U.S. patent application number 13/385411 was filed with the patent office on 2012-08-16 for flexible modular hierarchical adaptively controlled electronic-system cooling and energy harvesting for ic chip packaging, printed circuit boards, subsystems, cages, racks, it rooms, and data centers using quantum and classical thermoelectric materials.
Invention is credited to Lester F. Ludwig.
Application Number | 20120204577 13/385411 |
Document ID | / |
Family ID | 46635825 |
Filed Date | 2012-08-16 |
United States Patent
Application |
20120204577 |
Kind Code |
A1 |
Ludwig; Lester F. |
August 16, 2012 |
Flexible modular hierarchical adaptively controlled
electronic-system cooling and energy harvesting for IC chip
packaging, printed circuit boards, subsystems, cages, racks, IT
rooms, and data centers using quantum and classical thermoelectric
materials
Abstract
A system for adaptive cooling and heat gathering, the method
comprising a thermoelectric device capable of acting as a
thermoelectric cooler in a heat pump mode and as a thermoelectric
generator in a heat engine operating mode, a control system
received provided input signals and providing a control output; and
switching electronics controlled by the control output and
connected to the thermoelectric device, wherein the control system
controls the operating mode of the thermoelectric device responsive
to provided input signals.
Inventors: |
Ludwig; Lester F.; (Belmont,
CA) |
Family ID: |
46635825 |
Appl. No.: |
13/385411 |
Filed: |
February 16, 2012 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
61443701 |
Feb 16, 2011 |
|
|
|
Current U.S.
Class: |
62/3.3 |
Current CPC
Class: |
F25B 21/04 20130101;
F25B 47/006 20130101; H01L 23/38 20130101; H01L 35/28 20130101;
H01L 2924/0002 20130101; H01L 2924/00 20130101; F25B 21/02
20130101; H01L 2924/0002 20130101; F25B 2321/0212 20130101; G06F
1/20 20130101 |
Class at
Publication: |
62/3.3 |
International
Class: |
F25B 21/02 20060101
F25B021/02 |
Claims
1. A system for adaptive cooling and heat gathering, the method
comprising: a thermoelectric device capable of acting as a
thermoelectric cooler in a heat pump mode and as a thermoelectric
generator in a heat engine operating mode; a control system
received provided input signals and providing a control output; and
switching electronics controlled by the control output and
connected to the thermoelectric device, wherein the control system
controls the operating mode of the thermoelectric device responsive
to provided input signals.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] Pursuant to 35 U.S.C. .sctn.119(e), this application claims
benefit of priority from Provisional U.S. Patent application Ser.
No. 61/443,701, filed Feb. 16, 2011, the contents of which are
incorporated by reference.
COPYRIGHT & TRADEMARK NOTICES
[0002] A portion of the disclosure of this patent document may
contain material, which is subject to copyright protection. Certain
marks referenced herein may be common law or registered trademarks
of the applicant, the assignee or third parties affiliated or
unaffiliated with the applicant or the assignee. Use of these marks
is for providing an enabling disclosure by way of example and shall
not be construed to exclusively limit the scope of the disclosed
subject matter to material associated with such marks.
BACKGROUND OF THE INVENTION
[0003] In the normal course of operation, the electronic Integrated
Circuit ("IC") chips that are comprised by computer hardware
generate heat. In CMOS and related technologies, the heat
generation is a function of the rate of state transitions per unit
time, so as software tasks are handled there are frequent, rapid
increases and decreases of emitted heat from a given chip as
computers operate. Additionally, as heat builds up in CMOS and
other semiconductor devices, leakage currents typically increase,
creating yet more heat.
Data Centers
[0004] The amount of heat generated is considerable even when only
one computer is involved. In the case of server farms and other
data centers the heat generation problem assumes vast proportions.
Most of the heat in data centers is removed by costly and
relatively inefficient means that also consumes yet more energy and
generates yet more heat. Unlike many industrial processes where
waste heat is harnessed for site-based energy reuse of
power-cogeneration, most of the heat in data centers is simply
dissipated into the environment.
[0005] Providing a cost-effective, efficient and practical solution
to this IC-chip and electronic component generated heat is
therefore crucial. As cloud computing, search, download, and other
network services radically increase centralized computing demands,
creating needs for computation and data centers to becoming larger
and larger, the need for massive facilities increases, causing the
magnitude of the heat generation problem to become increasingly
urgent.
[0006] The increasing demand and trend to ever-larger and more
robust computer data centers necessitates that the industry
practice of utilizing cheap, passive traditional thermal design be
replaced with more advanced thermal technologies.
Leveraging Data-Center Turn-over to Rapidly Introduce Cooling and
Energy Harvesting Technologies into Data Centers
[0007] Fortuitously, new inventions (such as the present invention)
that address these problems all or in part can conveniently take
reliance upon Moore's Law (which observes that approximately every
18 months computer power doubles while the cost roughly halves) and
ongoing changes in computer server architecture. The resulting
combined forces of natural degradation and functional obsolescence
force computer hardware (data center hardware in particular) to
naturally be replaced or upgraded on a periodic basis. As each
hardware replacement cycle brings in new computer hardware, this
allows new cooling technologies to be introduced.
Approaches to Reduce Heat Production in Integrated Circuit
Operation
[0008] A large number and wide variety of approaches are currently
under research, development, and deployment to reduce heat
production in integrated circuit operation. A brief survey of these
very active areas can be found in [1], Chapter 1, and the material
there is summarized in the list below:
[0009] Dynamic Power Consumption [0010] Reducing Capacitance [0011]
Reducing Switching Activity [0012] Reducing Clock Frequency [0013]
Reducing Supply Voltage
[0014] Static Power Consumption [0015] Leakage currents arise from
the flow current across a transistor even when the transistor is in
an OFF state. [0016] Gate-oxide leakage is dependent on the
thickness of the oxide that is used to insulate the gate from the
substrate. As process technologies are decreasing, so is the
gate-oxide thickness. Higher k dielectrics will have to be used to
offset subthreshold leakage. Flow of current between the drain and
source of a transistor when the voltage is below threshold.
[0017] Circuit-Level Power Consumption in Integrated circuits
[0018] Transistor Reordering [0019] Half-Frequency and Half-Swing
Clocks [0020] Low-Power Flip-Flop Design [0021] Technology mapping
automates the process of producing a power-optimized circuit in
order to minimize the total power consumption. [0022] Bus Inversion
[0023] Crosstalk Reduction [0024] Low-Swing Buses [0025] Segment
the bus into multiple groups that allow the majority of the buses
to be powered down while only the active buses are in use. [0026]
Adiabatic circuits are a novel concept that reuses the electrical
charge dissipated from one wire and recycles it for use in another
wire [0027] replace the traditional shared-bus approach with a more
generic interconnect network.
[0028] Low-Power Memory Design [0029] Partitioning Memory [0030]
Specialized Power-Friendly Caches [0031] Filter cache [0032] Trace
cache [0033] Adaptive caches [0034] Drowsy cache
[0035] The present invention is not directed in these directions,
but rather as to what to be done with the electronic-component heat
that must be generated, regardless of its source or cause, The
present invention addresses this on several fronts: [0036] Improved
heat transport, aggregation, management; [0037] Component thermal
environment improvement; [0038] Consideration of the full heat
transport hierarchy; [0039] Adaptive opportunistic energy
harvesting; [0040] Leveraging reciprocal (heat transfer, heat to
electrical current conversion) properties of both classical
semiconducting thermoelectric devices and quantum-process
thermoelectric devices; [0041] Adaptively switching modes and/or
multiplexing between cooling mode, energy-harvesting mode, and
temperature sensing modes; [0042] In switching among modes and in
general operation, including consideration of and/or compensation
for the dynamic behavior of the thermoelectric devices employed;
[0043] Various control systems to manage local and system-wide
operation.
Computer System Cooling Technologies
[0044] A large number and wide variety of approaches are currently
under research, development, and deployment to reduce, manage, and
handle heat build-up in integrated computer systems and data
centers. A survey of the many well-known classical and contemporary
techniques for this these at the board and chassis level can be
found in [2] and the references therein. A representative treatment
of the many well-known and contemporary techniques for this these
at the data center level can be found in [3]-[4], these largely
involving forced air and chiller technologies. The present
invention provides economical practical near-term supplements,
enhancements, and alternatives to these approaches, including for
example the invention innovations listed in the previous subsection
and called out in bold font in FIG. 3.
Chip Cooling Technologies
[0045] A large number and wide variety of approaches are currently
under research, development, and deployment employing
thermoelectric devices. A survey of the many well-known classical
and contemporary techniques for these employing semiconducting
thermoelectric devices can be found in [5] and the references
therein. A brief treatment of techniques and properties of
quantum-well thermoelectric devices can be found in [6] and the
references therein. Treatment of techniques and properties of Avto
metal thermoelectric devices can be found in [7]-[10] and the
references therein. Additionally, micro-droplet microfluidic
cooling is also currently under research and development, some
employing some minor interworking with thermoelectric devices.
Treatment of such approaches employing planar (two-dimensional)
micro-droplet transport can be found in [1] and the references
therein, and approaches employing three-dimensional and
multiple-layer micro-droplet transport are taught in co-pending
U.S. Patent Application 61/599,643.
