CA1280164C - Electronic battery testing device - Google Patents

Abstract

ABSTRACT

An improved self-contained electronic device for testing storage batteries and other dc sources is disclosed. The testing device accurately performs small-signal measurements of the battery's dynamic conductance and provides for displaying either a numerical reading proportional to the battery's available power, or a qualitative assessment of the condition. Special design features permit powering all active circuit elements by means of common power terminals and allow circuit implementation with medium scale integrated (MSI) circuits.

Description

1~80i64 Technlcal Fleld Thls lnventlon relates to an electronlc measurlng devlce tor assesslng the ablllty of a storage battery or other dc source o~
electrlclty to dellver power to a load. ~ore speclflcally, lt relates 5 to lmproved apparatus of the type dlsclosed ln U.S. Patent ~,87~,91 1, ELECTRONIC BATTERY TESTING DEVICE, lssued to Kelth S.
Champlln, March 25, 1975, and of the type dlsclosed ln U.S. Patent 3,909,708, ELECTRONIC BATTERY TESTING DEVICE, lssued to Kelth S.
Champl ln, September ~0, 1975.

~gund Art Storage batterles are employed ln many appllcatlons requlrlng electrlcal energy to be retalned for later use. Most commonly, they 15 are employed ln motor vehlcles utlllzlng lnternal combustlon englnes. In such appllcatlons, energy stored ~y "charglng" the ~attery durlng englne operatlon ls later used to power llghts, radlo, and other electrlcal apparatus when the englne ls stopped. The most severe demand upon the battery of a motor vehlcle ls generally made 20 by the self-starter motor. Typlcally, several k~lowatts of power are requlred by the self-starter motor to crank the englne. Fallure to satlsfactorlly accompllsh thls task, partlcularly ln cold weather, ls usually the flrst lndlcatlon o~ battery deterloratlon or trou~le wlth the charglng system. Clearly, a simple measurement that .

~<~30~4 accurately assesses a battery's a~lllty to supply power is of conslderabte value Prlor to the pu~llcatlon ot U.S. Patents 3,873,911 and ~,909,708, a ~attery's ablllty to supply power was customarlly assessed by means o~ a load test. A load test sub~ects a battery to a 5 heavy dc load current havlng a predetermlned value dlctated by the battery's ratlng and temperature. After a prescrlbed tlme lnterval, the battery's termlnal voltage under load ls observed. The battery ls then consldered to have ~passed" or "~alled" the load test accordlng to whether thls termlnal voltage ls greater than, or less than, a lO partlcular prescrlbed value. Although the load test has been wldely used tor many years to tleld-test storage batterles, lt possesses several serlous dlsadvantages. These lnclude:
1 Currents drawn are very large and theretore requlre apparatus that ls heavy and cumbersorne.

2. Because ot these large currents~ conslderable ~sparklng" can occur at the battery termlnals lr the test apparatus ls connected or dlsconnected under load condltlons. Such nsparklng~ ln the presence o~ battery gasses can cause an exploslon wlth potentlally serlous in~ury to the operator.

3. A load test leaves the battery in a signlficantly reduced state of charge and therefore less capable o~ cranklng the englne than before the test was per~ormed.

4. Slnce the battery's termlnal voltage drops contlnuously wlth tlme durlng the load test, the test results are lmpreclse and 25 greatly dependent upon the sklll or the operator.

~0164 5 Load test results are not repeatable slnce the test ltself temporarlly polarlzes the ~attery. Such test-lnduced polarlzatlon slgnltlcantly alters the lnltlal condltlons of any load tests pertormed subsequently.
A practlcal alternatlve to the common load test ls taught ln U.S. Patent ~,873,91 l and U.S. Patent 3,909,708. Both o~ these patents dlsclose electronlc apparatus tor accurately assesslng a battery's condltlon by means ot small-slgnal ac measurements o~ lts dynamlc conductance. These two patents teach that a battery's 10 dynamlc conductance ls directly proportlonal to lts dynamlc power;
the maxlmum power that the battery can dellver to a load. ~ynamlc conductance ls theretore a dlrect measure ot a battery's electrlcal condltlon. Vlrtually mllllons ot battery measurements pertormed over the course ot thlrteen years have tully corroborated these 15 teachlngs and have proven the valldlty or thls alternatlve testlng method.
In comparlson wlth the load test method ot battery appralsal, the dynamlc conductance testlng method taught ln U.S. Patents 3,873,9 l l and 3,909,708 has many advantages. For example, 20 dynamlc conductance testlng utlllzes electronlc apparatus that ls small and llghtwelght, draws very llttle current, produces vlrtually no ~'sparklng" when connected or dlsconnected, does not appreclably dlscharge or polarlze the battery, and ylelds very accurate, reproduclble, test results.
Two electronlc battery tester embodlments are dlsclosed ln U.S. Patent 3,873,9l l; each ot whlch accurately determlnes a ~0164 battery~s dynamlc conductance and provldes the operator wlth a numerlcal readlng that ls dlrectly proportlonal to thls quantlty. The flrst embodlment comprlses a brldge clrcult that ls brought to ~alance by the operator to obtaln the numerlcal readlng. The s preterred second embodlment provldes the operator wlth a dlrect readout that may be dlsplayed numerlcally on a dlgltal or analog meter. ~he operatlng prlnclples o~ tne pre~erred, dlrect-readlng, second embodlment ot the lnventlon taught ln U.S. Patent 3,873,9l l are based upon the theory of hlgh-galn feedback ampll~lers.
o U.S. Patent 3,909,708 llkewlse dlscloses two electronlc battery tester embodlments. However, ~rom the operator's polnt ot vlew, thelr operatlon more closely resembles the operatlon ot a tradltlonal load-test apparatus than does operatlon ot elther ot the numerlcal-readlng em~odlments dlsclosed ln U S Patent ~,87~,9l 1 .
15 Rather than obtalnlng a numerlcal measurement, the operator makes prellmlnary adJustments to knobs on the panel ot the apparatus;
settlng them to the electrlcal rat~ng and temperature o~ the battery undergolng test. The dlsclosed apparatus then employs small-slgnal measurements o~ dynamlc conductance to slmply ascertaln whether 20 or not the battery ls capable o~ dellverlng an amount o~ power approprlate to the battery's ratlng and temperature. Accordlngly, the two embodlments dlsclosed ln U.S. Patent 3,909,708 provlde slmple "pass-tall~ battery condltlon lnformatlon, Just as does conventlonal load test apparatus. However, they accompllsh thls 25 result wlthout drawlng a large current trom the battery and are therefore not sub~ect to the serlous dlsadvantages of a load test.

301~i4 Just as wlth the second embodlment dlsclosed tn the earller patent, the operatlng prlnclples of the second, preterred, embodlment dlsclosed ln U.S. Patent 3,909,708 are based upon the theory o~ hlgh-gatn ~eed~ack amplltlers.
Both preterred embodlments of electronlc battery testlng 5 devlces, the second embodlment dlsclosed ln U.S. Patent ~,873,9 l l and the second embodlment dlsclosed ln U.S. Patent 3,909,708, are ~ased upon feedback ampllfler prlnclples. The orlglnal lmplementatlons ot these electronlc ~attery testlng devlces ~oth utlllzed contemporary solld state devlce technology. Such technology was, however, llmlted to only dlscrete devlces such as blpolar translstors and dlodes, and small-scale lntegrated (SSI) clrcult verslons ot slngle-element monollthlc operatlonal amplltlers.
Great advances have been made ln sol~d-state lntegrated 15 clrcult (IC) technology durlng recent years. In partlcular, hlgh performance complementary metal-oxlde-semlconductor (C~10S) and blpolar r~edlum-scale lntegrated (~1SI) clrcults, such as dual and quad operatlonal amplltlers, have ~ecome a~undantly avalla~le at very low prlces. Therefore, important advantages, of both technlcal 20 and economlc natures, could currently be reallzed by exploltlng thls newer, more advanced, solld-state devlce technology ln the electronlc ~attery testlng art Unfortunately, a num~er of deslgn consideratlons preclude the slmple lntroductlon of the newer IC technology lnto the teedback-25 ampllrler type o~ electrontc ~attery tester clrcultry dlsclosed ln 1~ ~3O~6L~

U S. Patents 3,873,9 l t and ~,909,708. Foremost among theseconslderatlons are the var~ous pro~lems lmposed ~y the tact that the commerclally avallable C~OS and ~lpolar MSI iCs do not provlde separate pln-outs tor supplylng power to the lndlvldual elements on the chlp. However, the orlglnal dlscrete-element feedbac~ amplltler 5 deslgns relled heavtly upon the avallablllty of such separate power connectlons. In partlcular, the orlglnal deslgns requ~red separate connectlons ~or supplylng power to dlt~erent actlve devlces ln order to lmplement "tour-polnt probeU archltecture and ttlere~y ellmlnate the spurlous reslstance ot the connectlng leads and ~attery contacts 10 ~rom the measurements; ln order to reallze a preclsely-leveled osclllator voltage and thereby obtaln lncreased measurement accuracy; and ln order to lmplement synchrono~ls detectlon ot the arnpllfled osclllator slgnal and thereby suppress measurement errors caused by spurlous plckup ot hum and nolse. Accordlngly, 15 ma~or changes ln the bas~c deslgn o~ the electronlc battery tester embodlments would be requlred before one could reallze any ot the potenttal techn1cal and economlc ~enetlts assoclated wlth the newer, more ettlclent and more cost-eftectlve, IC technology.

