Automatic Battery Analyzer Including Apparatus for Determining Presence of Single Bad Cell

This invention relates to apparatus for detecting the presence of a single defective cell in a multi-cell storage battery of the type normally used in automotive applications and to apparatus for automatically analyzing such a battery and including apparatus for detecting the presence of a single defective cell.

There has in the past been much effort at developing apparatus for determining the condition of lead-acid storage batteries of the type normally used for starting, lighting, and ignition in the automotive field and, in particular, for determining the acceptability of such batteries for use under adverse conditions such as the starting of an automobile engine at low temperatures. It is desirable that such apparatus be as simple as possible to use and involve the exercise of a minimum amount of judgment by the operator so that it may be operated by personnel with a relatively low level of skill or experience. It is further important that such apparatus be as accurate as possible; in applications where the apparatus is used to determine the acceptability of a battery already in service to a consumer, the same apparatus is likely to be used with batteries which are both in and out of applicable warranty provisions. It is also important that the period of time required to analyze a battery not be unreasonably long and it is further desirable that the analysis be performed without first requiring the time consuming step of recharging the battery on batteries in as wide a variety of different conditions as possible.

Apparatus for automatically analyzing the condition of automotive batteries are known in the prior art. Such apparatus and various of the desideratum relevant thereto are disclosed in the U.S. patent application of C. E. Frailing and T. J. Dougherty, Ser. No. 863,925, filed Dec. 23, 1977, entitled "Automatic Battery Analyzer" and assigned to the assignee of this application, now U.S. Pat. No. 4,193,025 which issued on Mar. 11, 1980. The disclosure of that application is hereby incorporated by reference. While the apparatus shown therein represented a significant improvement over what had been known theretofore, it has now been found that further improvements on such apparatus can be made. In particular, it was found that apparatus constructed in accord with the referenced patent application can to an inconclusive result when a battery was first tested in an undesirably high percentage of cases requiring that the battery be put through a time consuming recharging operation and then tested again. Further, it was found that the accuracy of battery analyzers could be increased over that attained with apparatus constructed in accord with the referenced patent application.

Thus, it is an object of this invention to provide apparatus for detecting the presence of a single defective cell in a multi-cell storage battery.

It is an object of this invention to provide apparatus for detecting the presence of a single defective cell in a lead-acid, multi-cell storage battery of the type commonly used for lighting, starting, and ignition in the automotive field.

It is an object of this invention to provide apparatus for automatically determining the acceptability of lead-acid, multi-cell storage batteries of the type commonly used for lighting, starting, and ignition in the automotive filed and including apparatus for detecting the presence of a single defective cell in the battery under test.

It is an object of this invention to provide apparatus for automatically determining the acceptability of lead-acid, multi-cell storage batteries which are useful for determining the acceptability of batteries over wider ranges of different battery conditions without the need of recharging the battery than have been obtained with prior automatic battery analyzing apparatus.

It is an object of this invention to provide apparatus for automatically determining the acceptability of lead-acid, multi-cell storage batteries which do not require an unnecessarily prolonged period of time to make the desired determination.

It is an object of this invention to provide apparatus for automatically determining the acceptability of multi-cell storage batteries which meet the aforementioned objects and are further characterized by simplicity of design, ease of operation, and economy of construction.

Further and additional objects of this invention will be apparent from this description, the drawing, and the appended claims.

FIGS. 1 and 2 are graphs useful for explaining the operation of the preferred embodiment of this invention;

FIG. 3 is a diagram partly in block form and partly in a schematic form of an automatic battery analyzer including the preferred embodiment of this invention;

FIG. 4 is a schematic diagram of the temperature sensing circuit of the apparatus of FIG. 3;

FIG. 5 is a schematic diagram of the load bank of the apparatus of FIG. 3;

FIGS. 6a-6c show the brass connector of the apparatus of FIG. 3 for connection to the negative terminal of the battery under test;

FIGS. 7a-7e are a flow diagram of the apparatus of FIG. 3; and

FIGS. 8 and 9 are flow diagrams of alternate embodiments of this invention.

There is disclosed in the drawings an automatic battery analyzer including the preferred embodiment of this invention and intended for analyzing 6 cell, 12 volt batteries of the lead-acid type normally used for automative lighting, starting, and ignition applications. As will be explained in greater detail hereinafter, the disclosed battery analyzer performs essentially four separate tests on the battery being tested in a prescribed order:

1. The open circuit voltage of the battery is measured, and if less than 10.80 volts the battery is determined to be unacceptable. This test primarily rejects batteries which have one or more severly shorted cells, leaks between one or more cells, or an open electrical circuit in the battery (usually at one of the welds connecting the cells together).

2. A load test is performed under which the battery is placed under a sufficient load to deliver a current equal to approximately one-half its cold current cranking rating for a period of twenty seconds. If the battery terminal voltage does not fail below a predetermined value, which value is corrected in accord with the temperature of the battery being tested, during that twenty second period, the battery is assumed to have passed the load test and to be acceptable and no further tests are performed on it.