[0046] As background, FIG. 1a depicts an exemplary computer
processor chip fitted with a traditional air-cooled finned heat
sink. In practice the depicted cooling fins can be much larger than
depicted here. One or more fan(s), each within or attached to a
chassis which envelopes an associated computer processor chip, are
used to force incoming air through and/or over the heat sink. The
forced air absorbs heat radiating from the fins of the heat sink,
transporting it away from the chip and thus preventing a higher
degree of heat buildup. FIG. 1b shows an arrangement like that of
FIG. 1a but fitted with a dedicated fan to increase the air flow
through the air-cooled finned heat sink. FIG. 1c shows an
arrangement like that of FIG. 1b but with a thermoelectric cooler
layer provided to increase the heat flow from the Integrated
Circuit chip to the air-cooled finned heat sink. FIG. 1d shows an
arrangement like that of FIG. 1c but with a dedicated fan to
increase the air flow through the air-cooled finned heat sink.
[0047] FIGS. 1e-1g depicts various alternative chip cooling
arrangements wherein heat pipes are used to transport heat away
from a computer processor chip. FIG. 1e depicts an exemplary heat
pipe system employed in a laptop computer. FIG. 1f depicts an
exemplary heat pipe system wherein the heat pipe connects
transferred heat produced at the chip to a fan-cooled heat-sink.
FIG. 1g depicts an exemplary arrangement heat pipe system wherein a
principal heat sink is connected via heat pipes with expansion heat
sinks.
Hierarchy of Heat Transfer in Data Center Environments
[0048] As yet further background, FIGS. 2a-2g illustrate a
hierarchy of environments involved in heat transfer.
[0049] FIG. 2a shows a group of computer chips and related
components (for example, voltage regulator components), for
example, on a common printed circuit board. These are the primary
source of heat generation, although heat is also generated by other
elements such as power supply transformers and circuitry, fans,
compressors, chillers, pumps, etc. included in cooling systems.
[0050] FIG. 2b depicts the two or more such groups of computer
chips and related components, for example, sharing a common printed
circuit board.
[0051] FIG. 2c illustrates two or more printed circuit boards,
either as in FIG. 2a or FIG. 2b, which together are comprised by a
common subsystem. FIG. 2d illustrates two or more subsystems, such
as depicted in FIG. 2a or otherwise, which together are comprised
in a common chassis. Such a chassis can be part of, for example, a
computer server.
[0052] FIG. 2e shows a plurality of chassis comprised by a cage.
Such arrangements are used in commercial "blade server" product
configurations.
[0053] FIG. 2f depicts a plurality of cages, such as depicted in
FIG. 2e or otherwise, fitted into a rack. Such rack configurations
are used endemically in data centers. FIG. 2g represents a cluster
of racks as commonly used in data centers.
OVERVIEW OF THE INNOVATION
[0054] The invention comprises a collection of interworking
innovations. These include: [0055] Systems and methods for
combining Peltier-effect heat transport and Seebeck-effect energy
harvesting for use at thermal interfaces in heat gathering and
transfer structures; [0056] Novel structures for multiple-mode
thermoelectric devices providing heat transfer, heat-to-electricity
conversion, temperature flux measurements for use in interfacing
integrated circuit packages and in creating active heat pipes;
[0057] Arrangements in the above permitting simultaneous mixed-mode
operation [0058] Automatic control for optimizing multiple mode and
mixed-mode operation in local or hierarchical contexts; [0059]
Pulse-width modulation and other duty-cycle control to prevent
Peltier cooling induced condensation; [0060] Use of traditional,
contemporary, and emerging quantum-process and nanomaterial
techniques for radical efficiency improvements in Peltier-effect
heat transport and Seebeck-effect thermoelectric energy harvesting;
[0061] Configurable, reconfigurable, or real-time-controlled
selective operation of combined Peltier-effect heat transport and
Seebeck-effect thermoelectric energy harvesting; [0062] Systems and
methods for a modular-structure heat gathering and heat transfer
infrastructure designed, for example, to work with existing
familiar board, blade, cage, rack, data center, and building
infrastructure, the systems and methods supporting optional
advantageous additional features of: [0063] Closed heat transport
systems within each module terminating in a thermal interface which
can be coupled to external heat transport stage, air cooling,
energy transformation, or other alternatives; [0064] Each module
equally usable in isolation or as part of a hierarchy, allowing
wide range of gradual phase-in deployments, trials, and strategies;
[0065] Internal energy harvesting within the module; [0066]
Hierarchical heat-gathering structures with dry thermal-transfer
interfaces between pairs of closed level-internal cooling fluid and
heat pipe structures within modules at each layer and fan backup
for isolated operation or parent-level failure recovery; [0067]
Hierarchical heat-to-electricity energy harvesting structures with
provisions for both local use of heat-harvested electricity as well
as provisions for exporting power into hierarchical or peer
arrangements; [0068] Hierarchical control structures capability of
working in isolation or coordinating with other control systems in
hierarchical or peer arrangements.
[0069] FIG. 2h depicts a representation implying the innovation can
be utilized when a data center occupies one or more whole or
partial floors of a building and/or a vertical riser within a
building. FIG. 2i depicts a representation implying the innovation
can be applied when a data center occupies an entire building. FIG.
2j depicts a representation implying the innovation can be utilized
when a data center is comprised of a campus of buildings.
[0070] FIG. 3 depicts an exemplary non-limiting, non-characterizing
view of various aspects of the invention. The invention combines
many new innovations (denoted by bolded font) together with novel
utilizations and adaptations of known art (denoted by unbolded
font).
SUMMARY OF THE INVENTION
[0071] Features and advantages of the invention will be set forth
in the description which follows, and in part will be apparent from
the description, or may be learned by practice of the invention.
The objectives and other advantages of the invention will be
realized and attained by the structure particularly pointed out in
the written description and claims hereof as well as the appended
drawings.
Aspects of the Invention Involving Heat Gathering at IC-Chip and
Board Level
[0072] In an embodiment, the invention provides for the use of
microfluidic micro-droplet heat transport at the IC-chip package
level.
[0073] In an embodiment, the invention provides for the use of
microfluidic micro-droplet heat transport at the multilayer Printed
Circuit Board ("PCB") level.
[0074] In an embodiment, the invention provides for the collection
of heat from at least one integrated circuit package.
[0075] In an embodiment, the invention provides for the collection
of heat from at least a plurality of computer systems comprised by
a data center.
[0076] In an embodiment, a thermal interface to at least one chip
is used to collect heat and direct it to a heat sink in thermal
contact with a circulating cooling fluid.
[0077] In an embodiment, the circulating cooling fluid possesses a
high heat-carrying capacity.
[0078] In an embodiment, the invention provides for thermal
interfaces with a circulating cooling fluid to be designed to
easily connect mechanically with an associated chip.
[0079] In an embodiment, the invention provides for the thermal
interface to be designed to direct heat to a fan where air
convection can be utilized to remove the heat.
[0080] In an embodiment, the invention provides for the thermal
interface to be designed to feed the heat into a heat transfer
interface that utilizes a circulating coolant to remove the heat,
or to both.
[0081] In an embodiment, the invention provides for one or both
sides of the thermal interface to be constructed of thermoelectric
materials to most efficiently collect the heat on one side and to
most efficiently utilize the cooling fan or fluid on the other
side.
[0082] In an embodiment, the invention provides for one portion of
the chip-generated heat to be spatially transferred and another
portion of the heat to be energy harvested.
[0083] In an embodiment, the invention provides pulse-width
modulation and other duty-cycle control to prevent Peltier cooling
induced condensation and icing.
[0084] In an embodiment, the invention provides for the use of
high-efficiency thermoelectric devices comprising quantum-well
materials.
[0085] In an embodiment, the invention provides for the use of
high-efficiency thermoelectric devices comprising Avto Metals.
Aspects of the Invention Involving Energy Harvesting
[0086] In an embodiment, the invention provides for concerted
effort to convert as much heat to electricity at the local chip
level as possible.
[0087] In an embodiment, the invention provides for the placement
of thermoelectric material inside, on top of, on the bottom of or
around a chip package.
[0088] In an embodiment, the invention provides for repeated
hierarchical steps of heat transfer from thermal sources, conducted
through a heat exchange or other thermal interface, and transferred
to a thermal sink.
[0089] In an embodiment, the hierarchy can comprise use of the heat
gathered at a thermal sink at one hierarchy level to serve as the
heat of the heat source in an adjacent level in the hierarchy.
[0090] In an embodiment, at any one or more places in the
hierarchy, energy harvesting operations can be introduced.
[0091] In an embodiment, energy harvesting operations convert heat
into electricity.
[0092] In an embodiment, electricity created by energy harvesting
operations is used to provide power for current or future heat
transfer operations.
[0093] In an embodiment, an energy harvesting operation improves
the efficiency of the inventive cooling system.
[0094] In an embodiment, each energy harvesting operations improve
the effectiveness of the inventive cooling system.
Aspects of the Invention Involving Combining Energy Harvesting and
Heat Transport
[0095] In an embodiment, the invention provides for the use of
reciprocal thermoelectric devices capable of operating in either a
thermoelectric cooler or a thermoelectric electric current
generator as determined by imposed thermal conditions and
electrical connections to the reciprocal thermoelectric device.
[0096] In an embodiment, the invention provides for at least one of
the thermoelectric devices can serve as a temperature sensor
[0097] In an embodiment, the invention provides for the mode of a
given thermoelectric device is switched over time. As one example,
a given thermoelectric device can be a thermoelectric cooler one
moment and a temperature sensor at another moment. As another
example, a given thermoelectric device can be a thermoelectric
electric current generator one moment and a temperature sensor at
another moment. As yet another example, a given thermoelectric
device can be a thermoelectric cooler one moment and a
thermoelectric electric current generator at another moment. As
still another example, a given thermoelectric device can be a
thermoelectric cooler one moment, a temperature sensor at another
moment, and a thermoelectric electric current generator at yet
another moment.
[0098] In an embodiment, the invention provides for a control
system that selects the mode of operation of at least one
reciprocal thermoelectric device, the selection made responsive to
the state of the system, time, a measurement condition, or some
combination of these.
