NZ551097A

NZ551097A - Battery management apparatus

Info

Publication number: NZ551097A
Application number: NZ551097A
Authority: NZ
Inventors: Darren Grant Lucinsky
Original assignee: Power Man Technologies Nz Ltd
Priority date: 2006-11-06
Filing date: 2006-11-06
Publication date: 2009-09-25
Also published as: WO2008056316A1

Abstract

A method of charging lead acid batteries is provided. The method includes, in order, the steps of: a) applying a first applied voltage to a battery until the open terminal voltage of the battery exceeds a first target voltage, and then disconnecting the first applied voltage, b) applying no charging voltage to the battery for a time exceeding 30 seconds, and monitoring the open terminal voltage of the battery until the voltage falls below a second target voltage, c) connecting a second applied voltage lower than the first applied voltage to the battery until the battery reaches a second target voltage lower than the first target voltage, d) connecting a third applied voltage lower than the second applied voltage to the battery for an extended period of time. An apparatus for performing the method is also provided.

Description

<div class="application article clearfix" id="description"> 551097 *10055140880* OW°JeaLP^Psrty 15 Bec Of N,z. 2008 £ I ved NEW ZEALAND PATENTS ACT, 1953 COMPLETE SPECIFICATION BATTERY MANAGEMENT APPARATUS We, POWER MANAGEMENT TECHNOLOGIES NZ LIMITED, a company duly incorporated under the laws of New Zealand of Level 1, The Square Centre, Cnr Main Street and The Square, Palmerston North, New Zealand, do hereby declare this invention to be described in the following statement: - / - 551097 -2- FIELD OF THE INVENTION The present invention relates to an apparatus for battery management, and in particular, but not solely, lead-acid battery charging methods and methods for determining lead-acid batten,- health. 5 BACKGROUND TO THE INVENTION A sealed lcad-acid (SLA) battery is an electrochemical dcvice. The following describes a typical sealed lead acid batter)'. The positive plates are a grid frame of lead-tin-calcium alloy and hold 10 porous lead dioxide as the active material. The negative plates ate also a grid frame of lead-tin-calcium alloy and hold spongy lead as the active material. The electrolyte is diluted sulphuric acid. The plates are electrically separated by a micro porous absorbed glass mat which is chemically stable in the electrolyte. The battery converts stored chemical energy to electric energy: 15 Discharge Pb02 + Pb + 2H2SO+ O PbS04 + PbS04 + 2H,0 Charging These types of SLA batteries arc called negative electrode recombination types. In the final stages of charging, oxygen is generated at the positive plates and is absorbed into the surface of 20 the negative plates and consumed. Over-charging leads to excessive oxygen generation and too much oxygen is absorbed into the negative electrode. I'his contaminates the negative electrode and damages the charge capacity of the battery permanently. An SLA battery must be charged prior to use. A charger matched to the application of the 25 battery is required to preserve the life of a battery. Battery life is determined by such factors as ambient temperature, the applied charging voltage, manufacturing materials, construction, charging times and load profile. Of these factors, charge time and load profiles are the most detrimental to battery life if not carefully considered. 30 The charging of SL A batteries is usually by several well known charging systems. The charging systems that exist are, constant voltage (CV), constant voltage/constant current (CV/CC), two step constant voltage (2SCV) or trickle charge (TC). The particular charging technique used typically depends on the particular battery or the battery application. 551097 -3 - These charging systems do not test the battery voltage directly. Instead, the charging systems often wait until the battery voltage has reached a specific level before determining whether charging should continue. This means the battery voltage measurement is usually taken while the battery remains connected to the charger. 5 These charging systems do not test for the battery health. In addition, it is possible to permanendy damage the batteries using these charging methods. There are two types of charging regimes, a regime is chosen based on battery application. These 10 are main power source (cycle use) and back-up power source (stand-by use). Main power source (cycle use) is to use the batter}- by repeated charging and discharging in turn. There are typically two well known methods of charging for main power source applications. These are the constant voltage (CV) charging method and the constant voltage and constant 15 current (CC/CV) charging methods. The constant voltage charging method is to charge the battery by applying a constant voltage across the battery terminals. The battery is charged at the cycle charge voltage (a battery parameter usually specified by the manufacturer). Charging is complete when the charge current 20 continues to be stable for three hours. Figure 4 illustrates examples of charge time varying with different charging currents. Combinations of output voltage and current from the charger result in a variety of charging profiles. Figure 5 illustrates typical characteristics of constant voltage charging. 25 Under the constant voltage constant current charging method, the charging voltage is applied to the battery and the charging current is limited, for example to 0.4 CA. CA is a battery parameter specified by the manufacturer. Figure 6 illustrates a typical example of charging voltage and current profile over time using this method. The initial charging voltage is reduced to limit the 30 charging current. The battery is considered charged when the charge current of the battery falls below a predefined limit. The limit is determined by the design of the battery. The charge current will never reach zero due to internal battery self discharge. The need for a compensating charge current 35 negates the effect of this discharge to maintain voltage level. 551097 -4- Back-up power source (stand-by use) is where a battery system is provided as a secondary source of power. An application load is normally supplied with power from a mains power source. The battery system is maintained so it can supply power to the load in case of main power failure. 5 There are three well known methods of charging for stand-by applications. These are the two-step constant voltage charging method, the compensating (trickle) charging method and the float charging method. The two-step constant voltage charging method first applies a cycle charge voltage limited at 0.15 10 CA until the charge current reaches a preset magnitude (unspecified by manufacturer, determined bv design). A trickle charge voltage is then continuously applied until the battery is used. The trickle charge voltage is a battery parameter specified by the manufacturer. Figure 7 illustrates a battery voltage and current profile using this type of method. 15 The trickle charging method is typically used when a batter}' is not normally connected to any load. Figure 8 illustrates an example schematic of this arrangement. The trickle charge voltage is applied to the battery continuously to compensate for self discharge. 20 The trickle charge voltage should also be adjusted for batter}? temperature. Figure 9 illustrates a plot of a recommended trickle charging voltage verses battery temperature. Typically, a manufacturer will provide a warning that states "small differences in charging voltage can result in a significant difference in battery life" and "charge voltage should be controlled 25 within a narrow range and with litde variation for a long period". 'There are also serious lifespan issues if the battery is used over a broad temperature range using this method. Figure 10 illustrates a plot of battery lifespan verses temperature. Batteries operating in temperatures above 45°C result in a serious reduction in battery life span, regardless of the charging method applied.. Temperatures below the optimum 25°C may also reduce battery lifespan, regardless of the 30 charging method applied. Figure 11 shows results from manufacturer testing at a battery operating temperature of 50°C. The high operating temperature has a detrimental effect to battery lifespan. A high battery temperature while connected to a load will result in decreased battery output current and 35 decreased output voltage and total capacity. 551097 -5- The float charging method is a variation to the trickle charging method where the battery is connected across the load in a standby mode. Figure 12 illustrates a schematic of this arrangement. A charging voltage is applied to the battery continuously to compensate for self 5 discharge. The power source (rectifier) is capable of supplying both the load and the battery with sufficient current to both operate the load and charge the battery. An advantage of this method is that the load gets a continuous (uninterrupted) power supply should the rectifier or power supply to the rectifier fail for any reason. 10 A disadvantage with this method includes variations in charge current due to variations in the load current requirements. In addition, the battery can be run flat when required to power the load in the event of a mains power failure. This can permanendy damage the battery. The battery may appear as a short upon re-activation of the mains power. The load will not get power and the battery may dangerously overheat. 15 Another disadvantage of this method is the load must be supplied with the battery charging voltage. This may cause issues with heat dissipation in the load power circuitry. An uninterruptible power supply (UPS) is designed to specifically provide AC mains support for 20 sensitive equipment, for example computers and medical equipment. UPS apparatus have a battery or batteries that supply a DC voltage. The DC voltage is converted to AC by an AC inverter. Typically, the AC inverter converts the 13C voltage into one of the two world wide AC voltage standards: 240VAC@50 Hz or 120VAC@60Hz. 25 UPS apparatus typically use valve regulated lead acid (VRLA) or wet cell batteries. The batteries are used either internally within the apparatus housing (physically small batteries, usually VRL A types) or external to the apparatus housing and connected with cabling (physically large batteries, large capacity VRLA and all wet cell types). 30 The batteries are charged by temperature compensated CV or CV/CC chargers. No battery health or capacity checks are performed with the exception of a "fault" when the battery has died completely. Faults are derived from either the charger drawing too much current or a battery voltage measurement reading too low. 35 551097 -6- In very large UPS installations, for example those used by telecommunication companies, the batteries are housed in a separate room as they are very large capacity and potentially very dangerous (fire, explosion, leaking acid, electric shock). 5 Companies such as BTECH and EATON POWERWARE have developed battery health testing methods as part of their corporate UPS solutions. The test derives information about the battery capacity by measurement of the battery impedance. However, the impedance testing requires rigorous mathematical processing. Such systems are therefore often linked back to a local computer or data centre for processing using licensed software. This type of battery health 10 testing system is prohibitively expensive and is well beyond the budget of most UPS user's requirements. SUMMARY OF THE INVENTION 15 It is therefore an object of one aspect of the present invention to provide a lead-acid battery charging system that goes at least some way toward overcoming the disadvantages associated with prior art battery chargers, or provides the public with a useful choice. It is an object of another aspect of the present invention to provide a method of determining the 20 health of a lead-acid battery, or at least provides the public with a useful choice. In a first aspect the invention consists in a method of charging lead acid batteries comprising, m order, the steps of: a) applying a first applied voltage to a battery until the open terminal voltage of said battery 25 exceeds a first target voltage, and then disconnecting said first applied voltage, b) applying no charging voltage to the battery for a time exceeding 30 seconds, and monitoring the open terminal voltage batteries of said battery until the voltage falls below a second target voltage, c) connecting a second applied voltage lower than said first applied voltage to said battery 30 until said battery reaches a second target voltage lower than said first target voltage, d) connecting a third applied voltage lower than said second applied voltage to said battery for an extended period of time. 551097 -7- Preferably if a charging voltage is applied to said battery, said charging voltage is intermittently briefly disconnected until said battery voltage is measured and reconnected when said battery voltage has been measured. Preferably step (b) includes measuring the open terminal battery voltage while no charging 5 voltage is applied. Preferably step (a) is reinitiated if said battery voltage falls below a fourth target voltage during step (d). Preferably step (c) is reinitiated if said battery voltage falls below a fourth target voltage during step (d) and the time spent at step (d) has exceeded 24 hours. 