WO 2003034568 A1
A power charger which charges and provide conditioned power from at least one of a super-capacitor (8) and a secondary battery (13). The type of energy device is user selectable, such that the user can select super-capacitors because of their much longer charge/discharge life and low maintenance requirements, while another user can chose secondary batteries because of their higher stored energy, while a third user may select a combination of the two.
DUAL CAPACITOR/BATTERY CHARGER FIELD OF THE INVENTION
The present invention is directed to an electrical charging device that can recharge either or both of secondary batteries or capacitors, and which implements various intelligent charge-discharge methods for automatically preventing the onset of memory
effects in NiCd batteries. A power factor correction circuit for use in a power pack is also disclosed.
BACKGROUND OF THE INVENTION New electro-chemical capacitors ("super-capacitors" or double-layer capacitors) are currently available in sizes that can compete in some applications with small NickelCadmium (NiCd) and Nickel-Metal-Hydride (NiMH) batteries. Secondary (rechargeable) batteries store energy in the dissociation and recombination of particular chemical compounds. Super-capacitors store energy in the electric field of ionic compounds in close proximity.,
Secondary batteries offer higher energy density than super-capacitors. For example, there exist conventional 2.3 N, 100 Farad super-capacitors that are approximately the size of two "AA" batteries. The energy stored in such a super-capacitor is provided by the well-known formula: CN /2 = 1 OOE (2.3N)'-/2 = 264 Joules. Standard "AA" NiCd batteries have a rating of 1.2N at 0.6 Ampere-hour, so two "AA" batteries
have a total capacity of 2(1.2V)(0.6A-l r)(3600sec/hr) =5184 Joules.
However, while conventional batteries have a significantly higher energy storage density, super-capacitors have superior peak current capability. For example, the
same super-capacitor discussed previously has a peak current rating of over 20 Amperes
while the two-"AA" ΝiCd pack has a current rating of only 2.4 A. Super-capacitors also
have a superior charge-discharge rated life. A typical super-capacitor can be charged and
discharged over 100,000 times while the lifetime of a typical ΝiCd battery is only 500-
1000 charge-discharge cycles.
Super-capacitors also do not suffer from the "memory effect" that occurs in
ΝiCd batteries. The memory effect is a well-known characteristic of ΝiCd batteries in which repeated partial discharges of the battery cause part of the energy in the battery to become inaccessible during discharge. This is a reversible change in the battery and can
be corrected by periodically fully-discharging or "deep discharging" the battery. Deep
discharges are generally performed by removing the battery from the battery-powered
device, discharging the battery, and then recharging it on a dedicated external battery charger. However, this solution is not always practical. Alternatively, deep discharge can be performed by running the powered equipment until the battery is fully discharged.
However, this solution is also not suitable in many situations because the battery may die
unexpectedly, and thus operation of the powered device will ve unreliable. Although
battery "lifetime" indicators may provide some warning, they often operate unreliably when there is an existing memory effect which must be countered. In addition, some electronic devices may be damaged as the battery voltage drops toward the deep-discharge
Another difference between batteries and super-capacitors is the operating
voltage range. Batteries have a narrow operating voltage range. For example, 90% of the
energy in a NiCd battery is supplied over a voltage range of 1.35 to 1.05V. In a conventional 2.3V supercapacitor, this range is from 2.3V to 0.7V. Thus, batteries are much easier to use in powering systems because the voltage is generally constant and the
battery output is about the same at the end of the discharge cycle as at the beginning.
Because of the various technical differences between batteries and super- capacitors, each device must be charged in specific, often incompatible, ways. Thus, hardware designers must select which rechargeable device to use in powering equipment,
making tradeoffs between the longer charge/discharge life and low-maintenance
requirements of super-capacitors and the higher stored energy density and more constant
output voltage of secondary batteries. The selection is important, particularly if the equipment includes a built-in charger.
It is an object of the invention to provide a power charger which charges
and provides conditioned power from either or both of a super-capacitor and a secondary
battery. The choice of energy device would preferably be at the user's discretion, so that one user could select super-capacitors because of their much longer charge/discharge life and low-maintenance requirements, while another user could choose secondary batteries
because of their higher stored energy, while a third user may select a combination of the
two. It is another object of the invention to provide an embedded charging
system in which the charger is included as part of a larger system performing some
function with the stored energy, and where the charger includes an on-board automatic
system for performing periodic deep-discharge maintenance of NiCd batteries to remove
the memory effect.