[0099] In an embodiment, the invention provides for a control
system to include consideration of the dynamic behavior of at least
one type of thermoelectric device.
[0100] In an embodiment, the invention provides for a control
system to include compensation for the dynamic behavior of at least
one type of thermoelectric device.
[0101] In an embodiment, the invention provides for a control
system that selects the mode of operation of at least one
reciprocal thermoelectric device to include consideration of and/or
compensation for the dynamic behavior of the reciprocal
thermoelectric device.
Aspects of the Invention Involving Heat Migration out of a
Subsystem
[0102] In an embodiment, the invention comprises one or more
heat-aggregating system and/or one or more heat-aggregating
subsystems.
[0103] In an embodiment, the invention provides for heat that can
not be efficiently or effectively harvested for energy to be
dispersed via fan(s) at one or more suitable location(s) within the
system.
Aspects of the Invention Involving Modular Hierarchical
Structure
[0104] In an embodiment, the invention provides for a modular
product hierarchy that can be designed to meet market need and
demand.
[0105] In an embodiment, the invention provides for heat that can
not be efficiently or effectively harvested for energy to be
dispersed via fan(s) at one or more suitable location(s) within the
system.
[0106] In an embodiment, the invention comprises one or more
heat-aggregating system and/or one or more heat-aggregating
subsystems.
[0107] In an embodiment, the invention comprises a daisy-chain heat
transfer arrangement employing closed systems of circulating fluids
with dry thermal interfaces among them for use in a hierarchical or
peer arrangement.
Aspects of the Invention Involving Interconnection of Heat Transfer
Subsystems within a Cooling Hierarchy
[0108] In an embodiment, the invention provides for the collection
of heat from at least a number of computer chips forming a
computing system.
[0109] In an embodiment, each module (board, chassis, cage, rack,
rack cluster, etc.), comprises at least two separate closed
circulating fluid cooling systems that are thermally linked by
thermal-transfer coupling elements.
[0110] In an embodiment, thermal-transfer coupling elements
comprise pressure-contact.
[0111] In an embodiment, thermal-transfer coupling elements
comprise fastener arrangements.
[0112] In an embodiment, thermal-transfer coupling elements
comprise a threaded structure.
[0113] In an embodiment, thermal-transfer coupling elements
comprise a quick-lock structure.
Aspects of the Invention Involving Product Evolution and Phased
Deployment
[0114] In an embodiment, the invention provides for a modular
product hierarchy that can be designed to meet market need and
demand.
[0115] In an embodiment, the invention provides for phased
replacement as required due to the end of operating life, adequate
degradation, or functional obsolescence.
[0116] In an embodiment, until such replacement or upgrade is
enacted, the invention provides for incremental implementation via
incremental retrofit of computers, chip(s) within individual
computers, cages, racks, etc.
[0117] In an embodiment, modular features used to implement
scalability of the innovation can be implemented in such a way that
each modular level can operate in a stand-alone mode, for example,
relying on backup fans to expel excess heat. This can also provide
a failsafe backup for heat dispersion should some part of a
hierarchical deployment fail.
[0118] In an implementation or a deployment, aspects of the
invention can be deployed or implemented at any one or more levels
as determined appropriate in a given situation or management
decision.
Aspects of the Invention Involving Hierarchical Control
Structures
[0119] In a further aspect of the invention, the invention provides
a hierarchical multiple-level control system comprises a plurality
of subsystems, each with their own associated control system, each
of which (1) can operate in isolation and (2) can be interconnected
or networked with additional subsystems associated with other
hierarchical levels.
[0120] In a further aspect of the invention, a hierarchical
multiple-level control system comprises a plurality of subsystems,
each with their own control system, that can operate in isolation,
but when interconnected or networked with additional subsystems
associated with other hierarchical levels, each subsystem will
assume their respective role in the hierarchy with respect to
(those) additional subsystems.
[0121] In an embodiment, the invention provides for hierarchical
multiple-level control system to include linear control systems,
therein permitting the additive control of at least one controller
state variables of one subsystem by control signals generated by or
associated with at least one other subsystem.
[0122] In an embodiment, the invention provides for hierarchical
multiple-level control system to include bilinear control systems,
therein permitting the multiplicative control of at least one
controller state variables of one subsystem by control signals
generated by or associated with at least one other subsystem.
[0123] In an embodiment, the invention provides for hierarchical
multiple-level control system to include bilinear control systems,
therein permitting both (1) additive control of at least one
controller state variables of one subsystem by control signals
generated by or associated with at least one other subsystem and
(2) multiplicative control of at least one controller state
variables of one subsystem by control signals generated by or
associated with at least one other subsystem.
[0124] In an embodiment, the invention provides for hierarchical
multiple-level control system to include synthetic hysteresis.
BRIEF DESCRIPTION OF THE DRAWINGS
[0125] The above and other aspects, features and advantages of the
present invention will become more apparent upon consideration of
the following description of preferred embodiments taken in
conjunction with the accompanying drawing figures.
[0126] FIG. 1a depicts an exemplary computer processor chip fitted
with a traditional air-cooled finned heat sink.
[0127] FIG. 1b shows an arrangement like that of FIG. 1a but fitted
with a dedicated fan to increase the air flow through an individual
heat sink.
[0128] FIG. 1c depicts an exemplary heat pipe system employed in a
laptop computer.
[0129] FIG. 1d depicts an exemplary heat pipe system wherein the
heat pipe connects transferred heat produced at the chip to a
fan-cooled heat-sink.
[0130] FIG. 1e depicts an exemplary arrangement heat pipe system
wherein a principal heat sink is connected via heat pipes with
expansion heat sinks.
[0131] FIG. 2a shows a group of computer chips, for example, on a
common printed circuit board.
[0132] FIG. 2b depicts the two or more such groups, for example,
sharing a common printed circuit board.
[0133] FIG. 2c illustrates two or more printed circuit boards,
either as in FIG. 2a or FIG. 2b, which together are comprised by a
common subsystem.
[0134] FIG. 2d illustrates two or more subsystems, such as depicted
in FIG. 2a or otherwise, which together are comprised in a common
chassis.
[0135] FIG. 2e shows a plurality of chassis comprised by a
cage.
[0136] FIG. 2f depicts a plurality of cages, such as depicted in
FIG. 2e or otherwise, fitted into a rack.
[0137] FIG. 2g represents a cluster of racks as commonly used in
data centers.
[0138] FIG. 2h depicts a representation implying the innovation can
be utilized when a data center occupies one or more whole or
partial floors of a building and/or a vertical riser within a
building.
[0139] FIG. 2i depicts a representation implying the innovation can
be applied when a data center occupies an entire building.
[0140] FIG. 2j depicts a representation implying the innovation can
be utilized when a data center is comprised of a campus of
buildings.
[0141] FIG. 3 depicts an exemplary non-limiting, non-characterizing
view of various aspects of the invention.
[0142] FIG. 4a depicts a general thermodynamics passive heat
transfer process from a hot body to a broader environment.
[0143] FIG. 4b depicts a heat pump arrangement for active heat
transfer process from a hot body to a broader environment. Energy
is applied over time to the heat pump (amounting to applied work)
and consumed in the heat-pumping process.
[0144] FIG. 4c depicts a heat engine arrangement for active heat
transfer process from a hot body to a broader environment. Energy
is harvested over time by the heat engine (amounting to harvested
work) and consumed (at least in part) by external processes.
[0145] FIG. 5a depicts an exemplary thermal integration and
transfer abstraction.
[0146] FIG. 5b depicts an exemplary thermal resistive
abstraction.
[0147] FIG. 5c depicts an exemplary thermal resistive series.
[0148] FIG. 5d depicts an exemplary thermal diode abstraction.
[0149] FIG. 6a shows heat generated from a wafer within an
exemplary computer chip package directed through a thermal
interface provided on top of the chip. Heat is transferred into a
heat sink where the heat can be dissipated using a circulating
cooling fluid.
[0150] FIG. 6b depicts an alternative situation where the heat
generated from a wafer within the computer chip is directed through
a thermal interface provided on the top of the chip. Although not
explicitly shown, such a thermal interface can be incorporated into
the chip packaging by the chip manufacturer.
[0151] FIG. 6c depicts an exemplary arrangement wherein a computer
chip is provided with a thermal interface comprising cooling fins.
Although not explicitly shown, such a thermal interface can be
incorporated into the chip packaging by the chip manufacturer.
[0152] FIG. 6d depicts an exemplary fluidic cooling conduit in
thermal contact with a thermal interface. Although not explicitly
shown, such a thermal interface can be incorporated into the chip
packaging by the chip manufacturer.
[0153] FIG. 7a provides an example depiction of the Peltier effect
wherein an electric current is used to create a temperature
gradient in an arrangement involving dissimilar materials. In this
figure the arrangement is depicted in terms of N-type and P-type
semiconductor materials, but the Peltier effect also pertains to
(and was originally discovered in the form of) junctions of
dissimilar metals.
[0154] FIG. 7b provides an example depiction of the Seebeck effect
wherein a temperature gradient is used to create an electric
current in an arrangement involving dissimilar materials. In this
figure the arrangement is depicted in terms of N-type and P-type
semiconductor materials, but the Seebeck effect also pertains to
(and was originally discovered in the form of) junctions of
dissimilar metals.
[0155] FIG. 7c depicts an electrical symbol that will be used for
an electrical instance of an arrangement involving an individual
junction of dissimilar materials, the arrangement configured to
provide one or more thermoelectric functions employing phenomenon
such as the Peltier effect, Seebeck effect, or other adapted,
related, or alternative thermoelectric effects. An instance of such
an arrangement will be termed a "thermoelectric device."