10 Preferably said first applied voltage is a charge cycle voltage. Preferably said charge cycle voltage is about 2.45V/cell. Preferably a first current does not exceed the manufacturer's recommended current limit and said first current is dependent upon said applied voltages. Preferably said first current is 0.4CA. 15 Preferably said first target voltage is between 2.35V/cell and 2.45V/cell. Preferably said second target voltage is about 2.25V/cell. Preferably said second applied voltage is between 2.30V/cell and 2.40V/cell. Preferably said third target voltage does not exceed a trickle charging voltage. Preferably said third target voltage is not more than 2.25V/cell. 20 Preferably said third applied voltage is about 2.25V/cell. Preferably said fourth target voltage is about 2.15V/cell. Preferably a load is said connected to said battery, after step (d), to facilitate the discharge of said batter)'. Preferably said battery is connected to said load until the battery voltage falls below a fifth target 25 voltage. 551097 - 8- Preferably wherein said fifth target voltage is about 1.90V/cell. Preferably step (a) is reinitialised when said fifth target voltage is met. In another aspect the invention consists in a method of determining the health of a lead-acid batter)? of a known type, comprising the steps of: 5 a) applying a charging voltage to said battery, b) disconnecting said charging voltage when the open terminal voltage of said battery rises above a first voltage limit, c) reapplying said charging voltage when said open terminal voltage of said battery falls below a second voltage limit, 10 d) determining the health of said battery by comparing a measure related to the voltage peaks counted against a known profile for the battery type. Preferably said charging voltage is greater than said first and said second threshold voltages. Preferably said charging voltage is approximately 14V. Preferably said first voltage limit is about 13V. 15 Preferably said second voltage limit is about 13.3V Preferably said first voltage limit equals said second voltage limit. Preferably while if said charging voltage is applied to said batten', said charging voltage is intermittently briefly disconnected while said battery voltage is measured and reconnected when said battety voltage has been measured. 20 Preferably said battery voltage is measured every 2 seconds. Preferably said battery voltage is measured after said charging has been disconnected for a first period of time. Preferably said first period of time is 1.9 seconds. Preferably said peak count and limit voltages is indicative of battery health. 25 Preferably when said peak count is high, determining that said battery is an unhealthy battery. 551097 -9- Preferably when said peak count is low, determining that said battery is a healthy battery. In another aspect the invention consists in a method of testing the health of a lead-acid battery of a known type comprising: intermittently applying a charging mode to charge up said battery, the charging mode 5 having a cycle at a first frequency including a charging period in which voltage is applied and a second period in which the open circuit battery voltage can be measure, said first frequency being much faster than said intermittent application of said charging mode, monitoring the open circuit terminal voltage of said batter}', determining battery health based on comparing a measurement relating to the number of peaks in said monitored open 10 circuit terminal voltage against a profile for the battery type. In another aspect the invention consists in a method of testing the health of a lead-acid battery of a known type by comparing a measure related to the voltage peaks counted during voltage response testing to a known profile for the battery type. Preferably said voltage response testing comprises: 15 a) applying a charging voltage to said batter}?, b) disconnecting said charging voltage when the open terminal voltage of said battery rises above a first voltage limit, c) reapplying said charging voltage when said open terminal voltage of said battery falls below a second voltage limit, 20 d) determining the health of said battery by comparing the voltage peaks counted to a known profile for said batter}? type. Preferably said charging voltage is greater than said first and said sccond threshold voltages. Preferably said charging voltage is approximately 14V. Preferably said first voltage limit is about 13V. 25 Preferably said second voltage limit is about 13.3V. Preferably said first voltage limit equals said second voltage limit. 551097 - 10- Preferably while if said charging voltage is applied to said battery, said charging voltage is intermittently briefly disconnected while said battery voltage is measured and reconnected when said battery voltage has been measured. Preferably said battery voltage is measured every 2 seconds. 5 Preferably said battery voltage is measured after said charging has been disconnected for a first period of time. Preferably said first period of time is 1.9 seconds. Preferably said voltage response testing is indicative of battery health. Preferably when said voltage response is fast, determining that said battery is an unhealthy 10 battery. Preferably when said voltage response is slow, determining that said battery is a healthy battery. BRIEF DESCRIpTION OF THE DRAWINGS A preferred embodiment of the inventions will be described with reference to the accompanying 15 figures. Figure 1 is a block diagram of the battery management apparatus and its components. Figure 2 is a block diagram of multiple preferred battery management apparatus in use with various communication interfaces. 20 Figure 3 is a block diagram of the physical components of the battery management apparatus and how they are connected together. Figure 4 is a graph of constant current charge characteristics by current showing voltage and current versus time. Figure 5 is a series of graphs showing characteristics of a constant current charger, and 25 corresponding graphs showing the voltage and current profile of a battery. Figure 6 is a graph for constant-voltage constant-current characteristics showing charge current and charge current versus time. Figure 7 is a graph of charging characteristics of a two step constant voltage control charger showing battery voltage and charge current versus time. 30 Figure 8 is a schematic representation for two step voltage and trickle charge system model. Figure 9 is a graph of compensated voltage values versus temperature. 551097 - 11 - Figure 10 is a graph of the influence of temperature on the service life of a battery showing service life versus temperature. Figure 11 is a graph of the service life characteristics at 50°C (122°F) compared with 25°C and 20°C of a battery using discharge time duration versus lifespan period. 5 Figure 12 is a schematic representation of a float charge system model. Figure 13 is a graph of open circuit voltage versus residual capacity for 25°C (77°F). Figure 14 is a graph of cycle life versus depth of discharge for a battery. Figure 15 is a graph of voltage response for a battery in good condition (new battery) utilising the invention showing battery voltage versus time. 10 Figure 16 is a graph of voltage response for a battery in poor condition (old battery) utilising the invention showing battery voltage versus time. Figure 17 shows a diagram of the components in a preferred battery health monitoring system. Figure 18 shows a flow diagram of steps taken to determine battery health. Figure 19 shows the flow diagram of figure 18 in further detail. 15 Figure 20 illustrates an example plot of open circuit battery voltage during health testing. Figure 21 shows a graphical analysis of battery health data. Figure 22 illustrates a hierarchal arrangement of several connected battery management apparatus. 20 DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS The preferred apparatus can be used for telecommunications type applications. Such applications include the power supplies located in cell phone repeater sites. Preferred apparatus for this application includes both solar charging capability and mains supply capability. However, the 25 inventions disclosed are applicable to other applications. The charge power source and load type, and any aspects of the apparatus adapted for these purposes are not part of the invention. Described in this specification is an apparatus that enables preferred battery management techniques. A preferred battery charging method and a preferred battery health monitoring 30 method will also be described. The apparatus monitors the state of charge in one or more batteries in a particular location. The apparatus also monitors the health of one or more batteries in the particular location. The apparatus also charges one or more batteries in a particular location using knowledge of battery 35 health and state of charge. 551097 - 12 - Figure 1 illustrates the battery management (BM) apparatus according to one embodiment of the present invention. The apparatus includes a housing 1 whereby the term external is defined as being external to the housing. 5 The apparatus includes a series of temperature sensors 2, 3 to measure the temperature of a battery under test or being charged. An environmental temperature sensor 4 is connectable to the apparatus. Each of the temperature sensors is connectable to the apparatus by connector 15. Connector 15 contains protection device 51 to isolate the internal components from potentially 10 harmful electrical mishaps, such as a power shortage. Internal components that may be sensitive to electrical mishaps include microcontroller 39. Connected to microcontroller 39 is digital input 40 and analogue input 41. Each of the inputs 40, 41 is interfaced outside the housing 1 by connectors 16 and 17 respectively. Again, each 15 connector 16, 17 is isolated by a protection device 52, 53. A master slave interface 42 is also interfaced out the housing 1 by a connector 18. Connector 18 is again isolated by a protection device 54. Connected to microcontroller 39 is digital output 46 and analogue output 47. F.ach of the outputs 20 46, 47 is interfaced outside the housing 1 by connectors 19 and 20 respectively. Again, each connector 19, 20 is isolated by a protection device 55, 56. A communications interface 48 is also interfaced out the housing 1 by a connector 21. Connector 21 is isolated by a protection device 57. Connectors 16,17, 18 allow each of the inputs 40, 41, 42 to connect to other devices external to the apparatus housing 1. 25 A solar power source 5 and an AC mains power source 6 are connectable to die apparatus housing 1 by connectors 9 and 10 respectively. The connector 9 allows the solar power source 5 to connect to a solar power sensing circuit 28 which is internal to the housing 1. Similarly, connector 10 allows the mains power source to connect to an AC power sensing circuit 29 which 30 is also internal to the housing 1. The solar power sensing circuit 28 is connected to a charge pump 22. The output of the charge pump 22 is connected to a charge pump sensing circuit 30. The AC power sensing circuit 29 is connected to an AC rectifier 23. The output of rectifier 23 is connected to an AC rectifier sensing 35 circuit 31. 