U.S. patent application number 14/574176 was filed with the patent office on 2015-09-17 for adaptive synchronous switching in a resonant converter.
The applicant listed for this patent is AVOGY, INC.. Invention is credited to Alberto Doronzo, Antoin Russell.
Application Number | 20150263640 14/574176 |
Document ID | / |
Family ID | 54070071 |
Filed Date | 2015-09-17 |
United States Patent
Application |
20150263640 |
Kind Code |
A1 |
Russell; Antoin ; et
al. |
September 17, 2015 |
ADAPTIVE SYNCHRONOUS SWITCHING IN A RESONANT CONVERTER
Abstract
An embodiment of a resonant converter includes having resonant
circuitry having inductive and capacitive elements configured to
create electrical resonance when an input voltage is applied and a
synchronous rectifier coupled between at least a portion of the
resonant circuitry and an output of the resonant converter. The
synchronous rectifier includes a diode, and an electrical switch.
Control circuitry is configured to operate the electrical switch
such that the electrical switch is turned on when there is
substantially no voltage across the diode and current flow in the
diode is positive in a direction from anode to cathode.
Inventors: |
Russell; Antoin; (Berkley,
CA) ; Doronzo; Alberto; (San Jose, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
AVOGY, INC. |
San Jose |
CA |
US |
|
|
Family ID: |
54070071 |
Appl. No.: |
14/574176 |
Filed: |
December 17, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14213361 |
Mar 14, 2014 |
|
|
|
14574176 |
|
|
|
|
Current U.S.
Class: |
363/126 |
Current CPC
Class: |
H02M 3/1588 20130101;
Y02B 70/1425 20130101; H02M 2001/0035 20130101; H02M 2001/0058
20130101; Y02B 70/1433 20130101; H02M 7/217 20130101; Y02B 70/1475
20130101; Y02B 70/10 20130101; H02M 1/32 20130101; Y02B 70/1466
20130101; Y02B 70/1491 20130101; H02M 2001/0064 20130101; Y02B
70/16 20130101; H02M 7/06 20130101; H02M 3/33592 20130101 |
International
Class: |
H02M 7/06 20060101
H02M007/06 |
Claims
1. A resonant converter comprising: resonant circuitry having
inductive and capacitive elements configured to create electrical
resonance when an input voltage is applied; a synchronous rectifier
coupled between at least a portion of the resonant circuitry and an
output of the resonant converter, the synchronous rectifier
comprising: a diode, and a first electrical switch; first control
circuitry configured to operate the first electrical switch such
that the first electrical switch is turned on when there is
substantially no voltage across the first diode and current flow in
the first diode is positive in a direction from anode to cathode; a
second electrical switch, comprising the only electrical switch
coupled between the input voltage and at least a portion of the
resonant circuitry; and second control circuitry configured to
operate the second electrical switch such that the second
electrical switch is turned on when there is substantially no
voltage across the second electrical switch.
2. The resonant converter of claim 1 wherein the first electrical
switch comprises a GaN transistor.
3. The resonant converter of claim 1 wherein the first control
circuitry further comprises: a current transformer coupled to a
current of at least a portion of the resonant circuitry; an
inverter amplifier coupled to the diode; and an AND gate configured
to perform a Boolean AND function using an output of the current
transformer and an output of the inverter amplifier.
4. The resonant converter of claim 1, wherein the resonant
converter is a non-isolated resonant converter.
5. The resonant converter of claim 1, wherein: the resonant
converter is an isolated resonant converter having a transformer
coupling an input stage with an output stage; and the output stage
comprises the synchronous rectifier.
6. The resonant converter of claim 1, wherein: the output of the
resonant converter has a positive rail and a negative rail; and the
synchronous rectifier is coupled in series between the resonant
circuitry and the positive rail of the output of the resonant
converter.
7. The resonant converter of claim 1, wherein: the output of the
resonant converter has a positive rail and a negative rail; and the
synchronous rectifier is coupled in series between the resonant
circuitry and the negative rail of the output of the resonant
converter.
8. A resonant converter comprising: resonant circuitry having
inductive and capacitive elements configured to create electrical
resonance when an input voltage is applied; a first synchronous
rectifier and a second synchronous rectifier, wherein each of the
first synchronous rectifier and the second synchronous rectifier
comprise: a diode; and an electrical switch in parallel with the
diode; first control circuitry configured to operate the first
synchronous rectifier and the second synchronous rectifier such
that, for each of the first synchronous rectifier and the second
synchronous rectifier, the electrical switch is turned on when
current flow in the diode is positive in a direction from anode to
cathode; and a resonant electrical switch, comprising the only
electrical switch coupled between the input voltage and at least a
portion of the resonant circuitry; and second control circuitry
configured to operate the resonant electrical switch such that the
resonant electrical switch is turned on when there is substantially
no voltage across the resonant electrical switch.
9. The resonant converter of claim 8 wherein the electrical switch
of either or both of the first synchronous rectifier and the second
synchronous rectifier comprises a GaN transistor.
10. The resonant converter of claim 8 further comprising a current
transformer having: a primary winding configured to conduct a
current of at least a portion of the resonant circuitry; a first
secondary winding configured to drive the electrical switch of the
first synchronous rectifier; and a second secondary winding, having
an opposite polarity of the first secondary winding, configured to
drive the electrical switch of the second synchronous
rectifier.
11. The resonant converter of claim 10 wherein: the first secondary
winding is configured to drive the electrical switch of the first
synchronous rectifier via a first driver; and the second secondary
winding is configured to drive the electrical switch of the second
synchronous rectifier via a second driver.
12. The resonant converter of claim 10 wherein the second secondary
winding is further coupled to an output of the resonant
converter.
13. The resonant converter of claim 8 wherein: an output of the
resonant converter has a positive rail and a negative rail; and the
first synchronous rectifier is coupled in series between the
resonant circuitry and the negative rail of the output of the
resonant converter.
