US20130187468A1

US20130187468A1 - Uninterruptible power supply control in distributed power architecture

Info

Publication number: US20130187468A1
Application number: US13/475,091
Authority: US
Inventors: Sangsun Kim; Cornelius Benedict O'Sullivan; Shane R. Nay
Original assignee: Google LLC
Current assignee: Google LLC
Priority date: 2012-01-24
Filing date: 2012-05-18
Publication date: 2013-07-25
Also published as: EP2807718A4; CN104067480A; US9166423B2; WO2013112361A1; TW201342770A; WO2013112429A1; US9214833B1; TW201340549A; CN104067480B; DE202013012507U1; CN104205562A; SG11201403954TA; DK2807719T3; US9035609B1; US20130187471A1; TWI547061B; US9843221B1; EP2807719B1; EP2807719A1; EP2807719A4

Abstract

Aspects of the disclosure relate generally to uninterruptible power supply (“UPS”) units for systems requiring back up power. The UPS units include driving circuitry for controlling charging and allowing discharging of a battery. The driving circuitry includes a controller as well as a pair of switches. The MOSFETs includes a charging and a discharging MOSFET in series with the battery operating as a bidirectional switch. When the UPS unit is connected to an AC power supply, the controller regulates the current through the charging MOSFET switch based on feedback from a feedback device in order to charge the battery. If AC power is lost, the controller goes into saturation, switching the discharging MOSFET to a on condition and allowing the battery to discharge. The transitions from AC to battery power and vice versa are automatically achieved via the controller.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application claims the benefit of the filing date of U.S. Provisional Patent Application No. 61/590,060 filed Jan. 24, 2012, the disclosure of which is hereby incorporated herein by reference.

BACKGROUND

Various systems utilize battery backup supply systems, such as uninterruptible power supply (“UPS”) units. The UPS units include batteries that are charged during periods when the system is being powered by an external power source. If the power source is lost, the batteries are used to power the system's load. A typical system may use two different power supplies, one to run the components (such as devices in a server array) and another to charge the batteries. This arrangement may also require that the system has separate charging and discharging circuits for the UPS units. These dual arrangements can be relatively costly to produce, set up, and power.

