound bath and some impurities were released. Microprobe analyses showed that the concentration of iron fell to the original value. The sample was then treated again with plasma for 10 minutes. After this treatment, a part of the surface became clean, while most of the surface was still covered with a layer of impurities. The microprobe analyses showed that the clean part of the surface consisted of silver and about 10% of copper (which is actually the structure of the bulk), and the dirty part of the surface consisted of iron, silver, copper, sulfur and chlorine. After repeating the plasma cleaning and the ultrasound cleaning for four times, all the surface became clean. The microprobe analyses showed only silver and cooper, while the concentration of other elements was below the detection limit of the microprobe which was about 01. at.% (Figure 6), X ray <■!:>■:i\v ) ! V'j Fig. 6. EDX spectrum of the surface layer of a silver coin after successful cleaning in hydrogen plasma and ultrasound bath. Fig. 5. EDX spectrum of the surface layer of the silver coin before discharge cleaning. Before the plasma treatment, a silver coin was analyzed with EMPA (Figure 5). Apart from oxygen and carbon, which cannot be detected with our microprobe, the layer of impurities of the surface consisted of silver, iron, copper, silicon, sulfur, and chlorine. 4. Discussion The experiments described above showed that active particles produced in hydrogen plasma react with the impurities bonded to the surface of the samples. Since the density of plasma is low it is obvious that hydrogen ions do not contribute much to the cleaning efficiency. It is probably atomic hydrogen which reacts with the surface impurities. Hydrogen atoms react with chemically bonded oxygen, chlorine, and sulfur to form molecules, which are easily pumped from the discharge vessel: OH, H2O, HCI, H2S, etc. These gases are then trapped by the liquid nitrogen cooled trap. The appearance of iron in the impurity layer after the plasma treatment can not be due to a deposition of iron since plasma was not in a contact to any material composed of iron. A small amount of iron was presented within the layer of impurities already before the 173 Informacije MIDEM 28(1998)3, str. 171-174 M. Mozetič, B. Praček: Interaction of Hydrogen Plasma with Corroded Silver Surface treatment. The original concentration of iron is rather low. During the discharge treatment, most of other impurities were removed from the surface, so the relative concentration of the iron in the surface layer was increased substantially. Iron cannot be removed from the surface with mild hydrogen plasma cleaning. Further-more, a layer of iron on the surface probably causes intensive recombination of atomic hydrogen on the surface and thus prevents successful removal of other impurities from deeper layers. Luckily, the layer of iron is weakly bonded to the surface, so it can be successfully removed mechanically either by the use of a needle or even by the use of ultrasound treatment. Once the layer of iron is removed, further removal of other impurities in hydrogen plasma can take place, so that after repeating the discharge cleaning and the ultrasound cleaning for several times, the surface of the samples became free of any impurities. 5. Conclusion Discharge cleaning experiments on silver samples were carried out in low pressure weakly ionized hydrogen plasma. It was shown that most impurities can be removed from the samples by treatment in hydrogen plasma with the density of 2.1016 rrr3 and the electron temperature of 6 eV. In the case of a thin layer of impurities, the samples were analyzed with AES depth profiling and it was shown that a layer of chemically bonded oxygen, chlorine, sulfur, carbon and silicon can be completely removed in ten minutes. Discharge cleaning of well corroded silver coins took nearly an hour. In this case, itwas shown that iron, which was also presented in the layer of impurities, could not be removed by treatment in hydrogen plasma. Furthermore, it prevented sucessful removal of other impurities. A successful procedure of removal iron from the surface was found to be the ultrasound tretament. By repeating the discharge and ultrasound cleaning procedures it was possible to remove a thick layer of impurities as well. References /1/ F Waelbroek, K J Dietz, P Wienhold, J Winter, I AliKhan, H Markens and E Rota, J Nucl Mater 93&94 (1980), 839. /2/ Y Sakamoto, Y Oshibe, K Yano and H Oyama, J Nucl Mater, 93&94, 333 (1980). /3/ Y Matsuzaki, N Suzuki and T Hirayama, Jap J Appl Rhys, 25. 