551097 - 13 - One or more batteries are located external to the apparatus housing 1 and connectable to the housing via a connector. Shown in Figure 1 are a first battery 7 and a second battery 8. First battery 7 and second battery 8 are connected to housing 1 by connectors 11 and 12 respectively. 5 First battery 7 is then electrically coupled to a sensing circuit 32. Similarly, the second battery 8 is connectable to a second sensing circuit 33. Each of the power sources and the one or more batteries are connected to a battery switch 37. Also connected to the battery switch 37 is a series of outputs. An external solar load 58 is 10 connectable to the apparatus housing 1 via connector 13. Connector 13 allows the external solar load 58 to connect to a solar load 27. The solar load 27 is then connected a solar load sensing circuit 34. The solar load sensing circuit 34 is connected to battery switch 37. An external load 59 is also provided externally to the apparatus housing 1. The external load 59 is 15 connectable to a main input 26 via connector 14 that is mounted in the housing 1. The main output 26 is connected to a main output sensing circuit 35. Battery switch 37 also connects to the main output sensing circuit 26. A charge regulator 25 is provided internal to the housing 1. A regulator temperature sensor is 20 located proximate the charge regulator 25. The charge regulator is connected to the battery switch 37 via a charge regulator sensing circuit 36. Each of the sensing circuits 24 - 36 and battery switch 37 are interconnected and interfaced to microcontroller 39 via connections internal to the housing. It may follow that the sensing 25 performed by each of the sensing circuits by be incorporated into the microcontroller 39, and that the internal interconnections may comprise a signal multiplexing technique, or an individual connection, A microcontroller power supply is connected between the battery switch 37 and the 30 microcontroller 39. A series of configuration switches 43 can be mounted on the housing 1. Each configuration switch is configurable to allow components internal to housing 1 by an external device or user. 551097 - 14- Also internal to the apparatus housing 1 are a memory device 44, a real time clock 45 and a display driver 49. The display driver 49 connects to status indicators 50 mounted on the housing 1 such that they are viewable from the outside of the housing. The memory 44, clock 45 and the driver 49 are each connected to the microcontroller 39. 5 The AC mains power source 6 or the solar panel 5, or both, can be used to power the apparatus. The AC rectifier 23 is required when the mains power supply is an alternating current. The apparatus is adapted to use solar power in remote areas where telecommunications equipment is often located, and mains power is not readily available. 10 It is often preferable for an AC mains power source 6 and a solar panel 5 to both be used. In such situations, power can be drawn from the solar panel 5 when sunlight is available and from the mains source when sunlight is not available. 15 The apparatus includes a charge pump 22 between the solar panel connector 9 and the battery switch 37. The charge pump 22 for the solar power source 5 enables the power supplied from one or more connected solar panels to be converted to a power source that can be readily connected to other nominal electronic devices. For example, the output voltage from a solar panel 5 will be low at dusk and dawn. The charge pump 22 converts the output voltage into a 20 high voltage and low current power source that is more suited for battery charging systems and digital electronics. The preferred BM apparatus is designed to be powered by an external solar panel 5 or AC mains 6. Either or both power sources can be connected when or where available. 25 A single battery can be used when solar power is not available but AC mains power is available. However, reliability is reduced if this battery fails due to there being no backup battery. The apparatus for facilitating solar power may include a predetermined load for solar intensity 30 determination. The use of a resistive load is useful for determining the maximum instantaneous power available, rather than only what power is being consumed. The predetermined standard resistive load 27 for solar intensity determination is preferably a high power resistive load. The high power load may alternatively be located outside the apparatus. EXT S load 58 in Figure 1 illustrates the high power load be located outside of the apparatus housing 1. Locating the load 35 outside the housing may be necessary for heat dissipation purposes. 551097 - 15 - The high powered load 27, 58 is required such that heat is dissipated when accurate voltage and current measurements arc made across the load, for example by the solar load sensor 34. The resistive load 27, 58 may be, for example, a 9.97 ohm resistor with a 10 Watt rating. Solar 5 intensity is accurately determined by measuring the voltage drop across the accurately known load. The micro power supply 38 provides power to the microcontroller 39. The micro power supply 38 is, for example, a switch mode power supply. 10 Preferably the external power output 14 is connectable to any suitable load 59. The load 59 may be any type of clectrical device that requires a controlled source of power. For example, the load 59 may be telecommunications equipment. 15 Preferably the battery or batteries 2, 3 are a rechargeable type as is commonly used in communications equipment. For example, the battery may be a lead-acid battery. The batteries may also me an array of separate cells, for example, 4x 12V batteries to supply 48V for telecommunications equipment. 20 Battery switching electronics are required when there is more than one battery to be connected to the BM apparatus. The switching electronics facilitate the connection and disconnection of one or more batteries from the power input and from the load output of the apparatus. The switching electronics may comprise, for example, a relay or power transistor type device. Battery switch 37 as shown in Figure 1 is a semiconductor switching network that facilitates the required switching. 25 The microcontroller 39 is the heart of the BM apparatus. The microcontroller 39 is designed to connect to each of the various components in the apparatus and provide an overseeing control. The microcontroller 39 contains internal software, or firmware that is used to control the apparatus. The operation of the firmware and interconnection of the microcontroller 39 is 30 discussed in detail later in the specification. Preferably the electronic memory 44 is connected to the microcontroller 39. The memory 44 facilitates the temporary storage and backup of data that the microcontroller may sample over the course of operation. The electronic memory 44 can be any standard type of memory cell 44 35 designed for interconnection to a microprocessor. Further, the memory cell 44 may not always be 551097 - 16- required where the apparatus is connected to a central mainframe or control station, or is in constant communication with a master controller. Visual status indicators 50 may be provided on the exterior of the apparatus. In this way the 5 status of the inputs and outputs may be quickly visually identified. Appropriate driver circuits 49 may be provided between the status indicators and the microcontroller 39. For example, visual status indicators 50 may comprise LED's and the controller circuits may comprise a transistor or specific LED driver IC, such as an LM3914. 10 The real time battery back up clock 45 provides an independent record of time. This often proves useful in the event of power failure that causes the remainder of the control circuitry to shut down. The real time battery back up clock 45 may then be used to diagnose the time of a fault. Incoming and outgoing information sent through the communication interface 48 can also be tagged with time and date information. The coordination and timing of health testing is also set 15 by the real time clock 45. Digital and analogue inputs (40 and 41) and digital inputs and outputs (46 and 47) are provided to ease the connection of various external components to the microprocessor. The inputs and outputs are typically used to connect external sensors and displays to the apparatus. 20 The voltage and current sensing electronics are used to accurately measure ingoing and outgoing electrical characteristics that may be used, for example, to charge a battery. Battery charging systems are complex and therefore charging parameters need to be monitored. 25 The voltage sensing is performed by calibrated operational amplifiers feeding a scaled output voltage into an analogue to digital converter input on the microcontroller 39. Current sensing is performed by measuring the voltage drop across a known resistance and using ohms law to calculate current. Alternatively, the current can be calculated by a scaled current 30 output from a switching device applied across a known resistance. The temperature sensors 2, 3, 4 arc operatively connected together for use external to the apparatus. The operative connection also ensures that each temperature sensor is supplied with power at all times. The temperature sensors are designed to be placed relative to a battery to 35 accurately monitor the battery temperature during a charge or discharge cycle. The temperature 551097 - 17- sensor chain terminates in a single connector 15 of the housing 1. Figure 9 illustrates how a battery lifetime can be reduced when charging occurs outside a preferable temperature range. The communications interface 48 may facilitate wired or wireless communication that allows an 5 external device to connect and remotely communicate with the apparatus. The communications interface typically facilitates a direct connection to the microcontroller 39 to, for example, update internal code or receive data stored in the microprocessor memory or read status information such as faults, ok, or transmitting. Such status information may also be displayed on the visual status indicators 50. 10 Preferably, the communication interface 48 is one of RF, POTS, VPN, TCP/IP, GPRS, CDMA, RS-232, RS-485, or any data communication interface of a similar type. Preferably the communications interface 48, alone or together with an associated program in the 15 microcontroller 39, allows a remotely located user to control, store, delete and modify the status of the battery management microcontroller 39 and its peripherals. In addition, stored data can be downloaded from the memory cell 44 for re-processing, re-sale, and further analysis. The communications interface 48 may also allow access to data indicating the status of the 20 batteries, the solar panel 5 and the AC Mains 6. For example, the status information can be used to indicate hardware faults. The communications interface 48 may also allow a remotely located user to adjust the date and time of the real time clock 45. 25 The communications interface 48 may also allow a remotely located user to upgrade the apparatus firmware or reconfigure connections. Preferably the communications interface 48 uses the TCP-IP protocol the apparatus emulates a 30 web server to allow remote access by any web browser or web capable device. The user logs into the apparatus using a web browser to interrogate the battery management microcontroller 39. The apparatus also has a master/slave interface 42 to allow connection to other compatible controllers. The master/slave interface allows several of the devices to coordinate the testing and 35 circuit switching of batteries together and function and report as a group. The entire group of 551097 - 18- devices can also optionally communicate through one master communication interface, such as the communications interface 48 or a dedicated communications interface 42 for connection to other devices. 5 The microcontroller 39 includes a microprocessor and associated memory storing an executable program (or programs) as firmware. The microcontroller firmware is executed by the microprocessor to manage functions of the device, including internal upgrades, battery charging, battery management and batter}' health checks. 