[0156] FIG. 7d depicts an electrical series connection of a
plurality of thermoelectric devices. (In the literature such an
electrical arrangement combined with an associated thermal
arrangement is called a "thermopile.")
[0157] FIG. 7e depicts a plurality of electrical series connections
of a plurality of thermoelectric devices.
[0158] FIG. 7f depicts an exemplary physical array of
thermoelectric cells such as the example provided in FIGS.
7c-7e.
[0159] FIG. 8a depicts an example accounting of various
material-based and junction-based electrical resistance aspects
inherent in a thermoelectric device.
[0160] FIG. 8b depicts a series-resistance aggregation of these
various material-based and junction-based electrical resistance
aspects inherent in a thermoelectric device into a composite
equivalent electrical resistance.
[0161] FIG. 8c depicts an abstract electronics representation of
the equivalent electrical resistance of a thermoelectric device or
thermoelectric cell operating in Peltier mode.
[0162] FIG. 8d depicts an abstract electronics representation of
voltage emf generated by Seebeck effect in a thermoelectric device
or cell and the associated (Thevenin-equivalent) series resistance
of the thermoelectric device or thermoelectric cell, the resulting
arrangement connected to a load resistance resulting in a
current/flowing around the resulting electrical loop.
[0163] FIG. 9a (adapted from
http://knol.google.com/k/andre-szykier/thermo-electric-energy/3sqds0076vq-
oz/2#) depicts a graphical representation of the relationship
between reduced current efficiency and relative current density for
several example materials applicable for use as one of the
dissimilar materials in a thermoelectric device.
[0164] FIG. 9b (adapted from
http://knol.google.com/k/andre-szykier/thermo-electric-energy/3sqds0076vq-
oz/2#) depicts a constant-temperature linear current-versus-voltage
("Ohms Law") and quadratic power-versus-current ("Joule's Law")
curves for an exemplary Bi2Te3 thermopile.
[0165] FIG. 9c (adapted from
http://knol.google.com/k/andre-szykier/thermo-electric-energy/3sqds0076vq-
oz/2#) depicts compatibility factors versus temperature for several
example n-type semiconducting materials applicable for use as one
of the dissimilar thermoelectric materials in a thermoelectric
device. Similarly,
[0166] FIG. 9d (adapted from
http://knol.google.com/k/andre-szykier/thermo-electric-energy/3sqds0076vq-
oz/2#) depicts compatibility factors versus temperature for several
example p-type semiconducting materials applicable for use as one
of the dissimilar thermoelectric materials in a thermoelectric
device.
[0167] FIG. 10a (adapted from [7]) depicts how in traditional
Peltier effect thermoelectric devices simple heat conduction heat
flow (thicker vertical arrows) returns heat that has been
transported in the opposite direction by electron flow (thinner
vertical arrows).
[0168] FIG. 10b (adapted from [7]) illustrates the phenomenon in
traditional Seebeck effect thermoelectric devices wherein simple
heat conduction heat flow (thicker vertical arrows) provides a
dominant path for heat flow, leaving far less heat to actually
drive the electron flow (thinner vertical arrows) that creates a
Seebeck effect electric current.
[0169] FIG. 10c depicts how thermoelectric process electron
transfer across a thermally-isolating physical gap prevents
efficiency-reducing heat-transfer in a thermoelectric device
operating in Peltier (heat transfer) mode.
[0170] FIG. 10d depicts how thermoelectric process electron
transfer across a thermally-isolating physical gap prevents
efficiency-reducing heat-transfer in a thermoelectric device
operating in Seebeck (heat-to-power conversion) mode.
[0171] FIG. 11a depicts a mechanical symbol that will be used for a
physical instance.
[0172] FIG. 11b depicts an exemplary physical arrangement of
thermoelectric devices in a linear side-by-side array.
[0173] FIG. 11c depicts an exemplary physical arrangement of
thermoelectric devices in a side-by-side matrix array.
[0174] FIG. 11d depicts an exemplary physical arrangement of
thermoelectric devices in a stacked "sandwich" array.
[0175] FIG. 11e depicts an exemplary physical arrangement of
thermoelectric devices in a 3D array.
[0176] FIG. 12, adapted from example [11], depicts an example
representation of the classic "Bottoming Cycle" employed in on-site
manufacturing electrical energy co-generation that can be adapted
to serve the electrical energy generation situation in data
centers.
[0177] FIG. 13, adapted from [12] depicts an example representation
of the total energy-flow budget relevant for actual and
economically-ascribed value of embodiments of the invention as a
green technology.
[0178] FIG. 14a depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so that each cell is electrically connected
to at least two switching transistors, the first switching
transistor connecting to a power source and the second switching
transistor connecting to a load, or load interface circuit, to
which the thermoelectric cell provides power.
[0179] FIG. 14b depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so that each cell is electrically connected
to at least two switching transistors, the first switching
transistor connecting to a power source and the second switching
transistor connecting to a measurement circuit.
[0180] FIG. 14c depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so that each cell is electrically connected
to at least two switching transistors, the first switching
transistor connecting to a load, or load interface circuit, to
which the thermoelectric cell provides power, and the second
switching transistor connecting to a measurement circuit.
[0181] FIG. 14d depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so that each cell is electrically connected
to at least three switching transistors, the first switching
transistor connecting to a power source, the second switching
transistor connecting to a load, or load interface circuit, to
which the thermoelectric cell provides power, and the third
switching transistor connecting to a measurement circuit.
[0182] FIGS. 15a-15j depict a collection of switch mode states
which can be periodically attained in a mutually exclusive fashion
according, for example, to a periodic state transition map with
periodic behavior determined by parameters such as at least one of
frequency, duty cycle, period, or duration in each state.
[0183] FIG. 16a depicts an example arrangement wherein an array of
thermoelectric cells, physically arranged for example in ways such
as the examples depicted in FIGS. 11b-11e, are interfaced with
arrangements of switching transistors, for example in ways such as
the examples depicted in FIGS. 14a-14d, such that each arrangement
of switching transistors selects modes of operation for at least
one thermoelectric cell, the modes including at least measurement,
heat transfer, and heat-to-electricity energy harvesting. The
switching transistors can be controlled by a control system.
[0184] FIG. 16b depicts a variation of the example arrangement of
FIG. 1a wherein the control system is provided input signals from
one or more additional sensors, for example a remote temperature
sensor.
[0185] FIG. 16c depicts a variation of the example arrangement of
FIG. 1a wherein there is no measurement mode and the control system
is provided input signals from one or more additional sensors, for
example a remote temperature sensor rather than measurement signals
from the switching transistor array.
[0186] FIG. 17 (adapted from [14]) depict a representation of the
graphical determination of pole locations for a representative
thermoelectric device.
[0187] FIG. 18a (adapted from [14]) depicts a representation of the
unit-step change in current for a representative thermoelectric
device.
[0188] FIG. 18b (adapted from [14]) depicts a representation of the
unit-step change in heat load for a representative thermoelectric
device.
[0189] FIG. 18c (adapted from [14]) depicts a representation of the
unit-step change in ambient temperature for a representative
thermoelectric device.
[0190] FIGS. 19a-19c (adapted from [14]) depict a representation of
the normalization of time in relation to the cold junction
temperature for various values of normalized quiescent electrical
current.
[0191] FIG. 20a (adapted from [14]) depicts a representation of the
unit-step in input power for a representative thermoelectric
device.
[0192] FIG. 20b (adapted from [14]) depicts a representation of the
unit-step in ambient temperature for a representative
thermoelectric device.
[0193] FIG. 21a (adapted from [14]) depicts a representation of the
unit-step in load resistance for a representative thermoelectric
device.
[0194] FIG. 21b (adapted from [14]) depicts a representation of the
unit step in load back-emf for a representative thermoelectric
device.
[0195] FIG. 22a depicts an example "top cover" or "top socket"
arrangement for making thermal connections to an integrated circuit
package. The example "top cover" or "top socket" arrangement can be
made to be mechanically compatible with a traditional electrical
socket for the integrated circuit package, or can be designed
together with an associated form of electrical socket for the
integrated circuit package.
[0196] FIG. 22b depicts an example arrangement wherein the example
"top cover" or "top socket" of FIG. 22a comprises a thermoelectric
array. The "top cover" or "top socket" arrangement can include
switching transistors and one or more of a control system and
additional sensors.
[0197] FIG. 23a depicts an example arrangement wherein a first side
of an active thermoelectric device is attached atop an integrated
circuit package and in thermal-transfer contact with one or more of
the integrated circuit wafer and heat conducting elements of the
integrated circuit package. The active thermoelectric device can
include switching transistors and one or more of a control system
and additional sensors.
[0198] FIG. 23b depicts the example arrangement of FIG. 23a wherein
the second side of the active thermoelectric device is attached and
in thermal-transfer contact with a thermal interface to a cooling
fin arrangement.
[0199] FIG. 23c depicts the example arrangement of FIG. 23a wherein
the second side of the active thermoelectric device is attached and
in thermal-transfer contact with a thermal interface which is in
turn in thermal-transfer contact with a fluidic cooling
arrangement.
[0200] FIG. 23d depicts the example arrangement of FIG. 23a wherein
the second side of the active thermoelectric device is attached and
in thermal-transfer contact with a thermal interface to a fluidic
cooling arrangement.
[0201] FIG. 24a depicts a thermal interface provided by an article
of integrated circuit packaging wherein the thermal interface is
located on the top of the integrated circuit packaging.
[0202] FIG. 24b depicts an alternative or supplemental thermal
interface provided by an article of integrated circuit packaging
wherein the thermal interface is located on the bottom of the
integrated circuit packaging.