~ummary ot the Inventlon The lmproved electronlc ~attery testlng devlce ln accordance wlth the present lnventlon lncorporates the ~unctlons ot both of the 25 two earller-dlsclosed feedback-type electronlc battery testlng 1~016~

devlces ln a slngle embodlment. By vlrtue of novel deslgn lnnovatlons dlsclosed hereln below, tne need ~or separate power connectlons to dlf~erent actlve devlces ls ellmlnated, thus permlttlng the successful lntegratlon of CMOS and ~lpolar ~ISI ICs lnto a practlcal battery tester lmplementatlon. Thls results ln the reallzatlon o~ a very slmple electronlc battery testlng devlce that ls relatlvely lnexpenslve to manufacture, but whlch provldes a very hlgh degree ot measurement accuracy. Dlsclosed lnnovatlons, whlch clrcumvent the need tor separate connectlons to supply power to dltterent actlve elements, lnclude:
o l. A novel technlque ~or ln~ectlng the osclllator slgnal lnto the reedback ampll~ler~s lnput clrcult The dlsclosed tn~ectlon technlque permlts utlllzlng a common power source tor both the osclllator and the teedback amplltler. Nevertheless, lt malntalns suttlclent lsolatlon ~etween the ampll~er's lnput and output 15 clrcults to allow the spurlous reslstance ot the ~attery leads and contacts to be ettectlvely ellmlnated trom the measurements by means of ~tour-polnt probe~ archltecture.
2. A novel, preclsely-leveled, osclllator lmplementatlon employlng an operatlonal ampllfler, a C~10S bllateral analog swltch 20 and a zener dlode. Thls slmple clrcult utlllzes an unregulated, common, power supply ~ut provldes an output slgnal of preclsely malntalned amplltude; thus permlttlng hlgh measurement accuracy.
3. A novel synchronous detector lmplementatlon employlng a C~lOS bllateral analog swltc~ along wlth an operatlonal ampll~ler 25 employed as an lntegrator. Thls slmple clrcult also obtalns power ~Ol~i through common connectlons whlch supply power to all other actlve devlces. It ls theretore powered contlnuously and thus dlffers tundamentally trom the synchronous detector clrcult dlsclosed earller ln U.S. Patent 3,909,708 WhlCh requlres that lts power source be lnterrupted perlodlcally at the osclllator rrequency.
Nevertheless, the new synchronous detector lmplementatlon provldes accurate, llnear detectlon of the ampllrled osclllator slgnal whlle effectlvely suppresslng externally generated hum and nolse that ls uncorrelated wlth the slgnal generated by the battery tester's lnternal osclllator.
o The lmproved electronlc battery testlng devlce hereo~ can be used tor obtalnlng elther a qualltatlve or a quantltatlve assessment ot a wlde varlety ot dc energy sources. In addltlon to automotlve-type batterles, the lnventlon can be used to test many other dc energy sources such as other types ot lead-acld batterles as well as 15 nlckel-cadmlum batterles; llthlum batterles; solar ~atterles; ruel cells; thermo-etectrlc generators; thermlonlc generators; and magneto hydro-dynamlc generators. The lnventlon hereot Is wldely appllcable to testlng SUCh dc energy sources by vlrtue ot lts slmpl~clty, lts sarety~ lts accuracy, lts ease or operatlon, and lts 20 lOw cost.

Brlet Descrletlon or the Drawlngs Flg. I ls a slmpl~fled block dlagram ot an lmproved electronlc 25 battery testlng devlce ln accordance wlth the present lnventlon.

~301~4 Flg. 2 ls a slmpllfled schematlc dlagram of a sectlon of the block dlagram ot Flg. l dlscloslng four-polnt pro~e~ archltecture for lnterconnectlng the hlgh-galn amplltler, the osclllator, and the battery undergolng test.
Flg. 3 ls a stmplltled schematlc dlagram, slmllar to that dlsclosed ln Flg. 2, but showlng connectlons requlred ~or provldlng osclllator power ln accordance wlth the teachlng of U.S. Patents ~,873,9 l l and 3,909,708.
Flg. 4 ls a slmpllfled schematlc dlagram, slmllar to that dlsclosed ln Flg.2, but showlng connectlons tor provldlng osclllator power ln accordance wlth the present lnventlon.
Flg. 5 ls a slmpll~led schematlc dlagram employed ln analyzlng measurement errors due to loop-coupllng lntroduced by the slgnal InJectlon clrcult ot flg. 4.
Flg. 6 ls a schematlc dlagram ot an osclllator clrcult provldlng a preclsely leveled output slgnal ln accordance wlth the present lnventlon.
Flg. 7 ls a schematlc dlagram ot a contlnuously-powered synchronous detector clrcult ln accordance wlth the present lnventlon.
Flg. 8 ls a set of plots showlng voltage wave~orms and tlmlng relatlonshlps at varlous locatlons ln the synchronous detector clrcult of Flg. 7.
Flg. 9 is a schematic diagram of an ad~ustable dc ampllfler and output meter clrcult ln accordance wlth the present lnventlon.
Flg. 10 IS a complete schernatlc dlagram ot an lmproved Il o~64 electronlc battery testlng devlce for testlng 1 2-volt automotlve batterles ln accordance wlth the present lnventlon.

petalled DescrlDtlon Referrlng tlrst to Flg. 1, a slmpllfled block dlagram of an lmproved electronlc ~attery testlng devlce ln accordance wlth tne present lnventlon ls dlsclosed. Slgnals representatlve of the slgnal at output 10 ot hlgh-galn ampll~ler cascade 12 are ~ed back to lnput 20 of hlgh-galn ampllfler cascade 12 ~y means of two teedback o paths; lnternal teedbackpath 14andexternal feedbackpath 16.
Internal teed~ack path 14 lncludes low pass fltter (LPF) 18 and eeds a slgnal dlrectly back to lnput 20 ot hlgh-galn amplltler cascade 12. The purpose ot lnternal teedbac~ path 14 and low pass fllter 18 ls to provlde large dc teedback but very llttle ac teedback 15 ln order to tlx the operatlng polnt of hlgh-galn amplltler cascade 12 and lnsure lts dc stablllty wlthout appreclably reduclng lts ac voltage galn. External feedback path 16 contalns reslstlve network 22 and feeds a slgnal back to the battery undergolng test 24.
Summatlon clrcultry 26 com~lnes the resultlng slgnal voltage 28 20 developed there~y across battery 24 wlth a 100 Hz perlodlc s~uare-wave slgnal voltage 30 provlded ~y osclllator 32 through reslstive : attenuator network ~4. The resultlng composlte slgnal 36 ls capacltlvely coupled to lnput 20 of hlgh-galn ampllfler cascade 12 y means o~ capacltlve coupllng network 38.
As ls tully expla~ned ~elow wlth reterence to Flg. 2, the ; 12 :: :

016~

voltage at output 10 of hlgh-galn ampllfler cascade 12 comprlses a constant dc blas component along wlth an ac slgnal component that ls proportlonal to the dynamlc conductance of the battery undergolng test 24. The constant dc b~as component ls lgnored whlle the ac slgnal component ls detected and accurately converted to a dc slgnal 5 voltage by synchronous detector 40 comprlslng analog swltch 42 and lntegrator 44. Synchronous detector 40 functlons by perlodlcally turnlng analog swltch 42 on and ott by means ot a slgnal derlved from osclllator ~2 and communlcated to the control lnput of analog swltch n through synchronlzatlon slgnal path 46. The resultlng perlodlcally-swltched slgnal ls then smoothed by lntegrator 44 . By vlrtue or the swltchlng ln synchronlsm wlth the slgnal generated by osclllator ~2, the dc slgnal at output 48 ot lntegrator44 ls proportlonal to the level ot any ac slgnal component at output 10 or ampllfler cascade 12 that ls tully correlated wlth the slgnal 15 generated by osclllator 32. However, lt ls not ettected by any spurlous ac slgnal components, such as ac hum and nolse, that are uncorrelated wlth the perlodlc slgnal generated by osclllator 32.
The smoothed dc slgnal at output 48 ot lntegrator 44 ls passed through ad~usta~le reslstlve network 50 and applled to the lnput of 20 dc-coupled operatlonal ampllfler 52. Feedback path 54 o~
operatlonal ampllfler 52 contalns dc mllllameter 56. Accordlngly, the readlng of dc mllllarneter 56 ls proportlonal to the dc slgnal level at the output 48 of lntegrator 44, and hence to the dynarnlc conductance of battery 24; whlle the constant ot proportlonallty ls 25 determlned by the value of reslstlve network 50.