253 (1986). /4/ S J DeBoer, V L Dalai, G Chumanov and R Bartels, Appl Phys Lett 66, 2528 (1995). /5/ J S Montgomery, J P Barnak, C Silvestre, J R Hauser and J R Nemanich, in Ultraclean semicondustor processing technology and surface chamical cleaning and pasivation, ed by M Liehr, M Heyns and M Hirose, Mater Sci Soc, Pittsburg, 279 (1995). /6/ I Šorli, W Petasch, B Kegel, H Schmid and G Liebl, Inform. Midem 26 (1996) 35. /7/ I Šorli, W Petasch, B Kegel, H Schmid, G Liebl and W Ries, Inform. Midem 26, (1996)113. /8/ S Veprek, J T Elmer, C Eckermann and M Jurčik - Rajman, J Electrochem Soc, 31, 29 (1986). /9/ M J de Graaf, R Severens, M J F van de Sanden, D C Schram, H J M Meijers and H. Kars, J Nucl Mater, 200, 380 (1993). /10/ J Patscheider and S veprek, Studies in Conservation 31, 29 (1986). /11/ D Korzec, FWerner, ABrockhaus, J Engemann, T P Schneider and R J Nemanich, J Vac Sci Technol A13, 2074 (1995). /12/ M Mozetič, M Kveder, A Pregelj and P Paulin, Infor Midem 23, 112 (1993). /13/ H S Tae, S J Park, S H Hwang, K H Hwang, E Yoon, K W Whang and S A Song, J. Vac Sci Technol B13, 908 (1995). /14/ F Brecelj, M Mozetič, K Zupan and M Drobnič, Vacuum 44. 459 (1993). /15/ Swift and Schwar, Electrical Probes for Plasma Diagnostics , lliffe Books, London (1969). /16/ M Mozetič, M Kveder, M Drobnič, A Paulin and A Zalar, Vacuum 45, 1095 (1994). /17/ L E Davis, N C MacDonald , P W Palmberg, G E Riach and R E Weber, Handbook of Auger Electron Spectroscopy, 2nd ed, Physical Electronics Industries, Eden Prairie, MA (1976). Dr. Miran Mozetič, d i pi. ing. Borut Praček, dipl.ing. Institute of Surface Engineering and Optoelectronics Teslova 30, 1000 Ljubljana, Slovenia tel.: +386 61 126 45 92 fax: +38661 1264593 Prispelo (Arrived): 9.10.1998 Sprejeto (Accepted): 12.10.1998 174 UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana REACTIONS ON CATALYTIC PROBE SURFACE DURING OXYGEN PLASMA TREATMENT OF POLYETHER S U LP H O N E Miran Mozetič, Plasma Laboratory, Institute of Surface Engineering and Optoelectronics, Ljubljana, Slovenia Keywords: catalytic probes, nickel catalytic probes, surface reactions, plasma treatment, oxygen plasma, polyether sulphone, surface activation, plasma etching, plasma dry etching, RF discharges, RadioFrequency discharges, T-t temperature-time characteristics Abstract: Experiments on the behavior of a nickel catalytic probe during activation of the surface of polyether sulphone in oxygen plasma are described. The temperature of the probe mounted 30 cm apart from inductively coupled RF oxygen plasma was measured for the case of empty discharge vessel and the case a sample with the dimensions of 8 cm x 1.2 cm x 0.4 cm was mounted in the middle of the discharge coil. It was found that both the maximum temperature and the first time derivative of the probe was much higher in the case of loaded discharge vessel than in the case of empty vessel. Both effects were described in terms of a higher probability for recombination of radicals such as CO on the probe surface than the recombination of neutral oxygen atoms. Reakcije na površini katalitične sonde med plazemsko obdelavo polieter sulfona Ključne besede: sonde katalitične, sonde katalitične nikljeve, reakcije površinske, obdelava s plazmo, plazma kisikova, polieter sulfon, aktiviranje površine, jedkanje plazemsko, jedkanje suho plazemsko, RF razelektritve radiofrekvenčne, T-t karakteristike temperatura-čas Povzetek: Prikazujemo obnašanje nikljeve katalitične sonde med aktivacijo površine polieter sulfona v kisikovi plazmi. Merili smo temperaturo sonde, ki je bila nameščena 30 cm od induktivno sklopljene RF kisikove plazme v primeru prazne razelektritvene komore in v primeru, ko smo v sredino razelektritvene tuljave namestili vzorec z dimenzijami 8 cm x 1.2 cm x 0.4 cm. Ugotovili smo, da sta tako navečja temperatura, ki jo doseže sonda, kot tudi prvi časovni odviod temperature sonde po vklopu RF generatorja, precej večja v primeru, ko je v razelektritveni komori nameščen vzorec, kot v primeru, ko je komora prazna. Oba pojava smo tolmačili z večjo verjetnostjo za rekombinacijo radikalov kot je CO na površini sonde, kot je verjetnost za rekombinacijo nevtralnih atomov kisika. 1. Introduction Oxygen plasma has become widely used for low temperature treatment of organic materials in research laboratories and industry. Technologies based on application of oxygen plasma include degreasing of different components /1/, plasma dry ashing /2/, activation of polymer surfaces for painting/printing /3/, and plasma ashing of biological samples /4/. Oxygen plasma for surface treatment of different samples is created in low-pressure discharges, such as microwave discharge (both non-magnetized and ECR modes), RF discharge (inductively and capacitively coupled) and a variety of DC discharges. Discharges are usually created in vacuum chambers made of a material with a low recombination coefficient for the reaction 20 O2, such as metals, which form thin films of stable oxides, glasses and ceramics. For different application, oxygen plasma with different parameters is used. By changing the type of the discharge, the discharge power, dimensions of the discharge vessel, pressure and pumping speed, one can obtain oxygen plasma with different parameters, i.e. the density and energy distribution of various particles created in plasma. In any case, the plasma is thermodinamical non-equilibrium. In most cases, the energy distribution function of each type of particles is close to Maxwellian, so the temperature is defined according to the rules of statistical mechanics. The electron temperature in plasma is usually