3. A voltage bounce back test is performed by which the battery open circuit voltage is measured at the end of a fifteen second period following the completion of the load test. If the battery terminal voltage at the end of that period exceeds a predetermined value, which value is also corrected in accord with the temperature of the battery being tested, the battery is assumed to have passed the voltage bounce back test indicating that it is unacceptable and no further tests are performed on it.

4. A constant resistance load test is performed under which the battery is loaded with a fixed resistance and the variation of the battery terminal voltage with time over a given period is analyzed to determine the presence of a defective cell in the battery. In the embodiment disclosed, a 0.1 ohm. resistive load is placed across the battery for twenty seconds, and the battery terminal voltage is measured at the ends of each of the fifth through the twentieth seconds. These separate determinations are then made. (a) If the measured battery terminal voltage should increase by 0.17 volts or more during the sixth through the fifteenth seconds of that time period, the battery is assumed to have at least one defective cell and to be unacceptable. (b) If the battery is not yet determined to be unacceptable at the end of the fifteenth second, its terminal voltage is compared to a temperature corrected, standard battery voltage at the end of the fifteenth second and if the terminal voltage exceeds that standard, the battery is assumed to be acceptable and merely in need of being recharged. (c) If the acceptability of the battery is still not determined at the end of the fifteenth second, then if the battery terminal voltage should rise by 0.17 volts or more during the sixteenth through the twentieth second, the battery is assumed to be unacceptable; if the battery is not thus determined to be unacceptable, it is assumed that the condition of the battery is essentially indeterminate and the battery must be recharged to a fully charged condition and retested. (In the case of a battery known to be fully charged, only the open circuit voltage and load tests of items 1 and 2 above are performed.

In the prior application Ser. No. 863,925 referred to above, a battery analyzer was disclosed which included apparatus for performing a load test and a voltage bounce back test similar to those described in items 2 and 3 above; those tests will not be discussed in detail herein. FIGS. 1 and 2 of the drawing are graphs showing the relationship between terminal voltage and time for a plurality of six cell, lead-acid storage batteries having a nominal 12 volt terminal voltage for the first 20 seconds following the placing of a 0.1 ohm., resistive load across the battery terminals. In FIG. 1, each of the batteries was known to be an acceptable one but to have been discharged to some degree, whereas in FIG. 2, each of the batteries was known to have at least one defective cell. In analyzing the graphs of both FIGS. 1 and 2, it is assumed that the voltages measured during approximately the first five seconds of the discharge are essentially unreliable because of the effects of an initial surge of battery current and heating of the resistive load during this time period. Each of the batteries represented by curves 1-5 of FIG. 1 was able to hold an approximately constant terminal voltage during the fifth through the twentieth seconds of the discharge and was found to be an acceptable battery but in need of recharging before it should be returned to service. The condition of the battery represented by curve 6 of FIG. 1 having a terminal voltage which decreased significantly during the fifth through the fifteenth seconds was found to be essentially indeterminable without first recharging it. Each of the batteries represented by curves 7-11 of FIG. 2 had a terminal voltage which increased significantly, i.e., had a positive rate of change, during at least a portion of the fifth through the twentieth second time period and was found to have at least one defective cell.

There is disclosed in the remaining figures of the drawing apparatus for analyzing lead-acid, automotive, starting, lighting and ignition batteries making use of the observations referred to in the preceeding paragraph. FIG. 3 is a block diagram of an automatic battery analyzer apparatus incorporating this invention and FIGS. 4 and 5 are schematic diagrams of particular circuits of the apparatus of FIG. 3. The apparatus of FIGS. 3-5 as actually constructed included provisions for analyzing portions of an automobile other than the battery. As those other provisions form no part of the present invention, they are not included in the following description to the extent feasible to avoid obscuring the present invention.

The automatic battery analyzer disclosed in FIG. 3 includes a microprocessor controller circuit 100, first, second, and third peripheral interface adapters (PIA's) 102, 104, and 106, respectively, for interfacing various input and output circuits to the microprocessor 100, and two programmable read only memory circuits 108 and 110 for storage of the program used to control the operation of the microprocessor 100. The microprocessor 100 is an MC6802 and each of the PIA's is an MC6821 available from Motorola Semiconductor Products Inc., Austin, Texas; extensive information concerning the interconnections, use, and programming of these devices is available from Motorola Semiconductor Products Inc. Equivalent devices manufactured by others could be used and corresponding information concerning such devices is available from their respective manufacturers. The memories 108 and 110 are known by the industry standard designation 2716 and are each 8 bits by 2048 words; they are commercially available from a number of different manufacturers. Only the interconnections of the address bus (a sixteen line bus indicated as AO-A15 in FIG. 4), the data bus (an eight line bus indicated as D0-D7), and the reset line between the microprocessor, the PIA's, and the memories are indicated in FIG. 3 as the remainder of the necessary interconnections will be apparent to those familiar with the published information concerning these devices. Further detail is unnecessary and would unnecessarily complicate this description and perhaps obscure the invention. Similarly, only the connections between the PIA input/output data lines and control lines and the various input and output devices are shown in FIG. 3.

