Mark Krinker Department of Electrical Engineering and Telecommunication Technology, City College of Technology, CUNY, New York, USA

IN-CIRCUIT TESTING CAPACITORS. NEW APPROACHES AND INSTRUMENTS. Infinitely Remote Point Method of Measurement. Part 1. Abstract Quality Q (merit factor) of capacitor is an ultimate characteristic by which one judges about status of the capacitor during operation of electronic equipment and at troubleshooting. The new in-circuit Method drastically relived checking and finding defective capacitors and inductors in electronic equipment. Unlike known in-circuit methods, the proposed one works with only one terminal of the capacitor, or does without the terminals of the capacitor at all. This approach is based on a new way of injecting probing current inside the capacitor.

1. General Conception of the Method The base of the Method was set forth in US Patent 6,198,290, 2001. The following is a further development of the Method. This papers stresses on physical aspects of the method. Details of the related instrumentation will be considered in Part 2. Unlike known methods, both out-of-circuit and in-circuit ones [1], which are based on passing capacitive current between terminals of the capacitor, the current one employs a different approach There are two versions of the method [2]. They have one common feature: a probing current is being injected through the jacket of the capacitor and then passes through the stuffing of the capacitor, collecting the information about its quality. 1.1 Two-Terminals Version of the Method The new method to test capacitors in a circuit is based on replacing conventional way of flowing alternating measurement current through the circuit "meter-terminal of capacitor-plate-dielectric-another plate-terminal-meter" for a new approach: meterenvelop of capacitor-dielectric-plate of capacitor-terminal of capacitor-meter. This allows avoiding influence of any circuit connected to capacitor in parallel and makes direct measurement in a circuit possible.

This version employs two-terminal measurement instrument. Comparative Fig.1 illustrates this approach. Part a illustrates the conventional incircuit method, based on passage of high-frequency AC current through the capacitor, while b illustrates one of the versions of the new approach: injecting the probing current through the body of capacitor, rather than its two terminals. We have to note that shown in Fig.1b implies capacitors with a metal jacket. This can be both electrolytic and non-electrolytic ones. For this purposes, a tip-like probe can be employed. Metal surface of the jacket plays a role of a measurement electrode. For capacitors with an insulating jacket, the probe has a developed conductive surface as this was shown in the Patent. Said surface initiates a displacement current, which probes the media inside the capacitor.

Fig.1 Comparison of conventional in-circuit method for testing capacitors, a, and new measurement technology, b. Employing Two-Terminal version of the Method for in-circuit testing capacitors. Measurement (probing) current in Fig.b is injected through the metal jacket and then passes through the body of the capacitor.

For the insulting jacket capacitors, the upper probe has a developed conductive surface to initiate the displacement current. 1.2 One-Terminal Version of the Method Another version of the method does not use apparent closed circuit of passing measurement current at all. Instead of that, measurement current passes the following way: "metermeasurement electrodeenvelope of capacitor--dielectric--plate

terminal of capacitor--total electric capacitance of all conductive environment, coupled to the terminal. This new approach enabled to create principally new, oneelectrode measurement method and instruments. The nature of formation of measurement probing current is a displacement current, caused by polarization processes both in the inter-plate dielectric and said total conductive environment. This version employs one-terminal measurement instrument. Comparative Fig.2 illustrates One-Terminal version of the Method.

Fig.2. Comparison of the conventional in-circuit method for testing capacitors, a, and One-Terminal version of the new measurement technology, b

Like Fig.1, this illustration implies conductive jacket capacitors. For the insulated jacket capacitors, the probe has a developed conductive surface. 1.3 Adequacy of the Method To be successful, new method has to be adequate with the commonly accepted. This is furnished by the following: Both the losing and saving energy in capacitors depend on physical processes in a media inside the body of the capacitor. Consequently, disregarding the way of passage probing current through the capacitor, a lost-to-saved energy ratio (i.e. tangent of dielectric loss angle, tg=1/ ) remains the same. =1/Q =1/

The first mentioned approach is realized by means of applying AC voltage across envelop of capacitor and its either terminal. This creates a probing current, whose characteristics depend on physical processes inside the capacitor being tested.

Following from the stated above, tg tt = tg et (1)

where tt means classical terminal-to-terminal method, while et means the new, envelop-to-terminal approach. Following from the equity of dielectric losses ratios, with regard to tg = CR , we have C tt Rtt = C et Ret (1a)

1.4 Electric Equivalent of One-Terminal Version of the Method. An Infinitely Remote Point. In the One-Terminal conception, the displacement current to probe capacitors is formed due to involvement of total conductive surface of circuitry environment. Fig.3illustrates this in a simplified form.

