WO2000074200A1

WO2000074200A1 - Battery charging and discharging system

Info

Publication number: WO2000074200A1
Application number: PCT/GB2000/002038
Authority: WO
Inventors: Alan Henry Weinberg
Original assignee: Alan Henry Weinberg
Priority date: 1999-05-27
Filing date: 2000-05-26
Publication date: 2000-12-07
Also published as: GB9912462D0

Abstract

A battery charge and discharge regulator includes a battery charge section Ch for providing a connection between a current source (200) and a rechargeable battery (203) in order to recharge the battery; and a power supply section M for connecting the current source (200) to a load (203) via a main bus (302) in such a way that the current source is constrained to supply current substantially at the voltage of the main bus. Moreover the battery is also charged without tying down the voltage of the charging sections. This is achieved by incorporating in the charging section a voltage converter (206) capable of allowing the current source to charge the battery at a voltage independent of that of the battery. A current controller (298) is also provided to ensure that the current source when supplying current always supplies some of this current to the main bus via a diode (205), thus maintaining the current source at the main bus voltage, bar the diode drop.

Description

BATTERY CHARGING AND DISCHARGING SYSTEM

This invention concerns the interface of a rechargeable battery to a current source such as a solar cell array (SA) . The invention particularly envisages the use of a Sequential Switching Shunt Regulator (S³R) as described by US 4186336 (Weinberg) (Reference 1) and the applicant's earlier application WO 99/28801 (Reference 2) but it can be used to improve the design of any application that uses a SA and rechargeable battery.

Such circuits are used, for instance, m providing electrical power to apparatus which is dependent on solar power but which also includes a battery to store energy. This stored energy provides power when the solar power is not available or during a peak power demand, when the solar-generated power alone is not sufficient. A prime example is satellites which of course require power all the time, even when in shadow (eclipse) . The switching between the two sources necessitates additional components, resulting m an additional mass for the system; moreover it entails inherent inefficiencies m the charging and discharging process, which m turn means that the battery ana SA have to be larger than they would have been without: these losses. The background of systems of the type in question will now be described. A typical SA characteristic is shown m Fig. 1 and is αependent upon temperature, sunlight and the age of the SA. From point A, the short-circuit point (V_SA=0 and

I_SA=I_SCι to a point D, the current is approximately constant as the voltage increases. From D to F both current and voltage vary. From to F to G, the voltage open-circuit point (V_SA=V_0C and I_SA=0) , the voltage increases slightly with decreasing current (about 10% depending on the SA cells used) . At some point E m the transition region the maximum power of the SA is produced. This point is known as the Maximum Power Point (MPP) of the SA characteristic. The voltage at this point will be referred to as V_MPP and the current as I_MPP •

One prior-art method used to charge the battery from the SA is to connect the SA directly across the battery as in Fig. 2. A single dedicated SA section 400 charges the battery 203 as required. Here the open-circuit voltage of the SA at point G (V_oc) must always be greater than that of the battery to maintain battery charge. Since the battery and SA are highly variable in voltage when in use the battery voltage must be designed to be well below that of the SA open- circuit voltage V_oc. This normally means operating on the constant-current part A to D of the SA characteristic. This is shown by a typical line 31 on Fig. 1. Working on this part of the SA characteristic also has the advantage that, with the current constant, it is easier to control the state of charge of the battery.

The disadvantage of this simple method of battery charge is that the voltage of the SA is always fixed (clamped) to that of the battery, as shown by line 31 of Fig. 1 which might be considerably less than the open-circuit voltage V_oc. This wastes the potential power (power=I_SA*V_SA) that the SA can supply since it could work at a higher voltage and supply the same current. The maximum power (V_MPP*I_MPP) that the SA can supply is somewhere between the points D and F on the SA characteristic (point E) . The power wasted will be approximately the ratio of the voltage of the battery to the voltage at the maximum power point V_κpp (since the currents at V_B and V_MPP are approximately the same) . For example, if V_B is one half of V_MPP then approximately half the maximum potential power of the SA is wasted .