14. The resonant converter of claim 8 wherein: an output of the
resonant converter has a positive rail and a negative rail; and the
first synchronous rectifier is coupled in series between the
resonant circuitry and the positive rail of the output of the
resonant converter.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The present application is a continuation of U.S. patent
application Ser. No. 14/213,361, entitled "ADAPTIVE SYNCHRONOUS
SWITCHING IN A RESONANT CONVERTER," filed on Mar. 14, 2014. The
entire disclosure of the above application is hereby incorporated
by reference, for all purposes, as if fully set forth herein.
BACKGROUND
[0002] Power electronics are widely used in a variety of
applications. Power adapters having power electronic circuits are
commonly used to convert the form of electrical energy, for
example, from AC to DC, from one voltage level to another, or in
some other way. Such devices can operate over a wide range of power
levels, from milliwatts in mobile devices to hundreds of megawatts
in a high voltage power transmission system. Despite the progress
made in power electronics conversion systems, there is a need in
the technology for advanced systems architecture and methods of
operating the same to achieve high efficiencies and improve on
size, weight, and complexity of the power electronic devices and
its applications.
SUMMARY OF THE INVENTION
[0003] The present invention relates generally to power electronic
converters. More specifically, the present invention relates to
resonant converter and adaptive control circuitry. Embodiments may
utilize techniques including (1) synchronous switching on resonant
circuit primary switches, (2) synchronous switching on output
synchronous rectifier drive circuits, (3) operating the resonant
converter in "burst mode" to maintain zero-voltage switching under
light to heavy load conditions, and/or (4) active voltage clamping
to minimize unnecessary energy clamping that can lead to added
component dissipation and reduced power converter efficiency.
[0004] An embodiment of a resonant converter, according to the
disclosure, includes resonant circuitry having inductive and
capacitive elements configured to create electrical resonance when
an input voltage is applied a synchronous rectifier coupled between
at least a portion of the resonant circuitry and an output of the
resonant converter. The synchronous rectifier includes a diode and
an electrical switch. Control circuitry is configured to operate
the electrical switch such that the electrical switch is turned on
when there is substantially no voltage across the diode and current
flow in the diode is positive in a direction from anode to
cathode.
[0005] An embodiment of a method of providing electrical power
conversion, according to the disclosure, includes providing a
resonant converter with resonant circuitry having inductive and
capacitive elements to create electrical resonance when an input
voltage is applied to the resonant circuitry. The method further
includes rectifying an output voltage of the resonant converter
using a synchronous rectifier coupled between at least a portion of
the resonant circuitry and an output of the resonant converter. The
synchronous rectifier includes a diode and an electrical switch.
The method also includes operating the electrical switch such that
the electrical switch is turned on when there is substantially no
voltage across the diode and current flow in the diode is positive
in a direction from anode to cathode.
[0006] Another embodiment of a resonant converter includes resonant
circuitry having inductive and capacitive elements configured to
create electrical resonance when an input voltage is applied, and a
first synchronous rectifier and a second synchronous rectifier.
Each of the first synchronous rectifier and the second synchronous
rectifier includes a diode and an electrical switch in parallel
with the diode. Control circuitry is configured to operate the
first synchronous rectifier and the second synchronous rectifier
such that, for each of the first synchronous rectifier and the
second synchronous rectifier, the electrical switch is turned on
when current flow in the diode is positive in a direction from
anode to cathode.
[0007] Numerous benefits are achieved by way of the present
invention over conventional techniques. Methods provided herein
enable an AC-DC converter to operate efficiently while maintaining
a desired output power level from light to heavy loads. By
attaining high efficiency, power system thermal requirement is
reduced and power density is significantly increased. Moreover, the
disclosed techniques can assist in preserving the integrity of the
switching elements when operating at high voltages and/or
frequencies. Disclosed techniques can apply to both isolated and
non-isolated resonant converters. These and other embodiments of
the invention, along with many of its advantages and features, are
described in more detail in conjunction with the text below and
attached figures.
BRIEF DESCRIPTION OF THE DRAWINGS
[0008] FIG. 1 is a schematic diagram illustrating a non-isolated
resonant converter, according to one embodiment;
[0009] FIG. 2 is a drawing of waveforms for V.sub.S, I.sub.S1, and
Drive 1 of the non-isolated resonant converter of FIG. 1;
[0010] FIGS. 3A and 3B are schematic diagrams illustrating types of
feedback that can be used to inform the control circuitry;
[0011] FIGS. 4 and 5 show waveforms of V.sub.S and Drive 1,
illustrating burst mode according to one embodiment;
[0012] FIG. 5 is a flow diagram illustrating the functionality of
an adjustable-output electrical adapter, according to another
embodiment;
[0013] FIG. 6 shows waveforms illustrating how Drive 1 reduces
T.sub.ON to reduce output power, and how this can result in
insufficient current to drive V.sub.S back to zero;
[0014] FIG. 7 is a schematic diagram illustrating an isolated
resonant converter, according to one embodiment;
[0015] FIG. 8 is a simplified illustration of a transformer that
can provide leakage inductance, according to one embodiment;
[0016] FIGS. 9-12 are schematic diagrams illustrating various
embodiments of an output stage of an isolated resonant
converter;
[0017] FIGS. 13-15 are schematic diagrams illustrating various
embodiments of circuitry for driving synchronous rectifiers, such
as those shown in previous figures; and
[0018] FIGS. 16 and 17 are flow diagrams illustrating embodiments
of methods of providing electrical power conversion.
[0019] FIG. 18 is a schematic diagram showing an example controller
circuit for providing S.sub.1 control, according to one
embodiment.
[0020] FIGS. 19-21 are schematic diagrams showing embodiments of
active clamp circuits.
[0021] FIG. 22 is a schematic diagram showing active clamp usage in
another embodiment.
[0022] In the appended figures, similar components and/or features
may have the same reference label. Further, various components of
the same type may be distinguished by following the reference label
by a dash and a second label that distinguishes among the similar
components. If only the first reference label is used in the
specification, the description is applicable to any one of the
similar components having the same first reference label
irrespective of the second reference label.