10 The firmware cycles though the following operations: a) Setup BM microcontroller peripheral features. b) Determine BM microcontroller start-up condition. This may be any one of: user reset, timer reset, brown-out, first time start, wake up from sleep and upgrade restart. Reload, reset or 15 initialise firmware variables based on the start-up condition. Output the start-up condition a stored output log, or send immediately over the communications interface 48. c) Check for remote command input from the communication interface. Apply remote commands immediately, store actions in output logs and acknowledge over the communication interface. 20 d) Check the firmware upgrade flag. If the upgrade flag is set, execute the boot loader firmware section. e) Take voltage, current and temperature readings using sensor circuits 28 to 36 and temperature sensors 2, 3, 4 and 24. (This step is described in more detail in appendix A). f) Calculate additional required parameters from the readings taken in step (e). (This step is 25 described in more detail in appendix B). g) Apply parameter rules and tests to determine status information. (This step is described in more detail in appendix C). h) Source switching. Based on status information or an overriding remote command, the batteries may be switched in or out of a circuit. For example the program may respond to status 30 violations, battery faults, charging faults, in-use faults, timer expiry, remote commands, component failures or rule violations (e.g. temperature, etc). (This step is described in more detail in appendix D). i) Apply any new charging and in-use modes of operation as a result of step (h). j) Report the new status information on the communication output interface. 551097 - 19 - k) Put the microcontroller into low power mode (sleep) until the node controller or some other means, (for example alarms, timers, rule violations, external interrupts) activates (wakes up) the microcontroller. m) Go to step b). 5 Microcontroller firmware - Boot loader operation The boot loader is a wholly independent small (code size) microcontroller firmware activity. The boot loader is not accessed during normal operation and only accessed when especially required. 10 The boot loader takes control of peripherals and reloads the main firmware over the existing code. This allows field upgrades of the firmware via remote control. The boot loader cycles though the following operations: 15 n) The boot loader indicates it has been started via the communication interface and requires a verification command byte from the communication interface. Failure to receive the correct code will result in an upgrade error condition and this will be reported back to the communication interface. The microcontroller will then reset the firmware upgrade flag and return to step (e). 20 o) The communication interface then receives a value (series of bytes) which is a count of the total number of bytes that will be sent for the upgrade. This data is echoed back to the communication interface for verification. p) The communication interface then receives a series of bytes called a record block. If the checksum is incorrect, or insufficient bytes are sent (within a timed period), an error is reported 25 back to the communication interface. The program then waits for the record block to be resent without error. q) The size of the record block is cumulatively added to a received bytes total, then the record block is disassembled and the code data segment is written to the correct memory location within the microcontroller. 30 r) If the total number of received bytes does not equal the total for the upgrade, the program returns to step (p). s) If the total number of received bytes matches the total for the upgrade, the program resets the firmware upgrade flag, returns to step (a) of the main firmware program. 35 Microcontroller firmware - Boot loader firmware upgrade process: 551097 -20- The upgrade process must exit correctly (resetting the firmware upgrade flag) once a firmware upgrade has started. 5 If a firmware upgrade attempt is unsuccessful, the firmware program will keep restarting the boot loader at step (d) until a successful upgrade is completed upon a microcontroller reset. The boot loader provides a mechanism for remote upgrade of the main firmware. In addition, the boot loader firmware itself can be upgraded by a similar process. The steps to upgrade the 10 firmware are as follows: 1. The original boot loader firmware is activated and downloads temporary boot loader firmware to replace the main firmware program. 15 2. The apparatus is reset and starts running the temporary boot loader firmware in the main program area. The temporary boot loader firmware stored by the mam program area downloads new boot loader firmware into the boot loader firmware section replacing the original boot loader version. The last action of temporary boot loader firmware stored in the main program area is to set the firmware upgrade flag. 20 3. The apparatus is reset and immediately starts the new boot loader version stored in the boot loader area which downloads the mam firmware again into the main program area replacing the temporary boot loader version. 25 Shown in Figure 1 are the battery charging system components of the apparatus used to facilitate a preferred battery charging method. Specifically, these components are the microcontroller 39, a power supply, charge regulator 25 and a battery switch 37. A first battery 7 and second battery 8 are shown as the batteries to be charged. 30 The power supply may be renewable (or intermittent) source such as the solar panel 5 or AC mains source 4 and AC rectifier 23. A device catering for a solar power source usefully incorporates charge pump 27 to change the voltage output to useful levels, and standard load 27 to calibrate the solar panel output. 551097 - 21 - The microcontroller 39 receives a range of inputs. The environmental temperature input 4 is connected to a pair of temperature sensors first battery temperature sensor 2 and second battery temperature sensor 3. Each batter}' temperature sensor 2, 3 is placed proximate a battery 7, 8 to be chargcd. In this way, the microcontroller 39 is continuously informed of the temperature of 5 each batter}'. It can be appreciated that more temperature sensors are connectable in the same configuration when there are more than two batteries to be charged. The temperature data read by the sensors is storable in the memory cell 44 together with time and date information provided by the real time clock 45. The stored data may then be used in 10 determining battery health or alter charging parameters. In addition, the communications interface 48 can be used to transmit the temperature data to a remote location for data analysis purposes. 15 During operation, the apparatus monitors the sensors for any operating parameters that fall outside normal operating conditions. For example, abnormal operating conditions may be indicated by: zero voltage reading on sensor 29 indicating an input power failure, low voltage reading on sensor 29 indicating a brown-out or excessively cold or hot measurements of the temperature sensors 2 and 3, indicating adverse operating temperatures of the batteries. The 20 apparatus produces an alert output signalling the type of problem and/or action taken should such an operating condition occur. Single Battery Systems 25 Systems that only have one battery will either have the battery on charge (OC) or in use (IU). This is more typical with the AC Mains power source option. The apparatus is configurable for use with a single battery. For this application to work reliably, the input power should ideally be the AC Mains. The battery can be disconnected from the load 30 as in the trickle charge system model shown in Figure 8. This allows isolated charging and testing to be done on the battery without affecting the load. The battery can also be disconnected from the load under specific circumstances such as over heating, dead shorts, testing or servicing. 551097 -22 - A single battery system allows the load to maintain a constant voltage regardless of battery condition or connection or reconnection to the charging circuit. I'his minimises circuit noise and power dissipation due to variation of voltages and currents within the load and charging circuits. 5 Complete discharge of the battery is also undesirable as this will lead to reduced battery lifespan as illustrated in Figure 14. Preferably the microcontroller 39 is programmed to disconnect a battery (in a controlled fashion using alarm outputs) before imminent battery failure, thereby saving the battery and ensuring that the load will be re-powered safely when the input power source is restored at a later date. 10 Multiple Battery Systems When using solar panels, it is typical to use at least two batteries to increase reliability and ensure a continuous output DC supply. This is due to low sunlight and at night time situations. For 15 example, a first battery can be used to supply power while a second battery is being charged. When the first battery becomes flat, the second battery is then used to supply power while the first battery is charged. In a remote location, the apparatus typically utilises a solar power source. The apparatus has to 20 provide output power to the load, have a backup power source and charge the batteries during light conditions. The apparatus is designed to provide a continuous power supply to a load using two batteries. The apparatus uses the solar panel to charge the batteries. Multiple batteries can also be used in mains power systems for added capacity and redundancy. 25 Continuous float charging can lead to extremely reduced battery life. Using a two battery system, one battery can be charged up to float mode, and when the second battery obtains float mode, the first battery can be used or discharged to allow the charging cycle to repeat from the start. 30 It is possible to extend the number of batteries in the switched system beyond two. To accommodate a more general control system with multiple batteries, the apparatus can be duplicated and split into operating modes. One apparatus is set up in a master mode and the remaining apparatus are set up in a slave mode. Figure 22 illustrates this arrangement. 551097 -23 - The master mode device la coordinates which batteries are connected to the load and which batteries are being charged or tested. Only one master mode device is required per battery gfoup. llie master device la is operatively connected to each of the slave mode devices lb, lc via the master slave interface 42. Using a 7-bit addressing scheme allows groups of 127 slave devices. 5 There is no theoretical limit to the quantity of slaved apparatuses that can be accommodated with either firmware modifications, or another tier of master mode control. The master device signals all the alert conditions via the communication interface 48. 10 Each devicc in slave mode or stand alone mode determines the condition of an individual battery. The device can switch the bat tery in and out of either the charging or the load circuit by controlling battery switch 37. This allows for isolated battery testing and true unloaded open circuit battery terminal voltage testing. 15 Determination of current battery condition can be reported back to a master mode device using the master slave interface 42. The device has visual status indicators so that fault conditions can be instantly recognised by sendee personnel. In the master slave arrangement there can be one device in slave mode per battery. The apparatus 20 in slave mode is therefore able to record specific details of each individual battery. For example, some specific details are: Battery identity, inception date, cycle counts, fault history, maximum and minimum parameters, temperature details, owner or user information, location, etc. This allows a remote user to interrogate specific batteries at any time for asset management purposes and metrics on say remaining lifespan. 25 Where the device has two batteries connected, the program preferably selects the least charged battery for charging. In systems where a solar panel provides the charging voltage this test is useful due to the large solar intensity variations during sunlight hours. The preferred apparatus makes improved use of whatever solar energy is available. 