1~0164 By ut~llzlng an approprlate tlxed reslstance value ln reslstlve networ~ 50 and then callbratlng mllllameter 56 ~n unlts proportlonal to the battery's dynamlc conductance, the embodlment dlsclosed ln Flg. 1 w~ll emulate the dlrect readlng battery tester dlsclosed ln U.S.
Patent 3,873,91 l. In addltlon, as ls shown below w.lth reterence to 5 Flg. 9, the reslstance value ot reslstlve network 50 whlch brlngs the readlng o~ dc mllllameter 56 to a partlcular flxed value ls dlrectly proportlonal to the dynamlc conductance ot battery 24 . U.S. Patent ~,909,708 rurthermore dlscloses that the dynamlc conductance ot a battery that ls capable of dellverlng l 009~ o~ lts rated power ls essentlally proportlonal to tts ratlng ln conventlonal battery ratlng unlts such as ampere-holJrs (AH) or cold-crank amperes (CCA).
Hence, by llnearly callbratlng reslstlve network SO ln battery ratlng unlts, and then deslgnatlng "pass" and "fall" reglons on the tace o~
mllllameter 56, the embodlment dlsclosed ln F~g. l wlll also 15 emulate the "pass-tall" battery testlng devlce dlsclosed ln U.S.
Patent ~,909,708. Accordlngly, by employlng a sw~tch to select elther a flxed-valued reslstlve network SO or an ad~ustable-valued network 50 that ls llnearly callbrated ln battery ratlng unlts, and then provldlng both a llnear scale and "pass-tall" reglons on the tace 20 t mll~lameter 56, one can reallze each o~ the tunctlons ot the two earller-dlsclosed electronlc battery testlng embodlments wlth a slngle devlce.
Referrlng next to Flg. 2, a slmplltled schematlc dlagram ot a sectlon of the block dlagram of Flg. l ls dlsclosed. Operatlonal 25 amplltler Al along wlth its dc blaslng reslstors Rl, R2, and R3, and 1~30164 translstor Ql connected as an em~tter followerl comprlse hlgh-galn ampl lf ler cascade l 2 of Flg. l . In addltlon, reslstors R4 and R5 along wlth capacitor C3 comprlse low pass fllter l 8; reslstor R6 comprlses reslstlve network 22; and capacltors C l and C2 comprlse capacltlve coupllng network 38. Battery 24 ls represented ~n Flg. 2 5 by lts equlvalent clrcult comprlslng a battery emf VB ln serles wlth an lnternal battery reslstance Rx. The perlodlc square-wave slgnal presented to summatlon clrcultry 26 ~y osclllator 32 at output 30 o~
reslstlve attenuator 34 ls represented by voltage Vln ln flg. 2.
Summatlon clrcultry 26 comprlses the sertes lnterconnectlon o~
voltage Vin and the voltage developed across battery 24 as sensed by the two connecttons C and D contactlng battery 24.
Stlll reterrlng to flg. 2, dc blas condltlons wlll tlrst be derlved. The dc blas voltage at the nonlnvertlng (~) lnput ot operatlonal amplltler Al ls establlshed by the voltage dlvlslon 15 between reslstors Rl and R2. The lnput lmpedance o~ operatlonal ampll-ler Al can be assumed to be much larger than reslstance R3.
Under such circumstances, the dc voltage across R3 ls negllglble and the dc voltage at the nonlnvertlng lnput, measured wlth respect to the negatlve termlnal of the battery, ls equal to V0 (R 1 ~R2) ( l ) Res~stors R4 and R5 provide an lnternal dc reedback path ~rom the emltter ot al to the lnvertlng (-) lnput ot Al. The resultlng ~3016~

negatlve dC ~eedback along wlth the very hlgh galn of the ampll~ler cascade causes the lnvertlng (-) lnput of Al to assume the same dc blas voltage as the nonlnvertlng (+) lnput. By agaln assumlng the lnput lmpedance ot operatlonal ampllfler Al to ~e very large, one tlnds that vlrtually no voltage drop occurs across reslstors R4 and 5 R5. Accordlngly, the emltter o~ Ql assumes the same dc blas voltage as the lnvertlng lnput o~ Al. The dc component of the output voltage ls there~ore Vout(dc) - V0 (Rl~R2) (2) The dC blas analysls carrled out above shows that translstor Ql operates as a class-A emltter tollower amplltler and has a dc blas current glven ~y o, Vout (dc) , V~ R2 (3) R6 (R l ~R2)R6 In addltlon to the dc blas component glven by equatlon (2), the output voltage VWt atso contalns an ac slgnal component. The low-pass ftlter comprlsed of C3, R4 and R5 effectlvely attenuates ac 20 output slgnals and prevents them frorn passlng through the lnternal dc feedback path. Accordlngly, the ac component of the output slgnal wlll be essentlally determlned by the negatlve teedback pl~ovlded by the external ~eedback c~rcuit.
Referrtng agaln to Flg. 2, one sees thàt an ac current 1~0~64 proportlonal to the ac slgnal component of Vout ls passed through the ~attery by means of an ac feed~ack-current 10QP comprlslng translstor Q 1, reslstor R6, battery reslstance Rx, and conductors leadlng to battery contacts at A and B. Thls ac teedback current ls equal to VOut (ac) (4) The resultlng ac slgnal voltage developed across the battery o reslstance Rx ls (lf Rx)~ Thls ac voltage ls sensed at battery contacts C and D and added ln serles to the ac slgnal voltage Vln dertved trom the square-wave osclllator output. The composlte ac slgnal voltage ls then capacltlvely-coupled to the dltterentlal lnput ot Al by means ot the two coupllng capacltors Ct and C2.
The lnput voltage-senslng loop comprlses battery reslstance Rx along wlth ac slgnal voltage V~n, capacltors Cl and C2, the dltterentlal lnput ot operatlonal amplltler Al, and conductors leadlng to ~attery contacts at C and D. One sees that the lnput voltage-senslng loop and the output feed~ack-current loop are 20 separate trom one another ~ut are coupled together ~y virtue ot thelr one shared element -- the battery reslstance Rx.
In vlew o~ the ac negatlve feed~ack and the very large ac galn of the biased amplltler cascade, the total ac slgnal voltage applted to the dlfferentlal lnput of operatlonal ampllfler Al ls essentlally 25 zero. Hence, the ac slghal voltage developed across Rx ls very nearly 1~0164 equal ln magnttude, but opposlte ln slgn) to the applled ac slgnal voltage Vln. Accordlngly, one can wrlte lf Rx ~ ~Vln (S) 5 Comblnlng equatlons (4) and (5) and solvlng tor VOut(ac) leads to VOut(ac) = ~ (R6 Vln ) Gx (6) where G~ ls the battery's dynamlc conductance measured ln Slemens One sees from the ac analysls carrled out a~ove that the 5 magnltude ot the ac slgnal component at the output ot the clrcult dlsclosed ln Flg. 2 ls dlrectly proport10nal to the dynamlc conductance, Gx, ot the battery 24 undergolng test.
Flg. 2 dlscloses that the battery testlng apparatus makes tour separate connectlons to the battery 24 undergotng test. Two of 20 these connectlons, at A and C, lndependently contact the posltlve (~) termlnal ot ~attery 24. The other two connectlons, at B and D, lndependently contact the negatlve (-) termlnal. Thls speclal four-conductor contactlng arrangement constltutes "~our-polnt probe"
archltecture. Its purpose ls to effectlvely lsolate the output 25 feedbac~-current loop trom the lnput voltage-senslng loop except ror t~e deslred coupllng provlded by the one shared element, Rx. The "four-polnt probe" archltecture descrl~ed hereln ls a solutlon to the severe measurement problem that results trom the tact that the lnternal reslstance ot a typlcal automotlve-type battery ls extremely small (- 0.005 ohms) compared wlth the spurlous s reslstance of pract~cal battery contacts and connectlng wlres --whlch often total several ohms. In practtce, "four-polnt probe"
connectlons to the battery may slmply comprlse temporary connectlons lmplemented wlth speclal two-conductor sprlng cllps of the type dlsclosed ln U.S. Patent ~,873,911.
o As ls dlsclosed above, the ac teedback current lf passes through the battery by means o~ the "tour-polnt pro~e~ contacts at A
and B. Equatlon (4) dlscloses that the value ot current lf ls determlned completely by the ratlo ot VOut(ac) to R6. Accordlngly, several ohms o~ addltlonal spurlous reslstance lntroduced lnto the 15 teedback-current loop by the leads and contacts at A and B wlll not alter the relatlonshlp between lf and VOut(ac) and wlll theretore not e~tect measurement accuracy. As ls turther dlsclosed above, the ac feedback voltage developed across the battery ls lndependently sensed at "four-polnt probe~ contacts C and D. Slnce the lmpedance 20 of the ampllfler's lnput circult ls of the order of many thousands of ohms, a few ohms of addltlonal spurlous reslstance ln the lnput voltage-senslng loop at contacts C an D wlll llkewlse nave negllglble effect on the measurements. If, however, contact were made to elther battery termlnal at a slngle contact polnt, any 25 spurlous contact reslstance and lead wlre reslstance would be 3016~