DETAILED DESCRIPTION
[0023] The present invention relates generally to power electronic
converters. More specifically, the present invention relates to an
adaptive resonant converter and associated control circuitry. The
disclosed embodiments have the capability to adapt to internal and
external changes and operate under varying line, load,
environmental and component parameters and yet preserve very high
efficiency ranging from no load to heavy loads. Such a power
converter can be utilized, for example, in AC-DC power converter to
power any of a variety of electronic devices such as laptop
computers, USB-powered devices, and the like with very high power
densities. Techniques detailed herein can apply to both isolated
and non-isolated resonant converters.
[0024] Among other things, four methods and techniques are applied
in the disclosed embodiments as follows. 1) A synchronous switching
technique is used on resonant circuit primary switches. By
utilizing zero-voltage switching, embodiments of the invention can
provide for greatly-reduced switching losses. 2) A synchronous
switching technique is also used on output synchronous rectifier
drive circuits. Control circuitry may monitor voltage and/or
current across certain switches in the primary and/or secondary
circuits to enable such efficient switching. 3) Embodiments may
include operating the resonant converter in a "burst mode" to
maintain that zero-voltage switching under light to heavy load
conditions. This function is automatically adjusted to compensate
for variations in line and load, also environmental and/or
component parameter changes. Further, the "burst mode" function can
be user programmable in order to adapt to application and preserve
the premium efficiencies. 4) An active voltage clamping circuit is
used to minimize unnecessary energy clamping that leads to added
component dissipation and results in reduced power converter
efficiency. This is especially the case when isolation transformer
in resonant converters is purposely designed with high leakage
inductance for integration and reduced component counts. The
relationship between primary resonant inductance (Lp) and
transformer leakage inductance (Llk) is:
Lp=(LlkNp.sup.2)/Ns.sup.s (1)
for a 1:1 transformer ratio Lp=Llk. Leakage inductance is selected
for maximum power transfer from primary to secondary circuits.
[0025] Advanced high density power electronic packaging can be
applied to reduce significantly loop inductance of power stages and
switching losses especially in high frequency resonant converters.
Power switching interconnections techniques can have benefits in
sensing, thermal management, and EMI containment. A fabricated
integrated structure by means of integrated process flow can be
replaced by an assembled one.
[0026] Further, part or all techniques disclosed can be applied to
a Power Factor Correction (PFC) or active rectifier circuit feeding
a resonant converter.
[0027] Below is description of the above-referenced circuits and
techniques.
Synchronous Switching
[0028] Control circuitry may monitor voltage and/or current
across-circuit switches to enable synchronous zero-voltage
switching. There are many different ways to sense voltage or
current. Embodiments are not particular to a specific method in
which voltage or current sensing is accomplished. For example,
current can be measured by hall-effect sensors, precision resistors
with or without active circuits for isolation, or current
transformers. Sensing primary current alone may not be a good
representative of output resonant current for switching purposes.
There is a significant phase shift between the primary current and
secondary current that can vary with load, temperature and
component parameter changes.
In some cases one current sensor can be utilized to monitor
resonant current and predict current through two or more switches
in isolated and non-isolated resonant circuits. Control may use the
desired voltage and current feedbacks to determine best turn-on
time of a particular switch.
[0029] FIG. 1 is a schematic diagram illustrating a non-isolated
resonant converter 100, according to one embodiment. The embodiment
shown in FIG. 1 and elsewhere herein are provided as non-limiting
examples. One of ordinary skill in the art would recognize many
variations, modifications, and alternatives to the components
provided herein.
[0030] In FIG. 1, the non-isolated resonant converter 100 is
configured to take an input voltage V.sub.IN, which may be an AC,
DC, or Rectified AC voltage source, and provide an output voltage
V.sub.O, which can be greater than the input voltage V.sub.IN.
Operation of the non-isolated resonant converter 100 is determined,
in part, by the operation of Drive 1, which drives electronic
switch S.sub.1, modulating the switch S.sub.1 to meet power
requirements, thereby making the non-isolated resonant converter
100 self-adapting. Inductive elements L.sub.B and L.sub.R and
capacitive elements C.sub.P and C.sub.R enable electrical resonance
in the non-isolated resonant converter 100. C.sub.P is the lumped
circuit capacitance at the node V, which may include the parasitic
capacitance of the semiconductor switch S.sub.1 and any other
electrically connected stray capacitances at that node such as the
L.sub.B, L.sub.R associated capacitances, and the circuit loading
of the L.sub.R, C.sub.R network. Synchronous rectification at the
output is provided by switch S.sub.R and diode D.sub.B, and output
power is provided across a load resistance R.sub.L. Additional
detail regarding synchronous rectification and Drive 2 are provided
herein below.
[0031] To enable the non-isolated resonant converter 100 to operate
in high-power applications at a high switching frequency,
specialized components can be utilized. For example, in some
embodiments, the transistors and diodes utilized in the switches
can be devices based on a wide bandgap material, such as GaN or
silicon carbide (SiC). This can enable the non-isolated resonant
converter 100 to operate at higher voltages, higher temperatures,
and higher frequencies than solutions using traditional
silicon-based devices. However, any semiconductor device can be
used.
[0032] Specialty magnet materials and geometry may be used to
advance the high frequency operation of the resonant circuits when
isolation transformers are used. Advanced material may also be used
for other magnetic components when non-isolated and isolated
topologies are utilized.
[0033] The values of the various components utilized in the
non-isolated resonant converter 100 can vary, depending on desired
functionality, manufacturing concerns, and/or other factors.
[0034] FIG. 2 is a drawing of waveforms for V.sub.S, I.sub.S1, and
Drive 1 of the non-isolated resonant converter 100 of FIG. 1,
provided to help illustrate the operation of the non-isolated
resonant converter 100. The wave form of Drive 1 shows how Drive 1
can operate switch S.sub.1 to undergo several on/off cycles having
switching periods including a time the switch is turned on,
T.sub.ON, and a time the switch is turned off, T.sub.OFF.