30 As solar intensity increases, so does charge current. As a consequcnce, the on charge (OC) battery terminal voltage also increases. According to the preferred program executed by the microcontroller, the charging is switched to the battery with the lower terminal voltage oncc the difference between the OC battery terminal voltage and the in use (IU) battery terminal voltage 551097 -24- exceeds a predetermined "max difference". The apparatus may also provide an output (to the communication interface 48) stating that a max battery difference condition has been triggered. It can be appreciated that the minimum requirements are: microcontroller, a charge switching 5 device, voltage sensing, and an error indicator. The microcontroller needs to be able to sense the battery voltage and to switch a charge voltage to and from the battery. The microcontroller can use a single indicator will indicate battery heath. An example of such an application would be for use in an emergency light system for buildings. 10 The preferred charging system of the present invention will now be described. Charging According to one aspect of the invention the microcontroller program implements a four step 15 charging method. The four step charging method is a cyclic charging system that uses intelligent (microprocessor) control. The four step charging method uses the ability to time, read voltage values, adjust the battery charger voltage, possibly read the charge current (if not limited elsewhere) and control the 20 switching of the battery to the charger and switching of the battery to the load. The microcontroller program implements the preferred charging method according to the following steps: 25 1. Max current: Set the battery charger voltage to 2.45V/cell (no more than the cycle charge voltage) with the current limited to 0.4 CA so as not to exceed manufacturer specifications. The charge voltage is varied so that the charge current does not exceed the current limit. 30 Max current mode finishes when the battery terminal voltage reaches some preset value between 2.35V/cel1 and 2.45V/cell, or a timer expires. The timer may be set at, for example, between 2 and 5 hours. 2. Pause: Switch off the battery charger. Read the battery terminal voltage. Wait until the 35 battery terminal voltage falls below a preset level, ideally below 2.25V/cell, or a timer expires. 551097 -25 - The timer may be set at, for example, between 5 seconds and 2 minutes. Optionally, the program may wait a further preset time to rest the battery before going to the next charging mode. 3. Set V: Set the battery charger to a new charge voltage level between 2.30V/cell and 5 2.40V/cell also limited to no more than 0.4 CA. Set voltage mode continues until the battery terminal voltage reaches 2.25V/cell (no more than the trickle charge voltage) or a timer expires. The timer may be set at, for example, between 2 and 8 hours. 10 4. Float: Set the battery charger to a new charge level of 2.25V/cell. Periodically release the battery from charging and test the battery terminal voltage. If the battery terminal voltage falls below 2.15V/cell, or charge current falls outside parameters, 15 or a timer expires, return to step 1 (or step 3 based on time in Float mode). The timer may be set at, for example, between 8 and 24 hours. The above stated current limits are empirically determined from design testing. A sudden current rise or fall would suggest some form of failure. The rate of change of voltage over time is 20 therefore observed for consistency. Eventually the battery will settle below the 2.15v/cell level. If the battery voltage setdes quickly, for example, in less than 24 hours, then step 1 is reinitiated as the battery is not holding its charge. If the battery voltage settles over a time greater than 24 hours, step 3 is required only to top up the charge of the battery. 25 The charge current parameters are based around the CA parameter specified by the manufacturer. For example a 1.2A/hr battery has a CA of 1.2Ahr. Therefore 0.4 CA is 0.4 x 1.2A/hr = 480mA. Charge currents range from 0.1CA to 0.4CA. Charging above 0.4CA can seriously stress the battery by over heating. Over heating may lead to internal mechanical failure in the battery cell structure. Trickle and compensation currents are between 0.15CA and 0.05CA. 30 The specified cell voltages are for VRLA oxygen recombination type batteries. The electrochemical nature of lead, sulphuric acid and oxygen is related to these voltages and has been verified by chemical proof. There are differences between cell voltages between different rechargeable battery types. Lead/Acid, NiMH and Li-Ion for example are 2.25v, 1.2v and 3.6v 35 respectively. 551097 - 26 - Each of the steps has been specifically determined to ensure maximum battery hfespan. It can be appreciated that one or more steps in the preferred charging process could be negated; however, a considerable reduction batter}' lifespan will result. These steps are highly heuristic and long term 5 statistical evaluation will lead to refinements. The addition of step 2 allows the battery to recover from the high current charging of step 1. The slightly higher voltages used in step 3 (below the initial charge voltage but above the float charge voltage) allow the battery cells to absorb as much charge as possible before going into 10 Float mode. Float mode voltages are very close to those of trickle charge voltages, so caution is taken with regard to how long float mode should occur, depending on the applied charging voltage and ambient and battery temperature. 15 According to another aspect of the invention the device may implement a five step charging method where multiple batteries are connected. Continuous float charging can lead to extremely reduced battery life. A fifth step in the charging 20 method is available when it is possible to switch the load to an alternate batter}7 or some form of external supply. The preferred five step charging method includes the steps of the preferred four step charging method, and one extra step. This step is applied after float charging for a preset period of time. 25 According to this step, the controller connects the batter}' to a load and then returns to step 1 when the battery terminal voltage falls below 1.90V/cell. The level of 1.90V/cell is for complete cell discharge, this may not be desirable in some circumstances, in which case a higher end voltage may be set. There may be other reasons why 30 the battery is to be switched back to step 1. Complete cell discharge is undesirable due to circumstances where a batter}' may already be exhibiting faults. In a remote location a battery may not be immediately replaceable. Complete battery discharge may over stress the battery a result in failure. The application may be left 35 vulnerable to power loss. 551097 -27 - The five step charging method is suited to systems with two or more batteries, as one of the batteries is periodically discharged. 5 Charge Cycle Period Each charging step in the above methods involves a series of short duration charging cycles. Each charging cycle includes a first period where the battery is connected to the charging supply and a second period where the battery is disconnected from charger. 10 The period of the charging cycle is long enough to apply a small charge current, but short enough to dctect possible problems (or alert conditions) in time before battery damage occurs. For example, the cycle period may be approximately 2 seconds (2.000s). 15 The program provides for a brief delay (for example 100ms) between isolating the battery and reading the open circuit battery terminal voltage. For the remaining part of the cycle the battery is on charge. The delay between isolation and voltage readings allow the battery voltage to settle slightly. 20 Without a delay, the open circuit battery terminal voltage reading is essentially the same as the applied charge voltage. It is the open circuit battery terminal voltage which gives an indication of the battery? chargc capacity (Figure 13), not the battery terminal voltage when connected to a load. Even with the delay, there can still be an oscillation effect on the open circuit battery terminal 25 voltage (Figure 16). The microcontroller program implements a set of charging override rules. The preferred program implements a rule that detects batteries with high internal impedance. The rule is triggered where a voltage rise within a preferred time period is greater than a predetermined value. 30 During the application of a battery charge the voltage may rise too quickly, this triggers a "max delta voltage" alert. Fast voltage rise times are typical of aging batteries due to increased internal impedance allowing a greater voltage variation at the on charge battery terminals. 551097 -28- The program records the changing voltage (delta V) during charging. The program determines parameters relating to battery lifespan from the recorded voltage data and open circuit battery voltages. For example, the program can determine internal impedance, charge capacity and potential cell damage. 5 Optionally, a "max delta voltage" alert happens during step 1 (Max current) of the charging process, this would suggest that the charge current is too high, the program changes to step 2 (Pause) of the charging method. This allows the battery voltage to stabilise and the microcontroller program then attempts to charge the battery at a reduced current and a lower 10 charge voltage (Step 3 - Set V). Optionally, when "max delta voltage alert" occurs during step 3 (Set V), this would suggest that there are serious problems with the battery condition (high internal impedance), the program changes to step 4 (float) of the charging method. 15 The preferred device includes additional novel battery health monitoring functions. The implemented charging method includes a short duration periodic charging cycle in which the controller tests the battery terminal voltage shortly after the charge voltage is removed. The 20 inventor has observed that over a number of cycles this produces a voltage oscillation at the battety terminals that can vary from a few millivolts to the total difference between the normal batter}' terminal voltage and the applied charge voltage. The open circuit battery terminal voltage is used to determine current battery charge status. As 25 the charging system is cyclic by design, the batter}' is being periodically disconnected from the charger every 2 seconds to obtain the open circuit battery terminal voltage. The disconnection and reconnection of the charge voltage causes the open circuit voltage of the battery to oscillate (Figure 16) with a battery in poor condition. 30 The amplitude of the oscillation is related to the charge current and the battery impedance. The period of oscillation is only dependent on the charge cycle period. A variety of cycling times between 0.5 and 10 seconds were tested and very similar magnitudes of terminal voltage oscillation occurred in each case (at the same charge current and voltage). 