Yet another object of the invention is to provide a charger which
automatically detects the storage element type and automatically or manually adjusts the
recharging parameters to accommodate the particular storage element.
It is a further object of this invention to provide a "pre-conditioning"
element that narrows the apparent charge and discharge voltage range of super-capacitor
BRIEF DESCRIPTION OF THE DRAWINGS The foregoing and other features of the present invention will be more
readily apparent from the following detailed description and drawings of illustrative
embodiments of the invention in which:
FIG. 1 is a block diagram of a charging system with removable energy
packs according to an embodiment of the invention; FIG. 2 is an illustration of a particular embodiment of the circuits in the
charger unit of Fig. 1; and
FIG. 3 is a circuit diagram of a power factor correction circuit. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring now to Figure 1 , one embodiment of a charging and power
system according to the invention comprises Charger Unit, 1 and one or more energy
sources, such as a Battery Pack 2, and a Capacitor Pack 3. A particular application may use
only the battery pack, only the capacitor pack, either pack, or a combination. Preferably,
the battery and capacitor packs 2, 3 are removable and have a respective mating plug 12,
17 which can be connected to a suitable receptacle 5 in the charger unit 1 and which includes appropriate terminals for providing charging current to the energy pack and for
extracting stored energy from the pack. Multiple receptacles 5 may be provided in the
charger unit for accepting more than one energy source at a time. In such a variant, mixed
energy sources may be used. Thus, for example, a user may attach a battery pack 2 and a capacitor pack 3 to the charger unit. As discussed below, the charger is configured to
detect which type of energy pack is attached to receptacle 5 and adapt the functionality of
the charge and/or discharge circuitry accordingly.
As shown, the Charger unit 1 comprises at least one power supply element
6 that converts input power from external connector A, into the appropriate voltage levels
required to operate the charger unit. The power produced by the power supply element 6,
is routed to a logic circuit 8 which controls the basic operation of the actual charging
circuitry 10 that in turn provides controlled power (received from the power supply element 6) to recharge the energy pack 2, 3. The logic circuit 8 can also control the operation of the output control 11. Preferably, logic circuit 8 is comprised of hardwired logic elements, either discrete or as part of an integrated structure, such as a programmable
logic array. Alternatively, logic circuit 8 may be implemented in programmable devices. A type selector circuit 9 is also provided. Type selector circuit 9 receives
inputs which are used to determine the type of energy storage pack attached to the charger
unit. In a preferred embodiment, each class or type of energy storage pack is assigned a
unique type identifier, such as one or more type identifier jumpers 16 or 22 placed on
terminals of the respective mating plug, 12 or 17 or another physical distinction. When the
energy pack is attached to the charger unit 1, the particular combination of open, grounded,
and/or jumpered terminals are sensed and used to determine the energy pack type.
Alternatively, the type of battery pack connected to charger unit 1 can be determined by
using a circuit which senses the charge characteristic of the attached device. In particular, batteries have a relatively constant charge voltage while capacitors increase voltage
linearly with increasing charge. In yet a further embodiment, a manual switch can be
toggled by the user to identify the type of energy source attached.
A signal indicating the type of energy storage pack that is attached to the
charger unit 1 is then provided to the logic circuit 8. In response, logic circuit 8 uses this
specific information to control the charging characteristic of the charging circuit 10.
Although the logic circuit 8 and type selector circuit 9 are shown as separate elements, this is primarily for ease of understanding. It should be understood that the functionality of the
two elements can be combined into a single circuit. The stored energy from the energy pack 2, 3 is routed through output control circuit 11. The output control circuit 11
switches power from the energy storage pack to the external connector 4 upon receiving
control signals from the logic circuit 8. The external load is mated to the external connector 4. Following is a brief description of the functionality of the a preferred embodiment of the Logic and Charger controller 8. The functions can be implemented by discrete logic circuits or via programmable logic.