Fig. 3 shows a more complex single SA section used to charge the battery as before but now with a series switch 207 between the SA and the battery and a diode 205 between the SA and the bus. This series switch 207 is used to switch the battery charging current on and off . The diode provides a path for the SA current not used for battery charge to supply the main bus which has the load connected to it . The path from the battery to the bus is not shown. When the switch is ON the SA voltage is clamped, as previously, to that of the battery (line 31 in Fig. 1) and the diode 205 is reverse-biased. When the switch 207 is OFF the SA voltage rises until the diode 205 conducts, clamping the SA voltage to that of the main bus which is assumed to be at a constant higher voltage (regulated by some other means) than that of the battery. One can thus see that by adding a switch and a diode the current used for charging the battery can be diverted (by switching off the switch 207) to supply the load. Also this current will be supplied at a higher voltage than that of the battery, i.e. the main bus voltage, and therefore more power can be generated from the SA for this mode of operation. It should be noted that in practice several of the above sections would be used in parallel to charge the battery. This allows the design to be modular, making it better in terms of failure modes, expandability and current handling (smaller currents) . Such a technique is described in WO 99/28801 mentioned above, in which by voltage domain control the charge sections' current can be diverted to the main bus, which allows them to operate at their maximum- power-point voltage, or other voltage higher than that of the battery, to supply additional transient power to the load. This is illustrated in Fig. 11 of WO 99/28801. Although this method can operate near the MPP of the solar array when supplying power to the bus it does not utilise the full power of the SA when in battery charge mode because the SA is still connected across the battery, and hence is voltage-clamped to the battery, as in Fig. 2.

The invention aims to provide a still more efficient, less complex and lower-mass solution for the electronic power equipment needed to charge and discharge the battery.

According to the invention there is provided a battery charge and discharge regulator including: a battery charge section for providing a connection between a current source and a rechargeable battery in order to recharge the battery; and a power supply section for connecting the current source to a load via a main bus in such a way that the current source is constrained to supply current substantially at the voltage of the main bus; characterised in that the charging section incorporates a voltage converter capable of allowing the current source to charge the battery at a voltage independent of that of the battery; and in that the charge and discharge regulator further comprises a current controller to ensure that the current source when supplying current always supplies some of this current to the main bus.

The power supply section can advantageously include a diode in the connection line from the current source to the load. The presence of this diode means that the current source can operate substantially at the main bus voltage when supplying power to the main bus, and hence at the maximum available efficiency, even when it is also charging the battery, but is isolated from the bus as regards reverse currents, thus preventing failure propagation if a short-circuit occurs in the voltage converter, for instance. The voltage converter can be any suitable kind, though it is preferably a switching converter; it can be either of the buck or of the boost type. For examples of DC/DC converters reference may be made to The Electronics Handbook, ed. J C Whitaker, CRC Press/ IEEE Press 1996, pages 991-1002.

The controller for ensuring that the current source always supplies current to the mam bus can use a sensor that detects a predetermined condition, and on its detection disconnects the current source from the battery. A suitable condition would be that the current I_MB from the source to the main bus approaches zero. This ensures that the current source cannot become clamped to the battery. The sensor can be simply a voltage detector connected across an element in the power supply section, advantageously across the diode as mentioned above, and can be appropriately amplified.

The aim is to operate the SA section essentially at the mam bus voltage, when supplying a low impedance voltage source like a battery; to this end one controls the charge current into the battery so that it is less than that of the output current of the current source. This means that some current always flows through the diode and therefore the current source is voltage- clamped to the ma bus voltage V_MB (less 1 diode drop)

For a better understanding of the invention embodiments of it will now be described, by way of example, with reference to the accompanying drawings, m which:

Fig. 1 shows a typical solar array (SA) characteristic;

Fig. 2 shows a prior-art method of battery charging using a SA; Fig. 3 shows a battery charge method used m reference 2 ; Fig. 4 illustrates the first embodiment of the invention;

Fig. 5 shows a second embodiment of the invention ,-

Fig. 6 illustrates a third embodiment of the invention:

Fig. 7 shows the third embodiment m more detail;

Fig. 8 shows a variant of the third embodiment of the invention for the case where the voltage of the SA (V_SA) is less than the voltage of the battery (V_B) ; Fig. 9 shows a further variant of the third embodiment for the case of V_SA < or > V_B;

Fig. 10 shows a complete power system using the third embodiment ;

Fig. 11 shows a prior art system for an unregulated (battery) ma bus; and

Fig. 12 shows a complete power system applying the invention to the case of an unregulated (battery) ma bus .