551097 -29 - In addition, while testing the prototype device on a selection of aged batteries, the oscillation effect was least noticeable on the newest batteries. New batteries had few, if any, oscillation peaks (Figure 15). As the batteries became older, peaks started to appear in the open circuit battery terminal voltage graphs at various points. The first peak would occur when the battery is 5 first connected to the charger. As the battery aged, additional peaks occurred in the graph (typically equally spaced period wise), with gradually increasing amplitude. Eventually with the oldest of batteries, the oscilladon effect was almost a sinusoidal wave as is shown in Figure 16, with a full peak occurring each charge cycle. The peak to peak magnitude of 10 the voltage swing is only about IV with an aging 12V battery. In a good battery, the application of a charging voltage (i.e. higher than the open circuit battery-terminal voltage) to a new battery (low internal impedance) produces a charging current and an increase in the battery terminal voltage by a small amount. Each time the charger is disconnected 15 and the battery terminal voltage is tested the battery terminal voltage quickly settles down to a value close to the original open circuit battery terminal voltage. The output graph of the battery terminal voltage is almost flat. Figure 15 illustrates this behaviour in a young battery in good condition. A slight positive incline is apparent due to charging 20 In figure 15, thick line 151 is the battery terminal voltage. Thin line 152 is the charging voltage. Each graduation on the time scale represents 2 seconds, and the total illustrated duration is 50 seconds, or 25 charging cycles. The battery cell structure starts to fail as the battery starts to age. The failing cell structure leads 25 to increased internal impedance and consequently larger voltage swings between the open circuit battery terminal voltage and the battery terminal voltage when the charger is applied. Tracking the open circuit terminal voltage after a charge cycle shows a delay in the effect of this voltage. Figure 16 illustrates the terminal voltage over time of an older battery in poor condition. Again, the thick line 153 is the battery terminal voltage. Thin line 154 is the charging voltage. This 30 suggests that the battery internal impedance is in fact a complex impedance of the form R + jX The impedance of a battery is of the form R + jX, where R represents the real battery resistance and jX represents the complex frequency based impedance. The batter)' impedance creates a delay (or lag) in the battery terminal voltage when the charge voltage is both applied and also 35 released. This method allows the impedance to be measured. Variation in the impedance over 551097 -30- time can be used to determine the internal state of the battery cell structure and an indication of its "health". Impedance comparisons between batteries that are deemed to be end-of-life allow some quantitative measurement of the remaining lifespan of the battery. 5 According to one aspect of the present invention, battery health is determined without determining the battery impedance. Instead, battery health is determined by observing the effects of the applied charging voltage on the battery terminal voltage. Batteries with large oscillating voltage response characteristics (such as illustrated in figure 16) tend to be in very poor condition and their "health" is consequently reported as "bad". 10 The preferred embodiment of the barter}' health determination method and program will now be described. Battery health monitor apparatus 15 Figure 17 illustrates a simplified overview of the components of the battery health monitor apparatus. A batter}' 100 is connected to a battery health monitor apparatus 103 via a switch 104. The apparatus 103 contains components that perform particular functions used to determine battery health. In particular, these functions are the measurement of the open circuit battery 20 voltage by a voltage monitor device 101, and the application of a charging voltage to the battery by charging device 102. Preferably the method is implemented in the device described with reference to figure 1, and the components in apparatus 103 are a subset of the components of the main battery management 25 apparatus. The measurement of the battery voltage is performed by the voltage monitor device 101, which is preferably a high impedance voltmeter. In this way, the battery voltage measured is an accurate measurement of the open circuit battery voltage. The voltmeter may be an analogue to digital 30 converter on a microprocessor, or another industry standard voltage measuring device. The application of a charging voltage is facilitated by the charging device 102. The charging device 102 can be any device that will supply an appropriate range of DC voltage and currcnt for charging a battery under test. For example, a 12 V lead-acid battery may require a charging voltage 35 of 12 to 15V DC, while the appropriate charging currcnt may be from 100mA to 40A. 551097 - 31 - Preferably the charging device 102 is the charge regulator 25 as shown in Figure 1. In addition, the charging cycle of a lead acid battery can be sophisticated. Therefore the charge 5 voltage and current supplied by the charging device 102 to the battery 100 may be complex and time varying, under control of processor 105. The power supplied by the charging device 102 to the battery 100 is preferably derived from an AC mains supply 106. Standard power supply electronics can be used to perform the AC to DC 10 conversion and filtration. Alternatively, the power supplied by the charging device 102 may be derived from any other suitable power source, for example, a solar panel. A switch 104 is provided to connect the battery to the charging device 102 or isolate the battery from the battery health monitor system 103. The switch 104 is preferably an actively switched 15 device such as a relay or a semiconductor switch such as a BTS660 or a BTS555. Preferably the switch 104 is the battery switch 37 as shown in Figure 1. A microprocessor is ideally suited to control the operation of the various circuit components. In addition, a microprocessor is ideally suited to process the data gathered by the circuit 20 components. Preferably processor 105 is adapted to control circuit operation and sample circuit data. The processor 105 is connected to the charging device 102, die voltage measurement device 101 and the switch 104. Preferably the processor 105 is a microprocessor or similar device. Preferably the processor 105 is the microcontroller 39 shown in Figure 1. 25 Figures 18 and 19 illustrate an overview of the preferred method used to monitor health of a battery connected to the health monitor system 103. The microprocessor 105 processes a series of steps that facilitate the determination of battery health. Through the process illustrated in Figure 18 the microprocessor 105 performs a rapid charging 30 cycle. The rapid cycle includes a charge voltage application portion and an open circuit terminal voltage measuring portion. The cycling between portions is of a sufficiently high frequency such that it does not interfere with the overall process. The battery open terminal voltage is monitored even when the charge is 551097 -32- normally being applied. Open terminal voltage can be continuously monitored when normally not being applied. For example, while step 123 of Figure 19 is satisfied. At step 110, the method initiates and the processor 105 reads the open circuit battery voltage 5 from the voltage monitor device 101. The program proceeds to step 114 if the open circuit battery voltage is less than the high voltage limit 111. At step 114 the processor 105 connects the charging device 102 to the battery 100 using switching the switch 104. The charging device 102 applies a charging voltage 112 to the battery 10 100. The charging voltage 112 is higher than the open circuit battery voltage and the high voltage limit 111. The charging voltage is supplied to the battery for a particular length of time. This time may be predetermined or variable. In the preferred form, the time is set by input variable 113, and will be 15 referenced at time A. Preferably time A is at least the time taken for the open circuit battery voltage to respond to the applied charging voltage. That is, a batter)' will respond to an applied voltage only after the voltage has been applied for a finite amount of time. The amount of time is typically 5 to 20ms, depending on the health of the particular battery. 20 At step 115, a cyclic check for the elapse of time A and the battery voltage limit is performed, or if other problems are detected. The charging voltage is disconnected upon the elapse of time A., or if the connected battery terminal voltage exceeds a high voltage limit 111, or if other problems are detected. An example of another problem includes excessive battery temperature. 25 The high voltage limit 111 represents a desired open circuit battery voltage that is somewhere between the nominal output voltages expected from a battery, and the applied charging voltage. For example, if the nominal battery output voltage is 12V and the applied charging voltage is 14V, a suitable high voltage limit is approximately 13V. 30 A 12V battery that is supplied with a 14V potential will rise in voltage until an equilibrium charging voltage is reached. Similarly, a battery charged by a greater voltage potential and having a resting or nominal open circuit voltage of 12V will fall in voltage if the greater charging potential is disconnected. 551097 -33- At step 116, the processor 105 disconnects the charging voltage from the battery 100 via switch 104 and the program operation is paused at step 118 for a second time period. The second time period time may be predetermined or variable. In the preferred form, the time is 5 set by input variable 117, and will be referenced at time B. The pause is to allow the open circuit battery voltage to settle. A natural phenomenon is the fluctuation of the open circuit battery voltage immediately following disconnection of a higher voltage potential. A voltage measurement taken immediately following disconnection of a charging voltage may not represent the accurate open circuit battery voltage. 10 Time B is determined from knowledge of the microprocessor speed and knowledge of the time taken for fluctuations in the open circuit battery voltage following disconnection of the higher voltage potential to subside. Time B is typically in the range of 50 to 200ms. The program operation proceeds to step 119 upon the elapse of time B. 15 At step 119, the processor 105 reads the open circuit battery voltage from the voltage sensor device 101. The open circuit voltage of the battery under test is then observed. Preferably, observation of the battery voltage includes monitoring and recording the open circuit battery voltage for a predefined length of time. For example, the open circuit battery voltage may be 20 sampled at 10ms intervals. Preferably the discrete voltage measurements are stored in a buffer. The buffer can be a memory cell in microprocessor 105. Alternatively, the buffer may be facilitated by a local or remote computer system. The stored data is later used to determine the health of a battery under test. 25 The program then proceeds to step 119 where the open circuit battery voltage is measured. Figure 19 illustrates the program operation of step 119 in further detail. The program operation of step 119 in Figure 18 initiates at step 120 in Figure 19. The open 30 circuit voltage of the battery under test is measured and stored in the buffer. A peak finding algorithm is applied to the voltage samples stored in the buffer at subsequent program step 123. The peak finding algorithm is to isolate the occurrence of any maxima in the stored voltage measurements, and when they occurred. 35 551097 -34- A threshold voltage variable 122 is input to the program step 123. To ensure false peak voltages are not recorded, the measured voltage maxima are compared to the threshold voltage variable 122. The voltage maxima should be above the threshold voltage 122 to be labelled a voltage peak. 5 The threshold voltage variable 122 represents the maximum peak-to-peak open circuit battery voltage ripple that can occur in a charge cycle before a peak is recorded. For example, in a 12volt battery system, a suitable threshold voltage would be approximately 300mV. The threshold voltage can be determined several ways. For example, it could be a certain voltage above previous voltage minima or calculated from the history of sampled data. 10 If a peak in the buffered voltage samples is detected, the peak is recorded at step 125 and the program resumes operation at step 120. The quantity7 of peaks recorded and the test time period over which they occurred is then used to determine the health of the battery under test 100. 15 The program operation shown in Figure 19 is subsequently ended at step 126 and the program operation shown in Figure 18 is resumed. The steps shown in Figure 18 are then looped and the open terminal battery voltage is cyclically sampled. The program operation shown in Figure 19 is terminated at step 121 if no voltage maxima occur. 20 The program operation shown in Figure 19 is terminated at step 124 if the measured voltage maxima are less than the threshold voltage 122. The program operation shown in Figure 19 is terminated at step 126 when a voltage peak is recorded. The program operation of Figure 18 is subsequently resumed at step 110. 