The Logic circuit and charger controller 8 is the circuit that controls the
overall operation of the charger unit 1. It monitors the input power, and when appropriate, enables the charging circuits to recharge the battery or capacitor pack. In the specific application where the energy storage is used for emergency operation (as opposed to the
primary power source of a portable device), the controller also enables the energy storage
pack to provide power to the load when the input power is absent or invalid. Portable
operation can be achieved, in one embodiment, by adding a switching element between the input power supplies and the power connection to the external device. When input power is valid, the controller enables the switching element and the load operates from the input
The charging functionality of the controller is implemented by monitoring voltages and currents at the energy pack via the energy pack connections. These
connections serve as sense leads as well as power leads. Preferably, the logic and
controller 8 and charging circuit 10 have separate charging functions for battery packs and
for capacitor packs, which functions are enabled or disabled according to the type of energy pack in use.
A variety of battery charging algorithms may be used, including fast charge with voltage shut-off followed by trickle charge, negative-delta- vee termination, timed fast
charge with trickle charge, and negative slope dee-vee-dee-tee. It should be apparent to
those skilled in the art that other battery charging schemes could also be implemented in accordance with this invention and that the particular charging method may vary according
to the type of device being charged.
Charging a capacitor pack 3 requires a slightly different scheme. Whereas batteries have a substantially constant voltage characteristic, capacitors have a linear voltage characteristic. To keep from over-heating the charging circuit when the capacitor
voltage is very low (large voltage drop across the current source), the charging power is
artificially lowered. This is accomplished by either reducing the charging voltage or
reducing the charging current. Both methods can be implemented using techniques which are well known to those skilled in the art. One preferred way to reduce the charging power
is by pulse- width-modulation of the charging circuit. This pulse- width-modulation lowers
the duty cycle and the average power dissipated by the charging circuit. Typically the
period of the modulation is 1-5 seconds because the charging circuit usually has a large
thermal mass on the heatsink. When the capacitor pack is charged up to the rated voltage, fast charge is terminated and fixed- voltage trickle charge holds the pack at full energy.
The type selector circuit 9 is used to enable or disable the various charging
modes by causing different software parameters or modules in the logic and charger controller 8 to execute if the controller is executed in firmware. If the controller is . executed using discrete circuitry, the outputs of the type selector circuit are used to select
appropriate voltage setpoints or resistor dividers and enable and disable certain circuit blocks.
In the figure, counter CI includes a first and a second counter. The first counter is incremented after a discharge and charge of an energy storage pack. Logic circuit 8' is used to monitor the first counter, and is also used to generate an output control
signal when the first counter equals a predetermined number. This output control signal is
supplied to output control circuit 11.
Using output switch T3 and current limit T4, the output control circuit 11
switches power from the energy storage pack to the external connector 4 (shown in Fig. 1)
when the output control signal from the logic circuit is received. In the preferred embodiment, the output switch is a transistor, and the current limiter is a transistor
connected in parallel to a resistor R4.
In embodiments of the invention, the logic circuit 8' comprises logic gates
Gl and G2, which monitor the output signals of circuits VI and V2 which are provide a output charging signal based on the difference between reference voltage #1 and the input
voltage applied to circuits VI and V2. When the logic circuit 8' determines that it is
possible to charge the energy storage pack, a high level signal (an output charging signal,
logic level 1) is provided to charging circuit 10. In an embodiment, the charging circuit 10
comprises a fast charging constant current source T5, a reference voltage (i.e., a Fast
Charge Endpoint Voltage comprising Zener Diode Zl and transistor T6), and a trickle charging resistor R5.
Switch T2 is controlled by the output charging signal. Upon receipt of this
signal, which also functions as a charge inhibit signal, the second counter is enabled.
After each discharge of the energy storage pack, the second counter is incremented. When a second predetermined number is reached, the discharge circuit 10 further enables charging of the energy storage pack such that deep discharge maintenance on the energy
storage pack is performed periodically. In preferred embodiments switch T2 is a transistor. If the controller is executed using discrete circuitry, the outputs of the
selector are used to select appropriate voltage set points or resistor dividers and enable and
disable circuit blocks in the logic and charger controller 8. The selector circuit 9 is used to
enable or disable the various charging modes by causing different software parameters or
modules in the logic and charger controller 8. In embodiments of the invention, selector
circuit 9 comprises a plurality of pull-up resistors Rl and R2 which are used to indicated the type of energy pack is attached to the charge 1. Based on the configuration of the
terminals which connect the energy pack to the charger unit 1, the selector circuit 9
provides outputs A and B to the logic and charger control circuit 8 for use in determining the type of energy pak in use.