[0035] As shown, when switch S.sub.1 is on, the voltage at the node
V.sub.S reduces and the current I.sub.S1 begins to ramp up,
reaching a peak current I.sub.P when Drive 1 turns the switch
S.sub.1 off. When switch S.sub.1 is off, resonance between L.sub.B,
C.sub.P, and equivalent impedance of L.sub.R and C.sub.R impedance
occurs. Current flows into C.sub.R where, in some embodiments, a
voltage of up to three times the input voltage V.sub.IN can be
achieved. However, V.sub.S then reduces to a lower voltage than
V.sub.IN due to the reversal of the current I.sub.S1. The frequency
of the resonance will be according to the equation:
Fr := 1 2 .pi. L C ( 2 ) ##EQU00001##
Where L and C are the Thevenin equivalent inductance and
capacitance at node V.sub.S.
[0036] One beneficial feature the non-isolated resonant converter
100 is that the voltage waveform created by the switching of
S.sub.1 and the resonance created between L.sub.B, L.sub.R, and
C.sub.P allows for zero-voltage switching of S.sub.1. That is,
S.sub.1 is switched when the voltage across S.sub.1 is at or near
zero volts, which greatly reduces the switching loss in the
capacitance Cp. The voltage across S.sub.1 returns to zero volts
when the magnitude of the I.sub.S1 is great enough, and the voltage
is held at zero volts through the action of the parasitic or
intentionally included S.sub.1 parallel diode or the subsequent
turn on of S.sub.1 at an optimal time T1.
[0037] For an optimal turn on of S.sub.1 at T1, the voltage V.sub.S
can be monitored either directly or via a similar representative
voltage waveform (e.g., transformer winding). For example, a
zero-voltage-detection circuit can be utilized such that when
V.sub.S is approaching or at zero volts, the next T.sub.ON
transition for S.sub.1 is started.
[0038] The increased efficiency from zero-voltage switching of
S.sub.1 node comes from the minimization of the energy in
capacitance C.sub.P which would otherwise be dissipated in S.sub.1
according to the formula:
EnergyCp:=1/2CpVs.sup.2 (3)
for each switching period.
[0039] By quickly turning on S.sub.1 as when the voltage is at or
near zero volts also reduces and/or eliminates the conduction in
the parallel diode (parasitic on intentional) to S.sub.1.
[0040] The zero-voltage-detection circuit and optimal turn-on of
S.sub.1 at T1 is also beneficial in that variations in the dynamic
or initial values of the resonant network (inductance, capacitance
or resistance) will change the shape of the resonant waveform and
also change the optimal turn-on start time T1. However, the
zero-voltage-detection circuit can adapt the switching waveform
every switching cycle to help ensure T1 is optimal, largely
independent of other circuit variations.
[0041] The efficient power transfer from the input to the output
(V.sub.IN to V.sub.O) of a non-isolated resonant converter 100, and
similar resonant converters, is therefore primarily governed by the
controlling of S.sub.1 on and off times while maintaining the
resonant operation to obtain zero-voltage switching. FIGS. 3A and
3B are simplified illustrations schematics of how feedback can be
used in the control circuit driving S.sub.1, in various embodiments
of a resonant converter.
[0042] FIG. 3A illustrates providing feedback in the system around
the power stage with voltage feedback. FIG. 3B, on the other hand,
illustrates providing feedback in the system around the power stage
with current feedback. It can be noted, however, that other
configurations may utilize a combination of both. For example,
embodiments may utilize power feedback (where
P.sub.O=V.sub.O*I.sub.O), in order to control the energy is
delivered to the output. As illustrated, the feedback signal in
some embodiments may be galvanically isolated (input to output) in
an isolated version of the controller scheme, with a signal
isolation circuit (such as an opto-coupler or signal
transformer).
[0043] Embodiments may potentially utilize a variety of different
methods for modulating S.sub.1 to achieve output regulation while
maintaining zero-voltage switching. Three such methods include
frequency modulation, on-time (T.sub.ON) modulation, and pulse
density modulation or "burst mode." This art includes a controlled
burst mode to maintain zero-voltage switching under varying
internal and external conditions. Details of this feature are
described below.
Synchronous Switching Output Stages
[0044] Synchronous switching may also be utilized on output stage
of a resonant power converter to further reduce losses and increase
efficiency. Synchronous rectification can provide benefits to
topologies such as flyback. Example benefits include: 1)
Magnetizing current can be negative, hence discontinuous conduction
mode is avoided and the output voltage is regulated even under no
load conditions; 2) zero-voltage switching can be achieved; and 3)
conduction losses of the rectifier are significantly reduced
specially at low voltage levels.
[0045] Below are examples with isolated and non-isolated
configurations.
[0046] As illustrated in FIG. 1, the non-isolated resonant
converter 100 is a simple embodiment of a converter in which
isolation is not provided. As illustrated in FIG. 7 and subsequent
figures, however, many variations on the simple design of FIG. 1
can be made, including using circuitry that provides isolation.
[0047] FIG. 7 is a schematic diagram illustrating an isolated
resonant converter 700, according to one embodiment. In this
embodiment, a transformer T.sub.1 provides isolation, and may
additionally provide voltage change as well, depending on desired
functionality. Double-ended arrows with dotted lines indicate
alternative configurations. Thus, the secondary winding of the
transformer T.sub.1 may be coupled either way, depending, for
example, on the desired phase in which the isolated resonant
converter 700 is to be operated, as described in more detail below.
In some embodiments, the L.sub.B inductor of FIG. 1 can be
complemented and/or entirely replaced by the transformer T.sub.1
magnetizing inductance L.sub.MAG. Additionally, as indicated, the
resonant inductor L.sub.R can be placed either on the input side
(L.sub.RB) or the output side (L.sub.RA). Alternatively, some
embodiments may include both.