several 10000°C (average kinetic energy of several eV), while the temperature of heavy particles is often close to the temperature of the inner wall of the discharge vessel. Due to inelastic collisions with hot electrons, the density of excited states of oxygen molecules in the discharge differs significantly from the values found in thermodinamical equilibrium gas. The density of electrons and positive ions is often within 1015 -10 7 rrf3 bracket, the density of negative oxygen ions is usually an order of magnitude lower, the density of neutral oxygen atoms is often around 1020 m"3 and sometimes even one or two orders of magnitude higher. The same applies for the density of ozone molecules (03). Many particles are found in excited states. Molecules are found in high rotational and vibrational states and also in single electron excited states. Excited states are usually relaxed by photon emission so plasma is an effective source of radiation ranging from IR to UV light. Several techniques have been used for determination of plasma parameters including electrical probes, magnetic probes, mass spectrometry, spectral intensities, line broadening, optical and ultra violet techniques, X -ray spectroscopy, far infra - red techniques, optical interferometry and microwave techniques /5/. For determination of the density of neutral atoms catalytic probes proved useful /6/. Many oxygen particles created in plasma react with the organic samples treated. Oxygen atoms, for instance, can react with the surface of hydrocarbons either by being trapped on the surface causing oxidation of the material or forming volatile molecules (such as CO and OH) which leave the surface. In any case, plasma parameters are altered by the presence of organic 167 Informacije MIDEM 28(1998)3, str. 175-179 M. Mozetič: Reaction on Catalytic Probe Surface During Oxygen Plasma Treatment of Polyether Sulphone samples in the discharge vessel. It is clear that the density of oxygen atoms in the presence of samples is lower than in an empty discharge vessel. Some atoms are lost since they remain in the surface layer of the samples, and more are lost because they form light molecules (radicals as CO and OH finally form stable molecules - carbon dioxide and water). The change of the density of neutral reactive particles can be monitored by the use of a catalytic probe. In the present paper we show how the presence of a polymer sample alters the behavior of a catalytic probe. 2. Experimental Experiments were performed in a glass discharge tube with the inner diameter of 36 mm and the length of 60 cm. Oxygen plasma was created at one side of the discharge vessel within a coil with the length of 15 cm, which was connected to an RF generator with the frequency of 27.12 MHz and the nominal power of 700 W. A catalytic probe was mounted on the other side of the discharge vessel. The probe is shown in Figure 1. The distance between the coil and the probe was 30 cm. The temperature of the probe after turn on of the RF generator was measured at different pressure. For the case of an empty discharge vessel, the temperature versus time is shown in Figure 2, 3 and 4. Fig. 1. Catalytic probe. 1 - nickel disc with the radius of 1 mm and the thickness of 0.2 mm, 2 - chromel - alumel thermocouple wires with the radius of 0.0625 mm and the length of 20 mm, 3 - thin glass tube with the outer diameter of 1.8 mm, 4 - kovar wire with the radius of 0.6 mm, 5 - glass tube with the outer diameter of 7 mm. After the experiments with the empty vessel, a sample was mounted in the middle of the RF coil. The sample was made of polyether sulphone (PES) of a rectangular shape with the length of 8 cm, the width of 1.2 cm and the thickness of 0.4 cm. The temperature of the probe during the first 10 s after turn on the RF generator is shown in Figure 5, 6 and 7. 3. Results Measurements of the temperature of the catalytic probe summarized in Figure 2 - 7 show that exotermic reactions take place on the probe surface when the RF generator is on. For the first few seconds after turn on the generator, the temperature increases linearly with time, and after some time, a constant temperature is obtained. The time needed for reaching the steady temperature depends on the pressure in the vacuum system. In the case of empty discharge vessel, the constant temperature is obtained in 20 s at the pressure of around 100 Pa, and in 40 s at the pressure of about 10 Pa. When the sample is mounted in the discharge vessel, the measurements of the temperature were performed only during the first 10 s so not in all cases the constant temperature is obtained. Still, comparison of Figures 4 and 7 show, that the time needed to reach the constant temperature of the probe depends on what's in the discharge vessel. In the case of the PES sample in the discharge vessel, the time needed to reach the constant temperature is at least for a factor of 2 smaller than for the empty discharge vessel. Time [s] Fig. 2. Temperature of the nickel catalytic probe in empty discharge vessel during the first 60 s after turn of the RF generator at low pressure. Time [s] Fig. 3. Temperature of the nickel catalytic probe in empty discharge vessel during the first 60 s after turn of the RF generator at medium pressure. 