Fig.3 Simplified electric equivalent of the one-terminal version of the new method. Unlike classical closed circuits, this one looks open.

Said conductive surface can be formed by a chassis and/or total surface of interconnected circuitry. Fig.4 shows a general, classic closed-circuit electric presentation of the unit diagram of the Fig.3. The measurement instrument comprises AC voltage generator and a meter. The meter is a generalization of any unit, which measures either of AC characteristics.

The measurement instrument forms AC electric field inside capacitor. Due to developed conductive surface coupled to the capacitor, said AC electric field forms the current, which alternatively charges and then discharges the conductive surface. This current flows through the capacitor, represented by connected in series C and R, performing probing function. The instrument analyzes characteristics of the current and converts them into output reading. Injecting probing current inside the capacitor can be done by two ways. First of them is a touching metal envelop of the capacitor with conductive tip connected to the measurement instrument, like its shown in Fig.3. This way works goodly for electrolytic and some types of non-electrolytic capacitors. For capacitors with a non-conductive envelop, the probe has to have a developed conductive surface as this was shown in the patent. The total conductive surface of the Fig.3 consists of two surfaces: total zero-potential surface of the measurement instrument and the conductive surface the tested unit. As follows from said above, said version is an one-terminal method. Fig.4 shows an electric equivalent of the Method according to closed circuit form.

Fig.4. Closed-circuit electric equivalent of one- terminal measurement conception of Fig.3. The Infinitely Remote Point coupling converts the open circuit of Fig.3 into classic closed circuit.

The circuit consists of AC voltage generator, which is coupled to RC circuit of a measurement instrument and a metal surface, a terminal and a capacitor being tested, which is coupled to a chassis with its other terminal. As applied to One-Terminal Method, the capacitor consists of internal resistor Ret, marked as R in Fig.4, connected in series with capacitance Cet (C in Fig.4). In terms of electrical equivalency, conductive surfaces of the measuring instrument and that of the tested unit can be shown as two capacitors C1 and C2 connected in series at an infinitely remote point P. The validity of this definition outcomes from a definition of the electric potential of a given point: this is a work done by a system to bring unit of a positive charge from infinity to the given point of the field. In the Method, energy of the AC generator moves electric charges back and forth through the capacitor from the probe

to infinity, because there are no a second pole like it takes place in conventional circuits. This is why we also call the Method as an Infinitely Remote Point Method. This definition does not necessarily mean that this point is located at infinity. Our purpose is to establish a correlation between set of parameters of traditional methods and that of the new approach. Ration of dielectric losses of combined series circuit, comprising internal resistor Ret, total capacitance of conductive external environment Cext and that internal envelop-to terminal capacitance C et , can be shown as tg ext = Ret Cext , because Cext << Cet due to fact that resulting capacitance of connected in series C1 and C2 (that is Cext) << C (that is Cet) From here, taking into consideration (1a), we have tg tt tg ext Cet / Cext (2)

where tg ext is a tangent of a dielectric loss angle measured with one-terminal measurement, and Cext is a total capacitance of all the conductive environment to which the capacitor is coupled. Equation (2) is very important for the Method because it bridges between traditional approach and the new one. Studying this method has shown that for majority of capacitors Cet~0.3-2.0 uF, Table 1, while capacitance of metal chassis represented by Cext vary within tens of pF. This allows us to use Cet / Cext as a known coefficient, which simplify the problem. Because of what a measurement instrument really reads is tg e , its value will be always lower than tg tt so, a scale has to be calibrated with regard to a ratio (2). Because Cext<<Cet, the major components which define the value of tg ext are Ret and Cext . tg ext = Cext Ret (3) While Cext remains invariable within a tested unit, Ret depends on C et which, in turn, is defined by a size of capacitor at invariable tg . Table 1 shows C et vs. total surface of capacitor S dependence. As seen from the table, for the majority of capacitors, variation of C et remains within the limited range of quantities. Only for high capacitance it differs as much as 2-4 times, that can be compensated technically. Because Cext is defined by total capacitance of the conductive environment of capacitor and capacitance of measurement instrument, there is always a chance to change the last one to compensate Cext.