[0048] The values of the inductive elements can vary, depending on
the input and/or output specifications of the converter. If, for
example, the output voltage is much lower than the input voltage,
the value of the resonant inductance can be reduced by the square
of the transformer turns ratio to achieve a much lower value of
inductance L.sub.RA in comparison to L.sub.RB. This can facilitate,
for example, the use of a much lower loss air-core inductor instead
of one with a magnetic core in position L.sub.RB, which would
introduce more loss in the inductance.
[0049] Inductance on either or both of the input side (L.sub.RB) or
the output side (L.sub.RA) can also be included in the circuit by
the addition of leakage inductance between the primary and
secondary winding. Leakage inductance increases through increasing
the physical separation between primary and secondary windings.
[0050] FIG. 8 is a simplified illustration of a transformer 800
that can provide such leakage inductance, according to one
embodiment. According to this embodiment, instead of winding
primary and secondary windings on the same core leg (e.g., on top
of each other) they can be wound side-by-side or on separate legs
to each other, to deliberately introduce a desired amount of
leakage inductance. Here, a magnetic core 810 is wound with a
primary winding 820 on one side, and a secondary winding 830 on the
other side, the magnetic core 810 conducting the magnetic flux 840
between the windings.
[0051] This method, when applied to the topology of FIG. 8 (as well
as other embodiments providing isolation), can have at least two
significant benefits. First, the components L.sub.RA and/or
L.sub.RB can be eliminated as physical components. And second,
galvanic isolation can be much more easily achieved between primary
winding 820 and secondary winding 830. This is the case where the
windings are wire-wound or embedded in a multilayer printed circuit
board. Further, because the windings are not stacked on each top of
other, the number of winding layers is reduced by half. This can
greatly reduce the cost of a multi-layer printed circuit board
(PCB), in embodiments in which PCB is used for constructing
windings. In embodiments not using PCB, manufacturing costs can
still be reduced because windings are relatively easy to wind and
the transformer requires no insulation tape.
[0052] In further reference to the isolated resonant converter 700
of FIG. 7, the polarity of the secondary winding 830 of the
transistor 800 may be in either direction, which can determine the
phase of S.sub.1 on which the power transfer happens through
L.sub.R (L.sub.RA and/or L.sub.RB) and C.sub.R.
[0053] Specialty magnet materials and geometry may be used to
advance the high frequency operation of the isolated resonant
circuits. Advanced material may also be used for other magnetic
components in circuit.
[0054] In the isolated resonant converter 700 of FIG. 7, the output
diode D.sub.O can be moved to a zero voltage reference (i.e.,
output GND) instead of the output positive voltage rail. This means
that it can be much easier to include a semiconductor switch
rectifier in addition to or in replacement of a diode rectifier
D.sub.O because the reference for the drive signal Drive 2 can be
zero volts.
[0055] FIG. 9 illustrates an embodiment of an output stage 900 of
an isolated resonant converter. This output stage can, for example,
be a variation on the output stage of the isolated resonant
converter 700 of FIG. 7. Here, as in FIG. 7, the secondary winding
of transformer T.sub.1 may be coupled in either orientation,
depending on desired functionality. In this output stage, 900
D.sub.O is located in the high side (positive voltage rail) in a
configuration similar to the non-isolated resonant converter. As
indicated previously, this configuration is possible, but it can be
more difficult to provide a drive waveform for a semiconductor
switch.
[0056] Semiconductor switches instead of or in parallel with diode
positions (as shown), known as synchronous rectifiers can reduce
the conduction losses of the switch. Such synchronous rectifiers
can be included in many applications where lower resistance is
desired. Additional detail regarding Drive 2 is provided herein
below.
[0057] Note that the position of the capacitor C.sub.R in FIGS. 7
and 9 can additionally or alternatively be across the diode D.sub.O
position. Because C.sub.R is typically much smaller than the output
capacitance, C.sub.R therefore forms an electrically equivalent
circuit when in series with the much larger output capacitance. The
resonant circuit discussed in previous embodiments therefore also
includes the parasitic capacitance of the diode D.sub.O.
[0058] FIG. 10 illustrates another embodiment of an output stage
1000 of an isolated resonant converter. Again, the secondary
winding of transformer T.sub.1 may be coupled in either
orientation, depending on desired functionality. This output stage
1000, however, includes a variation in which L.sub.R (which may
also be leakage inductance, as previously discussed) and C.sub.R
are in series instead of in parallel with the transformer
secondary.
[0059] The topology output stage 1000 can be beneficial because the
voltage on the rectifiers D.sub.O1 and D.sub.O2 (and/or their
semiconductor switch equivalents) is limited to approximately the
output voltage plus switch voltage drop while conducting. In
practice, for example, V.sub.S1 can be 3-4 times lower than V.sub.S
in previous embodiments. This can be beneficial because
lower-voltage-rating diodes and semiconductor switches can be used.
These components typically often have lower on resistance and lower
conducting voltage drop, thereby reducing heat and increasing
efficiency. In certain applications, such as high output current
applications, these benefits may justify the increased complexity
of output stage 1000.
[0060] FIGS. 11 and 12 show additional embodiments of output stages
1100 and 1200, respectively. The configurations illustrate
electrical equivalent variations of the output stage 1000 of FIG.
10. FIG. 10 shows how S.sub.R1 can be coupled in series between the
resonant circuitry and the positive rail of the output of the
resonant converter, where FIG. 11 shows how S.sub.R1 can be coupled
in series between the resonant circuitry and the negative rail of
the output of the resonant converter. FIG. 12 shows how output
capacitor C.sub.O can be split in two, and the secondary winding of
transformer T.sub.1 can be coupled in-between the two new
capacitors C.sub.O1, C.sub.O2.
Control for Synchronous Rectifiers
[0061] Synchronous switching can be used on output synchronous
rectifier drive circuits. By utilizing zero-voltage switching,
embodiments of the invention can provide for greatly-reduced
switching losses. Control circuitry may monitor voltage and/or
current across certain switches in the primary and/or secondary
circuits to enable such efficient switching.