According to a further aspect of the invention, the energy packs are configured for
use with a charger, such as the charger disclosed above. Returning to Fig. 1, each of the
energy storage packs 2 and 3 comprises at least a mating plug 12, 17 and an energy storage
device, such as battery 13 or capacitor 18. In addition, and according to a further aspect of the invention, the energy storage pack can be augmented with particular enhancements
specific to the type of storage device. Enhancements to the battery pack may be included
within the battery pack or within the charger, and generally these functions may be
separate from the charger circuitry or included with the charger circuitry. For example, in
addition to battery 13, the battery pack 2 can further comprises a controller 14 and a deep discharge maintenance system 15. Similarly, the capacitor pack 3 may include controller 19, and buck/boost regulators 20, 21. Other enhancements include the addition of circuitry
to count the number of times the pack has been charged and discharged and, particularly
for NiCd batteries, the duration from, or a time stamp of the time that the last deep
discharge occurred. Other contemplated enhancements include circuitry to measure the duration from, or a time stamp of the time, the battery was charged (to compensate for
self-discharge of the battery), and circuitry to indicate the time in service (to compensate
for aging batteries). In addition, circuitry or appropriate memory devices can be included within the pack to store a serial number identification.
Preferably, all of the stored information can be automatically extracted by
the charger unit and processed accordingly or passed to an external device. For example,
the device serial number can be automatically retrieved to identify the location of power
packs which have been recalled due to a newly discovered flaw. Such circuitry could be
composed of some or all of the following elements as either discrete circuits or software
elements: timers, voltage detecting circuits, counters, logic circuits, and memory elements.
Other elements may also be used as will be appreciated by one of skill in tlie art.
Particular embodiments of the function blocks in the battery pack 2 will now be discussed.
With reference to the block diagram of the battery pack 2 in Fig. 1, there are
several ways in which a deep discharge maintenance system 15 may be implemented. In a
first embodiment, system 15 comprises a circuit which selectively disables the recharge of
the battery in response to an input control signal. In a second embodiment, system 14 comprises a circuit, such as a transistor-based circuit, which discharges the battery into a
resistive load upon receipt of an input control signal and which stops the discharge when
the battery voltage, as indicated by an internal battery voltage measurement circuit, reaches
a predetermined level. The control signal can be provided by the controller 14, discussed
below, or a source external to the battery pack 2, such as the charger unit 1.
The functionality of the battery pack controller 14 is dependent on the
operation of tl e deep discharge maintenance system, if there is one, and also upon the intended uses of the battery pack. Potential uses include constant energy applications, such as emergency lighting and backup power for equipment such as emergency recorders and
communication systems. One anticipated use is within aircraft emergency systems.
A first controller embodiment is for use primarily when energy consumed
by a load is a constant, is known, and is less than a full discharge of the pack. The controller comprises a counter that is incremented after each discharge/charge cycle. A
logic circuit monitors the counter and determines when a first predetermined number of
cycles have occurred, preferably a number of cycles which are related to when a "memory
effect" is likely to occur. If the controller is driving a discharge maintenance system 15
which selectively disables charging of the battery, a second count is started. The second count is updated each time the pack is discharged. Recharging is inhibited during this discharge-inhibit state. When the second count reaches a second predetermined amount,
normal recharging resumes. The seconds preset number is selected to permit the normal
load cycles of the pack to discharge the pack nearly all the way, preferably including a
large safety margin. Thus, periodically, deep discharge maintenance on the pack is
In one embodiment, two separate counters are used. A discharge-inhibit
signal is provided to the maintenance system 15 whenever the first count equals the first
predetermined number of cycles. This signal also enables the second counter and disables subsequent increments of the first counter. When the second counter reaches the second predetermined level, both counters are reset.