25 Figure 20 illustrates an example plot of the open circuit battery voltage verses time when the program operation of the present embodiment is applied. The open circuit batter}7 voltage will begin to climb from a nominal resting voltage as the charging voltage is applied at step 114. The nominal resting voltage is approximately 12.7V for a 30 standard charged lead-acid batter}7, however, the nominal or resting voltage may be anywhere between a charged and discharged voltage for any given battery. The charging voltage 112 is applied until the open circuit battery voltage equals or becomes greater than the desired high voltage limit 111. In Figure 20, the high voltage limit 111 is reached 35 at time point 127. The charging voltage 112 is then disconnected in accordance with step 115. 551097 -35- The open circuit battery voltage is then monitored in accordance with step 119 for the occurrence of a voltage peak. At time point 128 a first peak in the open circuit voltage is detected and recorded. The open circuit battery voltage will then begin to fall naturally. At time point 129 the open circuit battery voltage drops below the high voltage limit 111 and the charging voltage 5 112 is subsequently reapplied. The open circuit battery voltage will continue to fall before beginning to rise again. At time point 130 the high voltage limit 111 is reached again and the charging voltage 112 is disconnected. A second voltage peak is detected and recorded at time point 131. A third, fourth and fifth voltage peak is also recorded at points 132, 133 and 134 respectively. 10 In addition, the high voltage limit 111 may comprise one or more high voltage limit voltages. For example, a positive going open circuit batter)' voltage may be subject to a first limit, and a negative going open circuit battery voltage is subject to a second limit. 15 The program operation is cycled a fixed number of times, for example, 150 cycles at regular time intervals. Regular time intervals may include a daily cycle. The number of peaks detected is recorded. The preferred program includes a termination step 135 in the loop of figure 18 to check whether the required number of tests have been performed. 20 The health of the battery under test is determined from the number of peaks recorded during the testing period. It follows that the battery health can be determined by recording the number of voltage peaks that occur during a constant test period. The inventor has determined that an unhealthy battery will show a greater number of peaks occur 25 during a given time span compared to the number of voltage peaks shown by a healthy batten'. Figure 21 illustrates a graph of recorded experimental data showing relative battery health versus the number of voltage peaks observed during a given time frame. This represents data for a particular battery make and model. Testing of additional lead acid batteries indicates that they 30 have similar response profiles but that the actual threshold frequencies for good and poor batteries are different. A healthy battery may exhibit no, a very low, or a slow open circuit fluctuation in response to the charging voltage. In contrast, a battery in poor health will have the tendency to exhibit a wide 551097 - 36 - fluctuation in open circuit voltages, or a fast voltage response. A battery in poor health will also show a relatively large number of voltage peaks. Other methods may also be implemented by processor 105 to determine battery health from the 5 plot of battery voltage verses time obtained such as shown in Figure 4. A Fourier transform of the voltage plot would produce a frequency spectrum of the fluctuating open circuit battery voltage. A spectrum that showed high frequency components would indicate an unhealthy battery, while a spectrum that showed low frequency components would indicate a healthy battery. 10 When testing a battery using the apparatus, comparison of voltage response information over time and between same model batteries allows a state of health to be determined. Voltage peaks above the threshold voltage 122 within the Voltage Response waveform are counted and increased counts confirm deteriorating health. Once a preset count is reached, an alarm can be 15 triggered. Count value ranges are determined from the charging parameters and more importantly the time a battery spends in sendee. To compare peak counts for battery condition requires that the battery testing parameters be consistent. 20 551097 -37- Appendix A: Expansion of Step (e) of the Firmware Section Take voltage, current and temperature readings. An expansion is as follows: e.l) In Use (IU) Backup Battery Readings Read the IU connected battery terminal voltage, output current and temperature. e.2) Input Power (IP) Source Readings Read the Solar Panel and AC Mains voltage and current. Connect an alternative power source to the main output (only necessary if Solar only). Read the solar panel voltage. Connect the solar panel to a known fixed load and read the voltage level. If necessary, reconnect the main output power to solar panel so daytime running is entirely from the solar panel. e.3) On Charge (OC) Battery Readings With the battery charger still connected to the OC battery. Read the OC connected battery terminal voltage, current and temperature. Disconnect the battery charger from the OC battery. Wait a small delay. (Delay can be adjusted for optimal performance.) Read the OC open circuit (disconnected) battery terminal voltage. e.4) Battery Charger (BC) Readings Read the open circuit BC output voltage (disconnected from OC battery). Reconnect the battery charger to the OC battery. Read the BC connected output voltage, current and ambient temperature. e.5) Main Output (MO) Readings Read the MO voltage, current and ambient temperature. Appendix B: Expansion of Step (f) of the Firmware Section Calculate additional required parameters from the readings taken in step (e). An expansion is as follows: f.l) In Use (IU) / Backup Battery Calculations Using IU connected battery terminal voltage and current readings calculate the power output (P = VI) of the IU battery in watts. Using previous cycle IU connected battery terminal voltage readings, determine delta V for this IU cycle. Using the IU connected battery terminal voltage reading, determine total delta V since the IU battery was first connected. 551097 -38- Using previous cycle IU connected battery7 current readings, determine delta I for this IU cycle. Using the IU connected battery current reading, determine total delta I since the IU battery was first connected. 5 Determine the IU maximum and minimum voltage, delta voltage, current and delta current readings. f.2) Input Power (IP) Source Calculations Using input power voltage and current readings calculatc input power draw in watts. If using a Solar Panel, determine the maximum solar panel output current (I = V/R). 10 Determine the solar luminous intensity and maximum available power which is proportional to the maximum Solar Panel output current. Determine the maximum and minimum voltage and current readings. f.3) On Charge (OC) Battery Calculations Using OC connccted battery terminal voltage and current readings calculate the power 15 input to the OC battery in watts. Using previous cycle OC connected batter}' terminal voltage readings, determine delta V for this OC cycle. Using the OC connected batter}' terminal voltage reading, determine total delta V since the OC battery was first connected. 20 Using previous cycle OC connected battery current readings, determine delta I for this OC cycle. Using the OC connected batter}' current reading, determine total delta I since OC battery was first connected. Using the OC open circuit battery terminal voltage, determine the OC battery charge 25 capacity. Determine the OC maximum and minimum voltage, delta voltage, current and delta current readings. £4) Battery Charger (BC) Calculations Using the BC disconnected voltage, determine the error voltage between the desired 30 charger voltage and the real output from the charger. Add the Error voltage value to an Error Sum total. Error Sum is the integral of the Errors and as such needs to be limited to avoid integrator wind up. Set the new BC voltage complete with the addition of error offsets. Determine the new error voltage. 551097 -39- Using BC connected voltage and current readings calculate the power output of the charger in watts. £5) Mam Output (MO) Calculations Using MO voltage and current readings calculate the power output from the RBM apparatus in watts. Determine the MO maximum and minimum voltage, current and temperature readings. Appendix C: Expansion of Step (g) of the Firmware Section Apply parameter rules and tests to determine BM status information. An expansion is as follows: g.l) In Use (IU) / Backup Battery Tests Determine IU battery status: If the IU batten^ terminal voltage is less than the MO voltage and similar to the IP connected voltage, main output supply is being provided by the input power source and the IU batteiT is being used for "Backup". Increment the "Backup" cycle counter. If the MO voltage is very closc to the IU batter}' terminal voltage, and the IP connected voltage is less than the IU battery terminal voltage, main output supply is being provided by the IU batter}' and the IU batter}' is "In Use". Increment the "In Use" cycle counter. Test to see if the IU battery is flat, dead or does not exist. g.2) Input Power (IP) Source Tests Determine from the IP open circuit voltage if input power exists and is providing a power source. Categorise the IP voltage: Very High, High, Normal, Low, Very Low, Dim, Very Dim and Dark. Categorise Solar Intensity: Very High, High, Normal, Low, Very Low, Dim, Very Dim and Dark. Categorise the day period: Dark, Dawn, Sunrise, Day, Sunset or Dusk. Determine if there is sufficient input power to supply the main output power exclusively from the input power means. g.3) On Charge (OC) Battery Tests Check there is a power source for the charger. If power is available, increment the "On Charge" cycle counter. If there is no power source increment the "Not Charging" cycle counter. Check that the IP voltage is sufficient for OC battery charging. 551097 -40- Check OC battery temperature, if it is too high, determine if the next charge mode is necessary. Excessive temperatures set a warning. Check OC battery open circuit terminal voltage, determine the battery charge capacity. If the OC battery has reached the target voltage and/or current, set the next charge 5 mode. If the OC timer has expired, set the next charge mode. g.4) Battery Charger (BC) Tests Check if the connected OC battery current (I) is too high for the current charging mode. If too high, adjust BC voltage until current is at or below the maximum allowed for the charge 10 mode. Check the BC ambient temperature, adjust voltages if necessary. Shut down if too hot. Check the number of charge cycles, determine if the next charge mode should be applied. g.5) Main Output (MO) Tests 15 Check the MO voltage to see that output supply is being provided. Appendix D: Expansion of Step (h) of the Firmware Section Source switching, an expansion is as follows: h.l) Remote Command: Based on any remote command input, switch batteries, input 20 sources and output sources as specified. h.2) In Use Failure: If the IU battery is flat, dead or does not exist, switch to a new IU battery. h.3) On Charge Full: If the OC battery has been fully charged, switch to a new OC battery for charging. 25 h.4) Charge Cycle Expiry: If the OC cycle timer expires, switch to a new OC battery for charging. h.5) Max (V) Difference: If the voltage difference between the OC and IU battery is too large, switch to the low voltage battery for charging. h.6) Max delta V: If the OC battery voltage increases too quickly, switch to the next charge 30 mode (or temporarily stop charging). h.7) Ringing: If the OC battery voltage peak to peak oscillation is too high, send an alert to the communication interface. Either the OC battery current and/or voltage is too high or most likely the OC batter)7 is starting to fail internally due to aging. Several levels of ringing alert are detected starting with low values and gradually increasing to the maximum possible difference 35 between the normal battery terminal voltage and the applied charging voltage. Critical levels are 551097 -41 - dependant on battery type, model and age and are set from known battery failure parameters (determined from testing). 551097 -42- </div>