Table 1. Parameters of electrolytic capacitors, related to the discussed method

Ctt, voltage Size, r*h,mm 1uF250V 6.5*12 100uF35V 6.5*12 47uF35V 6.5*11.5 2.2uF250V 8*12 10uF50V 10.5*5 2.2uF350V 12*10 220uF100V 12.5*25 100uF200V 16*26.5 2200uF50V 18*36 Cet,uF 0.407 0.668 0.417 0.286 0.435 0.465 0.398 0.75 1.833 Company Elite Elna GE Elna Elna JH Elna Elna Elna Total S, mm^2 623 623 602 804 408 1068 2453 3467 5087 Cet/S 6.54 10.73 6.93 3.56 10.66 4.36 1.62 2.16 3.6

This has to be said that Ret is nothing but an equivalent Series Resistance, ESR, measured here between envelop and terminal of the capacitor. Therefore, we can measure the conventional tangent of dielectric loss angle of capacitor, employing non-conventional one terminal technology not related to terminal of the capacitor at all. Fig. 5 shows an example of practical employment of the proposed one-terminal method.

Fig.5. In-circuit testing electrolytic capacitors with the new one-electrode technology and one of the related instruments

It might be well to point out contribution of the capacitance of an operator and its introduced energy loss. 1.6. Displacement Current Simulator Reading the measuring instrument, based on the one-terminal technology, depends on ESR (that is Ret) of the tested capacitor and availability of the displacement current passing through the body of capacitor from the meter to a chassis and other conductive environment. By default we consider that at least one terminal of the capacitor has a good contact with the chassis. Thats true for a wide class of capacitor applications. However, the less percentage of their employment is a case when no terminal of the capacitor is directly coupled to the chassis. In this case there is a resistance between the terminal and the chassis. For the measuring instrument, its equivalent to having increased ESR that can come to the false reading. To avoid this, conception of the Displacement Current Stimulator, DCS, was proposed. The DCS is any technical means that forms the efficient displacement current disregarding a quality of the real conductive environment of the capacitor. Examples of DCS are shown in Fig.6. The a-version is a conductive insulated body, contacting either terminal of the capacitor. The probing displacement current flows through the capacitor and the conductive insulated body rather that through the resistor, which can cause a false reading. The insulator around the body allows an operator to hold it in hands, increasing its capacitance (therefore, the displacement current) and protecting the circuit against introduced energy loss associated with the operator. The b-version employs a soft conductive surface like a steel wool, contacting terminals of the capacitor. Therefore, the probing displacement current simply bypasses the resistor.

Fig.6. Employing Displacement Current Stimulator, DCS, to avoid a false reading of the oneterminal instrument caused by a lack of proper contact between a terminal of the capacitor and the

chassis. Here: a a manual DCS having elastic contact for connecting to either terminal of the capacitor: b a pad-like DCS, which provides the effective probing current due to pretty developed surface. For the case b its enough to put the printed-circuit board (non-powered and discharged first!) on this DCS.

1.7. Advantages of the Method and Related Instruments When compared to existing in-circuit methods and instruments, the proposed one has the following advantages: Operation with one probe. The existing methods and in-circuit instruments require having two probes; No needs to contact terminals of the tested capacitor. The existing methods require connecting to the terminals of the capacitor, what is associated with disassembling the unit subject to the test. In the proposed method, a user simply touches a cap rather than a terminals of the capacitor; Expressivity of the method: just a couple seconds are needed to touch the envelop of the capacitor; Possibility to work in a powered unit: this is a first method and instrument allowing this, because a contour of the probing current mismatches with the operating current. This open the way for monitoring capacitors under exploitation. Working in the powered units, however, requires general safety measures; No needs to know capacitance of the capacitor. All existing in-circuit methods, based on ESR require knowing value of the capacitance: for instance ESR= 3Ohm is good for 1uF capacitor, but isnt acceptable for 1000 uF capacitor. Frequently, its impossible to read the value of capacitance without taking special measures, making the repair process hard. The absence of neediness to know the capacitance is based on the practical fact of that a transient capacitance Cet, involved in the measurement, practically does not follows its real capacitance Ctt, Table 1. In year 2001 the method and instrument were successfully tested in NASA Long-standing exploitation of the method and associated instrument in the computerrepair company confirmed high efficiency of this technology which saved a lot of time in repairing process. References 1. http://personales.ya.com/old-timers/eico955/eico955.htm 2. M.Krinker. Method to Detect Defective Capacitors in Circuit and Meters for That. US Patent 6,198,290. 2001.