Each of the PIA's 102, 104, and 106 includes A and B side eight bit, input/output data registers (PAO-PA7 and PBO-PB7), and A and B side two bit control registers (CA1, CA2, CB1, and CB2). Address lines AO and A1 are connected to register select inputs of each of the three PIA's 102, 104, and 106, and a PIA select circuit 112 has inputs coupled to address lines A2, A3, A7, and A15 and outputs coupled to device select inputs of the three PIA's 102, 104, and 106 such that decoding of the address lines occurs placing the PIA registers at the addresses shown in Table 1 hereinafter. Address lines A0-A10 are connected to address inputs of the memories 108 and 110 and a memory select circuit 114 has inputs coupled to address lines A13, A14, and A15 and outputs coupled to device select inputs of memories 108 and 110 such that decoding of the address lines occurs placing the memories at hexadecimal addresses FOOO-FFFF.

The input information to the automatic battery analyzer of FIG. 3 is supplied through the first PIA 102. The battery under test 114 has its negative terminal coupled through connector 12 and through a low resistance current shunt 116 to one terminal of a variable resistance load bank 118. The positive terminal of battery 114 is coupled to a second terminal of load bank 118. The variable resistance load bank 118 is connected to an operated by a set of load bank relays 120 and is effective to place a variable resistance load across the battery under test. The magnitude of that load is determined by which one or ones of the load bank relays 120 are closed. A differential amplifier circuit 122 is provided having its inputs coupled across the battery under test 114 and its output coupled to one signal input of an analog switch 124. A differential amplifier circuit 126 has its inputs coupled across the current shunt 116 and has its output coupled to another signal input of analog switch 124. The signals at the outputs of amplifier circuits 122 and 126 are responsive to the voltage across and the current through the battery under test, respectively, at any instant of time. A temperature sensing circuit 128 includes a thermocouple 130 physically located at the connector 12 coupled to the negative terminal of the battery under test; the circuit 128 will also be described in further detail hereinafter. An output of circuit 128 is responsive to the temperature of thermocouple 130 and is coupled to a third signal input of analog switch 124. The series connection of two variable resistances 132 and 134 and a potentiometer 138 is connected between a positive supply source +V and circuit ground with the tap on potentiometer 136 being coupled to a fourth signal input of analog switch 124. Potentiometer 136 is calibrated so that the user may adjust it to a position corresponding to the cold cranking rating of the battery under test whereas variable resistors 132 and 134 are used for calibration. The analog switch 124 is effective to develop at an output a signal responsive to that present at a selected one of its signal inputs and has input select inputs coupled to data lines PA0-PA2 of the A side of the third PIA 106. The output of analog switch 124 is coupled to the input of a three and one-half digit analog-to-digital converter 138. The data lines PB4-PB7 of the first PIA 102 are coupled to the digit select lines of analog-to-digital converter 138, the outputs of analog-to-digital converter 138 are coupled to the data lines PBO-PB3 of the first PIA 102, and the data update input terminal and the end of conversion output terminal of the analog-to-digital converter 138 are coupled to the control lines CB2 and CB1, respectively, of the first PIA 102. A voltage reference source 139 is also coupled to the analog-to-digital converter 138. Thus, signals representing separate digits of the battery current and voltage and connector temperature are applied to the microprocessor data bus by the first PIA 102 in response to the generation of proper signals by the first and third PIA's 102 and 106. Analog switch 124 and analog-to-digital converter 138 may comprise, respectively, MC14051 and MC14433 units available from Motorola Semiconductor Products, Inc. with appropriate associated circuitry.

A power up and reset circuit 140 is provided for resetting the microprocessor 100 and PIA's 102, 104, and 106 when power is initially applied to the battery analyzer. The circuit has an output providing a signal to the reset inputs of the microprocessor 100 and each of the PIA's 102, 104, and 106 of proper magnitude and duration to cause resetting of these units to their initial condition. A "Stop" switch 142 is coupled to the circuit 140 for manual actuation so that an operator may stop a test in progress by closing the switch and resetting the microprocessor and the PIA's. Two manually operated switches 144 and 146 labelled "Test" and "Retest" are coupled to a switch encoder circuit 148. Manual operation of these switches indicates either that the operator desires to perform an initial test on the battery or that he wishes to retest a battery previously tested for reasons which will be discussed subsequently. Encoder circuit 148 is effective to apply an appropriate code to data lines PA0-PA3 of first PIA 102 in response to the status of switches 144 and 146.