[0203] FIG. 24c depicts an alternative or supplemental thermal
interface provided by an article of integrated circuit packaging
wherein the thermal interface is located on one or more sides of
the integrated circuit packaging.
[0204] FIG. 24d depicts an alternative or supplemental thermal
interface provided by an article of integrated circuit packaging
wherein the thermal interface is located within the integrated
circuit packaging.
[0205] FIG. 25a represents a thermal interface that is built upon
the top of a computer chip.
[0206] FIG. 25b shows a thermal interface that is built on the
bottom of a computer chip.
[0207] FIG. 25c illustrates a thermal interface that is placed
inside a computer chip packaging and makes an appearance at the
edge of the chip packaging.
[0208] FIG. 25d depicts a thermal interface that is built around a
computer chip (i.e. the computer chip package is largely contained
within the thermal interface).
[0209] FIG. 26a (adapted from Adaptive Cooling of Integrated
Circuits Using Digital Microfludics by P. Paik, K. Chakrabarty, and
V. Pamula, 2007, ISBN 13: 978-1-59693-138-1) depicts a side view
representation of a microfluidic electrowetting micro-droplet
transport "chip" implementation fitted over an integrated circuit
package and in turn in thermal contact with an active cooling
element such as a thermoelectric cooler.
[0210] FIG. 26b (adapted from Adaptive Cooling of Integrated
Circuits Using Digital Microfludics by P. Paik, K. Chakrabarty, and
V. Pamula, 2007, ISBN 13: 978-1-59693-138-1) depicts a top view
representation of a number of micro-droplets being transported (via
electrowetted transport) through various straight and
right-angle-turn paths over a planar array of microelectrodes
comprised by such a microfluidic electrowetting micro-droplet
"chip."
[0211] FIG. 26c (adapted from [1]) depicts a general planar
microfluidic micro-droplet transport arrangement.
[0212] FIG. 26d (adapted from [1]) depicts a "side view" of
electrowetting transport of a micro-droplet through the
microfluidic transport arrangement via sequencing of potential
applied to microelectrodes 1, 2, 3, and 4.
[0213] FIG. 26e (adapted from [1]) depicts a "top view" motion of
the micro-droplet via electrowetting transport through the
microfluidic transport arrangement of FIG. 26d.
[0214] FIG. 27a and FIG. 27b (adapted from [1]) depict "top" and
"side" views of a first heat transfer contact (graded) modulation
scheme explained in [1].
[0215] FIG. 27c (adapted from [1]) depicts a second heat transfer
contact (on-off) modulation scheme explained in [1].
[0216] FIG. 28a (adapted from [1]) depicts a "confined system"
adaptation of the microfluidic electrowetting micro-droplet planar
microelectrode array and micro-droplet transport to implementations
using Printed Circuit Boards ("PCBs").
[0217] FIG. 28b (adapted from [1]) depicts an "open system"
adaptation of the microfluidic electrowetting micro-droplet planar
microelectrode array and micro-droplet transport to implementations
using Printed Circuit Boards ("PCBs").
[0218] FIGS. 29a and 29b (each adapted from [1]) depict example
routing paths of micro-droplets over the planar microelectrode
array.
[0219] FIG. 30 depicts an representation of the "top" or "bottom"
view an example array of microelectrodes, each microelectrode
rendered as conductor area on a Printed Circuit Board (PCB) and
provided with an associated electrically-conducting "trace" for
electrically connecting the microelectrode to voltage potential
control circuitry, and interspersed between some pairs of
electrodes a physical open hole suitable for a micro-droplet to
travel through.
[0220] FIG. 31 depicts a side-view representation of an example
two-layer micro-droplet transport arrangement with conduits linking
the two micro-droplet transport region. This Figure incorporates a
side-view of the arrangement like that depicted in FIG. 30. The
view shown in FIG. 30 would herein lie in the center facing
downwards and comprises additional microelectrodes; two of the
physical open holes suitable for a micro-droplet to travel through
depicted in FIG. 30 appear (in side-view) in FIG. 31 as the
conduits linking the two micro-droplet transport region. In the
depiction of FIG. 31, above the upper micro-droplet transport
region is a solid layer of PCB material punctuated with
thermally-conducting segments that conduct heat from the item to be
cooled into the upper micro-droplet transport region.
[0221] FIG. 32 (adapted from U.S. Patent Application 61/599,643)
depicts an example situation where wherein the momentum of the
micro-droplet is not suppressed (i.e., the micro-droplet is not
locked into position under the activated microelectrode for an
interval of time) and the micro-droplet continues moving a bit
beyond the immediate region crowned by the activated
microelectrode. Here the micro-droplet moves towards the opening of
the conduit joining the lower micro-droplet transport region and
the upper micro-droplet transport region.
[0222] FIG. 33 (adapted from U.S. Patent Application 61/599,643)
depicts transmission through a first conduit joining two
droplet-transport layers from a non-heat-gathering-layer to a
heat-gathering-layer wherein the transmission through the conduits
employs a component of capillary forces and electric fields from
distant microelectrodes.
[0223] FIG. 34 (adapted from U.S. Patent Application 61/599,643)
depicts depict transmission through the first conduit joining the
two droplet-transport layers wherein the transmission through the
conduits employs essentially only proximate microelectrodes.
[0224] FIG. 35 (adapted from U.S. Patent Application 61/599,643)
depicts example transmission through the first conduit joining the
two droplet-transport layers.
[0225] FIG. 36 (adapted from U.S. Patent Application 61/599,643)
depicts the attraction of the micro-droplet to a region immediately
to the right of the second conduit joining the two
droplet-transport regions via activation of the microelectrode
immediately to the right of the second conduit joining the two
droplet-transport regions.
[0226] FIG. 37 (adapted from U.S. Patent Application 61/599,643)
depicts a representation of heat transfer from the previously
heated micro-droplet to the electrical ground plane and further
into the material joined to the electrical ground plane wherein the
material joined to the electrical ground plane comprises a "global"
(large area) thermoelectric structure.
[0227] FIG. 38 (adapted from U.S. Patent Application 61/599,643)
depicts an expanding variation on the arrangement of FIG. 38
wherein the electrical ground plane depicted throughout earlier
figures is replaced by an extended array of local thermoelectric
structures.
[0228] FIG. 39 illustrates the concept that an arbitrary number of
thermal interface stages can be cascaded and arranged
hierarchically so as to remove heat from any number of computer
chips or other heat sources, transferring the heat via an
associated number of heat exchanges and subject to an arbitrary
number of energy harvesting operations. Energy harvesting can occur
in heat exchange and/or heat transfer steps as advantageous.
[0229] FIG. 40 illustrates the concept that heat from an arbitrary
number of computer chips or other heat sources can be transferred
through an arbitrary number of heat exchanges and heat transfers as
proves advantageous.
[0230] FIG. 41a represents a heat-aggregating subsystem wherein
heat generated by an integrated circuit chip is transferred via
circulating cooling fluid or heat pipes through a plurality of
levels of heat handling and aggregation. At one or more of these
levels heat can be either converted to electricity, eliminated via
a fan, or transferred via a heat exchange for further transfer of
the heat for harvest and reuse.
[0231] FIG. 41b represents a combination of any number of
subsystems to form a heat-aggregating system wherein each component
subsystem is comprised as described in FIG. 41a.
[0232] FIG. 41c shows multiple heat-aggregating systems wherein any
number of subsystems make up each of the heat-aggregating systems
and each subsystem is comprised as described in FIG. 41a. Multiple
heat-aggregating systems can be formed by combining any number of
computer cages, computer racks, rack clusters, data center floors,
data center buildings or data center complexes.
[0233] FIG. 42a depicts an abstract thermodynamic representation of
a passive heat transfer arrangement for the transfer heat from a
hotter heat transport system to a cooler heat transport system.
[0234] FIG. 42b depicts an abstract thermodynamic representation of
a heat transfer arrangement transfer heat from a hotter heat
transport system to a cooler heat transport system further
comprising a heat engine, generating energy or work such a system
can be used for opportunistic energy harvesting, for example under
the careful operation of a control system.
[0235] FIG. 42c depicts an abstract thermodynamic representation of
a heat transfer arrangement comprising a heat pump, using energy or
work to unidirectionally transfer heat from a hotter heat transport
system to a cooler heat transport system. Such an arrangement can
be used to improve the rate of heat transfer and prevent a "stall"
in heat transfer should both depoicted heat transport systems
operate at roughly the same temperature (for example, in heavy heat
situations).
[0236] FIG. 43a and FIG. 43b depict a two-stage heat transfer
arrangement that can be used to illustrate how the arrangement
represented by
[0237] FIG. 42b can further be used for opportunistic energy
harvesting, for example under the careful operation of a control
system. At times when there is extra heat build up, one of the
thermoelectric heat engines can be mode switched to act as a
thermoelectric heat engine.
[0238] FIG. 44a illustrates an arbitrary system (or subsystem)
being connected to another system (or subsystem) by pressure
contact.
[0239] FIG. 44b illustrates an arbitrary system (or subsystem)
being connected to another system (or subsystem) by
fastener-facilitated contact (for example using threaded fastener
arrangements).
[0240] FIG. 44c illustrates an arbitrary system (or subsystem)
being connected to another system (or subsystem) by a mating
arrangement (for example, spring-spread pins, friction pins,
twist-lock, etc.).
[0241] FIG. 45 (adapted from [15]) depicts an arrangement wherein a
thermoelectric device is introduced at the thermal interface
between two closed loop fluid cooling systems.