[0062] For output stage circuits shown in FIGS. 1, 7, and 9, the
output stage resonant circuit is parallel, and there is phase
difference between the current and voltage waveform. There is also
a phase difference between the S.sub.1 waveform and transformer
waveform and the output rectifier (D.sub.O) position. Therefore,
the S.sub.1 controller is unable to determine when to turn the
synchronous rectifier (S.sub.R) on and off.
[0063] Assuming that there is a parasitic or intended diode in the
S.sub.R position, an ideal condition for turning on switch S.sub.R
occurs when the diode voltage is minimized (conducting) and the
current flow in the diode is positive (anode to cathode). The
voltage information may not be enough on its own to operate S.sub.R
because once the S.sub.R turns on and voltage is minimized, it will
be difficult to determine when to turn S.sub.R off from the low
D.sub.O/S.sub.R voltage.
[0064] FIG. 13 is a schematic diagram illustrating a block level
solution to this problem. Here, a current transformer, AND gate,
inverter amplifier, and a driver to help ensure that S.sub.R is
turned on when the voltage is minimized (e.g., at or approaching
zero) and the current flow in D.sub.O/S.sub.R is positive (anode to
cathode direction). S.sub.R is turned off again when the current
flow is approximately zero. Here, the primary winding of the
current transformer is coupled to a current of at least a portion
of the resonant circuitry (e.g., I.sub.R, I.sub.DB, or I.sub.DO),
the inverter amplifier is coupled to a node of the diode D.sub.O,
and the AND gate is configured to perform a Boolean AND function
using an output of the current transformer (e.g., from the
secondary winding) and the output of the inverter amplifier. This
circuit can be used to provide the signal for Drive 2 in FIGS. 1,
7, and 9.
[0065] As for the circuits shown in FIGS. 10-12, because the
capacitor C.sub.R is in series with the D.sub.O element, it may
only be necessary to know when the current flow is positive in the
rectifier (anode to cathode). Accordingly, the circuit in FIG. 14
shows a solution for determining the control drive for S.sub.R1 and
S.sub.R2 (individually).
[0066] Alternatively, the current in C.sub.R may be sensed
bi-directionally with a single current transformer as the positions
S.sub.R1 and S.sub.R2 conduct in opposite phases either with two
anti-phase secondary windings. An example of such a circuit is
provided in FIG. 15. Here, the current in the secondary resonant
circuit is detected using first and second secondary windings of a
current transformer, which drive switches S.sub.R1 and S.sub.R2 via
drivers. When the current is positive, one rectifier is turned on,
and when the current is negative, the other is turned on.
Alternatively, a single current transformer secondary winding with
positive and negative current detection (not shown) can be
utilized.
[0067] FIG. 16 is a flow diagram illustrating a method of providing
electrical power conversion, according to one embodiment. The
functionality, in whole or in part, can be provided by hardware
and/or software, including the circuitry and other components
described in relation to FIGS. 1, 7, and 9-12.
[0068] At block 1610, a resonant converter is provided with
resonant circuitry having inductive and capacitive elements to
create electrical resonance when an input voltage is applied to the
resonant circuitry. Values of inductive and capacitive elements can
vary, depending on switching frequency, desired functionality,
and/or other factors.
[0069] At block 1620, voltage-monitoring circuitry is used to
determine when there is substantially no voltage across an
electrical switch coupled to the resonant circuitry. As illustrated
in FIGS. 1-6, the control of a switch (e.g., switch S.sub.1 of FIG.
1) can be based on the voltage across that switch. Switching
efficiency is optimized when the voltage is at or near zero.
Accordingly, at block 1630, the electrical switch is operated such
that the electrical switch is turned on when substantially no
voltage is detected across the electrical switch.
[0070] Optionally, at block 1640, the electrical switch is operated
in a mode in which the switch undergoes a plurality of on-off
cycles over a period of time before being turned off A "burst mode"
can allow the resonant converter to maintain output power while
enabling for zero-voltage switching.
[0071] FIG. 17 is a flow diagram illustrating a method of providing
electrical power conversion, according to another embodiment.
Similar to FIG. 16, the functionality shown in FIG. 17, in whole or
in part, can be provided by hardware and/or software, including the
circuitry and other components described in relation to FIGS. 1, 7,
and 9-12.
[0072] At block 1710, resonant converter is provided with resonant
circuitry having inductive and capacitive elements to create
electrical resonance when an input voltage is applied to the
resonant circuitry. Again, values of inductive and capacitive
elements can vary, depending on switching frequency, desired
functionality, and/or other factors.
[0073] At block 1720, an output voltage of the resonant converter
is rectified using a synchronous rectifier comprising a diode and
an electrical switch. Such rectification is provided in the
previously-discussed embodiments, for example, by switch S.sub.R.
As discussed above, switching efficiency for synchronous rectifiers
may be timed not only based on voltage, but on current as well.
Thus, at block 1730, the electrical switch is operated such that
the electrical switch is turned on when there is substantially no
voltage across the diode and the current flow in the diode is
positive in a direction from the anode to the cathode.
[0074] It should be appreciated that the specific blocks shown in
FIGS. 16 and 17 illustrate methods of providing electrical power
conversion according to two specific embodiments. Other embodiments
may include alternative and/or additional functionality.
Embodiments may further include functionality that is not
illustrated in FIGS. 16 and 17. Furthermore, steps may be added,
removed, and/or rearranged depending on the particular
applications. One of ordinary skill in the art would recognize many
variations, modifications, and alternatives.
[0075] It will be understood the examples and embodiments
describing "zero-voltage" switching may not operate switches at
exactly zero voltage. Different tolerances of components and
materials used in the circuitry can cause, for example, a
zero-voltage detector to vary in its detection of zero volts.
However, such a detector may detect a voltage of substantially zero
(i.e., substantially no voltage), where any existing voltage is,
within tolerances, considered zero volts for purposes of which it
is used.
[0076] It is also understood that the examples and embodiments
described herein are for illustrative purposes only and that
various modifications or changes in light thereof will be suggested
to persons skilled in the art and are to be included within the
spirit and purview of this application and scope of the appended
claims.