SUMMARY

One aspect of the disclosure provides an uninterruptible power supply unit. The uninterruptible power supply unit includes a first MOSFET switch and a second MOSFET switch. The uninterruptible power supply unit also includes a controller for controlling current through the first and second MOSFET switches. The controller is configured to cause the first MOSFET switch to operate in a linear region during charging. The uninterruptible power supply unit includes a battery in series with the first and second MOSFET switches. The first MOSFET switch is configured to charge the battery when operating in the linear region. The uninterruptible power supply unit also includes a protection block in communication with the first and second MOSFET switches, the protection block configured to switch the first and second MOSFET switches off when the battery is operating out of a predetermined range of operating conditions.
In one example, the uninterruptible power supply unit also includes a feedback device arranged in series with the first MOSFET switch. In this example, the feedback device is configured to provide feedback information to the controller, and the controller is also configured to control the current through the first MOSFET switch based on the feedback current. In addition, the feedback device includes a shunt resistor. The controller is configured to adjust a gate-to-source voltage of the first MOSFET based on the feedback information. The controller is also configured to switch the first MOSFET to the active condition by generating a voltage through a gate of the first MOSFET based on the feedback information.
In another example, the controller comprises an amplifier. In a further example, when the charging current falls below a threshold value, the second MOSFET switch is switched on such that the second MOSFET operates in a switching region, and the second MOSFET switch is configured to provide discharging current to the battery. In this example, the uninterruptible power supply also includes a second protection block configured to receive the feedback information from the feedback device and to switch the second MOSFET off when the battery is charging. The second protection block is configured to receive the feedback information from the feedback device and switch the second MOSFET switch on when the battery is discharging.
In another example, the protection block comprises a processor configured to switch the first and second MOSFET switches off based on an abnormal voltage of the battery. In yet another example, the protection block comprises a processor configured switches the first and second MOSFET switches off based on an abnormal current of the battery. In a further example, the protection block comprises a processor configured to switch the first and second MOSFET switches off based on an abnormal temperature of the battery. In another example, the first and second MOSFET switches are configured to operate as a bidirectional switch. In still another example, the controller is configured to adjust a gate-to-source voltage of the first MOSFET switch when the first MOSFET switch is operating in the linear region during charging. In yet a further example, the controller is configured to automatically transition the uninterruptible power supply unit from a charging mode where the battery is charged to a back up mode where the battery supplies power to a load external to the uninterruptible power supply unit. In another example, the controller is configured to automatically transition the uninterruptible power supply unit from a back up mode where the battery supplies power to a load external to the uninterruptible power supply unit to a charging mode where the battery is charged. In another example, when the charging current falls below a threshold value, the first MOSFET switch is switched on such that the first MOSFET operates in a switching region, and the first MOSFET switch is configured to provide discharging current from the battery.
Another aspect of the disclosure provides a backup power supply. The backup power supply includes a first uninterruptible power supply unit. The first uninterruptible power supply unit includes a first MOSFET switch and a second MOSFET switch. The first uninterruptible power supply unit also includes a controller for controlling current through the first and second MOSFET switches. The controller is configured to cause the first MOSFET switch to operate in a linear region during charging. The first uninterruptible power supply unit includes a first battery in series with the first and second MOSFET switches. The first MOSFET switch is configured to charge the first battery when operating in the linear region. The first uninterruptible power supply unit also includes a protection block in communication with the first and second MOSFET switches. The protection block is configured to switch the first and second MOSFET switches off when the first battery is operating out of a predetermined range of operating conditions. The backup power supply system also includes a second uninterruptible power supply unit comprising a second battery. The first uninterruptible power supply unit is configured to control the charging current, charge the first battery, and switch the first and second MOSFETs independently of the second uninterruptible power supply unit.
In one example, when the charging current falls below a threshold value, the first MOSFET switch is switched on such that the first MOSFET operates in a switching region. In this example, the first MOSFET switch is configured to provide discharging current from the battery.
A further aspect of the disclosure provides a system. The system includes a load configured to receive power from a power supply and an uninterruptible power supply unit in communication with the load. The uninterruptible power supply unit includes a first MOSFET switch and a second MOSFET switch. The uninterruptible power supply unit also includes a controller for controlling current through the first and second MOSFET switches. The controller is configured to cause the first MOSFET switch to operate in a linear region during charging. The uninterruptible power supply unit includes a first battery in series with the first and second MOSFET switches. The first MOSFET switch is configured to charge the first battery when operating in the linear region. The uninterruptible power supply unit also includes a protection block in communication with the first and second MOSFET switches, the protection block being configured to switch the first and second MOSFET switches off when the first battery is operating out of a predetermined range of operating conditions. When the charging current falls below a threshold value, the first MOSFET switch is switched on such that the first MOSFET operates in a switching region. The first MOSFET switch is also configured to provide discharging current from the battery to the load.
In one example, the uninterruptible power supply unit further comprises a feedback device arranged in series with the first MOSFET switch. In this example, the feedback device is configured to provide feedback information to the controller, and the controller is also configured to control the current through the first MOSFET switch based on the feedback current.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is an example diagram of power architecture in accordance with implementations of the disclosure.

FIG. 2 is a diagram of a UPS unit in accordance with implementations of the disclosure.

FIGS. 3A and 3B are example diagrams of various circuits of UPS units in accordance with implementations of the disclosure.

FIG. 4 is an example flow diagram in accordance with implementations of the disclosure.

FIG. 5A is an example of a server architecture in accordance with implementations of the disclosure.

FIG. 5B is an example of a network system architecture in accordance with implementations of the disclosure.