Three forms of output devices are driven by the second and third PIA's 104 and 106, the variable resistance load bank 118 previously described, a light emitting diode display matrix 160, and a three-and-one-half digit display 162. The data lines PA0-PA7 and control line CB2 of the second PIA 104 are each individually coupled through one of a set of optically coupled isolators and drivers 164 to a relay in a set of load bank relays 120 such that the signals on those lines are each operative to control the operation of one of the relays. The relays controlled by data lines PA0-PA6 determine the magnitude of a variable resistance load placed across the battery under test with each controlling the current flow through one of a set of seven resistances connected in parallel. A schematic drawing of the load bank 118 with the associated relay contacts of the load bank relays 120 as well as the shunt 116 is shown in FIG. 5. The positive terminal of the battery under test is coupled to one contact of each of eight normally open pairs of contacts 300, 302, 304, 306, 308, 310 312, and 314. The other contacts of each of those pairs is connected to one terminal of a resistor 318, 320, 322, 324, 326, 328, 330, and 332, respectively, the other terminal of those resistors being connected together to the shunt 116. The relative magnitudes of the six resistors 318-328 are in binary increments with resistors 328 having the least resistance, and the magnitudes of the three resistors 328-332 are equal. A relay coil 334 is coupled across resistor 330 and operates a normally open contact pair 336 coupled in parallel with the contact pair 314. The relays 300-312 are operated in response to the signals on data lines PA0-PA6 of the second PIA 104, respectively, whereas relay 314 is operated in response to the signal on control line CB2 of the same PIA. Relays 300-312 and their associated resistors are effective to place a variable resistance load across the battery under test, but relay 314 is effective to place a fixed resistance load across the battery.

The light emitting diode matrix 160 includes 8 separate series combinations of a resistor and a light emitting diode, the light emitting diodes 170, 172, 174, 176, 178, 180, 182, and 184 being labelled to indicate "Testing" or test in progress, "Recharge/Retest", "Good/Recharge", "Bad", "Good", "Load Test On", "Low Voltage", and "Polarity Reversed", respectively. Control lines CA2 and CB2 of the third PIA 106 are coupled to a decoder/driver circuit 186 to enable one of the vertical matrix lines of the matrix 160, only one such vertical matrix line actually being shown in the apparatus of FIG. 3. Data lines PB0-PB7 of the third PIA 106 are individually coupled through one of a set of drivers 188 to the light emitting diode/resistor series combinations of the matrix 160 in order to appropriately select a particular one of the diodes for energization.

The digit display 162 would normally not be used with the battery analyzer of FIG. 3, but may be used when it is desired to know the value of the voltage appearing at the output of the analog switch 124 at any particular time. A manually operated switch, not shown, is coupled to data line PA3 of the third PIA 106 for coupling that line to circuit ground when no such display is desired and to a positive potential when such a display is desired. The data lines PB0-PB7 are individually coupled through one of a set of drivers 190 of PIA 106 to the segment and decimal point select lines of the display 162 and the data lines PA4-PA7 are individually coupled through one of a set of drivers 192 to the digit select lines of the display 162.

FIG. 4 is a detailed schematic diagram of the temperature sensing circuit 128 used to develop a signal responsive to the temperature of thermocouple 130. The thermocouple 130 is comprised of a copper/constantan junction. The copper side of that junction is coupled through copper lines 200, 202, and 203, and through a resistor 204 to the inverting input of a differential amplifier 206. The constantan side of that junction is coupled through constantan lines 208 and 209 to another copper/constantan junction 210 and then through copper lines 212 and 214 and through a resistor 216 to the noninverting input of amplifier 206, that input also being coupled to ground through a resistor 218. Feedback is provided through a resistor 220 coupled from the output of differential amplifier 206 to its inverting input. Differential amplifier 206 is coupled to positive and negative supply voltages +V1 and -V1, respectively, and a capacitor 222 is coupled from the positive power supply for the differential amplifier to the frequency compensation input thereof. The output of differential amplifier 206 is coupled through a resistor 224 to the inverting input of another differential amplifier 226.

A temperature compensating network 228 is also coupled to the inverting input of differential amplifier 226. Network 228 includes a temperature transducer 230 having its power supply terminals coupled to circuit ground and through a resistor 232 to the negative supply voltage -V1. The output of the transducer is coupled through a resistor 234 to the inverting input of differential amplifier 226. The copper/constantan junction 210 is placed in physical contact with the temperature transducer 230 so that the transducer is responsive to the temperature of that junction. A reference diode with built-in heat source 236 (e.g., an LM399Z) is included having its anode coupled to circuit ground, its supply input for the heat source coupled to the positive voltage supply +V1, and its cathode coupled to the positive supply +V1 through a resistor 238 and to the inverting input of amplifier 226 through the series combination of a variable resistor 240 and a fixed resistor 242.