The invention aims to overcome the battery voltage clamp problem by making the battery charge sections operate at a voltage independent of the battery voltage, even for the battery charge case. This will require fewer SAs , which m turn reduces mass and cost. Fig. 4 schematically illustrates tne first principle involved. The SA 200 (again, crly one section is considered for simplicity) is operated on the constant -current part of its characteristic . A DC-to-DC (switching) converter 206 s placed series between the SA section and the battery, m a charging section Ch of the circuit, to isolate the SA from the constant voltage of the battery. The input current to this converter (I_B) can now be controlled, m such a way that it is essentially always less than the SA current (I_SA) . Once this condition is established some current must always flow through the diode 205, located m a power supply section M of the circuit, keeping it in conduction with a forward voltage of one diode drop (less than 1 volt) . This means that the voltage of the SA section is essentially at the same voltage as the voltage of the main bus 302 supplying the load 301.

There are two main cases to be considered for this main bus voltage. One is when it is regulated to be at a fixed constant voltage (as given by Reference 1) , given by point C on Fig. 1, and the other case is where it is regulated to be at the maximum power point of the SA (as given by Reference 2) . This is shown as point E on Fig. 1.

A problem that the above circuit has, however, is that the SA current and voltage vary with temperature, age and illumination which makes it difficult to measure or predict. Thus to ensure that I_B is always less than I_SA would appear to require a current sensor, for each of these variable- currents (i.e. I_SA and I_B in each SA section), together with some comparator circuit to compare them. This can be quite complex. The problem can be overcome by detecting with a current sensor shown schematically at 298 the current 1,^ flowing in the diode 205 to the main bus, and when this equals or approaches zero a signal is sent to the converter 206 to switch Off. This principle can be made even simpler by the second embodiment of the invention which is illustrated by Fig. 5.

For illustration purposes a simple DC-to-DC (battery charge) converter 206 of the buck type (step- down) is shown, but many other converter topologies can be used. This converter consists essentially of a switch 207, a diode 201 and an inductor 204. When the switch 207 is OFF, current flows though the diode 205 to supply the main bus voltage and the voltage of the SA at point S is one diode drop above that of the main bus. When the switch 207 is closed the current I_B increases at a rate (V_SA-V_B)/ ^' !_ , where L is the inductance 204, until it is equal to I_SA. At this point the current through the diode 205 is equal to zero and therefore the voltage across the diode drops to zero.

This zero-voltage condition is detected by the comparator 208 which changes state at its output to switch OFF the switch 207; when this happens all the SA current I_SA flows through the diode 205. This instantaneously prevents the SA voltage, point S, from falling below that of the main bus, as it would do in view of the effective connection directly to the battery. By this means the SA voltage is kept essentially at the voltage of the main bus for any positive value of SA current. The SA current therefore does not have to be measured or predicted. Also, at this instant of switch-off of 207, the stored energy in the inductor 204 causes the diode 201 to conduct and the inductor current to continue to charge the battery, but now with a current that decreases approximately at a rate of (V_B/L) . After some delay, the converter control signal will switch 207 ON again in order to repeat the cycle .