common to ~ot~ the teed-~ack current loop and the voltage-senslng loop and would there~ore add dlrectly to the measured value of Rx One sees that the "four-polnt pro~e" archltecture descrlbed a~ove separates the spurlous elements o~ the feedback-current loop from the voltage-senslng loop thUs perrnlttlng accurate ~attery 5 conductance measurements to be obtalned even though the lnterconnecttng leads and contacts may themselves have resistances that are l~undreds of tlmes larger than the battery's lnternal reslstance, Rx However, ln order tor such ~our-polnt probe"
archltecture to ~unctlon ettectlvely, a very hlgh degree o~ clrcult lsolatlon must exlst between the teedback-current loop and the voltage-senslng loop Otherw~se, spurlous slgnal voltages developed across the spurlous reslstances ln the feedback-current loop --voltages that are usually many tlmes larger than the mlcrovolt-slze ac slgnal developed across Rx -- wlll be coupled lnto the voltage-15 senslng loop and degrade measurement accuracy For many battery testlng appllcatlons, lt ls very advantageousto power the battery testlng apparatus by the battery undergotng test rather than requlre, lt to have lts own source o~ power In the clrcult dlsclosed ~n Flg 2, operational amplifler Al recelves lts 20 operatlng power from the battery undergoing test through power termlnal VA+~ connected to battery contact A, and power termlnal VA-. connected to battery contact 8 Slnce "rour-polnt probe"
archltecture places translstor Ql In contact wlth the ~eedback-current loop, the A and B contacts are used to power Al ln flg 2 25 Thls cho~ce ls dlctated by the lnherent coupllng that exlsts between 1~301~i4 Al and Ql along wlth the need for lsolatlng the feedback-current loop lrrom the voltage-senslng loop.
"Four-polnt probe" archltecture places the oscillator slgnal Vln ln the voltage-senslng loop. for a practlcal transtormerless osclllator clrcult, the osclllator~s output voltage ls establlshed wlth respect to one o~ lts power supply termlnals. Thus, the cholce of ~attery contacts to be used for powerlng the osclllator wlll be strongly lnfluenced by the need to provlde adequate clrcult lsolatlon between the feedback-current loop and the voltage-senslng loop.
Reterrlng next to Flg. ~, a schematlc dlagram slmllar to Flg. 2 ls dlsclosed lncludlng connectlons used for provldlng osclllator power accordlng to the teachlngs ~ound ln U.S. Patents ~,87~,91 1 and 3,909,708. The ac slgnal Vln ls seen to ~e establlshed between a slngle osclllator output termlnal and the osclllator clrcult's power supply termlnal, V0-. Thus, one o~ the osclllator's power supply termlnals ls also one ot lts output slgnal termlnals and must there~ore be ln contact wlth the voltage-senslng loop. Accordlngly, to avold coupllng the voltage-senslng loop to the ~eedback-current loop, the osclllator ln Flg. ~ recelves lts power through the voltage-- senslng contacts at C and D.
Powerlng the osclllator c~rcult by means of the voltage-senslng contacts has two disadvantages. Flrst, because of the very large ac ampllfler galn, signal levels ln the voltage-senslng loop are very small. Consequently, excess nolse generated by currents flowlng through the voltage-senslng contacts, lead wlres, and lnput clrcultry lntroduces serlous measurement problems. Second, 0i64 practlcal MSI lntegrated clrcults, such as dual and quad operatlonal amplltlers, do not provlde separate pln-outs tor lndlvldual elements on the chlp. Theretore, lf the osclllator ls to share ~lSI ICs wlth the ampllfler and detector, lt must ~e capable of belng powered from the same palr of battery contacts as the other actlve devlces.
Referrlng now to Flg. 4, a method ls dlsclosed for ln~ectlng a slgnal lnto the voltage-senslng loop ~y an osclllator powered rrom contacts ln the feed~ack-current loop wlthout lntroduclng excesslve loop coupllng. The osclllator of Flg. 4 develops an ac voltage Vosc between a slngle output termlnal and lts negatlve power supply o termlnal VO- . The clrcult ls powered by connectlons rom the osclllator's VO~ and VO- power termlnals to the A and B ~attery contacts, respectlvely. Thls places the osclllator's power termlnals dlrectly ln parallel wlth the power termlnals tor tt~e hlgh-galn ampllfler cascade, VA~ and VA-.
The osclllator ln~ects a slgnal Vln lnto the voltage senslng loop ~y means o~ lnJectlon reslstor R7 and voltage-vlewlng reslstor R8. The slgnal current passlng through reslstors R7 and R8 returns to the VO~ termlnal ot the osclllator ~y passlng through the D
contact and connectlng wlre, through the negatlve battery termlnal 20 ltself, and then through the B contact and connectlng wlre. Thus, the spurlous reslstances ot the B and D contacts and connectlng wlres wlll tend to couple the two loops and may therefore degrade measurement accuracy. However, as wlll be shown more clearly ~elow, lf the osclllator voltage Vos, ls made sufflclently large, the 25 two loops can be lsolated to such a degree that the errors lntroduced 01~4 by spurlous reslstances as large as several ohms wlll ~e negllglble.
Reterrlng next to Flg. 5, a slmpllfled schematlc dlagram ls dlsclosed whlch wlll now be employed to analyze measurement errors resultlng trom loop-coupllng tntroduced by the slgnal ln~ectlon clrcult ot Flg. 4. The tour reslstances RA, Rg, Rc, and RD ln 5 Flg. S represent the spurlous reslstances ot the tour lead-wlres and contacts at A, B, C, and D, respectlvely. As seen ln Flg. 5, the ac feed~acl~ current lf passlng through reslstor R6 spllts lnto two currents, lg and lD. Current lg passes through spurlous reslstance RB and enters the negatlve termlnal o~ the battery at contact B.
o Current lD passes through the osclllator clrcult, through reslstors R7 and R8, through spurlous reslstance RD, and enters the negatlve termlnal ot the battery at contact D. The two currents add together ln the battery. Thelr sum, lf, leaves the battery at contact A, passes through spurlous reslstance RA, and returns to the collector ot 15 tranS1StOr Q 1 .
By uslng an approprlate spllttlng tactor derlved trom the reslstances ot the two paths, one can show that the current lD ls proportlonal to lf and glven by [R7~R8 R3;RD~ (7) The total slgnal voltage at the dltterentlal lnput to the operatlonal ampllfler ls found by superposltlon of the voltage drops 1~801~4 ln the voltage-senslng loop due to the currents lf and lD, and the voltage Vln ln~ected lnto the voltage-senslng loop ~y the osclllator.
By vlrtue ot the very large ac galn, thls total ~nput s~gnal voltage ls essentlally zero. Thus, one can wrlte [R7 ~ R8 ~ R B ~ RD ~

Ellmlnatlng lD ~rom equatlons (7) and (8) and uslng equatlon (4) to express lf ln terms ot VOut(ac) leads to v0ut(ac) - R6 ( R8 ~ RD ) Vosc Rx (R7~R8~RB~RD)~RB(R8~RD) Equatlon (9) ~an be stmplltled by notlng that reslstors R7 and R8 are much larger than spurlous reslstances RB and RD. Thus, one can assume that R7 RB 1 (tO) Accordlngly, equat)on (9) can be wrltten ~Oi6 Vout(ac) - R6 R8 ( 1 1 ) vosc Rx (R7 ~ R8) + RB R8 Equatlon ( 1 1 ) ls the deslred result. It relates vout(ac) to Vasc and lncludes the effects o~ both the battery~s lnternal reslstance, Rx, and the spurlous lead-wlre and contact reslstance, Rg. Note that only one ot the tour spurlous reslstances wlll have a slgnltlcant et~ect on the measurernents. Under the assumptlon that Rx (R7 ~ R8) ~ > RB R8 ( 12) the second term ln the denomtnator o~ equatlon ( 1 1), the error term, can be neglected ln comparlson wlth the ~lrst term. Under these clrcumstances, equatlon ( 1 1 ) becomes Vout(aC) = r R8 ] [R6 ]
Vosc L R7 ~ R8 Rx whlch agrees wlth the earlter result glven ~y equatlon (6) slnce Vln = R7 R8 Vsc ( 14) Substltutlng equatlon ( 13) lnto lnequallty ( 12) ylelds the ~: 25 ~ , ' , " ~'; .

~30164 followlng suttlclent condltlon for neglectlng the error term ln the denomlnator ot equatlon ( l l ):

V~sc ¦ R6 RB ( I S) Vout(aC) s In the practlcal electronlc ~attery testlng devlce dlsclosed hereln below, the tollowlng approxlmate values apply:

V ~
o Vout(ac) - l volt R6 - 20 ohms Theretore, the magnltude ot the lett-hand slde ot lnequallty (lS) can be approxlmated ~y Vout(ac) ¦ ( l 6) The error term of equatlon ( l l ) lntroduces only a one percent measurement error when the left-hand slde ot lnequallty (15) ls lO0 tlmes larger than the rlght-hand slde. Therefore, a spurlous reslstance RB Ot one ohm or less wlll lntroduce a measurement error that does not exceed one percent . Furtherrnore, the error analysls above shows that the measurements are une~tected by spurlous : 25 , 1~01~

reststances RA and Rc; and are also uneffected by RD as long as lnequallty (10) ls satlsfled. Thus, for the osclllator slgnal ln~ectlon method dlsclosed ln Flg. 4, only the spurlous reslstance RB causes any potentlal degradatlon of measurement accuracy. ~loreover, as ls clearly shown by the analysls above, even the deleterlous effect of 5 RB can be effect~vely nulllfled wlth thls clrcult by chooslng V~sc to be sul'flclently large.
The osclllator clrcult produces a square wave output slgnal at a frequency of approxlmately 100 Hz. Although the exact frequency of osclllatlon ls not crltlcal, equatlon ( 1 1 ) dlscloses that VOut(ac) ls o dlrectly proportional to VOsc. Thus, for hlgh accuracy, the magnltude ot the osclllator slgnal mwst remaln very constant under all cond~tlons o~ voltage and temperature encountered ~n operatlon.
The second lnvent~on embodlment dlsclosed ln U.S. Patent 3,873,911 utlllzes an osclllator clrcult comprls~ng a dlscrete 5 operatlonal ampl~fler connected as a conventlonal astable multlvlbrator. The problems wlth uslng thls clrcult, and stlll attalnlng the hlgh measurement accuracy deslred ror the present lnventlon, are twofold. Flrst, the level ot the osclllator's output slgnal ls nearly proportlonal to lts supply voltage. Thus, wlth 20 osclllator power supplled by the battery belng tested, the accuracy ot conductance measurements ls very dependent upon the "surface charge" condltlons of the battery. Second, because of lmperfectlons that are always present ln the output clrcults o~ IC operatlonal ampllflers, the saturated maxlmum and mlnlmum output voltage 25 levels are lnevlta~ly otfset ~rom the power supply voltage levels, ~30i6~