[0242] FIG. 46 depicts the component layout of a daisy-chain heat
transfer arrangement employing closed systems of circulating fluids
for use in a hierarchical or peer arrangement. There can be any
number of circulating fluid systems from which heat can be pulled
from any number of heat-generating sources for energy harvest or
transfer; alternatively, any number of cooling fans can be utilized
so that heat transferred from any number of heat-generating sources
can be dispelled into the air.
[0243] FIG. 47 illustrates two layers of a hierarchical system with
a tree architecture that provides increasing degrees of
hierarchical aggregation at each sequentially lower level.
[0244] FIG. 48 illustrates a hierarchical system for heat gathering
compatible with the hierarchical arrangements depicted in FIGS.
2a-2j.
[0245] FIG. 49 illustrates a hierarchical system for
heat-to-electricity conversion compatible with the hierarchical
arrangements depicted in FIGS. 2a-2j
[0246] FIG. 50 illustrates a hierarchical control system compatible
with the hierarchical arrangements depicted in FIGS. 2a-2j.
[0247] FIG. 51 depicts a representation of an example hierarchical
multiple-level control system comprising N levels, each level in
the hierarchy comprising a single subsystem.
[0248] FIG. 52 depicts a representation of an example
strictly-layer parent-to-child and child-to-parent communications
between pairs of consecutive subsystem levels in the example
hierarchy depicted in FIG. 51.
[0249] FIG. 53 depicts a representation wherein more general
communications between pairs of subsystems in levels in the example
hierarchy is provided for. In one extreme, all subsystems can be
interconnected in a full-mesh topology.
[0250] FIG. 54 depicts a variation on the representation of FIG. 53
wherein additionally only some of the subsystems associated with
some of levels in the example hierarchy are present.
[0251] FIG. 55 depicts a variation on the representation of FIG. 52
wherein there are a plurality of subsystems associated with each
level in the example hierarchy.
[0252] FIG. 56 depicts a variation on the representation of FIG. 55
wherein there is at least one subsystem associated with each level
in the example hierarchy.
[0253] FIG. 57a depicts a representation of an example linear
control system accepting outside control and measurement inputs and
internal feedback paths.
[0254] FIG. 57b depicts a representation of an example variation on
the arrangement of FIG. 57a wherein additional inputs are provided
by other subsystems and additional outputs are provided to other
subsystems. Additionally, the representation provides for changes
to parameters and/or configuration of the controller responsive to
the presence or existence of other subsystems (in other layers of
the hierarchy, same layer of the hierarchy, etc.) as advantageous
in various implementations and embodiments.
[0255] FIG. 58a depicts a representation of an example bilinear
control system accepting outside control and measurement inputs and
internal feedback paths.
[0256] FIG. 58b depicts a representation of an example variation on
the arrangement of FIG. 58a wherein additional inputs are provided
by other subsystems and additional outputs are provided to other
subsystems. Additionally, the representation provides for changes
to parameters and/or configuration of the controller responsive to
the presence or existence of other subsystems (in other layers of
the hierarchy, same layer of the hierarchy, etc.) as advantageous
in various implementations and embodiments.
[0257] FIG. 59a depicts a representation of an example nonlinear
control system accepting outside control and measurement inputs and
internal feedback paths.
[0258] FIG. 59b depicts a representation of an example variation on
the arrangement of FIG. 59a wherein additional inputs are provided
by other subsystems and additional outputs are provided to other
subsystems. Additionally, the representation provides for changes
to parameters and/or configuration of the controller responsive to
the presence or existence of other subsystems (in other layers of
the hierarchy, same layer of the hierarchy, etc.) as advantageous
in various implementations and embodiments.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0259] In the following, numerous specific details are set forth to
provide a thorough description of various embodiments. Certain
embodiments may be practiced without these specific details or with
some variations in detail. In some instances, certain features are
described in less detail so as not to obscure other aspects. The
level of detail associated with each of the elements or features
should not be construed to qualify the novelty or importance of one
feature over the others.
[0260] In the following description, reference is made to the
accompanying drawing figures which form a part hereof, and which
show by way of illustration specific embodiments of the invention.
It is to be understood by those of ordinary skill in this
technological field that other embodiments may be utilized, and
structural, electrical, as well as procedural changes may be made
without departing from the scope of the present invention.
1. Heat Transfer Background
[0261] To begin, a brief thermodynamic framework is established.
FIG. 4a depicts a general thermodynamics passive heat transfer
process from a hot body to a broader environment. Here excess heat
travels (typically according to the Heat Equation) from hot bodies
to colder bodies (not depicted in FIG. 4a) or dissipates in the
cooler environment (as depicted in FIG. 4a) with no power or energy
is used to manage, direct, or control the flow of heat. Heat flows
until all bodies in the thermally-connected systems reach the same
uniform temperature.
[0262] FIG. 4b depicts a heat pump arrangement for active heat
transfer process from a hot body to a broader environment. Energy
is applied over time to the heat pump (amounting to applied work)
and consumed in the heat-pumping process. Thermoelectric cooling
(for example employing the Peltier process and analogous processes
employing Avto metals and quantum well materials) is an example of
such a heat pump arrangement. Additional considerations relating to
the energy and work applied to the heat pump (for example Thomson
effect, Joule heating, etc.) are not brought forth in this
representation.
[0263] FIG. 4c depicts a heat engine arrangement for active heat
transfer process from a hot body to a broader environment. Energy
is harvested over time by the heat engine (amounting to harvested
work) and consumed (at least in part) by external processes.
Thermoelectric electric current generation (for example employing
the Seebeck process and analogous processes employing Avto metals
and quantum well materials) is an example of such a heat engine
arrangement. Additional considerations relating to the energy and
work applied to the heat engine (for example Benedicks effect,
Joule heating, etc.) are not brought forth in this
representation.
[0264] Regarding heat flow, analogies can readily be made with
electrical currents and potentials. FIG. 5a depicts an exemplary
thermal integration and transfer abstraction employing abstractions
and operations familiar to electrical engineering. FIG. 5b depicts
an exemplary thermal resistive abstraction familiar to electrical
engineering. FIG. 5c depicts an exemplary thermal resistive series
familiar to electrical engineering.
[0265] FIG. 5d depicts an exemplary thermal diode abstraction
familiar to electrical engineering. This function can be realized
by a heat pump, for example a thermoelectric cooler
2. Heat Gathering and Transport at the Chip Level
[0266] The invention provides for and can use one or more of a
number of known and a number of new and novel practical ways for
heat gathering and transport at the chip level. Several of these
are presented in this section. Other variations, adaptations, and
additional approaches are provided throughout the rest of the
document.
2.1. Traditional Passive Thermal Handling for Integrated
Circuits
[0267] FIG. 6a shows heat generated from a wafer within an
exemplary computer chip package directed through a thermal
interface provided on top of the chip. Heat is transferred into a
heat sink where the heat can be dissipated using a circulating
cooling fluid.
[0268] FIG. 6b depicts an alternative situation where the heat
generated from a wafer within the computer chip is directed through
a thermal interface provided on the top of the chip. Although not
explicitly shown, such a thermal interface can be incorporated into
the chip packaging by the chip manufacturer.
[0269] FIG. 6c depicts an exemplary arrangement wherein a computer
chip is provided with a thermal interface comprising cooling fins.
Heat pulled from the computer chip can be transported and dispersed
by an associated cooling fan via air convection. Although not
explicitly shown, such a thermal interface can be incorporated into
the chip packaging by the chip manufacturer.
[0270] FIG. 6d depicts an exemplary fluidic cooling conduit in
thermal contact with a thermal interface. Although not explicitly
shown, such a thermal interface can be incorporated into the chip
packaging by the chip manufacturer.
2.2 Thermoelectric Devices for Cooling and Heat-to-Electricity
Conversion
[0271] Thermoelectric devices employ many effects, most of which
have a long history. The table below, adapted from [17] lays out
the four prominent thermoelectric effects and their dates of
established recognition.
TABLE-US-00001 Summary Schedule of the Thermo-Electricity (1916)
Homogenous Heterogenous (1 substance) (2 Substances) Heat Current
causes Electric Benedicks Seebeck Current (1916) (1821) Electric
Current causes Heat Thompson Peltier Current (1856) (1834)
[0272] FIG. 7a provides an example depiction of the Peltier effect
wherein an electric current is used to create a temperature
gradient in an arrangement involving dissimilar materials. In this
figure the arrangement is depicted in terms of N-type and P-type
semiconductor materials, but the Peltier effect also pertains to
(and was originally discovered in the form of) junctions of
dissimilar metals. French physicist Jean Peltier (1785-1845)
discovered this effect eight years after German physicist Thomas
Seebeck published research findings and analysis on thermoelectric
current generation known now as the Seebeck effect (to be discussed
shortly). Peltier studied electrical current flowing through pairs
of junctions of dissimilar metals and discovered that heat could be
transferred in a direction depending upon the direction of
electrical current passed through them. Also of note is that heat
generated via Joule (a.k.a. ohmic) heating loss (proportional to
the product of the square of the magnitude of the electrical
current and the electrical resistance of the metal and junction
arrangement) from an electrical current was absorbed if the current
was reversed. The Peltier heat transfer effect is proportional to
the magnitude of the electrical current.
[0273] Preceeding Peltier's work, German physicist Thomas Johann
Seebeck (1770-1831) observed that an electrical current is created
through the junction of two dissimilar metals when the junctions of
the two metals are at different temperatures, and that the effect
increases as the difference between the temperature increases.
Seebeck researched this effect over combinations of a selection of
elemental metal materials available in the 1820's (such as
antimony, bismuth, cadmium, cobalt, copper, gold, iron, lead,
manganese, mercury, nickel, palladium, platinum, silver, tellurium,
tin, and zinc) and arranged their presence in an ordered series.