177 M. Mozetič: Reaction on Catalytic Probe Surface During Oxygen Plasma Treatment of Polyether Sulphone_ Informacije MIDEM 28(1998)3, str. 175-179 Time [s] Fig. 4. Temperature of the nickel catalytic probe in empty discharge vessel during the first 60 s after turn of the RF generator at high pressure. Time [s] Fig. 5. Temperature of the nickel catalytic probe in loaded discharge vessel during the first 60 s after turn of the RF generator at low pressure. Time [s] Fig. 6. Temperature of the nickel catalytic probe in loaded discharge vessel during the first 60 s after turn of the RF generator at medium pressure. Time [s] Fig. 7. Temperature of the nickel catalytic probe in loaded discharge vessel during the first 60 s after turn of the RF generator at high pressure. From the curves of the probe temperature versus time (Figure 2 - 7) the fist time derivative after turn on the RF generator is calculated. The results are summarized in Figure 8. In the case of empty discharge vessel, the first derivative of the probe temperature is of the order of 10 K/s, while in the case of the discharge vessel with the sample it is an order of magnitude higher. In both cases, the derivative at low pressure increases with increasing pressure, reaches the maximum at the pressure of about 80 Pa and decreases with further increase of the pressure. 4. Discussion Measurements of the first time derivative of the probe temperature give valuable data on the state of the gas in the vicinity of the probe /7,8/. If the gas is in the state of thermodinamic equilibrium, the temperature of the probe will remain unchanged, as the discharge vessel near the probe remains at room temperature. The rate of heating of the probe is therefore a measure of the non-equilibrity of gas in its vicinity. The higher the first derivative, the higher non-equilibrium of the gas. In the case of pure gas, the first time derivative of the probe gives semi-quantitative values on the density of neutral oxygen atoms in the gas /7/: 4Mc dT p _L__ 0 vyWD7t:r2 dt ' where no is the O density near the probe, M the probe mass, cp the thermal capacity of the material the probe is made of, v the average thermal velocity of oxygen atoms, y the recombination coefficient for oxygen atoms on the probe surface, Wd the dissociation energy of oxygen molecule, r the probe radius and dT/dt first time derivative of probe temperature just after turn on of the RF generator. For the case the probe is made of nickel and its surface is covered with a thin film of nickel oxide, the constants in equation (1) are as follows: M = 5.6 x 10-6 kg, cp = 444 J/kgK, v = 700 m/s, y = 0.04, 177 InformacijejyilDEM 28(1998)3, str. 175-179 M. Mozetič: Reaction on Catalytic Probe Surface During _Oxygen Plasma Treatment of Polyether Sulphone Wd = 5.12 eV, r = 1 mm. Taking into account the measured values of dT/dt, which are about 10 K/s, the density of oxygen atoms in the vicinity of the probe is 1.4x1021 rrr3 what is a reasonable result. When a sample is placed in the discharge vessel, the density of oxygen atoms drops according to the discussion in chapter 1. However, the first derivative of the probe temperature did not drop at all. Moreover, the derivative increased for an order of magnitude, as can be seen in Figure 8. The result is certainly worth a discussion. 1 10 100 1000 p [Pa] Fig. 8. Time derivative of the temperature of the catalytic probe after turn on the RF generator at different pressure in the case of empty and loaded discharge vessel. In the case of empty discharge vessel, the probe is heated mainly due to recombination of neutral oxygen atoms on its surface. The probability of recombination (i.e. the recombination coefficient) for this reaction is rather low: y = 0.04 /9/ and the O concentration in the vicinity of the probe rather high (of the order of 1021 rrr3, as calculated above). When the sample is mounted into the discharge vessel, the density of neutral oxygen atoms drops as they react chemically on the sample surface forming molecules like OH and CO, So one would expect that the derivative of the probe temperature would drop as well. However, newly formed molecules may also chemisorb on the probe surface and react with other particles reaching the surface. The most probable reaction taking place on the probe surface is CO + O CO2. Other exothermic reactions may also occur on the probe surface. Due to all these reactions, the energy dissipation on the probe remains high although the density of O atoms is lowered. The energy dissipated on the probe surface at the reaction O + O O2 is similar to the energy dissipated at the reaction CO + O CO2, i.e. few eV. If the probability of both reactions is the same, the time derivative of the probe temperature will not change much for the cases of empty and loaded discharge vessel. The experimental results, however, showed that the derivative of the probe temperature increased for an order of magnitude when the sample was mounted into the discharge vessel. Therefore, one can conclude that the probability of oxidation of carbon monoxide on the probe surface is much higher than the probability of the recombination of oxygen atoms. 