DETAILED DESCRIPTION

Conventional UPS units often use a mechanical switch or relay in series with the batteries to disconnect the battery from the power source, such as a DC bus. Typically, these mechanical switches are not used to control the charging of multiple batteries as they are not able to regulate the individual charging current of each battery. An AC-DC power supply is able to limit the battery current of only a single UPS by regulating a common DC bus voltage. This also does not allow for individual control of the charging current. In such systems, the charging current is dependent upon the characteristics of the batteries within the UPS. In addition, the charging power of the UPS is unknown, so provisioning power for both an AC source and an AC-DC power supply cannot be adequately planned. Without limiting the battery charging current of individual UPS units, the AC-DC power supply must be over-rated to support the load and overall unknown charging power. Batteries that exceed the recommended charging current may heat up, thus shortening the battery's lifetime and causing safety concerns.
The configurations described herein disclose an active device in a UPS rather than a mechanical switch. For example, by utilizing two metal-oxide semiconductor field effect transistors (“MOSFET”) switches in series as a bi-directional switch, the MOSFETs may act as a disconnect device. For example, the MOSFETs may disconnect the battery or batteries from the DC bus in case of any faults detected within the UPS unit. The MOSFETs may also be used for discharging and charging, and thus, a separate battery charger or backup converter are not needed.
FIG. 1 is an example of a distributed power architecture 100 for a server system having a load and a plurality of UPS units 140. In this example, the architecture includes an AC power source 110 that supplies power to AC-DC power supplies 120. These power supplies may include, for example, one or more rectifiers. The AC-DC power supplies 120 provide power to a load 130. In this example, the load 130 may include a plurality of computing components.
The AC-DC power supplies 120 are also connected to the plurality of UPS units 140. As shown in FIG. 1, the UPS units 140 are arranged on a common DC distribution bus in parallel with the AC-DC power supplies 120 and the load 130. The UPS units 140 are used to ensure continued operation of the load 130 in the event of a failure of the AC power source 110 and/or AC-DC power supplies 120. The number of UPS units (N) used in the system may be determined based on the amount of backup power required to power the load for some pre-determined period of time.
FIG. 2 is an example of a UPS unit 140. In this example, the UPS unit 140 includes a housing 210, power terminals 220 to receive power from the AC-DC power supplies 120, and driving circuitry 230.
FIGS. 3A and 3B are examples of driving circuitry that may be used with the UPS unit 140. In these examples, the driving circuitry includes a controller 310, a battery pack 320 having one or more batteries, switches 340 and 342, and a feedback device 350. In addition, these circuits may also include a protection circuit 380 and a transistor 390 for fast off switching of the switches 340 and 342 based on temperature, voltage and current information associated with the batteries. As shown in these examples, the battery pack 320, the switches 340 and 342, and the feedback device 350 are arranged in series with one another. An additional protection circuit 360 may be associated with switch 342 for discharging.
The switches 340, 342 desirably comprise MOSFET switches. MOSFET switches are used to supply current for battery charging and discharging. The MOSFETS have different modes of operation. For example, a MOSFET have a switched mode of operation, including a “fully off” condition and a “fully on” condition. Another mode of operation is a linear region of operation where the drain-to-source voltage can be regulated by adjusting gate-to-source voltage. In this example, when operating in the linear region, the MOSFET allows a gate-to-source voltage of between 0 and 12 volts to pass through the MOSFET's gate. Whether a MOSFET are used as switches or in operated in their linear mode depends on whether the batteries are being charged (linear operation), discharged (on), or disconnected (off) from the load and the AC power supply.
The pair of MOSFET switches may be used for both the charging and discharging of the batteries. For example, MOSFET switch 340 can be used to control the charging of the batteries while MOSFET switch 342 can be used for discharging of the batteries. This combination of a charging MOSFET and a discharging MOSFET allows operation as a bidirectional switch.
The controller 310 may be, in one example, an amplifier configured to receive information from the feedback device. Based on the received information, the controller is able to automatically transition the UPS unit from using an outside power source to charge the battery to supplying power to a load. The feedback device 350 can include a shunt or current sense resistor. In the examples of FIGS. 3A and 3B, the feedback device senses current from one of the power terminals 220 and sends it to the negative terminal of the controller 310.