VO+ and VO-, ~y slgnlflcant amounts that depend on temperature. As a consequence, the osclllator's output slgnal level changes wlth temperature, thus lntroduclng a slgnlflcant temperature-dependent measurement error.
The second lnventlon embodlment dlsclosed ln U.S. Patent 3,909,708 utlllzes a dlfferent type of osclllator clrcult ln an attempt to solve the problems descrlbed above. Instead of an operatlonal ampll~ler, the clrcult employes two dlscrete translstors functlonlng ln a conventlonal astable translstor mult~vlbrator. Thls tends to ellmlnate the problem of the temperature-dependent output slgnal level that Is lnherent to IC operatlonal ampllfler multlvlbrators. In addltlon, the voltage supplled to the multlvlbrator ls regulated wlth a zener dlode Thls holds the osclllator's supply voltage, and hence lts output voltage, constant and tends to reduce the dependence of the measurements ~n the 15 battery's "surface charge".
Ne~ther of the problem solutlons employed ln U.S. Patent 3,909,708 can be employed wlth the present lnventlon, however. The use of dlscrete translstors ls, o- course, the antlthesls of utlllzlng ~1SI technology. Moreover, wlth ~1SI technology, lt ls not posslble to 20 separately regulate the power supplled to the osclllator slnce all actlve devlces ln the IC must recelve power from the same source.
Referrlng now to Flg. 6, a schematlc dlagram of an osclllator clrcult for produclng a preclsely-leveled output slgnal ln accordance wlth the present lnventlon ls dlsclosed. Operatlonal ampl~er A2 25 along w~th res~stors R9, R 1 O, R 1 1, R 12 and capacltor C4 comprlse a 30~4 conventlonal astable multlvlbrator clrcult. Posltlve feedback ls provl~ed by reslstor R 12 along wlth voltage dlvlder reslstors R9 and RtO. Negatlve feedback ls provlded by reslstor Rl 1 along wlth capacltor C4. As ls well known to one of ordlnary sklll ln the artl the output voltage of operatlonal ampllfler A2 alternately assumes a 5 maxlmum value near lts posltlve supply voltage, VA~, and a mlnlmum value near lts negatlve supply voltage, VA-. If R9 and R10 are equal, the output wavetorm of thls osclllatlon ls nearly symmetr~cal wtth an osclllatlon perlod, T, glven by T= (2Rll C4)1n{l~RR2 } (17) The synchronlzlng output o~ the astable multlvlbrator, Vs~c, ls connected to the control lnput o~ a C~10S bllateral analog switch Sl.
5 One o~ the slgnal termlnals ot analog swltch 51 ls held at a constant voltage, Vz, by a zener dlode Dl wh~ch ls supplled power through serles reslstor R 13. The other slgnal termlnal ot S 1 ls connected to one slde ot an output load reslstor R 14 whose other slde ls connected to the negatlve termlnal of dlode Dl. The osclllator 20 output slgnal, VOsc, ls developed across the output load reslstor R 14.
The operatlonal ampllfler power termlnals, VA+ and VA-, are connected ln parallel wlth the analog swltch power termlnals, Vs~
and Vs~, and toget~er comprlse the osclllator power supply termlnals, VO~ and VO-, of Flg 4. These common power connectlons 25 recelve power trom the battery undergolng test by means ot tee~ack-current loop connectlons at battery contacts A and B, respecttvely. As ls shown ln Flg. 6, the negatlve slde of the common-mode output slgnal voltage, vosc~ ls ln common wlth power termlnal VO- and hence wlth battery contact B.
Durlng the portlon Tl of tlme perlod T that the multlvl~rator 5 output ls at lts hlghest level, VSync(hl)~ the analog switch Sl ls turned "on"~ Assumlng that the "on" reslstance of Sl ls much less than Rl4, the output voltage then assumes lts hlghest value Vosc(hl)~vz . (l8) Durlng the portlon T2 O~ tlme perlod T that the multlvlbrator output ls at lts lowest level, VS~C(1O)~ the analog swltch S 1 ls turned "o~t".
Assumlng that the "otf" reslstance of S l ls much larger than R l 4, the output voltage ls essentlally pulled to zero by Rl4 so that V0sc(lo) ~ O . ( l 9) Accordlngly, the osclllator output voltage slgnal, Voec~ osclllates between Vz and zero and very closely approxlmates a per~ectly-20 leveled square wave havlng a constant peak-to-peak amplltude, Vz.
One sees from equatlon (13) that the peak-to-peak amplltude of the ac signal component at the output of the hlgh-galn amplltler cascade ln Flg. 4 ls theretore 301~4 I V ( ~ I [ R8 ~ [ R6] V (20) The zener dlode voltage, Vz, ls chosen to ~e 5. ~ volts to take advantage ot the very nearly zero temperature coertlc~ent that ls 5 characterlstlc ot zener dlodes havlng thls partlcular zener voltage.
Accordlngly, lVwt(ac)l wlll ~e very nearly lndependent ot both battery voltage and lnstrument temperature.
One sees that tt~e slmple preclsely-leveled osclllator clrcult dlsclosed ln Flg. 6 has propertles tnat are very nearly ~deal tor appllcatlon ln an electronlc ~attery testlng devlce. It can be powered from common power supply connectlons and dellvers a preclse output slgnal level that ls vlrtually lndependent ot ~Oth temperature and supply voltage. These speclal attrlbutes ot the slmple clrcult dlsclosed ln Flg. 6 contrlbute to the very hlgh 15 measure~ent accuracy t~at ls ac~leved wlt~ tne lnventlon nereor.
Reterrlng next to Flg. 7, a scnematlC dlagram ot a slrnple, yet very accurate, sync~ronous detector clrcult ln accordance wlth the present lnvent~on ls dlsclosed. The purpose ot thls clrcult ls to provlde a dc output slgnal, Vd~t, that ls preclsely proportlonal to the 20 peak-to-peak amplltude or the ac slgnal component ot the amplltler's output voltage, IVOu~(ac)I, whlle totally lgnorlng the dc ~las component, Vout(dc)~ A partlcular teature ot the clrcult ot Flg. 7 ls that lt ls vlrtually unresponslve to spurlous slgnal components, SUCh as ac hum and nolsej that are uncorrelated wlth t~e osclllator .
,~ ;'.

, ~01~'~

slgnal, Vogc The synchronous detector clrcuit thus permlts operatlon ot the battery testlng devlce ln electrlcally "nolsy~' envlronments wlthout requlrlng the extenslve use ot shleldlng --WhlCh would substantlally lncrease manufacturlng costs. As ln the case ot the osclllator clrcult dlsclosed above, the s~mchronous detector clrcult dlsclosed hereln ls capable ot belng powered trom a common source dellverlng power contlnuously to other actlve elements ln the clrcult. It ls theretore tully compatlble wlth modern t~151 IC technology.
The clrcult ot Flg. 7 dlscloses a C~OS bllateral analog swltch o S2 and operat~onal amplifler A3 along with reslstors R15, R16, R17, and capacltors C5 and C6. Operatlonal amplltler A~ along wlth reslstors R15, R17, and capacltor C6 comprtse an lntegrator clrcult.
The nonlnvertlng (~) lnput ot A 3 ls blased to the value ot the dc component ot Vout by means o- reslstor R 16 along wlth bypass 15 capacltor C5. The slgnal applled-to the lnvertlng (-) lnput ot A3 ls derlved rrom Vout and passes through reslstor R15 and analog SWltCh 52. Thls slgnal ls swltched "on" and "ott" at the osclllator trequency by vlrtue ot the synchronlzatlon slgnal, VSync~ that ls obtalned trom the clrcult ot Flg. 6 and applled to the control lnput of S2.
Just as ln the case of the osclllator clrcult dlsclosed Flg. 6, both the operatlonal ampllfler and the analog swltch of the detector clrcult ot Flg. 7 are powered through common connectlons to the feedback-current loop battery contacts at A and B. Both ot the two lnput slgna~s dlsclosed ln the clrcult ot Flg. 7, VO~Jt and Vsync~ are 25 common-mode slgnals establlshed wlth respect to the negatlve 1~ ~3O1~L~