5. Conclusions Experiments on the behavior of nickel catalytic probe placed in the discharge vessel were described. The time dependence of the temperature of the nickel catalytic probe was measured at different pressure for the case of empty discharge vessel and the vessel loaded with a sample of polyether sulphone of a rectangular shape with the length of 8 cm, the width of 1.2 cm and the thickness of 0.4 cm. The measurements showed that for the first few seconds after turn on the plasma source the temperature of the disc rised linearly with time, and after about 10sthe constant temperature was obtained. The constant temperature of the probe depended on pressure and whether the discharge vessel was loaded with the sample or not. For the case of the loaded vessel, the constant temperature of the probe rised over 1000 K at any pressure between 40 and 150 Pa, while at higher and lower pressure the constant temperature was less than 1000 K. For the case of the empty discharge vessel, the maximum temperature of nearly 550 K was obtained at the pressure of 80 Pa. The first time derivative of the probe temperature after turn on the plasma source was of the order of 10 K/s for the case of empty discharge vessel, and of the order of 100 K/s when the sample was mounted into the discharge vessel. In both cases, the derivative increased with increasing pressure up to the pressure of about 80 Pa where the maximum value of about 25 and 230 K/s, respectively, was reached, and decreased with further increase of the pressure. It was suggested that the higher rate of heating of the probe in the case of the sample placed in the discharge vessel was due to a higher probability of reaction of CO and OH molecules on the probe surface than the probability of recombination of oxygen atoms on the same surface. Acknowledgement The research was funded by the Ministry of Science and Technology of the Republic of Slovenia (Grant No L2 -0503- 1534-98). REFERENCES /1/ I. Sorli, W. Petasch, B. Kegel, H. Schmidt and G. Liebel , Inform. Midem 26 (1996), 35. /2/ B. Chapman, Glow Discharge Processes, J. Willey&Sons, New York (1980). /3/ I. Sorli, W. Petasch, B. Kegel, H. Schmidt, G. Liebel and G. Reis, Inform. Midem 26 (1996), 113. 178 M. Mozetič: Reaction on Catalytic Probe Surface During Oxygen Plasma Treatment of Polyether Sulphone_________________Informacije MIDEM 28(1998)3, str. 175-179 /4/ T. Mozetič, Extended abstracts 7th Joint Vacuum Conf., De- brecen(1997), 219. /5/ R.H. Huddlestoneand S. L. Leonard Plasma Diagnostic Techniques, Academic Press, New York (1965). /6/ M. Mozetič, M. Kveder, M. Drobnič, A. Paulin and A. Zalar, Vacuum 45 (1994), 1095. /7/ M. Mozetič, M. Drobnič, A. Pregelj and K. Zupan, Vacuum 47 (1996), 943. /81 H. Wise and B. J. Wood, Gas and surface reaction collisions, in Advances in atomic and molecular physics, ed. by D. R. Bates and I. Estermann, Academic Press, New York (1967). /9/ R. A. Hartunlan, W. P. Thompson and S. Safron, J. Chem. Phys 43 (1965), 4003. Dr. Miran Mozetič, dipl. ing. fiz., Laboratorij za plazmo Inštitut za tehnologijo površin in optoelektroniko, Teslova 30, 1000 Ljubljana Tel. 386 61 1264592, fax 386 61 1264593 e-mail: miran.mozetic@guest.arnes.si Prispelo (Arrived): 9.10.1998 Sprejeto (Accepted): 12.10.1998 179 UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana APLIKACIJSKI PRISPEVKI - APPLICATION ARTICLES v Novitete iz razvoja podjetja Iskre Žarnice Elvelux Uvod Podjetje Iskra Žarnice 10 let ni imelo novitet iz lastne proizvodnje. V zadnjih dveh letih pa je vložila kar precej sredstev v ustrezne kadre, opremo in v razvojna prizadevanja. Odprte in potrjene ima tri večje razvojne projekte z Ministrstvom za znanost in tehnologijo, Stanje tehnike v svetu V svetu obstojajo dve osnovni smeri v proizvodnji žarnic. Prva smer deluje z zastarelo opremo in številno delovno silo. Na posamezni proizvodni liniji, ki je praviloma namenjena eni vrsti žarnic, v več razredih napetosti, je v eni izmeni 10 do 60 zaposlenih. Mesečni osebni dohodki neposrednih delavk v taki proizvodnji so na nivoju 75 do 100 Dem, inženirjev pa 125 do 150 Dem. Cena proizvodov je zaradi minimalne amortizacije zastarele strojne opreme in nizke cene delovne sile nizka. Končna kontrola je osebna in nestabilna. Doseženi vakuum v žarnicah je za 10-krat ali več nižji od zahtevanega. Kvaliteta je slaba, uporabljani so ceneni materiali.Proizvodi niti zdaleč ne odgovarjajo standardom. Take avtomobilske žarnice hitro pregorijo. Svetijo celo manj kot 10 % od zahtevanih standardiziranih časov. Nikoli ne dosegajo nivoja kvalitete prve vgradnje pri proizvajalcu vozil. Uporabljajo se le v trgovanju kot rezervni deli. Žal kupci največkrat gledajo le na ceno in kaj malo na kvaliteto. Zato je možno žarnice take kvalitete kupiti tudi pri nas. Resne organizacije kot so Petrol in OMV jih ne nudijo. Zastarela oprema za proizvodno