The output of differential amplifier 226 is coupled both to the inverting input thereof through the series combination of two resistors 244 and 246 to provide feedback and to an output terminal point 248 through a resistor 250. The output terminal point 248 is coupled to one input of the analog switch 124. A protection circuit 252 is also coupled to output circuit point 248 to limit the magnitude of the potential at that circuit point and protect the input to analog switch 124 from excess potential. That network includes first and second diodes 254 and 256 having their cathode and anode terminals, respectively coupled to output circuit point 248 and their other terminals coupled to circuit points 258 and 260, respectively. Those latter two circuit points are each coupled to ground through resistor 262 and 264 with circuit point 258 being also coupled to the anode of a diode 266 having its cathode coupled to a second negative voltage supply -V2 of greater voltage (i.e., less negative) than the first negative voltage supply -V1 and circuit point 260 also being coupled to the cathode of a diode 268 having its anode coupled to a second positive voltage supply +V2 of lesser voltage than the first positive voltage supply +V1. The temperature sensing circuit 128 provides an output of approximately 20 millivolts per .degree.F. The solid state temperature transducer 230 compensates for changes in temperature of the reference junction 210. The reference diode 236 and associated resistors provide a constant, temperature stable, offset voltage.

The thermocouple 130 is physically located within a connector 400 attached to the negative post of the battery under test. The connector 400 is disclosed in FIGS. 6a-6c. The connector includes a connector block 402 having a circular portion 403 and a shank portion 405. A receptacle opening 404 on one side of circular portion 403 extends partway through the connector block for receiving the negative post or terminal of the battery under test. Receptacle opening 404 may be tapered so that the connector block 402 may easily be fit snugly onto the negative post of the battery under test. A thumbscrew 406 is provided which screws into a threaded opening 408 in the connector block 402. Opening 408 is disposed so that as thumbscrew 406 is advanced into it, a portion of the thumbscrew extends into the receptacle opening 404 so as to firmly attach the connector 402 to the battery negative post. An opening 410 is provided in circular portion 403 on the other side of receptacle opening 404 and extending through the connector block into communication with the receptacle opening. When the battery analyzing apparatus is connected to a battery having a threaded opening in the top of the negative post, such as are typical on side post automotive batteries, the thumbscrew 406 may be placed through opening 410 and advanced into the threaded opening on the battery post to secure the connector block 402 to the battery post.

The connector block shank portion 403 additionally has three openings 412, 414, and 416 at the base thereof and extending lengthwise through it. Opening 412 is sufficiently large to receive the electrical conductor which is coupled to the current shunt 16, whereas opening 414 need only be sufficiently large to receive the conductor coupled to the inputs of amplifier circuits 122 and 126. Opening 416 is of sufficient diameter to receive thermocouple 130 and its connecting wires 200 and 208 and retain the thermocouple 130 in thermal contact with the connector block 402. Opening 416 extends through the shank portion 405 into the circular portion 403 to permit the thermocouple 130 to be in thermal contact with the connector block 402 at a position relatively close to the receptacle opening 404. Threaded openings 418, 420, and 422 are provided in the connector block 404 and each is in communication with one of openings 412, 414, and 416, respectively, for receiving set screws to retain the associated conductors in the openings 412, 414, and 416. In an apparatus constructed in accord with this invention, the connector block had an overall length of 2.50 inches, a width of 1.00 inches, and the circular portion 403 and the shank portion 405 had thicknesses of 0.79 and 0.70 inches, respectively.

A flow chart disclosing the operation of the automatic battery analyzer of FIGS. 3-6 is shown in FIGS. 7a-7e, the presence of the same reference letters on different lines of those FIGS. 7a-7e indicating that those lines are connected together. As soon as the automatic battery analyzer is turned on by applying power thereto, a loop 500 is entered which successively measures and stores the temperature of the connector block 402, scans the pushbutton switches 144 and 146 to see if either one has been closed, and measures the potential across the battery connectors. As long as the voltage across the battery connectors is within the range of plus or minus 2.0 volts, it is assumed that the battery analyzer has not yet been connected to the battery under test and that the measured temperature of the connector block, T.sub.A, is the ambient temperature. When the measured voltage across the battery connector is outside the plus or minus 2.0 volt range, a second loop 502 is entered which skips the temperature measurement and storage steps of loop 500 since the connector block is no longer at ambient temperature but is at some temperature between the ambient temperature and the battery post temperature. Thus, after the connector is attached to the battery under test, the stored value of T.sub.A is temperature of the connector measured at the most recent time prior to connection of the connector to the battery.