Another advantageous feature, used in a third embodiment of the invention, is to make the DC/DC converter, placed between the SA section and the battery, a reversible one; that is, it has to be capable of supplying power to charge the battery from the SA in one direction and of supplying power from the battery to the main bus in the other direction. Th s enables the converter that would otherwise be needed tc provide power from the battery to the main bus to bε omitted. In sunlight, the SA sections charge the battery until a peak power load demand occurs; by the action of the main error amplifier (MEA) for the main bus voltage regulation and the domain control (see reference 2 for details) the battery charge current is decreased in order to supply extra main bus power. If this is not enough, the battery can provide additional discharge power through the reversible DC/DC converter to satisfy the load demand. In eclipse, the only source of power is the battery power and therefore the reversible DC-to-DC converter supplies battery discharge power to the main bus, normally at the voltage it had during the sunlight period. Most DC/DC converter topologies can be made reversible by replacing the power diode by an active switch. Fig. 6 shows the topology of Fig. 5 in its reversible form, including the active switch 501. Fig. 7 shows (schematically) in more detail the reversible DC/DC converter 701. The current magnitude and direction are controlled by the amplitude and sign of the MEA signal 24, which tells the circuit what (total) current is needed. The converter current is controlled by the current feedback loop, consisting of a comparator 90 that compares a proportion of the MEA signal with the voltage across a current sense resistor 81, in series with the inductor 204. By this means, the comparator senses when the inductor current is equal to the target value demanded by the MEA signal and controls the active switches, i.e. the power FETs 207 and 501, in order to maintain the MEA target current. This operation is interrupted by the comparator 208 to switch off the FET 207, if its current demand exceeds the current produced by the SA. The Nand gates 80 having opposite outputs are to make sure that when 207 is ON 507 is OFF, and when 507 is ON 207 is OFF. The comparator 208 could be replaced by any suitable amplifier such as a transistor. It acts to override the MEA feedback loop by limiting the current supplied to the battery from the SA. The advantage of this design is the following: a) The battery charge and discharge regulator is combined in one reversible DC/DC converter, which means fewer components than if two separate converters were used; b) This reversible DC/DC converter operates from a single section of the SA isolated by a diode 205 as shown m Figs. 6 and 7. This allows the design to be much simpler than one that has to operate from the mam bus. This is because failure modes when operating m this way are less critical to the mam bus. For example, the converter can fail by a short circuit to the battery (short circuit failure of 207) and this will result m one charge section being permanently connected to the battery, whereas, for a DC/DC converter that is connected directly to the mam bus or to the total SA, such a short would cause a total loss of the satellite unless some over-current protection is used at its input, which requirement complicates the design considerably. Another example of a fault is a short-circuit to ground at the input to the charge converter, which the example shown would cause a loss of one section of the total SA (typically 7% of the total power) whereas with the prior art this failure would cause total loss of the SA power and therefore the spacecraft. c) During sunlight, the charge section current can be diverted from battery charge to the ma bus and if this current is not sufficient to satisfy the load demand the battery can supply extra current by operating the DC/DC converter m battery discharge mode

In a fourth embodiment of the invention the battery voltage is above that of the mam bus. The above discussion has been for the case where the battery voltage is below tnat of the mam bus. Fig. 8 shows one well known type of converter that has a higher battery voltage. Here switches 601 and 607 are m antiphase, controlled by the comparator 208, and the inductor 604 is on the SA side of both switches.

In a fifth aspect of the invention the technique is applied to the case where the battery voltage can be below or above the ma bus voltage as its voltage varies . This needs a converter of the type shown in Fig. 9, having switches 901-904 switching in pairs in antiphase. The advantages of using a DC/DC converter that can interface with a battery whose voltage can be above or below the main bus voltage is that an independent choice can now be made of the main bus and battery voltages which can give a more efficient and lighter design.

Fig. 10 shows a complete power system showing the set of SA sections 200 charging the battery and the set of sections 200a used only for supplying the main bus. A typical satellite might require a 6 KW array, perhaps ten sections of 600 W each, with perhaps four of them charging the battery. The battery 203 is charged from the charge SA sections 200 through the DC/DC converters 701, connected in parallel, with a current determined by the battery control logic 124 and the signal of the MEA (Am, 57) . These battery charge SA sections are also used to satisfy transient peak power demands, by the use of domain control comparators 121 and 122 as explained previously and in Reference 2. The DC/DC converters 701 are of the type given in Fig. 7, that is, they are reversible and have the comparator 208 (as explained in the description of the second embodiment of the invention, Fig. 5) to keep the battery-charge SA at the voltage of the main bus. The main SA sections, 703, are regulated by a S³R controller 501 and a feedback amplifier, the MEA amplifier 57 (as explained in reference 1 and 2) , to control a main bus voltage supplying the load 89 and a main bus capacitor 53.

The sixth embodiment of the invention is to cover the case when it is an advantage to operate the load directly across the battery. This configuration is normally used when there is high transient peak power demand by the load. The voltage supplied to the load depends upon the charge or discharge state of the battery and gives a voltage variation of about 50%. The advantage of the technique is that no DC/DC converters with their efficiency and mass penalty are required between the battery and the load. This penalty can be very high for the case of high peak power demand, even if the duration of that power demand is short, because the converters must be sized for the peak power. The prior-art power system configuration is normally known as the unregulated bus and is shown m Fig. 11. The total SA 400 is connected together (i.e. no discrete SA sections) and all the DC-to-DC converters 151 are connected m parallel to supply power to the battery 203 and the load 301 during sunlight. The battery provides the power for transient peak power demand and eclipse operation.