In an alternate implementation, preferably for use when energy consumed
by a load is a constant, is generally known, and is less than a full discharge of the pack, a
microprocessor or similar digital controller, is configured to provide functionality similar to that discussed above. Advantageously, the use of a microprocessor permits the actual
energy consumed by the load to be monitored. The duration of the charge-inhibit state is
selected according to the measured energy consumption. In one embodiment, one or
several instantaneous power measurements are made and this information is used to calculate the second count stop value. Alternatively, power usage can be continuously
monitored. This embodiment allows a deeper discharge to occur since the actual power
drawn by the load is known. Further, the processor can monitor the dept of discharge of
the previous deep discharge, compare that with values stored in memory from laboratory
measurements and other deep discharge events, and determine when the batter pack has decreased capability due to aging or other effects. The microprocessor can also alter the
deep-discharge method in accordance to compensate for known aging effects. A
maintenance signal can also be generated when battery pack capacity has decreased to a preset level to indicate that service or replacement may be required. This signal can be a warning light, or other signal, on the battery pack itself or a signal which is recognized by
the charger unit and acted upon accordingly.
The design of controller 14 is somewhat different when used with the
second embodiment of the deep discharge maintenance system 15. In particular, when counters or equivalent programmable functions determine that a large number of partial
discharge/recharge cycles have occurred, which number is associated with occurrence of a
memory effect, the controller produces a control signal which initiates a discharge of the
battery to a known fully discharged state, approximately 0.9V/cell for NiCd, for example. The next time the battery is discharged, the memory effect is mitigated.
Preferably, several constraints are placed on the discharge. First, a deep- discharge is only initiated if it has been quite some time since the last normal discharge. For example, when the battery pack is used in an aircraft data recorder, it is preferable to
only discharge the pack if the aircraft has been sitting idle for 3-4 hours, for example, like
overnight. Second, in a preferred embodiment, the controller terminates deep-discharge
immediately if power is re-applied to the system, such as would occur when an aircraft is
being readied for flight. In addition, the controller preferably recognizes when a deep- discharge cycle is unsuccessful , e.g., as a result of either conditions 1) and 2) , and then repeats the deep-discharge cycle at an appropriate time.
Particular embodiments of the function blocks in the capacitor pack 3 will
now be discussed.
In a preferred embodiment, the capacitor pack 3 comprises a controller 19 and one or more buck/boost regulators 20, 21. When the capacitor voltage is near the load
voltage, conventional switching or linear power supply circuits are used to convert the
voltage being supplied to the capacitor pack from a normal battery charging voltage to the
optimal charging voltage for the capacitor pack. Since the capacitor charges from OV up to the full voltage, the charging voltage must also vary widely to prevent dissipating too
much power in the charger. Conventionally, the buck/boost regulator includes a cascaded
boost regulator and buck regulator, a single buck-boost regulator topology, or a forward
converter. Any of these topologies are adequate for this application. A similar buck/boost
regulator is used to convert the capacitor voltage back to the ideal load voltage.
When the capacitor voltage is much lower than the load voltage, only a buck converter is used to charge the capacitors. A boost converter is used to recover the
energy in the capacitor and convert it to the proper load voltage. The circuits can be implemented using simple regulators. When the capacitor voltage is much higher than the load voltage, only a
boost converter is used to charge the capacitor. A buck converter is used to recover the
energy in the capacitor and convert it to the proper load voltage. The circuits can be
implemented using simple regulators.
The capacitor pack controller 19 comprises voltage and current
measurement and limiting circuits. When charging power is applied, the controller sets one
or more charging voltage setpoint values for the charging buck/boost regulator(s). The setpoint values may change over time as the capacitor becomes charged fully. The final
setpoint voltage is the predetermined rated voltage of the capacitor pack (point of
maximum energy storage).
According to a further aspect of the invention, circuitry is added to the
energy packs 2, 3, to permit charging of non-optimal batteries or capacitors to a different
voltage and convert the stored voltage back to an expected voltage. Preferably, the
circuitry is contained within the energy pack. However, in an alternative embodiment it
can be included in the charging unit itself. This additional circuitry permits the charger to charge optimal batteries or capacitors as reliably and inexpensively as possible yet also
allows the basic system to work with other storage elements as they become developed.