When either of switches 144 or 146 is closed, a lamp test procedure 504 is performed by which all of the light emitting diodes 170-184 are simultaneously illuminated for a period of five seconds so that the operator may verify that each of them is functional. Another loop 506 is then entered which reads the battery voltage and, if that voltage is negative, turns on the light emitting diode 184 indicating to the operator that he has incorrectly connected the battery analyzer to the tester and that the positive and negative connectors should be reversed. If this occurs, the analyzer returns to the loop 500 when it is disconnected from the battery under test. After loop 506, the light emitting diode 170 is turned on to indicate that a test is in progress, the appropriate registers are set to zero, and the battery voltage is again read. If the battery voltage is less than 10.8 volts (for a nominal 12 volt battery) at the commencement of the test with no load applied to it, the battery is assumed to be unacceptable, light emitting diode 176 is turned on indicating that the battery under test is unacceptable, light emitting diode 182 is turned on indicating that the battery terminal voltage is excessively low, light emitting diode 170 is turned off indicating the test has terminated, and the analyzer returned to loop 502 and, after the bad battery is disconnected from the analyzer, to loop 500, in readiness for the next test.

If the battery voltage at the beginning of the test is greater than 10.8 volts, a procedure 508 is performed to determine the temperature of the battery under test. The nature of this procedure is more fully described and discussed in the copending application of G. A. Karnowski and D. V. Sriramamurty entitled "Storage Battery Electrolyte Temperature Measuring Apparatus", attorney's Ser. No. 126,172 filed date Feb. 29, 1980, and being filed contemporaneously herewith. First a fifty-five second time delay is introduced which, when added to the five second period of the lamp test routine 504, yields a total time delay for sixty seconds from the time the switch 144 or 146 was closed. At the end of that sixty second period the temperature of the connector block, T.sub.B, is measured. In the disclosed apparatus, two possible routines are used to determine battery temperature, T.sub.E, depending upon whether the difference between T.sub.A and T.sub.B is greater than 2.degree. F. If that difference is greater than 2.degree. F., the battery temperature is calculated as follows:

If the difference not greater than 2.degree. F., the battery temperature is calculated as follows:

After the battery temperature is determined, that temperature is used as an input to the testing procedure carried out. In the analyzer of FIGS. 3-7, load and bounce back tests similar to those used in the aforementioned Ser. No. 863,925 are used, but the third test used is one which detects the presence of a bad cell within the battery under test. As the first step in performing the load and bounce back voltage tests, the load and bounce back reference voltages are determined from the determined battery temperatures and stored in memory. The calculated reference voltages are referred to as ERFTMP and EBBTMP in FIG. 7 and Table 1. The relationships between the two reference voltages and battery temperature are shown in FIGS 4a and 4b of said application Ser. No. 863,925. At the initiation of the load test, light emitting diode 180 is illuminated to indicate that the load test is in progress and the cold cranking rating of the battery is read from the potentiometer 136 previously adjusted by the operator to correspond to the cold cranking rating of the battery under test. That rating is a value in amperes and is the current the battery is intended to be able to deliver for a period of thirty seconds at zero degrees Fahrenheit without the battery terminal voltage falling below 7.2 volts for a conventional, six-cell, twelve volt lead-acid storage battery. The relays in the load bank relays 120 are then closed to place the corresponding resistances across the battery under test and load it to deliver a current of approximately one-half its cold cranking rating. The battery current is monitored by measurement of the voltage across the shunt 116 and if the difference between the actual current and one-half the cold cranking rating is greater than approximately 9 amperes, the load is altered to bring the current closer to one-half of the cold cranking rating. The battery terminal voltage is measured during the load test. If at the end of twenty seconds that voltage exceeds the previously determined ERFTMP the battery under test is assumed to be acceptable, the load is removed from the battery, light emitting diodes 180 and 170 are extinguished, light emitting diode 178 is energized to indicate that the battery under test has been determined to be acceptable, and the tester returns to loops 500 and 502 in preparation for the next test. If the battery terminal voltage falls below ERFTMP at any time during the load test, the battery is assumed to have failed the load test, light emitting diode 180 is extinguished, and the load is removed from the battery. In the event that load test failure occured while the battery under test is being retested and is known to have been fully charged (i.e., the test sequence was initiated by closure of pushbutton switch 146), the battery is then known to be unacceptable. Light emitting diodes 176 and 170 are illuminated and extinguished, respectively, and the analyzer is returned to loops 501 and 502 in preparation for the next test. If the load test failure occurred during an initial test of the battery, the voltage bounce back test is performed by measuring the battery voltage fifteen seconds after termination of the load test. If that measured voltage exceeds EBBTMP, the battery has passed the bounce back voltage test but is assumed to be unacceptable, light emitting diodes 176 and 170 are illuminated and extinguished, respectively, and the analyzer is returned to loops 500 and 502 in preparation for the next test. If that measured voltage is less than EBBTMP, the condition of the battery is still undetermined and the test to determine the presence of a bad cell is performed.

The first step in the bad cell detection test is to clear a register labelled BATVL. A temperature corrected standard voltage, BATVTC, is then calculated as follows:

A fixed resistance load of 0.1 ohm then placed across the battery by closing the contact pair 314. After a five second delay, a loop procedure 520 is executed during which the voltage across the battery terminals is measured each second and the result of each voltage measurement is compared with the voltage measurement made the previous second.