The controller automatically detects the state of the bus voltage based on current feedback received from the feedback device 350. For example, when the DC bus voltage is greater than the battery voltage, the controller is in charging mode. In the charging mode, the controller regulates or limits the charging current through MOSFET switch 340 by adjusting the gate-to-source voltage of the MOSFET switch 340 based on current received from the feedback device 350. In one example, the controller 310 is desirably associated with a reference current value. This value can be set through a pulse-width modulation (PWM) signal 360 at the positive terminal of the controller 310. Thus, the reference current value is adjustable based on the needs of the system. For example, the AC-DC power supplies 120 must produce enough current to power the load 130 and the number of UPS units (N) for charging. Because the reference current for each UPS unit may be set, this allows for an accurate calculation of how much power is needed for the load and charging the UPS units.
In some examples, the reference charging current value is set very low in comparison to the discharging current needed to power the load. By using a relatively low charging current, the thermal stress on the charging MOSFET operating in the linear region is low as well. If the current through the charging MOSFET is too high, the MOSFET can heat up and fail. This can also reduce the power drain on the AC-DC power supplies 120.
When the terminals of UPS unit are initially connected to the power from the AC-DC power supplies 120, the controller receives charging current feedback from the feedback device 350. In response, the controller generates a gate voltage in order to activate the MOSFET switches. Using the 12 volt example, the controller 524 increases the gate voltage of the MOSFET switches to between 0 and 12 volts depending upon the reference current value. This can switch all of the MOSFET switches, or both 340 and 342, to the active condition and allow the battery pack 320 to charge.
In this case, the controller compares reference current value and the information from the feedback device, and adjusts the current through the MOSFET 340 in order to control the charging of the battery pack 320. The charging current feedback at the negative terminal, received from the feedback device 350, follows the current defined at the positive terminal in voltage.
When the charging current becomes a bit lower than the reference current value, the DC bus voltage will be very close to or the same as the battery voltage. At this point, the battery may be almost fully charged. In response to current feedback from the current sense device, the output of the controller may be saturated at the maximum gate voltage and the battery is float charged to keep the battery close to or at its fully charged level.
As noted above, the control circuitry 230 can also be used for discharging. If the power source 110 and/or AC-DC power supplies 120 fail, the power received at the terminals 220 of the UPS device will drop off. The DC bus voltage will be less than the battery voltage. This causes the charging current feedback to be significantly lower than the reference current value. The difference between the charging current feedback and the reference current value causes the controller's output to go into saturation and causes the MOSFET switches to go into the fully on condition. In other words, the MOSFET switches are no longer operating in the linear region. At this point, the controller is no longer controlling the charging of the battery pack 320, and the current from the battery pack can discharge and flow through the terminals 220 to power the load 130. Having the MOSFET switches in the fully on condition when the battery pack is discharging can also reduce conduction loss.
The battery pack can continue to discharge until the battery pack is fully discharged or until the power source 110 and/or AC-DC power supplies 120 have been restored. When the power source has been restored, the UPS unit can automatically transition from discharging to charging via the controller.
Returning to the example of FIG. 3A, when the power from the AC-DC power supplies 120 is restored, the charging current feedback causes the controller 310 to immediately regulate the charging current to the battery pack as described above.
The protection circuit 380 may be configured to turn off both MOSFET switches 340 and 342 in order to disconnect the battery from the DC bus. The protection circuit 380 may include a microcontroller, CPU, or any type of circuit that can sense the condition of the current, temperature or voltage of the battery. If one or more of these conditions is outside of a predetermined normal operating range (for example, operating at an abnormal voltage, current, and/or temperature), the protection circuit 380 may automatically switch the MOSFETs 340 and 342 to the off condition disconnecting the UPS from the AC-DC power supply and the load. The protection circuit may operate much faster to shut off the MOSFETs than the controller.
In some examples, the driving circuitry can also include the additional protection circuit 360 mentioned above for disconnecting the discharging MOSFET 342 during the charging of the battery pack. For example, based on feedback current from the feedback device 350, the protection circuit 360 may sense when the power supply is active. In response, the protection circuit 360 may send an off signal to the discharging MOSFET 342 so that no discharging current is allowed to flow during charging operation. If the power source 110 and/or AC-DC power supplies 120 fail, the DC bus voltage received at the terminals 220 of the UPS device will drop off. This causes the charging current feedback to drop off as well when MOSFET 340 is switched on. In this example, based on feedback current from the feedback device 350, the protection circuit 360 may sense when the power supply fails. In response, the protection circuit 360 may send an on or activate signal to the discharging MOSFET 342 so that the discharging current is allowed to flow during discharging operation.