power supply lead contactlng the battery at B.
Operatlon of the synchronous detector clrcult of flg. 7 wlll now be explalned by means of reterence to the tlmlng and wave~orm dlagrams o~ Flgs. 8a, 8b, 8c, and 8d.
Flg. 8a lllustrates the wavetorm of the common-mode Yoltage, Vout~ developed across reslstor R6 o~ the clrcult o~ Flg. 4 and applled to the slgnal lnput of the clrcult of Flg. 7. By vlrtùe of the 180 degree phase lnverslon performed by the hlgh-galn ampllfler cascade o~ Flg. 4, Vout assumes lts low value Vout(lo) durlng t~me- perlod Tl for whlch V0SC ls h~gh, and lts hlgh value VoUt(hl) durlng tlme perlod T2 for whlch Vosc ls low. Vout thererore osclllates about lts dc blas value Vout(dc) ~ V0 glven by equatlon (2). The peak-to-peak value ot the ac component o~ Vout ls seen to be ¦Vout(aC)¦ = {Vout(hl)--Vout(dc)} + ~VOut (dc)--vOut(lo)l (21 ) Because o~ the ac coupllng provlded by coupllng capacltors Cl and C2 at the lnput ot ampll~ler Al, the average excurslons ot Vout above and below lts dc blas value are equal. Accordlngly, the two shaded areas o~ Flg. 8a can be equated to yleld {VoUt(dc)--Vout(l)}Tl = {Vout(hl)--Vout(dC)}T2 (22) 1~30~64 Su~stltut~ng equatlon (22) lnto equation (21 ) ylelds ¦Vout(aC)l = [ lT 2 ] [Vout(dc) -Vout(lo)l (2~) Flg 8b lllustrates the waveform of the synchronlzatlon slgnal, YS~nC~ developed at the output of operatlonal amplltler A2 ln Flg 6 and applled to the synchronlzatlon lnput of the clrcult of Flg 7 Thls common-mode voltage slgnal osclllates between two voltage levels, VS~"c(hl) and VS~nc(lo)~ at the osclllator frequency of approxlmately 100 Hz The perlod ot the osclllatlon T~ ls there~ore approxlmately lOmllllseconds Now conslder the clrcult ot Flg 7 ln greater detall For slmpllclty, assume lnltlally that lntegratlon capacltor C6 ls zero The negatlve ~eedback provlded ~y reslstor R17 along wlth the hlgh galn ot operatlonal ampll~ler A3 ensures that the voltage at the lnvertlng (-) lnput ot A3 wlll be very nearly the same as the voltage at the nonlnvertlng (~) lnput 8ecause o~ the low-pass ~llterlng actlon of blas reslstor R16 and bypass capacltor C5, the voltage at the nonlnvertlng lnput, measured wlth respect to the negatlve power 20 supply termlnal VA-, ls slmply the dc blas component of the lnput voltage, Vout(dc)~ Therefore, the voltage at the lnvertlng (-) lnput ls llkewlse Vout(dc)~
Durlng tlme lnterval Tl, analog swltch S2 ls turned "on" by synchronlzatlon slgnal Vs~nc(hl) Dur~ng thls lnterval, operatlonal 25 ampllfler A~ serves as a slmple lnvertlng ampllfler wlth lnput ~ , 1~30164 s~gnal Vout - vout(lo) and output slgnal Vd~t = Vd~t(hi). A feedback current ls tlows through reslstor R l 7, through swltch S2, and through reslstor Rl5 as shown ln Flg. 7. Assumlng that the "on"
reslstance ot S2 ls small compared wlth RtS, one can wrlte thls current ln terms of the voltage drop across elther Rl5 or Rl7 as 5 follows:

s = {Vout(dC)--V0ut(lo)} = Vdet(hl) (24) o Comblning equatlons (23) and (24) leads to det [Tl tT2 ~ [ Rl53 1 out ¦ , (25) Equatlon (25) detlnes the hlgh output level Vd~(hl) seen ln Flg. 8c.
Durlng tlme lnterval T2) analog swltch S2 ls turned "o~" by synchronlzatlon slgnal VS~ o)~ Assumlng that the "or~" reslstance of S2 ls su~lclently large, one can assume that ls~ 0 so that no 20 voltage drop exlsts across Rl7. Accordlngly, the output o~ A3 assumes the same voltage, Vout(dc)~ that exlsts at both the lnvertlng input and the noninvertlng lnput of A3. Since the output voltage equals the voltage at the nonlnvertlng lnput, Vd~t ls zero. Thls zero value def~nes the low output level seen ln Flg. 8c.
Flg. 8c lllustrates the waveform ot the dl~terentlal-mode 1~3016~

output slgnal of the synchronous detector. One sees from the analysls detalled above that the output voltage ,Vd~, osclllates between the hlgh value Vd~t(hl), glven by equatlon (25), and zero. The average, or dc value of Vd~t, ls therefore r Tj det LT1 +T2 ~ det . (26) Comblnlng equatlons (25) and (26) to ellmlnate Vd~t(hl) leads to [~T~ I T2) ] [ 5 ]

The ettect of lntroduclng the lntegrat~on capacltor C6 lnto the 15 clrcult ot Flg. 7 ls lllustrated ln Flg. 8d. One sees that the average value of the detector output slgnal, Vd~(dc), ls uneftected by C6.
However, the lntegratlon capacltor smooths the varlatlons ln output voltage about the average value, thus reduclng the rlpple component of Vd~t. For a suttlclently large value of lntegratlon capacltor C6, 20 Vd~t ls slmply equal to Vd~t(dC) Equatlon (27) shows that the dc slgnal voltage at the dlfferentlal output of the synchronous detector ls dlrectly proportlonal to the common-mode ac slgnal component at the output of the hlgh-galn ampllfler. The analysls above dlscloses that the 25 constant of proportlonallty is not e~tected by ampll~ler galn or 1 ~3016~

osc~llator frequency. The relationshlp between VOut(ac) and Vd~t depends only upon the ratlo of two reslstance values, R l 5 and R l 7, and the osclllator symmetry ratlo (Tl/T2). Futhermore, lt ls very lnsenslt~ve to changes ln the symmetry ratlo when Tl and T2 are nearly equal. Thus, the detector clrcult dlsclosed ln Flg. 7 has 5 nearly ldeal characterlstlcs for appllcatlon to the accurate determlnatlon of battery conductance.
Moreover, the dlrect proportlonallty descrlbed by equatlon (27) only occurs tor slgnal components whlch are at the osclllator's exact frequency and are fully correlated wlth the osclllator slgnal.
All other slgnals, such as spurlous ac hum and nolse, wlll not contrlbute to the average value o~ Vd~ and wlll therefore be et~ectlvely removed trom the detector's output s~gnal by the smoothlng e~rect o~ the lntegratlon capacltor, C6. The clrcult dlsclosed ln F~g. 7 thus permlts obtalnlng very accurate battery 15 measurements ln electrlcally "nolsy" envlronments wlthout requlrlng that the testlng devlce be extenslvely shlelded and therefore expenslve to manu~acture. Important economlc bene~lts can consequently be galned throug~ the use o~ the synchronous detector clrcult dlsclosed ln Flg 7 ln a practlcal battery testlng 20 devlce. Furthermore, ~n contrast to the detector clrcult dlsclosed ln U.S. Patent ~,909,708 whlch requlres "chopped" dc power, the clrcult o~ Flg. 7 has no speclal power source requlrements and ls fully compatlble wlth modern MSI IC technology.
Referrlng next to Fig. 9, a schematlc dlagram o~ a slmple 25 ad~ustable dc ampll~ler and output meterlng clrcult ln accordance 30i6~

wlth the present inventlon ls dlsclosed The clrcult comprlses only operatlonal ampllfler A4, dc mllllameter mA, and varlable resistor Rl8 Iust as ln the case o~ all the other actlve devices, operatlonal ampllf~er A4 recelves lts dc power from common connectlons to the ~attery at the feedback-current loop contacts A and B
S The nonlnvertlng t+) lnput of operatlonal ampll~ler A4 ls connected to the nonlnvertlng (~) lnput of operatlonal ampllfler A~
ot Flg 7 by the negatlve slgnal lead of Vd~t Accordlngly, the nonlnvertlng lnput of A4 lS blased to the same dc level, Vout(dc)~ as the nonlnvertlng lnput o~ A3 By vlrtue o~ the negatlve feedback o lntroduced by the slgnal path through mllllameter mA, along wlth the hlgh galn of operatlonal ampll~ler A4, the lnvertlng (-) lnput of amplltler A4 assumes the same voltage level as the nonlnvertlng lnput The entlre dlt~erentlal lnput slgnal, Vd~t, theretore appears across the lnput reslstor R l 8 Slnce the current through the 15 mllllameter, Im, ls the same as the current through lnput reslstor R l 8, the meter current can be slmply calcu1ated by applylng Ohm's law to R t 8 The dc meter current ls theretore m R l 8 (28) One sees that the slmple clrcult dlsclosed ln Flg 9 provldes a dc current Im through the mllllameter that ls directly proportlonal ' to VdBt(dc) The constant of proportlonallty relat~ng Im to Vdeffdc) ls seen to be lndependent o~ the meter's lnternal reslstance and ls 25 determlned completely by the value o~ the lnput reslstor Rl8 Oi64 Equatlons (20), (27) and (28) can now be comblned to derive a slngle relatlonshtp relatlng the meter current, Im, to the osclllator's zener voltage, Vz, ~or the enttre electronlc battery testlng devlce dlsclosed ln the ~lock dlagram o~ Flg. 1. The resultlng equatlon ls wr~tten Im = r Tl T2 lr R6 R8 R17 1 Gx (29) VZ ~T1 +T2)2JL(R7 ~ R8) R15 R18 J

Equatlon (29) conflrms that the dc meter current ls dlrectly proportlonal to the battery's lnternal conductance Gx. ~oreover, the constant of proportlonallty ls slmply and preclsely determlned by the value ot Vz along wlth slx reslstances and the symmetry ratlo, (Tl/T2) . In practlce, (Tl/T2) ls very nearly one. Thus, Im 1 r R6R8R17 Gx (30) Vz 4 (R7~R8)Rt5R18 By uslng flxed reslstances and callbratlng the dc mllllameter ln unlts proportlonal to lnternal conductance, such as cold-cranklng amperes or ampere-hours, tne disclosed ~attery testlng devlce will emulate a dlrect readlng devlce of the type dlsclosed ln U.S. Patent 3,8731911. It wlll be apparent to one skllled ln the art that ln such an appllcatlon, the dc ampllfler and mllllameter could be replaced by any llnear dlsplay devlce, Such as a dlgltal meter, that ls capable . . .