Controlled Burst Mode
[0077] Embodiments may potentially utilize a variety of different
methods for modulating S.sub.1 to achieve output regulation while
maintaining zero-voltage switching, including a controlled "burst
mode." An overview of such modulation techniques are described
below.
[0078] Frequency Modulation.
[0079] With frequency modulation, the higher the frequency the
lower the value of I.sub.P, which means that switching frequency
can be used to regulate output power. However, frequency regulation
typically excludes the use of a zero-voltage detector, which
(zero-voltage detecting method) is in conflict with frequency
modulation because it can change the T.sub.OFF time (duty-cycle)
and consequently the switching period (and therefore
frequency).
[0080] Maximum T.sub.ON Modulation.
[0081] In accordance with maximum T.sub.ON modulation, S.sub.1 can
be modulated such that Ton has a maximum on time proportional to
1/V.sub.IN. That is, the higher the input voltage V.sub.IN, the
shorter the length of Ton. This helps ensure that the maximum power
transfer in the circuit is relatively constant with any variation
in V.sub.IN, as I.sub.P is closely related to T.sub.ON. Although
this may be the case for maximum power transfer (maximum output
load) further T.sub.ON modulation may be necessary to regulate the
output (voltage, current, or power) to lower/light output loads.
Additional details regarding optimal T.sub.ON modulation are
provided herein below.
[0082] Controlled Burst Mode.
[0083] A maximum time for T.sub.ON may be preferentially
proportional to 1/V.sub.IN. Here, however, the switch S.sub.1 is
driven on and off for burst intervals, rather than continually. In
this way the average power transferred is reduced.
[0084] FIGS. 4 and 5 show waveforms of V.sub.S and Drive 1,
illustrating burst mode according to one embodiment. FIG. 4
illustrates waveforms of a single series (or "burst") of on/off
cycles for S.sub.1. Output power can be maintained and/or adjusted
by adjusting the frequency of bursts. FIG. 5 illustrates an example
of how bursts be provided in succession to maintain a certain
output power. Additionally, as indicated above, T.sub.ON may be
adjusted to maintain a certain output power and zero-voltage
switching.
[0085] As FIG. 4 illustrates, the on time T.sub.ON of each on/off
cycle in a burst mode can be progressively longer. These increasing
T.sub.ON periods are labeled B1, B2, B3, and B4. By increasing the
lengths of T.sub.ON as Drive 1 progresses from B1 to B4, the
resonant network is able to progressively establish resonance for
each burst without overshoot. Without using such progressive
modulation, the initial resonant peaks of V.sub.S following B1, B2
etc. would be much higher and overshoot could result, which could
be harmful for the switch S.sub.1.
[0086] It will be understood that the waveforms of FIGS. 4 and 5
are provided for illustrative purposes. In practice, various
features of the illustrated waveforms, such as number of switching
periods in a single burst period, magnitude of V.sub.S and Drive 1,
the duty cycle of each on/off cycle, and the like can vary,
depending on the configuration, power requirements, and/or other
factors.
[0087] Further, the "burst mode" function can be made user
programmable in order to easily adapt to particular application and
preserve the premium efficiencies. It can be beneficial to initiate
burst mode when zero-volt has not been achieved in resonant
circuits. This is especially the case for a wide range applications
in which the converter operates to minimize power dissipation at
light loads to achieve high efficiencies. Without a mechanism to
detect zero-volt, switches can be damaged at light loads when
switching at very high frequencies.
[0088] In accordance with some embodiments, the switch S.sub.1 can
be a GaN transistor, such as a MOSFET, MESFET, and the like. In
such embodiments, the switch S.sub.1 can be modulated at much
higher frequencies than similar silicon-based switches.
Higher-frequency switching allows for a reduction in size of
magnetic and capacitive components, which can reduce the overall
size and cost of the power adapter. In some embodiments, for
example, the order of magnitude of the switching frequencies can be
in the megahertz, while bursting frequencies can be in the tens of
kilohertz.
[0089] T.sub.ON Modulation.
[0090] As a variation to the burst mode described above, T.sub.ON
can be controlled for lower output loads to achieve the required
set-point and regulation. (That said, some embodiments may us
T.sub.ON modulation together with other modulation techniques.)
Although T.sub.ON modulation can be successful in the output power
range of most applications, at lower output loads the T.sub.ON time
will be small. A smaller T.sub.ON time can result in a smaller
current I.sub.P. And in certain circumstances there may not be
enough circulating energy in the resonant network for V.sub.S to
return to zero-volts. FIG. 6 helps illustrate this dilemma, as well
as a solution that can be implemented, according to some
embodiments.
[0091] FIG. 6 shows waveforms illustrating how, using T.sub.ON
modulation, Drive 1 reduces T.sub.ON to reduce output power, and
how this can result in insufficient current I.sub.S1 (not shown) to
drive V.sub.S back to zero. Here, a zero-voltage detection circuit
can be used to recognize when zero-voltage switching fails to occur
and help the circuit prevent efficiency loss and potential damage
to the switch S.sub.1 that could result from switching when V.sub.S
is not at or near zero.
[0092] In the illustrated example, the zero voltage detection
signals can be monitored to determine when there is failure of
V.sub.S to return to zero. If, for example, the zero-volt-detection
signal is not received for a number of switching cycles, Drive 1 is
disabled for a period of time so that when Drive 1 is enabled
again, the circuit requires more instantaneous power to regulate
the average power to the required level. The increase in power
requirements enables Drive 1 to have longer T.sub.ON times, which
allow for zero-voltage switching again for the next number of
cycles.
[0093] In other words, a burst mode can be initiated cyclically
when zero-voltage switching of Vs is detected to have failed.
[0094] Methods for detecting zero-voltage enable adaptive control
to enhance burst mode operation and preserve converter efficiency
at all loads--especially light loads--and prevent potential switch
damage.