In the case of a normally operating battery, the voltage supplied to a fixed resistance load is a monotonically decreasing function of time after an initial period of discharge. Here the initial period is taken as five seconds. If after that initial period and while the constant resistance load is across the battery terminals the terminal voltage measured at any time exceeds the voltage measured the preceeding second, it is known that the terminal voltage is increasing, i.e., that its rate of change is positive. The first time the measured battery voltage exceeds the voltage measured in the previous second, the previously measured voltage is stored in the BATVL register. That stored voltage is assumed to be the minimum battery voltage. if while the constant resistance load is across the battery terminals, the measured battery terminal voltage exceeds the minimum battery terminal voltage by 0.17 volts, the battery is assumed to have a bad cell and thus to be unacceptable, light emitting diode 176 is illuminated, the relay contacts 314 are opened to remove the load from the battery, light emitting diode 170 is extinguished, and the tester returns to loops 500 and 502 in preparation for the next test.

Loop 520 is also exited after a ten second time period at point 524. If at that time the battery terminal voltage measured during the proceeding second exceeds the previously determined BATVTC, the battery is assumed to be acceptable but in need of a recharge before being returned to service; in this case light emitting diode 174 is illuminated to make the proper indication to the operator, relay contacts 316 are opened to remove the load from the battery, light emitting diode 170 to extinguished, and the tester returns to loops 500 and 502 in preparation for the next test. If at that same time the battery terminal voltage measured during the proceeding second does not exceed the previously determined BATVTC, a loop 526 is entered which is quite similar to the loop 520 and carries out the constant resistance load test for another five seconds while continuing to look for increases in the battery terminal voltage. If the battery terminal voltage should increase during that additional five second period and if the battery terminal voltage should exceed by 0.17 volts the minimum value thereof obtained during either that period or the previous 10 second period, the battery is again assumed to have a bad cell and to be unacceptable, light emitting diode 176 and 170 are illuminated and extinguished, respectively, the load is removed from the battery, and the tester returns to loops 500 and 502 in preparation for the next test. If during that additional five second period the battery is not thus determined to be unacceptable, then it is assumed that insufficient information concerning the battery is available to determine its condition and the battery must be recharged and retested to determine its acceptability. In this case, light emitting diode 172 is illuminated to give the proper indication, the load is removed from the battery, light emitting diode 170 is extinguished, and the tester returned to loops 500 and 520 in preparation for the next test.

A program for the 6802 microprocessor in assembly language to execute the program of the flow charts of FIGS. 7a-7e is at Table I as follows: ##SPC1## 15/977 20/240

It will thus be seen that there have been provided circuitry for completely automatically detecting the presence of a single defective cell in a multi-cell storage battery as well as a fully automatic battery analyzing apparatus which makes use of such a bad cell detecting circuit. The disclosed circuitry is effective to detect the presence of significant increases in the battery terminal voltage while the battery is subjected to a relatively constant load; it is most effective in detecting the presence of electrical shorts around or through the cell separators, the presence of unintended connections between adjacent battery straps, adjacent battery plates, or between a battery strap and plate, and is effective but to a lesser degree in detection the loss of electrical continuity or mechanical connection between the battery plates and straps, the presence of a broken strap, and the exisence of positive plate "polarization" in one of the battery cells. The combination of such circuitry with apparatus for performing a battery load test and a bounce back voltage test results in a battery analyzer which is more reliable than those available in the prior art and which gives a correct indication of the acceptability of the battery under test without the necessity of recharging and retesting the battery with more frequency than those available in the prior art.

In the particular embodiment disclosed, in order to detect the presence of a bad cell, the battery terminal voltage was analyzed to determine whether it increased by more than 0.17 volts in any time interval during the periods of 5 to 20 seconds following the application of a 0.1 ohm resistive load to the battery. The analysis may alternately be viewed as a determination of whether the rate average of change of the battery terminal voltage exceeded 0.017 volts/second at any time during the 5 to 15 second period or 0.011 volts/second at any time during the 5 to 20 second period. If different loads were used, correspondingly different voltage changes and/or rates of change might also be used. Further, many other techniques may be used for determining whether the time rate of change of the battery terminal voltage exceeds a predetermined value.

Flow charges disclosing the operation of alternative embodiments to the automatic battery analyzer of FIGS. 3-7 are shown in FIGS. 8 and 9. Each of FIGS. 8 and 9 are flow charts of the operation of only a portion of the battery analyzer and are intended to be read in conjunction with FIGS. 7a-7e to disclose the operation of the complete apparatus. The flow chart of FIG. 8 is intended to be substituted for the material within the dashed box 550 of FIG. 7e to disclose a first alternative embodiment of this invention and the flow chart for FIG. 9 is similarly intended to be substituted for the material within the dashed box 552 of FIGS. 7c and 7d to disclose a second alternative embodiment. In each case, the lines designated by letters in the FIG. 7 drawings are intended to join the similarly designated lines of FIGS. 8 and 9.