Flow diagram 400 of FIG. 4 is an example of some of the aspects described above. In this example, a UPS unit is connected to a power supply that powers a load at block 402. For example, UPS unit 140 may be connected to AC power supply 110 or AC-DC power supplies 120 which supply power to load 130. A controller receives charging current feedback from a feedback device at block 404. For example, the controller 310, which may include an amplifier, receives charging current feedback from the feedback device 350, which can comprise a shunt resistor. At block 406, the controller compares the charging current feedback to a reference current value. The controller then activates a charging MOSFET switch based on the charging current feedback at block 408. As described above, the reference current can be predetermined based on the power needs of the system. The controller then adjusts the current through the charging MOSFET based on the comparison as needed at block 410.
If there is no change in the power supply, for example, it has not been lost or failed at block 412, the charging controller continues to receive the charging current feedback at block 404. As described above, the controller compares the charging current feedback to a reference current value and adjusts the current through the charging MOSFET based on the comparison as needed at blocks 410 and 412, respectively.
Returning to block 412, if the power supply has been lost or has failed, the controller goes into saturation and causes a discharging MOSFET to switch to the fully on condition at block 414. At block 416, current through the discharging MOSFET provides power to the load until the power supply has been restored at block 418. Alternatively, current through the discharging MOSFET can provide power to the load until the battery pack has been fully discharged. When the power supply has been restored, the flow returns to block 404 where the controller again receives charging current feedback from a feedback device. The flow then continues as described above.
The aspects and features described above are especially useful in systems where power is not expected to be lost often for long periods. For example, in a developed area with a reliable power source, it may be fairly rare to have a power failure. In these systems, the UPS units do not have to be fully charged quickly, but can be slowly built up over longer periods, such as a few hours or a whole day. However, when a failure does occur, a typical server array may rely on a backup generator which typically takes on the order of a half minute to power up. During this period, the battery pack provides the discharging current to the load. The rate of the discharging current may be much higher than that of the charging current.
In addition, the driving circuits described above address complex charging and discharging issues with simple and cost effective solutions. For example, by using an amplifier to control the charging current in a single UPS device, the charging current in a plurality of different UPS devices is regulated without requiring external oversight or monitoring. The arrangement of the paired charging and discharging MOSFET switches both eliminate the need for separate battery charging power supplies and reducing the total cost of the system.
The UPS units described herein can be used in conjunction with various backup power systems. For example, these devices may be useful in telecom systems or server architectures. FIG. 5A is an example of a server architecture including a plurality of the UPS units described herein. In this example, the server 510 includes a rack 520, having a set of shelves 530, for housing the load 130 as well as the UPS units 140. The AC-DC power supplies 120 can be incorporated into the rack 510 (as shown in FIG. 5A) or can be at a different location, for example, as the AC power source 110 is shown in FIG. 5A.
The load 130 can include a variety of devices. For example, the load 130 can include a dedicated storage device, for example, including any type of memory capable of storing information accessible by a processor, such as a hard-drive, memory card, ROM, RAM, DVD, CD-ROM, or solid state memory. The load may include a preprogrammed load which draws power from the AC-DC power supplies 120 in order to test the operation of the server 510. The load 130 may also include a computer including a processor, memory, instructions, and other components typically present in server computers.
FIG. 5B is an example of a network system including the server architecture of FIG. 5A. For example, server 510 may be at one node of a network 540 and capable of directly and indirectly communicating with other nodes of the network. For example, these computers may exchange information with different nodes of a network for the purpose of receiving, processing and transmitting data to one or more client devices 550-52 via network 540. In this regard, server 510 may transmit information for display to user 560 on display of client device 550. In the event of a failure of the AC power source 110, the UPS units may allow the server 510 to continue communications with the other nodes without interruption.
As these and other variations and combinations of the features discussed above can be utilized without departing from the subject matter defined by the claims, the foregoing description of the embodiments should be taken by way of illustration rather than by way of limitation of the subject matter defined by the claims. It will also be understood that the provision of the examples disclosed herein (as well as clauses phrased as “such as,” “including” and the like) should not be interpreted as limiting the claimed subject matter to the specific examples; rather, the examples are intended to illustrate only one of many possible embodiments. Further, the same reference numbers in different drawings may identify the same or similar elements.