Claims

1. A method of charging lead acid batteries comprising, in order, the steps of: a) applying a first applied voltage to a battery until the open terminal voltage of 5 said battery exceeds a first target voltage, and then disconnecting said first applied voltage, b) applying no charging voltage to the battery for a time exceeding 30 seconds, and monitoring the open terminal voltage of said battery until the voltage falls below a second target voltage, 10 c) connecting a second applied voltage lower than said first applied voltage to said battery until said battery reaches a second target voltage lower than said first target voltage, d) connecting a third applied voltage lower than said second applied voltage to said battery for an extended period of time. 15

2. A method of charging lead acid batteries as claimed in claim 1, wherein if a charging voltage is applied to said battery, said charging voltage is intermittentiy briefly disconnected until said battery voltage is measured and reconnected when said batter}' voltage has been measured. 20

3. A method of charging lead acid batteries as claimed in claim 1 or claim 2, wherein step (b) includes measuring the open terminal battery voltage while no charging voltage is applied.

4. A method of charging lead acid batteries as claimed in any one of claims 1 to 3, wherein step (a) is reinitiated if said battery voltage falls below a fourth target voltage during step (d). 25

5. A method of charging lead acid batteries as claimed in any one of claims 1 to 3, wherein step (c) is reinitiated if said batteiv voltage falls below a fourth target voltage during step (d) and the time spent at step (d) has exceeded 24 hours. 30

6. A method of charging lead acid batteries as claimed in either claim 4 or claim 5, wherein said fourth target voltage is about 2.15V/cell.

7. A method of charging lead acid batteries as claimed in any one of claims 1 to 6, wherein said first applied voltage is a charge cycle voltage. 1 35 - 4 AUS 2009 1 551097 -43-

8. A method of charging lead acid batteries as claimed in any one of claims 1 to 7, wherein said charge cycle voltage is about 2.45V/cell.

9. A method of charging lead acid batteries as claimed in any one of claims 1 to 8, wherein a first current does not exceed the manufacturers recommended current limit and said first current is dependent upon said applied voltages.

10. A method of charging lead acid batteries as claimed in claim 9, wherein said first current is 0.4CA. 10

11. A method of charging lead acid batteries as claimed in any one of claims 1 to 10, wherein said first target voltage is between 2.35V/cell and 2.45V/cell.

12. A method of charging lead acid batteries as claimed in any one of claims 1 to 11, 15 wherein said second target voltage is about 2.25V/cell.

13. A method of charging lead acid batteries as claimed in any one of claims 1 to 12, wherein said second applied voltage is between 2.30V/cell and 2.40V/cell. 20

14. A method of charging lead acid batteries as claimed in any one of claims 1 to 13, wherein said third applied target voltage does not exceed a trickle charging voltage.

15. A method of charging lead acid batteries as claimed in any one of claims 1 to 14, wherein said third applied target voltage is not more than 2.25V/cell. 25

16. A method of charging lead acid batteries as claimed in any one of claims 1 to 15, wherein said third applied voltage is about 2.25V/cell.

17. A method of charging lead acid batteries as claimed in any one of claims 1 to 16, 30 wherein a load is connected to said battery, after step (d), to facilitate the discharge of said battery.

18. A method of charging lead acid batteries as claimed in claim 17, wherein said battery is connected to said load until the battery voltage falls below a fifth target voltage. 35 551097 -44-

19. A method of charging lead acid batteries as claimed claim 18, wherein said fifth target voltage is about 1.90V/cell.

20. A method of charging lead acid batteries as claimed in either claim 18 or claim 19, 5 wherein step (a) is reinitialised when said fifth target voltage is met.

21. A battery management apparatus comprising: a micro-controller, means for sensing an open terminal voltage of a battery and providing data 10 corresponding to the open terminal voltage as an input to the micro-controller, a charge regulator for receiving power from a power supply and applying a regulated voltage to a connected battery in accordance with and output of the micro-controller; and means for sensing charging current and providing data corresponding to the charging current as input to the micro-controller; 15 wherein said micro-controller is programmed to perform a method of charging lead acid batteries comprising the steps of: a) applying a first applied voltage to a battery until the open terminal voltage of said battery exceeds a first target voltage, and then disconnecting said first applied voltage, 20 b) applying no charging voltage to the batter}7 for a time exceeding 30 seconds, and monitoring the open terminal voltage of said batter)7 until the voltage falls below a second target voltage, c) connecting a second applied voltage lower than said first applied voltage to said battery until said battery reaches a second target voltage lower than said first 25 target voltage, d) connecting a third applied voltage lower than said second applied voltage to said battery for an extended period of time.

22. An apparatus as claimed in claim 21, wherein if a charging voltage is applied to said 30 battery, said charging voltage is intermittently briefly disconnected until said battery voltage is measured and reconnected when said battery voltage has been measured.

23. An apparatus as claimed in claim 21 or claim 22, wherein step (b) includes measuring the open terminal battery voltage while no charging voltage is applied. 35 /W V ^ - k AUG 2009 I - / 551097 -45-

24. An apparatus as claimed in any one of claims 21 to 23, wherein step (a) is reinitiated if said battery voltage falls below a fourth target voltage during step (d).

25. An apparatus as claimed in any one of claims 21 to 24, wherein step (c) is reinitiated if 5 said battery voltage falls below a fourth target voltage during step (d) and the time spent at step (d) has exceeded 24 hours.

26. An apparatus as claimed in either claim 24 or claim 25, wherein said fourth target voltage is about 2.15V/cell. 10

27. An apparatus as alcimed in any one of claims 21 to 26, wherein said first applied voltage is a charge cycle voltage.

28. An apparatus as claimed in any one of claims 21 to 27, wherein said charge cycle voltage 15 is about 2.45V/cell. 20

29. An apparatus as claimed in any one of claims 21 to 28, wherein a first current does not exceed the manufacturers recommended current limit and said first current is dependent upon said applied voltages.

30. An apparatus as claimed in claim 29, wherein said first current is 0.4CA.

31. An apparatus as claimed in any one of claims 21 to 30, wherein said first target voltage is between 2.35V/cell and 2.45V/cell.

32. An apparatus as claimed in any one of claims 21 to 31, wherein said second target voltage is about 2.25V/ cell.

33. An apparatus as claimed in any one of claims 21 to 32, wherein said second applied 30 voltage is between 2.30V/cell and 2.40V/cell.

34. An apparatus as claimed in any one of claims 21 to 33, wherein said third applied target voltage does not exceed a triclde charging voltage. 25 4 AU6 2009 551097 -46-

35. An apparatus as claimed in any one of claims 21 to 34, wherein said third applied target voltage is not more than 2.25V/cell.

36. An apparatus as claimed in any one of claims 21 to 35, wherein said third applied 5 voltage is about 2.25V/cell.

37. An apparatus as claimed in any one of claims 21 to 36, including a load and switching means operable by an output of the micro-controller to connect the load to the battery, and wherein the micro-controller is programmed to connect the ioad to the battery after step (d) to 10 facilitate the discharge of the battery.

38. An apparatus as claimed claim 37, wherein said battery is connected to said load until the battery voltage falls below a fifth target voltage. 15

39. An apparatus as claimed in claim 38, wherein said fifth target voltage is about 1.90V/celL

40. An apparatus as claimed in either claim 38 or claim 39, wherein step (a) is reinitialised when said fifth target voltage is met. O O 20 25