The series is structured so that thermally-induced electromotive
force ("emf") generated increases as the difference between the
positions of the metals in the series increases. The direction of
current flow at the hotter of the two junctions is from a metal
occurring earlier in the series to the metal occurring later in the
series. The ordering of the series turns out to be dependent upon
the temperature and impurities.
[0274] FIG. 7b provides an example depiction of the Seebeck effect
wherein a temperature gradient is used to create an electric
current in an arrangement involving dissimilar materials. In this
figure the arrangement is depicted in terms of N-type and P-type
semiconductor materials, but the Seebeck effect also pertains to
(and was originally discovered in the form of) junctions of
dissimilar metals.
[0275] FIG. 7c depicts an electrical symbol that will be used for
an electrical instance of an arrangement involving an individual
junction of dissimilar materials, the arrangement configured to
provide one or more thermoelectric functions employing phenomenon
such as the Peltier effect, Seebeck effect, or other adapted,
related, or alternative thermoelectric effects. An instance of such
an arrangement will be termed a "thermoelectric device." The
junction-side thermal interface is denoted with the thick bar, and
the electrical terminals and associated thermal interface are
denoted by the thinner pair of lines. Thermal conduction occurs
through the long open-rectangle bars. A voltage appears or is
applied across the electrical terminals (the terminals denoted as
in this diagram by the circle-symbols labeled "A" and "B"). A
current will flow between the electrical terminals if an electrical
load is connected to the electrical terminals.
[0276] FIG. 7d depicts an electrical series connection of a
plurality of thermoelectric devices. A voltage appears or is
applied across the electrical terminals (the terminals denoted as
in this diagram by the circle-symbols labeled "A" and "B"). A
current flows between the electrical terminals if an electrical
load is connected to the electrical terminals. (In the literature
such an electrical arrangement combined with an associated thermal
arrangement is called a "thermopile.")
[0277] FIG. 7e depicts a plurality of electrical series connections
of a plurality of thermoelectric devices. The invention provides
for individual thermoelectric devices to be physically arranged in
a wide range of ways including in a linear side-by-side array, a
side-by-side matrix array, a stacked "sandwich" array, a 3D array,
etc.
[0278] FIG. 7f depicts an exemplary physical array of
thermoelectric cells such as the example provided in FIGS. 7c-7e.
The invention provides for each cell to comprise any of an
individual thermoelectric device, a linear side-by-side array of
thermoelectric devices, a side-by-side sub-matrix array of
thermoelectric devices, a stacked "sandwich" array of
thermoelectric devices, a 3D array of thermoelectric devices,
etc.
[0279] The cells in the exemplary physical array of thermoelectric
cells such as the example provided in FIGS. 7c-7e can in turn be
electrically connected to separate circuits, electrically connected
in series, electrically connected in parallel, or electrically
connected in other topological arrangements. As described later,
the invention provides for thermoelectric cells to be connected
with one or more switching transistors.
[0280] Such electrical, electrical interconnection, and
matrix-layout arrangements of individual thermoelectric devices
such as those depicted in FIGS. 7c-7f have been used in various
commercial thermoelectric device products.
[0281] FIG. 8a depicts an example accounting of various
material-based and junction-based electrical resistance aspects
inherent in a thermoelectric device.
[0282] FIG. 8b depicts a series-resistance aggregation of these
various material-based and junction-based electrical resistance
aspects inherent in a thermoelectric device into a composite
equivalent electrical resistance.
[0283] FIG. 8c depicts an abstract electronics representation of
the equivalent electrical resistance of a thermoelectric device or
thermoelectric cell operating in Peltier mode.
[0284] FIG. 8d depicts an abstract electronics representation of
voltage electromotive force "emf" generated by Seebeck effect in a
thermoelectric device or cell and the associated
(Thevenin-equivalent) series resistance of the thermoelectric
device or thermoelectric cell, the resulting arrangement connected
to a load resistance resulting in a current I flowing around the
resulting electrical loop.
2.3 Optimizing Performance of Traditional Thermoelectric
Devices
[0285] Many books have been written relating to optimizing
performance of traditional thermoelectric devices, for example [6],
[13]-[18].
[0286] FIG. 9a, adapted from
http://knol.google.com/k/andre-szykier/thermo-electric-energy/3sqds0076vq-
oz/2#, depicts a graphical representation of the relationship
between reduced current efficiency and relative current density for
several example materials applicable for use as one of the
dissimilar materials in a thermoelectric device.
[0287] Current-Voltage-Power characteristics of a thermoelectric
process can be viewed in terms of a voltage drop varying linearly
with respect to electrical current due to the internal resistance
of the thermoelectric material discussed in conjunction with FIGS.
8a-8d. FIG. 9b, adapted from
http://knol.google.com/k/andre-szykier/thermo-electric-energy/3sqds0076vq-
oz/2#, depicts a constant-temperature linear current-versus-voltage
("Ohms Law") and quadratic power-versus-current ("Joule's Law")
curves for an exemplary Bi2Te3 thermopile. At zero current through
the thermoelectric device (no load on the terminals), the produced
voltage is maximized as there is no ohmic loss. However, since
power is the multiplicative product of current and voltage, no
power is produced in a zero current situation. At high current
values, the voltage drops to zero or below and the power produced
again is zero (and in fact can be negative, signifying the
consuming of power rather than producing power). For some value of
current between these extremes, the produced power will be
maximized.
[0288] Attention is now returned to further details of the Peltier
effect. The heat transfer process though the materials configured
to produce the Peltier effect (proportional to the current squared)
always works in opposition to Joule heating (proportional to the
current squared) caused by the electrical resistance of the
arrangement of configured materials. This causes reduced
efficiency.
[0289] To understand reduced efficiency further, it is noted that
the Peltier effect is a surface effect occurring at the junction
between two materials. The electrical resistance, which is the
source of the heat generation, involves several processes: these
include a volume-related component (involving the electrical
conductivity, cross-sectional surface area, and length of an
article of material through which current flows), a surface-area
component and inter-materials component involving the junction of
the dissimilar materials, etc. A design approach involves the
relative current density u defined as the ratio of the electric
current density to the heat flux from thermal conduction.
[0290] Returning attention now to the Seebeck effect, one method to
improve efficiency involves segmented together materials. These
techniques leverage the fact that thermoelectric properties
(Seebeck coefficient, electrical resistivity, thermal conductivity,
etc.) of materials vary with temperature. From that viewpoint, it
is undesirable (and in some situations not possible) to employ a
single material spanning a large temperature gradient. In
principle, segments of "thermoelectrically compatible" but somewhat
different materials can be joined so that a material performing
with high-efficiency at high temperatures is segmented with a
material performing with high-efficiency at low temperatures, and
aligning the segmented material to match the high-temperature and
low-temperature regions of the applied temperature difference. In
this way optimal materials are matched to places in the temperature
gradient through the thermoelectric device so that each material
operates in its optimally-performing temperature range. A key
attribute of thermoelectric compatibility stems from the fact that
the heat and electric charge must flow through the connected
materials. A metric of thermoelectric compatibility is the
so-called "compatibility factor, often denoted s. If the
compatibility factors differ by a factor of two or more, a given
value of the relative current density u can not be suitable for
both materials and segmentation will not be efficient.
[0291] FIG. 9c, adapted from
http://knol.google.com/k/andre-szykier/thermo-electric-energy/3sqds0076vq-
oz/2#, depicts compatibility factors versus temperature for several
example n-type semiconducting materials applicable for use as one
of the dissimilar thermoelectric materials in a thermoelectric
device. Similarly, FIG. 9d, adapted from
http://knoLgoogle.com/k/andre-szykier/thermo-electric-energy/3sqds0076vqo-
z/2#, depicts compatibility factors versus temperature for several
example p-type semiconducting materials applicable for use as one
of the dissimilar thermoelectric materials in a thermoelectric
device.
[0292] Alternatively, compatibility and segmented design can be
avoided by instead thermally cascading a plurality of
thermoelectric generators. In such a thermally cascaded approach,
each thermoelectric generator produces an independent electrical
current, which in turn allows independent values of relative
current density u, in each stage so as to optimize the u for the
thermal role of each stage. This, along with a number of electrical
and thermal complexities, crop up in such cascades. For example,
high temperature thermoelectric generator stages should not be
directly connected to an electrical load (due to Wiedeman Franz law
and Joule loss considerations). Various circuit and thermal
topologies can be designed to optimize performance of a cascade
against these concerns.
2.4 High Performance Quantum-Process Material Thermoelectric
Devices
[0293] A large number and wide variety of approaches are currently
under research, development, and deployment of materials with
high-performance thermoelectric properties. Two examples of these
are quantum-well and Avto metals. A brief treatment of techniques
and properties of quantum-well thermoelectric devices can be found
in [6] and the references therein. Treatment of techniques and
properties of Avto metal thermoelectric devices can be found in
[7]-[10] and the references therein.
[0294] A major problem with Peltier effect thermoelectric devices,
among others, is that while electrons transport heat in one
direction, the material itself provides a reverse heat flow
(through simple heat conduction) that returns much of the
transported heat.
[0295] FIG. 10a, adapted from [7], depicts how in traditional
Peltier effect thermoelectric devices simple heat conduction heat
flow (thicker vertical arrows) returns heat that has been
transported in the opposite direction by electron flow (thinner
vertical arrows). These processes make even the most optimized
Peltier effect thermoelectric devices made with traditional
thermoelectric materials have heat-transfer efficiencies of 5%-8%
of the theoretical Carnot heat-transfer efficiency upper limit.
This compares unfavorably with compressor-based cooling approaches
which typically have heat-transfer efficiencies of 45%.