For a high-power satellite and m order to get maximum power from the SA the input current to the DC- to-DC converters is adjusted by the logic control circuit 156 m such a way that the SA operates at its maximum power point. One of the major disadvantages of this system is that the DC-to-DC converters have to be designed so that any failure of their electronic components cannot present an overcurrent to the SA This failure case would take the SA away from its maximum power point and could result m the SA having not enough power to supply the load and charge the battery. This failure could cause a total loss of the spacecraft and must be avoided To provide protection against this failure it is normal for each DC-to-DC converter to have a current limiter connected to its input. If a component failure then occurs that results m an overcurrent at the input to one of the converters then this current limiter operates, to protect the SA from going under voltage, by limiting that current or switching OFF the failed DC-to-DC converter. The addition of this input current limiter complicates the design somewhat. Since the mass and cost of these converters is already high any possibility of reducing them is important .

The sixth embodiment of the invention avoids the use of this input current limiter for these DC-to-DC converters. This is illustrated by Fig. 12. The total SA is split up into sections, 200, and each section charges the battery through a DC/DC converter 701 of the type illustrated in Fig. 5 (buck) . As before, the buck converter is chosen for the case of the SA voltage always being above that of the battery. If it is below it, or can be either side of it, the appropriate converter should be used. The diodes 205 provide a path for the current not used for charging the battery and for supplying load current, to charge a capacitance 104. Connected to this capacitance is a converter 704 of the same type as the charging converters 701, but with the addition of input current protection 703 and redundant components, so that the failure of any one single component does not affect its operation significantly. An amplifier 801 is used in a feedback loop to vary the input current of this converter at the point 802 where the diodes 205 are connected to it, so as to control the SA voltage to be at its maximum power point, plus one diode drop due to the diode 205. The converter 704 also charges the battery.

With this system the SA will operate at its MPP if the load and battery charge power demand is high- enough. If this power demand is less than the maximum power the SA can produce then the current in the diodes 205 increases. This increase is absorbed by the converter 704 until its current limit operates, at which point it can absorb no more current . This forces the SA to operate above its maximum power point voltage on the SA characteristic (line E to G in Fig.l) . This reduces the output current of the SA until a balance is achieved where the SA power is equal to that demanded by the load. The only system impact is a slightly higher SA voltage (about 10%) .

Claims

CLAIMS :

1. A battery charge and discharge regulator including: a battery charge section (Ch) for providing a connection between a current source (200) and a rechargeable battery (203) in order to recharge the battery; and a power supply section (M) for connecting the current source (200) to a load (301) via a main bus (302) in such a way that the current source is constrained to supply current substantially at the voltage of the main bus; characterised in that the charging section incorporates a voltage converter (206) capable of allowing the current source to charge the battery at a voltage independent of that of the battery; and in that the charge and discharge regulator further comprises a current controller (298; 208) to ensure that the current source when supplying current always supplies some of this current to the main bus.

2. A regulator according to claim 1 and including a diode (205) in the connection line from the current source to the load.

3. A regulator according to claim 1 or 2 , in which the voltage converter (206) is a buck- or boost-type switching converter.

4. A regulator according to any preceding claim, in which the current controller (298) is arranged to sense a zero or near-zero current to the main bus, and then to disconnect the current source from the battery (203) .

5. A regulator according to claim 2 and 4, in which to sense the current the current controller senses the voltage drop across the diode (205) .

6. A regulator according to any preceding claim, in which the voltage converter (206) is adapted to work in reverse, supplying the main bus from the battery.

7. A regulator according to claim 4, in which the disconnection is performed by an active switch (207) .

8. A power supply system including a plurality of current sources (200) and a corresponding set of regulators according to any preceding claim.

9. A system according to claim 8, in which the current sources (200) are solar panels.

10. A system according to claim 9, and further including additional solar panels (200a) used for supplying the main bus but not for charging the battery.

11. A method for supplying a load with electrical power from a current generator, in which the generator can also be used to charge a battery, wherein a nonzero current is maintained at all times to the load.