linijo stane v nivoju 15.000 do 30.000 Dem. Druga smer je v visoki avtomatizaciji proizvodnje z vpeljano robotiko, pnevmatiko, hidravliko, senzoriko in elektroniko. Regulacija plinov in vakuuma je natančna in avtomatska. Oprema za novo proizvodno linijo, na kateri pa je mogoče delati le en tip žarnice, stane od 3 do 7 milijonov Dem. Na posamezni liniji sta največ do dva zaposlena v vsaki izmeni. Vsa končna kontrola je avtomatska. Firme s tako opremo proizvajajo žarnice vrhunske kvalitete v 3 izmenah dnevno in s po dvemi 12-urnimi izmenami v soboto in nedeljo. Linije delujejo cca 3-4 hitreje kot pri nas. Produktivnost ene moderne linije je med 3000 in 6000 žarnic kosov na uro. Izmet je minimalen, pod 2 %. Seveda pa je delovna sila v Nemčiji, Franciji in Italiji kar 4-krat dražja kot pri nas in amortizacija opreme zelo visoka. Oprema, ki jo poseduje podjetje Iskra Žarnice, je nekje vmes, žal močno bližje prvi skupini proizvajalcev. Najmlajša proizvodna linija je stara preko 22 let. Slične linije imajo na Poljskem in Češkem. Povsem nove razvojne usmeritve zahtevajo žarnice z daljšo življenjsko dobo. Serijsko se še ne vgrajujejo. Kako dalje v Iskri Žarnice Za najnižji nivo cen žarnic, ki prihajajo iz proizvodnje po prvi usmeritvi, je delovna sila v Sloveniji predraga. Zato smo se zavestno usmerili v kvaliteto prve vgradnje za proizvajalce vozil, kjer je trg sicer 5 x manjši, a so cene za vsaj 50 % višje. Tu proizvajamo za velike svetovne firme kot so Osram, Philips, Narva in Hella pod njihovimi zahtevami kvalitete in njihovimi imeni. Prav tako smo se usmerili v žarnice in na trge, ki zahtevajo spoštovanje IEC predpisov ter proizvajamo žarnice pod tujo blagovno znamko za zahtevne kupce pod njihovimi imeni. Ti so iz Nemčije, Francije, Italije in Španije in so nekoč proizvajali avtožarnice in jih je konkurenčna tekma v produl 35GHz) npn-transistors in combination with a proven 0.8 pm mixed-signal CMOS process. Design rules, process parameters with first simulation models for the npn-transistors and digital CMOS libraries are available for this new technology. DSP Technology at Austria Mikro Systeme Thanks to the compilation of an assembler software library, Austria Mikro Systeme now has a sophisticated DSP (Digital Signal Processor) technology for a wide range of applications at its disposal. DSP technology is based on the GEPARD signal processor family developed at Austria Mikro Systeme. The parameterizabie GEPARD architecture makes short-term provision of DSP cores possible which can be perfectly adjusted to individual applications. The development of circuits is supported by suitable simulation models for the Austria Mikro Systeme HIT kit (High Performance Interface Tool Kit) available for CAD tools from Mentor Graphics, Cadence Design Systems and Synopsys. A suitable assembler, a code simulator as well as a comprehensive, handoptimised assembler software library are available for programming of the GEPARD DSP cores. DSP technology takes the progressive digitalisation of mixed signal applications into account. Austria Mikro Systeme has for several years been setting milestones in digital signal processing in the development and production of mixed analogue-digital circuits. In addition to GEPARD DSP cores, highly-optimised, hardwired signal processor modules, such as digital filters, FFT and DCT modules, are applied. High performance A/D and D/A converters as well as tested analogue modules complement the product range of Austria Mikro Systeme. Application areas include coder-decoders, modems, circuits for audio and speech processing, as well as measuring and sensor signal analysis. The circuits are produced in submicron-CMOS technologies, and the transfer to subhalfmicron-CMOS technologies is currently being prepared. The GEPARD DSP core technology was presented at the leading European Conference in this line of industry, the "Design, Automation and Test in Europe Conference 1998" (D.A.T.E. '98), and awarded the "Best ASIC Prize". Schloss Premstatten A-8141 Unterpremstatten, Austria Fax: +43 (03136) 52 501, 53 650 Te.: +43 (03136) 500-0 E-mail: info@vertical-global. com http: llwww.wertical-globai.com IMEC News OCAPI, a new IC design environment This design environment tries to bring the circuit designer closer to the system designer and increase the design productivity of complex signal processing systems. A DECT transceiver ASIC was designed using OCAPI. OCAPI is based on object-oriented C + + and advanced synthesis methods that enable integrated systems design, simulation and synthesis of advanced signal processing. The system specification is automatically transferred from C + + to VHDL or a compilable C+ + code. OCAPI also enables to refine the design gradually. The object-oriented methodology offers a solution to the re-use of previously designed building blocks. OCAPI offers a number of benefits, such as mixed-mode design evaluation, state-of-the-art synthesis script, elaborated verification strategy, object-oriented modeling for hardware (resulting in faster design of new functions) and the single C++ design environment from