Non-optimal storage elements are storage elements having voltage and
charge combinations which do not lend themselves to operation at the desired output voltage and power level. For example, a small standard NiCd battery is the "AA" size rated at 600 mA-hr and 1.2 V. To build a 24V system, 20 of these batteries can be
combined in series. The rated capacity of the pack would be 600 mA-hr. However, if the application called for only a 120 mA-hr life, for size, weight, or cost considerations, a non- optimal battery pack, comprised of 4 cells in series, for example, could be used instead.
The 4 cells would nominally provide 4.8 V at 600 mA-hr. According to this aspect of the
invention, the additional circuitry steps-up the battery output voltage 5 times to 24 V,
reducing the overall capacity by a factor of 5 to the required 120 mA-hr output. The charging voltage, 28V typically for a 24V system, would be stepped down to 6V,
appropriate for a4.8V battery pack.
According to yet another aspect of the invention, particularly useful for
applications where the energy storage pack provides a known amount of energy to the load and the energy storage pack contains much more energy than that known amount to be
supplied, is to perform a "deep discharge" wherein the charger selectively declines to
charge the pack, instead of actively discharging it.
For example, a load requires an energy of 0.5 A-hr and the pack is rated at 5
A-hr. In a conventional system, each time the pack is charged and discharged by the load,
the pack is only cycled to the 10% discharged point (0.5/5), or 90% of full charge. If the
pack were a NiCd pack, then the memory effect would begin asserting itself after some
tens of cycles. According to the invention, when it is determined that a deep-discharge of the pack is appropriate, e.g., in response to the operation of discrete circuitry or software, the charger could prevent the recharge of the pack for some number of charge-discharge
cycles. As a result, the pack discharges to 90% on the first discharge, then 80% on the next
discharge, since it is not being recharged, then 70%, and so on down to 20%o, for example.
At this point, the maintenance deep-discharge cycle is completed. The regular charging circuitry is then reactivated and the battery pack is fully recharged to 100%. Of course, the
percentage of power drawn and thus the number of chargesuppressed discharge cycles will
vary depending on the power capacity of the battery, the power requirements of the load,
and the duration between charge cycles. In particular implementations, power is drawn from the charger unit
directly from the power units and the battery pack is used only as a backup power supply
in the event that the main power fails. In one particular environment, typically used in
power systems installed in commercial airlines, power is available form both 28 V DC mains and 120V AC mains. It would be advantageous to draw power from both of these
sources to produce a regulated DC output for use in charging power packs and in powering
A new power factor correction circuit for permitting this operation is shown in Fig. 3. Power factor correction circuits are often added to provide reduced current draw
and distortion in power supplies. According to this aspect of the invention, the circuit has
the capability to operate either from the AC mains 31 or a DC mains 33 (separate from the
AC mains). The circuit configuration shown has the following characteristics:
1) Power Factor Correction ("PFC") circuit 30 is configured to operate on the rectified secondary side of a power transformer 32. In contrast, conventional PFC
circuits operate on the primary power without a transformer.
2) The PFC circuit voltage reference is supplied from either tlie primary 35
or the secondary 35' of the transformer 32. In contrast, conventional PFC circuits use the input waveform (sinusoid or rectified sinusoid) as a reference. In the present embodiment, the PFC circuitry makes the input current follow a scaled version of the reference voltage. Using the secondary of the transformer is easier to implement but the primary is more
accurate since the control system corrects for transformer distortion.
3) The PFC circuit boosts the secondary DC voltage up to the final (output)
level. Conventional PFC circuits boost before the main regulator, so that the output of the
PFC has to be converted again before it can be used.
4) DC Mains system 33, when available, can be connected into the rectified
secondary side of the transformer 32. The PFC circuitry 30, 36 operates as a boost
regulator, boosting the DC up to the required output level. Advantageously, only one
boost regulator is needed, not two. If primary AC side sensing is used, additional circuitry is required to get DC reference. However, if secondary side sensing is used, no additional
circuitry is required.
5) Using a transformer on the AC mains instead of using a switching
regulator makes electromagnetic interference easier to deal with since the transformer can
be shielded and leakage inductance can be added to smooth out the primary current. While the invention has been particularly shown and described with reference to preferred embodiments thereof, it will be understood by those skilled in the art
that various changes in form and details may be made therein without departing from the
spirit and scope of the invention.