[0095] FIG. 18 shows an example controller circuit for providing
S.sub.1 control, according to one embodiment. A 555 timer is used
in a stable mode to generate circuit resonant frequency. R.sub.A,
R.sub.B and C values determine the frequency. The 555 timer's
output frequency and zero-volt detect signals are fed to an OR gate
which then it would trigger the timer output. A zero-volt detect
signal is necessary to set the timer output high. Output of the 555
timer is fed to Driver 1. Output voltage is also sensed and fed to
a comparator for voltage regulation. The 555 timer is therefore
reset in case the output voltage is higher than reference or other
protection signals are activated low.
Controlled Active Clamping
[0096] A controlled active clamping technique can be used to hold
peak resonant voltage at pre-determined levels in order to force
zero-volt switching and prevent possible switch damage due to
stress voltage, as well as eliminate unnecessary clipping that
leads to excess losses and converter inefficiency. In an isolated
converter, when transformer peak reset voltage is significantly
larger than the input voltage, a clamp circuit is activated at
predetermined peak voltage. Modulation can reduce excess loss in
the clamp circuit under varying load conditions. The modulation of
peak voltage allows for efficient power transfer and controllable
output voltage regulation.
[0097] Typically snubbing and clamping circuits, such as
resistor-capacitor-diode (RCD) circuits, are used on switches to
limit voltage spikes to reduce component stress. This leads to
extra circuit dissipation, and thus power savings can be realized.
In such circuits, a voltage spike is caused by the energy stored in
the transformer's leakage inductance of an isolated resonant
circuit, when the switch turns off and abruptly halts current flow
in the primary winding. The first step to reducing both the voltage
spike and the loss in the clamp is to design a transformer with
minimal leakage inductance, which may not be ideal for a resonant
converter. The resonance between this inductance and the parasitic
capacitance of the switch produce large voltage stress as well as
losses, therefore decreasing converter efficiency. The clamp
resistance can be increased to further reduce the loss, but doing
so also increases the magnitude of the voltage spike. During the
reset portion of the switching cycle, the reflected output voltage
is impressed across the clamp resistor leading to extra loss. Using
a higher voltage switch provides more margin for the voltage spike
and allows for a much larger resistor. However, an increased
voltage rating results in higher on-resistance which leads to lower
efficiency at high loads. When a controller is operating in burst
mode, the clamp circuit discharges between ON states. If the clamp
capacitor is too large, excess energy is stored and dissipated
during the OFF state. In some situations, the clamp capacitor may
not fully discharge before the next ON state begins.
[0098] Embodiments can utilize an active clamping technique rather
than an RCD clamp circuit. A non-dissipative LC plus clamp switch
circuit, for example, can to force the transformer leakage
inductance energy to oscillate on input as reactive power and/or
transfer the energy to load as real active power. In either case,
the energy is not dissipated in a resistor and the losses are
decreased. Benefits to an active clamp circuit include the ability
to transfer energy under wide line and load variations. The
technique is suitable for resonant circuits including Power-Factor
Correction (PFC) circuits. Transformer reset is accomplished with
an active clamp circuit consisting of switch and a capacitor
working with transformer leakage inductance. Active clamp circuit
works as a controllable current source, so as to regulate power
according to load variations.
[0099] This arrangement offers many benefits. For example, the duty
cycle can go higher than 50%, resulting in higher turns ratio,
lower primary currents and secondary voltages, and smaller output
inductor. Also, the voltage stress on the primary switch remains
relatively constant over the full input voltage range, leading to
better overall efficiency. In addition, zero-volt switching is
possible with this approach, which can lead to further size
reduction by increasing the switching frequency.
[0100] FIGS. 19-21 are schematics of examples active clamp circuit
that can be utilized in embodiments for isolated converters. Values
for the various components involved, alterations to the
architecture, and other variations can vary, depending on desired
functionality and will be understood by a person of ordinary skill
in the art. As shown, the circuits of FIGS. 19 and 20 utilize a
comparator and driver to determine when the active clamp switches
on, which, as described above, can occur at any of a variety of
desired voltages, depending on application and desired
functionality (e.g., 500 V, 800 V, etc.). The circuit of FIG. 20,
on the other hand, illustrates how clamp switch can be fed by a
winding of the transformer between nodes 2 and 3. Thus, the turn-on
voltage for the clamp switch can be determined by number of
windings between nodes 2 and 3. The circuit in FIG. 21 therefore
illustrates how active clamping can be done with passive
components.
[0101] FIG. 22 illustrates a technique applied to a PFC circuit
similar to the non-isolated circuit of FIG. 1. In a low-power,
low-current application, using a small magnetizing inductance to
achieve zero-volt switching can be more appropriate. Resonance
between the leakage inductance and the clamp capacitance takes
place when the transformer is reset. Magnetizing inductance is
designed together with the switching frequency in order to provide
zero-volt switching at high input voltage and keep the size and
losses in the transformer small.
[0102] When transformer peak reset voltage is significantly larger
than the input voltage, the clamp circuit can be activated at a
predetermined peak voltage. Modulation can reduce excess loss in
the clamp circuit under varying load conditions. Modulation of peak
voltage allows for efficient power transfer and controllable output
voltage regulation. Having the ability to control the peak voltage
level clamping also allows for zero-volt switching of S.sub.1 and
also the clamp switch. Clamp circuit and burst mode control force
zero-volt switching under varying load conditions, especially light
loads.
[0103] The voltage across S.sub.1 can be sensed using techniques
similar to the methods shown in FIGS. 19-2120. Other methods can
also be applied for sensing voltage. The sensed signal is compared
to a reference voltage by comparator in the clamping circuit. At a
predetermined OFF state voltage (i.e., a threshold voltage),
clamping switch is turned on, and excess resonant energy is put
back to the V.sub.BUS. The adaptive nature of the circuit makes it
possible to compensate for load and environmental variations to
achieve higher efficiencies.
[0104] Having described various embodiments of the invention, it
will be understood that the examples and embodiments described
herein are for illustrative purposes only and that various
modifications or changes in light thereof will be suggested to
persons skilled in the art and are to be included within the spirit
and purview of this application and scope of the appended
claims.
* * * * *