the highest untimed level to the bit-true, clock cycle true level. This completely new design environ- ment was applied for the design of a DECT transceiver ASIC which will significantly improve the signal quality between base station and mobile system. The chip processes DECT burst signals, received through a radio frequency front-end RF. These signals are equalized to remove the multi-path distortions introduced in the radio link. The chip has a 194 mm2 die area in 0.7 pm CMOS technology. New IMEC Industrial Affiliation Programs (HAP) in telecommunications High-speed wireless local area networks The objective of this program is to achieve an overall network capacity up to high bitrates. For this purpose, accurate design of all parts of the system, from high level networking down to the hardware implementation of the functional blocks, is required. At the moment, a prototype OFDM (orthogonal frequency division multiplexing) modem ASIC is under development and a 182 UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana demonstration setup based on the WLAN modem chip will be realized by end 1998. Target customers are system houses and microelectronics companies with an interest in WLAN or indoor communication. Multimedia image compression A first-of-a-kind MPEG-4 architecture will be developed, optimized for high performance and low power, in addition to the system-on-chip MPEG-4 architecture, tools supporting the design flow of this family of applications will be developed. Target customers for this program are system houses and multimedia component providers that want to extend their existing expertise with MPEG4 specific implementation knowledge. IMEC cooperates with industry to enhance ASIC and system design skills A program is being set up to increase the number of trained engineers, and to transfer state-of-the-art research experience in the development of embedded systems towards design groups from industry. This initiative tries to give an answer to a number of problems, such as the booming complexity of ICs, the rapid changes in design methodology for complex systems and the important shortage in the market of skilled engineers. The training program consists of 3 classes of training courses: - Use of standard EDA tools, currently available in the market (e.g. VHDL, synthesis, test-pattern generation, DSP systems...) - Concepts for the development of complex embedded systems, such as hardware/software codesign, C + + based hardware design and real-time operating systems. - Custom courses All regular courses are being announced on the website: www.imec.be/training/Welcome.html or Bart De Mey, Tel. +32 16 28 12 49 Fax +32 16 28 15 84 Postacademic Interuniversity Course in Telecommunications Nearly 500 students enrolled in this videoconference course. The initiative, started in September last year, is an excellent example of cooperation between Flemish universities, polytechnical schools and industry The program is spread over two years and intended for renewing and upgrading the knowledge of engineers working in industry The course can be followed simultaneously through videoconferencing at 7 places in Flanders and some sites in the rest of Europe. In the beginning of 1998, IMEC's infrastructure was expanded to enable other organizations with videoconferencing capabilities to participate as a site. For more information, please contact: Tanja Gouverneur, course coordinator Tel. +32 16 28 13 70, Fax: +32 16 28 15 10 e-mail gouverne@imec.be www.imec.be/seminars/itc New semiconductor assembly foundry settles in Flanders Anew European semiconductor assembly foundry company that will be active in advanced packaging technologies has been setup. Custom Silicon Configuration Services NV (CS2) has been established through IT Partners, the private IT oriented venture capital fund. The company has chosen its location at Zaventem, near Brussels, in the vicinity of IMEC. The new company CS2 will provide semiconductor assembly services to major European semiconductor manufacturers, and to US and Asian wafer fabrication plants with products destined for European Original Equipment Manufacturers. The nature of the business will be subcontract manufacturing with a high level of service. Market opportunities Today new packaging and assembly technologies are much more application specific and therefore need full flexibility, rapid cycle time and high reliability This is in contrast with older, higher volume technologies that are traditionally manufactured from a central base in the Far East. The close proximity of a subcontract packaging and assembly company will offer distinct advantages to traditional offshore operations in areas such as overall cost due to time savings and the presence of design services close to the end user. Assembly in Europe or the USA can offer a number of benefits to the semiconductor industry. In state-of-the-art microelectronics technology, labor rates are no longer the competitive edge, but rather flexible automation and rapid cycle times. Time is money especially for custom-designed services. "The extra cost