In the FIG. 8 embodiment, at the conclusion of the five second period of loop 526 rather than ascertaining that insufficient information is available to determine the condition of the battery and indicating that the battery must be recharged and retested, apparatus is provided for performing two further tests on the battery. Each of the two additional tests involve measuring the battery terminal voltage at 20 seconds following the application of the fixed resistance load across its terminals, correcting that measured voltage to compensate for the temperature of the battery under test, and normalizing that measured value to the terminal voltage which would be generated by a battery having a cold cranking rating of 500 amperes being in a condition smaller to that of the battery under test. That correction and normalization is carried out to arrive at a value of BATV20.sub.NC as follows: ##EQU1## CCR=Cold cranking rating of battery under test.

Following determination of BATV20.sub.NC, the load is removed from the battery and, as the first test, a determination is made of whether BATV20.sub.NC is less than 2.0 volts. If so, the lamp indicating that the battery should be recharged and retested is illuminated; this test is included because it is known that in some cases badly discharged but otherwise acceptable calcium and strontium lead storage batteries will have an abnormally low value for BATV20.sub.NC.

The second test of FIG. 8 is performed if the first test does not indicate that the battery should be recharged and retested and comprises a comparison between BATV20.sub.NC and the battery terminal voltage at the time 15 seconds after the fixed resistance load is removed from battery, i.e., 35 seconds after that load was applied to the battery. A 15 second delay is first executed and the battery terminal voltage is then measured and corrected for the temperature of the battery under test to arrive at a value for BATV35.sub.C as follows:

T.sub.E =battery temperature in .degree.F.

The value of BATV20.sub.NC is then compared with BATV35.sub.C according to the following inequality:

If the inequality is satisfied, the battery is assumed to be unacceptable. It has been observed that batteries which are either severely discharged or have a single defective cell may both have low values of approximately eleven volts for BATV35.sub.C, but among batteries with such low values of BATV35.sub.C, those having a defective cell will tend to have higher values for BATV20.sub.NC. This second test is intended to take this observation into account.

In the embodiment of FIG. 9 the procedure for testing the battery after it has been recharged is modified from that shown in FIG. 7 in two specific manners as disclosed in that FIG. 9. First, if the battery is being retested, the value of terminal voltage required at the end of the load test for the battery to be determined acceptable is increased over that used in the FIG. 7 embodiment by approximately 0.2-0.3 volts. Second, the voltage bounce back test is performed when the battery is retested. If the battery terminal voltage at the end of 15 seconds following the termination of the load test exceeds the calcualted value of EBBTEMP, the battery is assumed to be unacceptable and the lamp so indicating is illuminated. If, however, the battery terminal voltage at the same time is less than EBBTEMP, it is known that the battery was not in fact fully charged with the retest procedure was commenced, and that it should be fully recharged and retested. The lamp so indicating is then illuminated. The purpose at this procedure is to ascertain the existence of batteries being retested which were thought to be fully recharged but were not. A particular instance where this might occur is where strontium or calcium storage batteries are placed in parallel on the same charger with antimony batteries.

While particular embodiments of this invention have been described, it will be understood, of course, that the invention is not to be limited thereto since modifications may be made by those skilled in the art, particularly in light of the foregoing teachings. As an example, and in addition to the modifications previously discussed, in the disclosed embodiment the circuitry for performing the bad cell detector test analyzes the battery terminal voltage when a constant resistance load is placed across the battery; in the battery analyzer described, this means that the battery is subject to relatively high current drains during two separate occasions, during the load test and during the bad cell detection test. It might be found desirable to perform a test equivalent to the disclosed bad cell detection test but using a variable load instead of an essentially fixed load; this possibility might be found especially attractive in embodiments using a microprocessor such as the embodiment here disclosed as the appropriate program could be provided with the microprocessor would make any necessary calculations, conversions, or corrections so that even though a variable resistance load was in use, the test performed would be fully equivalent to that performed by the bad cell detector circuitry disclosed herein. In some instances it may even be found desirable to perform the bad cell detector test at the same time the load test is being executed. Further, while in the disclosed embodiment a 0.1 ohm load is applied to the battery for fixed time periods of fifteen seconds and five seconds, different loads and different time periods could be used. Likewise, while in the disclosed embodiment a rise in the battery terminal voltage while under load of more than 0.17 volts was considered to be indicative of a defective cell, it might be found desirable in the future to use a different voltage change for the criterion. Further, it might be found desirable to add to the battery analyzer disclosed herein the internal resistance test referred to in the aforementioned application Ser. No. 863,925.

It is contemplated by the appended claims to cover any such modifications as fall within the time spirit and scope of this invention.