0~64 of provldlng a numerical dlsplay proportlonal to V~et(dc).
Alternatlvely, by lettlng one of the six resistances be a varlable reslstance, callbratlng lt ln battery ratlng unlts such as cold-cranklng amperes or ampere-hours, and then arranglng mllllameter mA to deslgnate simple qualltatlve condltlons, the 5 dlsclosed devlce wlll emulate a "pass-fall" battery testlng devlce of the type dlsclosed ln U. S. Patent 3,909,708. It wlll ~e apparent to one skllled ln the art that ln such an appllcatlon, the mllllameter could be replaced by a varlety ot dlsplay means, such as colored llghts, that are capable ot lndlcatlng the qualltatlve condltlons.
o Moreover, lt can be seen rrom equatlon (~0) that lt Im ls brought to a partlcular tlxed value, such as the "pass-~all" polnt, by ad~ustlng one ot the reslstances ln the denomlnator -- R15 or Rl~
(or R7 under the condltlon R7 R8) -- the approprlate value ot the varlable reslstance wlll be dlrectly proportlonal to the battery's 15 conductance Gx. Hence, lt varlable reslstance R18 ln Flg. 9 has a llnear taper, lt can be llnearly callbrated ln conventlonal battery ratlng unlts -- such as ampere-hours or cold cranklng amperes --that are proportlonal to the conductance of a ~attery capable ot deltverlng lts full-rated power. Such ltnearlty of the ratlng scale 20 tmproves preclslon and Is a great convenlence for the operator.
Flg 10 dlscloses a complete schematlc dlagram of an Improved devlce for testlng 1 2-volt automotlve batterles ln accordance wlth the present lnventlon. Operatlonal amplltlers 100, 102, 104, and 106 comprlse four elements of an r1Sl quad operatlonal ampl~ler 25 lntegrated clrcult, ICl. Bllateral analog swltches 108 and 110 ~0164 comprlse two elements of a cr10s bllateral swltch lntegrated circutt, IC2. Both ICl and IC2 are powered by means of common connectlons, l l 2 and l l 4, to the battery undergolng test 24 through current-~eedback loop contacts l 16 and l l8, respectlvely.
Hlgh-galn ampllfler cascade 12 or flg. l comprlses operatl.onal s amplltler tO0 and npn translstor 120 connected as an emltter follower. Reslstor l22 conducts a dc blas voltage to the nonlnvertlng (~) lnput of operatlonal amplltler lO0 trom voltage dlvlder reslstors l 24 and l 26 whlch are connected to battery 24 through voltage-senslng contacts l 28 and l 30. The output voltage ot hlgh-galn amplltler cascade l 2 ls establlshed across external-path reedback reslstor 22. An lnternal feedback path comprlslng reslstors l ~2 and l 34 conducts the dc voltage at the common connectlon between the emltter of npn translstor 120 and reslstor 22 to the lnvertlng (-) lnput of operatlonal ampllrler l O0 . Reslstors l 32 and l ~4 along wlth capacltor l ~6 comprlse low-pass rllter l 8 ot Flg. l .
The ac slgnal voltage developed across ~attery 24 ls sensed at voltage-senslng contacts l 28 and l ~0 and added ln serles to an lnput signal voltage component establlshed across vlewlng reslstor 20 l~8. The resultant composlte ac slgnal voltage ls applled to the dlfferentlal lnput of operatlonal ampllfler lO0 by a capacltlve coupllng network comprlslng capacltors 140 and l42 A feedback current that ls proportlonal to the voltage establlshed across reslstor 22 passes through battery 24 by means or external teedback 25 path conductors 144 and l 46 along wlth cu~rent-tee~ack loop l~t ~301~4 battery contacts 1 16 and 1 18.
The ac lnput slgnal voltage establlshed across vlewlng reslstor 138 ls generated by a preclsely-leveled osclllator clrcult comprlslng operatlonal ampllfler 102, analog swltch 108, and zener dlode 148. Operatlonal ampllfler 102 along wlth reslstors 150, 152, 154, 156, and capacltor 158 comprlse a conventlonal astable rnultlvlbrator clrcult used to generate a square-wave synchronlzlng slgnal. Reslstor 160 supplles blas current to zener dlode 148. The synchronlzlng output of operatlonal ampllfler 102 connects to the control lnput of analog swltch 108. The two slgnal contacts o~
o analog swltch 108 lnterconnect the output of zener dlode 148 wlth the lnput o~ potentlometer 162. Potentlometer 162 provldes means to lnltlally ad~ust the level of the voltage slgnal outputted by analog swltch 1 08 SPST swltch 164 provldes ~or the selectlon ot elther o~ two levels o~ slgnal voltage and serves as a temperature compensatlon adJustment. Thls temperature compensatlon ad~ustment corrects for ~attery temperature and provldes means for obtalnlng lncreased accuracy when measurlng batterles at other than room temperature.
Wlth SPST swltch 164 tn the open posltlon, a current proportlonal to 20 the output voltage of potentlometer 162 passes through In~ectlon reslstor 166 and ls ln~ected lnto vlewlng reslstor 138 thereby developlng a slgnal voltage across vlewlng reslstor 1~8 Closure o~
swltch 164placesreslstor 1681nparallelwlthreslstor 166 thereby ~ncreaslng the level of stgnal voltage developed across 25 vlewlng reslstor I 38; as would be approprlate to measurlng a 1.2~0~64 battery that was at a reduced temperature It will be apparent to one of ordlnary sklll in the art that several alternatlve temperature compensatlon methods are avallable For example, the ternperature compensatlon ad~ustment could provlde more than two slgnal values;
or lt could be lmplemented wlth a conttnuous, rather than a dlscrete, 5 reslstance ad~ustment In addltlon, a temperature compensatlon ad~ustment could be lmplemented by varylng reslstances at other locatlons ln the battery tester clrcult as can be easlly recognlzed from an examlnatlon of equatton (~0) derlved hereln above Analog swltch 1 10 along wlth operat~onal ampllfler 104, o whlch ls connected as an lntegrator, comprlse synchronous detector clrcu~t 40 ot Flg 1 Reslstor 170 and ~ypass capacltor 172 comprlse a low-pass tllter wh~ch blases the non~nvertlng lnputs ot operatlonal amplitlers 104 and 106 to the voltage level of the dc blas component developed across reslstor 22 A slgnal current 15 derlved trom the total voltage at the common connectlon between reslstor 22 and translstor 120 passes through reslstor 174 and analog swltch 1 10 to the lnvertlng lnput ot operatlonal ampllfler 104 Th~s slgnal current ls perlod~cally lnterrupted at the osclllator frequency by vlrtue of the control lnput of analog swltch 1 10 belng 20 connected to the synchronlzlng output of operatlonal amplifler 102 Reslstor 176 prov~des negatlve dc feed~ack to operatlonal ampllfler 104 Integrat~on capacltor 178 serves to smooth the detected voltage signal outputted by operatlonal ampllfler 104 A current derlved from the detected slgnal voltage at the 25 output of operat~onal amplltler 104 passes throug~ mllllameter 180 ~0164 to the output of operatlonal ampllfler 106 by way of one of the two paths selected by SPDT swltch 182 Wlth swltch t80 ln posltlon 1, the meter current passes through flxed reststor 184 Under these condltlons, the dlsclosed lnventlon emulates a dlrect readlng battery testlng devlce havlng an output lndlcatlon that ls proportlonal to the 5 dynamlc conductance of battery 24 Wlth swltch 182 ln posltlon 2, the meter current passes through flxed reslstor 186 and varlable reslstor 188 Under these condltlons the dlsclosed lnventlon emulates a "pass-fall" battery testlng devlce havlng an ad~ustable ~attery ratlng scale that ls llnearly related to the settlng of 10 varlable reslstance 188 and a ratlng of ~set that ls determlned by the value or tlxed reslstor 186 As wlll be apparent to one of ordlnary sklll ln the art, several alternatlve llnear battery ratlng adJustment methods can be lmplemented As dlscussed hereln above, a llnear relatlonshlp wlll exlst between battery ratlng and the ad~ustment 15 reslstance lr any one ot the three reslstances ln the denomlnator o~
equatlon (30) ls chosen to be the ad~ustment Thus, one could choose to select and vary reslstor 174 tR15) lnstead of reslstor 188 (R18) Alternatlvely, lf reslstor 168 was not employed for temperature compensatlon, one could choose to vary reslstor 166 (R7) under the 20 condltlon that the lnJectlon reslstor 166 ls much larger than the vlewlng reslstor 138 (R8) A 11st of component types and values for the lmproved electronlc battery testlng devlce dlsclosed ln Flg 10 follows 30~6L~

REFERENCE NUMBER ~2MPONENT
$emtconductor Dev1ces 100,102,104,106 ICl - L~324N
108,110 I C2 - CD4066B
120 TIP~lC Power Transistor 148 1 N52~ 1 B Zener Dlode S Reslstors - Ohms (1 /4-W unless s~ecltled) 22 22 - 5 Watts 138,186 100 188 500 Varla~le 124,126,160 lK
162 5K Trlmpot 122,176 47K
lC 170,174 1 OOK

132,134,1 50,1 52 t l~leg 168 1.5 Meg Ca~ac1tors - 1'1td 158 0.022 136,140,142, 172 0.47 ~Jleter 180 1 mA dc mllllameter 2G Although a speclfic mode for carrylng out the present lnventlon has been hereln described, it ls to ~e understood that mad~flcatlon and varlatlon may be made without departlng from what ls regarded to be the sub~ect matter of the lnventlon. For example, vlsual dlsplay means have been speclflcally dlsclosed : ~ `

1~301~4 nerein above. However, the output of the disc~osed electronic battery testing circuit could alternatively be monitored by a voltage senslng devlce that responds to a drop in output slgnal level by soundlng an audlble alarm, causing a visible display, or by switching partlcular equlpment to an alternative power source. I~loreover, the 5 clrcult output could be monltored ~y a computer speclflcally programmed to respond appropriately to the level of the output slgnal. The range of potential computer responses would be virtually unllmlted and is restricted only by the imaginatlon of the computer progammer. These, and other varlatlons are belleved to be o well withln the scope of the inventlon and are lntended to be covered ~y the appended clalms.