for high-end products to be assembled or packaged in the Far East, amounts to upwards of 15%", says Steve Lerner, CEO of CS2. "The total packaging and assembly cycle time can be weeks for Asian plants, with the actual manufacturing time of maybe 4-5 days, but CS2 can deliver within 72 hours (maybe even 12 hours for small lots of a few hundred), a service that can't be matched by others halfway around the world." Finally, subcontracting, or outsourcing as a part of the overall electronics industry, shows a healthy growth market as shown by the forecast in Fig. 1. I SMC - - I 1S1V ¡20 90 1 9(.i7 111111 MM 9000 8054 60(10 ;...... -1969 MBS HkB : hH| IMSl 3000 S|i|i8i fllllil gB£S| ■(SS| ■■■I 0 .......................................' ' ..... ...EAA^J............................... mi im ms w? 2 wo Z&QI Fig. 1: Evolution of subcontracting in the electronics industry (source: Electronic trends Publication) 183 UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana Technology evolution From Fig. 2 it is clear that the following years will be characterized by a significant growth in array packages such as ball grid arrays (BGA) (64% per year) and chip-scale packages (CSP) (215%). In addition, flip-chip technology, where the bare die is attached to a substrate upside down after the attachment of solder bumps, will grow at a compound annual rate of 26% in the period 1996-2001. CS2 will focus on area-array technologies, such as BGA, Flex based area arrays, CSP and multi-chip modules using both wirebonded and flip-chip interconnect technologies. BOA Bi-.rt> (ilifj C'up-sc.-ili- Totaal Fig. 2: Worldwide IC packaging market forecast (in units for 1996-2001 (source: Dataquest)) Custom Silicon Configuration Services The new company will provide services to European manufacturers that include design and characterization of packages, reliability testing, assembly and selected R&D. "The strategic alliance with IMEC allows us to leverage a worldclass engineering and research organization with our extensive experience in the assembly services", Steve Lerner said. The company was established by the financial backing of IT-Partners, the private IT-oriented venture capital fund (see Newsletter 19), with approximately 10 million USD in the first round. CS2 is currently building the first module of its facility, expects packaging by the beginning of 1999 and will be operating at full capacity within one year of starting production. At that time, the company's size will be about 60 people. Subsequent expansions are planned and will be a function of market demand. The chief executive officer of CS2, Steve Lerner, is a former managingdirectorofAmkorAnam Europe and has over 15 years of experience in the semiconductor subcontract assembly industry. In the heart of Europe The location of Zaventem is less than 15 miles from either Brussels or Leuven and close to the airport. In fact, the time between airport and factory is only 10 minutes! The central location in Europe renders CS2 in close proximity to major European semiconductor manufacturing plants for its potential customers (such as Alcatel Microelectronics, Philips, Siemens and SGS Thomson) and to IMEC for joint R&D work. In addition, the region of Flanders is known for its high quality of life, the multi-lingual and multi-cultural environment, the strong technical talent readily available and the progressive financial community (such as IT-Partners). CS2 NV Hermes Center Hermesstraat 4a, B-1930 Zaventem Tel. +32 2 720 00 00, Fax: +32 2 720 00 01 e-mail: steve.lerner@cs2.be NEW MURATA PRODUCTS Reflow SMD sounder This new Murata piezoelectric sounder is fully reflow compatible and can be soldered like the rest of your SMD components. Another component that you can place and forget. The sound pressure level is 75dB minimum at 10 cm, from a 3V peak-to-peak 4kHz square wave input. Maximum input voltage is 25V p-p and electrostatic capacitance is 14nF at 1 kHz. The sounder is specified for operation over the temperature range -20 to 70°C and perfect for providing on-the-board-sound. Today, microprocessors are used for microwave ovens, air conditioners, cars, toys, timers and many other warning tasks. Externally driven sounders are frequently employed in digital watches, electronic calculators, telephones, and office equipment like note book computers and print- ers. They even appear in automotive equipment. (->ne manY applications 184 UDK621.3: (53+ 54+ 621 +66), ISSN0352-9045 Informacije MIDEM 28(1998)3, Ljubljana The Murata PKMC 16E4000-TY ideal for these functions, being compact 16 mm square with a 2.7 mm maximum height. Packaging is on trays in boxes 1,200 pieces. SPS capacitor selection Switched mode power supply design engineers are perfectly capable of choosing the right capacitor for their circuit; but it takes time. Here we take you straight to the right leaded or chip capacitor and save you the job of having to look them up. For more details or special applications please contact your usual Murata service point. Primwy SecOTdaiy Descripfion i Symbol No. Ht::h VC'kicik cypco-o: C6, CC C9, C9, CIO ,