Claims

1. An uninterruptible power supply unit comprising:

a first MOSFET switch and a second MOSFET switch;

a controller for controlling current through the first and second MOSFET switches, the controller being configured to cause the first MOSFET switch to operate in a linear region during charging;

a battery in series with the first and second MOSFET switches, the first MOSFET switch configured to charge the battery when operating in the linear region; and

a protection block in communication with the first and second MOSFET switches, the protection block configured to switch the first and second MOSFET switches off when the battery is operating out of a predetermined range of operating conditions.

2. The uninterruptible power supply unit of claim 1, further comprising a feedback device arranged in series with the first MOSFET switch, the feedback device being configured to provide feedback information to the controller, wherein the controller is further configured to control the current through the first MOSFET switch based on the feedback current.

3. The uninterruptible power supply unit of claim 2, wherein the feedback device includes a shunt resistor.

4. The uninterruptible power supply unit of claim 2, wherein the controller is configured to adjust a gate-to-source voltage of the first MOSFET based on the feedback information.

5. The uninterruptible power supply unit of claim 2, wherein the controller is configured to switch the first MOSFET on by generating a voltage through a gate of the first MOSFET based on the feedback information.

6. The uninterruptible power supply unit of claim 1, wherein the controller comprises an amplifier.

7. The uninterruptible power supply unit of claim 1, wherein when the charging current falls below a threshold value, the second MOSFET switch is switched on such that the second MOSFET operates in a switching region, and the second MOSFET switch is configured to provide discharging current to the battery.

8. The uninterruptible power supply unit of claim 7, further comprising a second protection block configured to receive the feedback information from the feedback device and to switch the second MOSFET off when the battery is charging.

9. The uninterruptible power supply unit of claim 7, further comprising a second protection block configured to receive the feedback information from the feedback device and switch the second MOSFET switch on when the battery is discharging.

10. The uninterruptible power supply unit of claim 1, wherein the protection block comprises a processor configured to switch the first and second MOSFET switches off based on an abnormal voltage of the battery.

11. The uninterruptible power supply unit of claim 1, wherein the protection block comprises a processor configured switches the first and second MOSFET switches off based on an abnormal current of the battery.

12. The uninterruptible power supply unit of claim 1, wherein the protection block comprises a processor configured to switch the first and second MOSFET switches off based on an abnormal temperature of the battery.

13. The uninterruptible power supply of claim 1, wherein the first and second MOSFET switches are configured to operate as a bidirectional switch.

14. The uninterruptible power supply of claim 1, wherein the controller is configured to adjust a gate-to-source voltage of the first MOSFET switch when the first MOSFET switch is operating in the linear region during charging.

15. The uninterruptible power supply of claim 1, wherein the controller is configured to automatically transition the uninterruptible power supply unit from a charging mode where the battery is charged to a back up mode where the battery supplies power to a load external to the uninterruptible power supply unit.

16. The uninterruptible power supply of claim 1, wherein the controller is configured to automatically transition the uninterruptible power supply unit from a back up mode where the battery supplies power to a load external to the uninterruptible power supply unit to a charging mode where the battery is charged.

17. The uninterruptible power supply of claim 1, wherein when the charging current falls below a threshold value, the first MOSFET switch is switched to on such that the first MOSFET operates in a switching region, and the first MOSFET switch is configured to provide discharging current from the battery.

18. A backup power supply comprising:

a first uninterruptible power supply unit, the first uninterruptible power supply unit comprising:

a first MOSFET switch and a second MOSFET switch;

a controller for controlling current through the first and second MOSFET switches, and wherein the controller being configured to cause the first MOSFET switch to operate in a linear region during charging;

a first battery in series with the first and second MOSFET switches, the first MOSFET switch is configured to charge the first battery when operating in the linear region; and

a protection block in communication with the first and second MOSFET switches, the protection block being configured to switch the first and second MOSFET switches off when the first battery is operating out of a predetermined range of operating conditions; and

a second uninterruptible power supply unit comprising a second battery;

wherein the first uninterruptible power supply unit is configured to control the charging current, charge the first battery, and switch the first and second MOSFETs independently of the second uninterruptible power supply unit.

19. The backup power supply of claim 17, wherein when the charging current falls below a threshold value, the first MOSFET switch is switched to on such that the first MOSFET operates in a switching region, and the first MOSFET switch is configured to provide discharging current from the battery.

20. A system comprising:

a load configured to receive power from a power supply;

an uninterruptible power supply unit in communication with the load, the uninterruptible power supply unit comprising:

a first MOSFET switch and a second MOSFET switch;

a protection block in communication with the first and second MOSFET switches, the protection block being configured to switch the first and second MOSFET switches off when the first battery is operating out of a predetermined range of operating conditions;

wherein when the charging current falls below a threshold value, the first MOSFET switch is switched on such that the first MOSFET operates in a switching region, and the first MOSFET switch is configured to provide discharging current from the battery to the load.

21. The system of claim 19:

wherein the uninterruptible power supply unit further comprises a feedback device arranged in series with the first MOSFET switch, the feedback device being configured to provide feedback information to the controller, and

wherein the controller is further configured to control the current through the first MOSFET switch based on the feedback current.