Claims (40)

1 An electronic device for testing a direct current energy source comprising:
a. high-gain amplifier means receiving dc power through amplifier dc power terminals;
b. internal voltage-feedback means, including low-pass filter means, interconnecting the output and the input of said high-gain amplifier means;
c. external current-feedback means, including feedback resistor means, conducting a current from the output of said high-gain amplifier means through said direct current energy source;
d. oscillator means producing a periodic oscillator signal, said oscillator means receiving dc power through oscillator dc power terminals connected directly in parallel with said amplifier dc power terminals;
e. voltage summing and coupling means adding a voltage derived from said periodic oscillator signal to the voltage across said direct current energy source, and capacitively coupling the sum voltage to the input of said high-gain amplifier means;
f. detector means producing a dc output signal in response to the ac voltage component across said feedback resistor means; and, g. display means responsive to the level of said dc output signal.

2. An electronic device as in claim 1 wherein said display means displays numbers proportional to said dc output signal.

3. An electronic device as in claim 2 wherein said numbers are cold-cranking ampere numbers.

4. An electronic device as in claim 1 including temperature compensation means for adjusting the level of said dc output signal in accordance with the temperature of said direct current energy source.

5. An electronic device as in claim 1 including rating adjustment means for changing the level of said dc output signal in accordance with the electrical rating of said direct current energy source, and said display means displays qualitative conditions of said direct current energy source.

6. An electronic device as in claim 5 wherein said rating adjustment means comprises a variable resistance, and said electrical rating is described in numbers that are linearly related to the resistance value of said variable resistance.

7. An electronic device as in claim 6 wherein said electrical rating is described in cold-cranking ampere numbers.

8. An electronic device as in claim 6 wherein said electrical rating is described in ampere-hour numbers.

9. An electronic device for testing a direct current energy source comprising:
a. high-gain amplifier means;
b. internal voltage-feedback means, including low-pass filter means, interconnecting the output and the input of said high-gain amplifier means;
c. external current-feedback means, including feedback resistor means, conducting a current from the output of said high-gain amplifier means through said direct current energy source;
d. oscillator means producing a periodic oscillator signal;
e. voltage summing and coupling means adding a voltage derived from said periodic oscillator signal to the voltage across said direct current energy source, and capacitively coupling the sum voltage to the input of said high-gain amplifier means;
f. detector means, including an analog switch means and a continuously-powered operational amplifier means, said analog switch means conducting a signal from the output of said high-gain amplifier means to the input of said continuously-powered operational amplifier means while being periodically turned on and off in synchronism with said periodic oscillator signal; and, g. measuring and displaying means deriving a display signal from the output of said continuously-powered operational amplifier means and providing a visual display in response to the level of said display signal.

10. An electronic device as in claim 9 wherein said measuring and displaying means displays numbers proportional to said display signal.

11. An electronic device as in claim 10 wherein said numbers are cold-cranking ampere numbers.

12. An electronic device as in claim 9 including temperature compensation means for adjusting the level of said display signal in accordance with the temperature of said direct current energy source.

13. An electronic device as in claim 9 including rating adjustment means for changing the level of said display signal in accordance with the electrical rating of said direct current energy source, and said measuring and displaying means displays qualitative conditions of said direct current energy source

14. An electronic device as in claim 13 wherein said rating adjustment means comprises a variable resistance, and said electrical rating is described in numbers that are linearly related to the resistance value or said varible resistance.

15. An electronic device as in claim 14 wherein said electrical rating is described in cold-cranking ampere numbers.

16. An electronic device as in claim 14 wherein said electrical rating is described in ampere-hour numbers.

17. An electronic device for testing a direct current energy source comprising:
a. high-gain amplifier means;
b. internal voltage-feedback means, including low-pass filter means, interconnecting the output and the input of said high-gain amplifier means;
c. external current-feedback means, including feedback resistor means, conducting a current from the output of said high-gain amplifier means through said direct current energy source;
d. oscillator means including analog switch means, constant voltage means, and synchronizing means, said analog switch means conducting a signal from said constant voltage means to the output of said oscillator means while being switched on and off at a periodic rate by said synchronizing means, thereby producing a periodic oscillator signal at said output of said oscillator means;
e. voltage summing and coupling means adding a voltage derived from said periodic oscillator signal to the voltage across said direct current energy source and capacitively coupling the sum voltage to the input of said high-gain amplifier means, f. detector means producing a dc output signal in response to the ac voltage component across said feedback resistor means; and, g. display means responsive to the level of said dc output signal.

18. An electronic device as in claim 17 wherein said display means displays numbers proportional to said dc output signal.

19. An electronic device as in claim 18 wherein said numbers are cold-cranking ampere numbers.

20. An electronic device as in claim 17 including temperature compensation means for adjusting said dc output signal in accordance with the temperature of said direct current energy source.

21. An electronic device as in claim 17 including rating adjustment means for setting the level of said dc output signal in accordance with the electrical rating of said direct current energy source, and said display means denotes qualitative conditions of said direct current energy source.

22. An electronic device as in claim 21 wherein said rating adjustment means comprises a variable resistance, and said electrical rating is described in numbers that are linearly related to the resistance value of said variable resistance.

23. An electronic device as in claim 22 wherein said electrical rating is descrlbed in cold-cranking ampere numbers.

24. An electronic device as in claim 22 wherein said electrical rating is described in ampere-hour numbers.

25. An electronic device for testing a direct current energy source comprising:
a. high-gain amplifier means receiving dc power from power-supplying contacts contacting the direct current energy source;
b. internal voltage-feedback means, including low-pass filter means, interconnecting the output and the input of said high-gain amplifier means;
c. external current-feedback means, including feedback resistor means, conducting a current from the output of said high-gain amplifier means through said direct current energy source by means of said power-supplying contacts;
d. oscillator means receiving dc power from said power-supplying contacts and producing a periodic oscillator signal current, e. voltage summing and coupling means, including a pair of voltage-sensing contacts contacting said direct current energy source, and viewing resistor means, said summing and coupling means adding the voltage across said viewing resistor means to the voltage across said voltage-sensing contacts and capacitively coupling the summed voltage to the input of said high-gain amplifier means;
f. voltage injection means passing said periodic oscillator signal current from the output of said oscillator means through said viewing resistor means thereby producing a periodic signal voltage thereacross;
g. detector means converting the ac voltage component across said feedback resistor means to a dc output signal; and, h. display means responsive to the level of said dc output signal.

26. An electronic device as in claim 25 wherein said display means displays numbers proportional to said dc output signal.

27. An electronic device as in claim 26 wherein said numbers are cold-cranking ampere numbers.

28. An electronic device as in claim 25 including temperature compensation means for adjusting said dc output signal in accordance with the temperature of said direct current energy source.

29. An electronic device as in claim 25 including rating adjustment means for changing the level of said dc output signal in accordance with the electrical rating of said direct current energy source, and said display means displays qualitative conditions of said direct current energy source.

30. An electronic device as in claim 29 wherein said rating adjustment means comprises a variable resistance, and said electrical rating is described in numbers that are linearly related to the resistance value of said variable resistance.

31. An electronic device as in claim 30 wherein said electrical rating is described in cold-cranking ampere numbers.

32. An electronic device as in claim 30 wherein said electrical rating is described in ampere-hour numbers.

33. A system for assessing the ability of a direct current supply, having an internal resistance, to deliver power to a load including:
a time-varying voltage source having a pair of dc power-receiving terminals;
a time-varying current generating means having a pair of dc power-receiving terminals connected in parallel with the dc power-receiving terminals of the time-varying voltage source, the current generating means operably connected to the supply and producing an output directly proportional to the time-varying voltage source and inversely proportional to the internal resistance;
means for sensing the current generating means output and producing an output signal responsive thereto; and means for responding to the level of the output signal.

34. The system as set forth in claim 33 wherein the means for responding includes a computer.

35 The system as set forth in claim 33 wherein the means for responding provides a numerical display with numbers linearly related to the output signal level.

36. The system as set forth in claim 33 including means for adjusting the output signal level in accordance with a selected rating, wherein the means for responding indentifies qualitative conditions of the supply.

37. A system for assessing the ability of a direct current supply, having an internal resistance, to deliver power to a load comprising:
a current-feedback loop including current-feedback contacts operably contacting the supply;
a voltage-sensing loop including a viewing resistor and voltage-sensing contacts operably contacting the supply;
a time-varying voltage source receiving dc power from the current-feedback contacts and injecting a signal current through the viewing resistor in the voltage-sensing loop;
a time-varying current generating means receiving dc power from the current-feedback contacts and generating a current proportional to the signal voltage across the viewing resistor and inversely proportional to the internal resistance of the supply;
means for sensing the current generating means output and producing an output signal responsive thereto; and means for responding to the level of the output signal.

38. The system as set forth in claim 37 wherein the means for responding includes a computer.

39. The system as set forth in claim 37 wherein the means for responding provides a numerical display with numbers linearly related to the output signal level.

40. The system as set forth in claim 37 including means for adjusting the output signal level in accordance with a selected rating, wherein the means for responding identifies qualitative conditions of the supply.