[0296] A related situation affects the efficiency of Seebeck effect
thermoelectric devices. Once again, as electrons transport energy
the material itself provides a heat flow path (through simple heat
conduction) through the material. Efficiency is greatly reduced
because most of the heat is transported through the materials
within the thermoelectric device, leaving far less heat to actually
drive the migration of electrons to create a Seebeck effect
electric current.
[0297] FIG. 10b, adapted from [7], illustrates this issue in
traditional Seebeck effect thermoelectric devices wherein simple
heat conduction heat flow (thicker vertical arrows) provides a
dominant path for heat flow, leaving far less heat to actually
drive the electron flow (thinner vertical arrows) that creates a
Seebeck effect electric current.
[0298] Typical Seebeck effect thermoelectric devices convert
approximately 10 percent of thermal energy to electricity. Even
with this low energy conversion ratio, available Seebeck effect
thermoelectric devices can produce useful ranges of voltages and
currents. For example, commercial devices available in development
kits from Custom Thermoelectric, Inc., 11941 Industrial Park Road,
STE 5, Bishopville, Md. 21813, 443-926-9135
(http://www.customthermoelectric.com/index.htm) provide voltages,
currents, and power quantities of value in powering computer
technology in acceptable physical sizes and formats:
TABLE-US-00002 Custom 150.degree. C. hot/50.degree. C. cold
300.degree. C. hot/30.degree. C. cold Thermoelectric, P I V P I V
Dimensions (mm) Inc. Part Number (Watts) (Amps) (Volts) (Watts)
(Amps) (Volts) Max.degree. C. L W H 1261G-7L31-04CQ 1.0 0.6 1.7 5.1
1.3 3.9 300 1.6 1.6 1.3 1261G-7L31-05CQ 1.2 0.9 1.3 7.5 2.0 3.6 300
1.6 1.6 1.3 1261G-7L31-10CX1 3.2 1.7 1.9 15 3.5 4.2 300 56 56
.17
[0299] Returning to the matter of efficiency, in both Peltier and
Seebeck modes vastly important efficiency limitations result from
heat conduction through the same material that implements the
desired thermoelectric process. One way to conquer this is to
somehow facilitate electron transfer while blocking heat transfer.
In terms of traditional material science, this has not yet been
attainable. However, the currency of thermoelectric process is
electron transport. Electrons can certainly traverse physical
separation gaps (between electrodes) that do not carry heat (as in
electron vacuum tubes used in early-to-mid 20th century
electronics). Additionally, electrons live in a world dominated by
quantum effects, and a variety of quantum effects, including
tunneling and standing wave resonance structures, that can be
induced by nanofabrication techniques. This suggests there could be
improvements made to thermoelectric process using additional
techniques, including physical separation gaps and nanofabricated
structure that induce quantum effects.
[0300] In fact there is at least one practical and commercially
viable approach of this sort employing so called "Avto metals"
which change electronic properties of a material by etching surface
patterns using available nanotechnology methods. Peltier processes
employing Avto metals appear to be able to reach heat-transfer
efficiencies of greater than 50% of the theoretical Carnot
heat-transfer efficiency upper limit, and Seebeck processes
employing such materials and techniques appear to be able to reach
conversion rates of 20%-23%. These technologies are described in
U.S. Pat. Nos. 7,658,772; 7,642,467; 7,589,348; 7,566,897;
7,427,786; 7,419,022; 7,351,996; 7,323,709; 7,253,549; 7,220,984;
7,208,021; 7,169,006; 7,166,786; 7,140,102; 7,124,583; 7,074,498;
7,005,381; 6,971,165; 6,876,123; 6,869,855; 6,774,003; 6,720,704;
6,651,760; 6,531,703; 6,495,843; 6,417,060; 6,281,514; 6,281,139;
6,239,356; 6,229,083; 6,214,651; 6,117,344; 6,089,311; 5,994,638;
5,981,866; 5,981,071; 5,810,980; 5,722,242; 5,699,668; 5,675,972
and the TRN article "Chips turn more heat into Power" available at
http://www.tmmag.com/Stories/2001/121901/Chips_turn_more_heat_to_power.su-
b.--1 21901.html (visited Feb. 15, 2011). FIG. 10c depicts how
thermoelectric process electron transfer across a
thermally-isolating physical gap prevents efficiency-reducing
heat-transfer in a thermoelectric device operating in Peltier (heat
transfer) mode. Similarly, FIG. 10d depicts how thermoelectric
process electron transfer across a thermally-isolating physical gap
prevents efficiency-reducing heat-transfer in a thermoelectric
device operating in Seebeck (heat-to-power conversion) mode.
2.5 Structured Physical Arrangements for Pluralities of
Thermoelectric Devices
[0301] FIG. 11a depicts a mechanical symbol that will be used for a
physical instance.
[0302] FIG. 11b depicts an exemplary physical arrangement of
thermoelectric devices in a linear side-by-side array.
[0303] FIG. 11c depicts an exemplary physical arrangement of
thermoelectric devices in a side-by-side matrix array.
[0304] FIG. 11d depicts an exemplary physical arrangement of
thermoelectric devices in a stacked "sandwich" array.
[0305] In an embodiment, a stacked "sandwich" array can be
implemented in an extended-length format which can be used as an
active version of a heat pipe.
[0306] FIG. 11e depicts an exemplary physical arrangement of
thermoelectric devices in a 3D array.
2.6 Summary of Thermoelectric Device Technology
[0307] At this point a few important facts that can be taken away
from the preceding discussion: [0308] Thermoelectric devices can be
operated in Peltier (heat transfer) mode and in Seebeck
(heat-to-power conversion) mode; [0309] In Peltier mode,
traditional thermoelectric devices made with traditional
thermoelectric materials have heat-transfer efficiencies of 5%-8%
of the theoretical Carnot heat-transfer efficiency upper limit
(comparing unfavorably with compressor-based cooling efficiencies
of 45%); [0310] In Seebeck mode, traditional thermoelectric devices
made with traditional thermoelectric materials convert
approximately 10 percent of thermal energy to electricity; [0311]
Early work leveraging various techniques, including physical
separation gaps and nanofabricated structures inducing quantum
effects on electrons, shows that Peltier mode heat transfer
efficiencies can be improved by a factor as high as 10 and Seebeck
mode conversions can be improved by a factor of 2 or more; [0312]
Such material and techniques will continue to favorably evolve and
performance metrics will continue to be improved; [0313] Even with
10 percent conversion rates, commercially available Seeback devices
are already available providing voltages, currents, and power
quantities of value in powering computer technology in acceptable
physical sizes and formats.
[0314] The present invention next provides a number of novel
innovations for leveraging the above into systems, methods, and
evolution strategies for technologies and products that provide a
flexible environment for cooling, thermal management, and
heat-to-electricity energy harvesting which (in various forms with
evolution paths among these forms) will be valuable for near-term
and future computer devices and data centers.
[0315] To begin, various exemplary structured physical arrangements
for pluralities of thermoelectric devices are first considered.
3. Multimode Thermoelectric Devices Combining Energy Harvesting and
Heat Transport Functions
[0316] In general, thermoelectric devices are reciprocal in that
they can operate in either a thermoelectric cooler or a
thermoelectric electric current generator as determined by imposed
thermal conditions and electrical connections to the reciprocal
thermoelectric device. Further, it is noted that when acting as a
thermoelectric electric current generator, a voltage is produced,
the same voltage that is used in thermocouples (a specialized
thermoelectric device) for the measurement of temperature. Thus,
thermoelectric devices can additionally serve as a temperature
sensor.
[0317] In general, thermoelectric devices can be optimized in their
design to best serve specific applications (for example,
temperature measurement, thermal cooling, electric energy
harvesting, etc.) Alternatively, thermoelectric devices can also be
optimized in their design to best two or all three of these
modalities.
[0318] In an embodiment, the invention provides for the use of
reciprocal thermoelectric devices capable of operating in either a
thermoelectric cooler or a thermoelectric electric current
generator as determined by imposed thermal conditions and
electrical connections to the reciprocal thermoelectric device.
[0319] In an embodiment, the invention provides for at least one of
the thermoelectric devices can serve as a temperature sensor.
[0320] In an embodiment, the invention provides for the mode of a
given thermoelectric device is switched over time. As one example,
a given thermoelectric device can be a thermoelectric cooler one
moment and a temperature sensor at another moment. As another
example, a given thermoelectric device can be a thermoelectric
electric current generator one moment and a temperature sensor at
another moment. As yet another example, a given thermoelectric
device can be a thermoelectric cooler one moment and a
thermoelectric electric current generator at another moment. As
still another example, a given thermoelectric device can be a
thermoelectric cooler one moment, a temperature sensor at another
moment, and a thermoelectric electric current generator at yet
another moment.
[0321] In an embodiment, the invention provides for a control
system that selects the mode of operation of at least one
reciprocal thermoelectric device, the selection made responsive to
the state of the system, time, a measurement condition, or some
combination of these.
[0322] When used to generate electricity in an environment that
also consumes electricity (such as in the computers of a data
center), the resulting situation is akin the well-established
industrial plant approaches to on-site electrical energy
co-generation. There are various forms of this, notably so called
"Topping Cycle" and "Bottoming Cycle" approaches (see for example
[11]). Of these, the "Bottoming Cycle" depicted in FIG. 12, adapted
from example [11], can be adapted to serve the electrical energy
generation situation in data centers. Here, the "raw material" in
and "manufactured products" out comprise abstractions such as data,
services, and interactive session transactions. Such a model
provides a framework for both control system design and for the
overall evaluation of the efficiency, effectiveness, and value of
various design options of the present invention.
[0323] Further as to this, the non-idealness, relative
efficiencies, material costs, operating energies, etc. all
contribute to figures of merit such as lifetime total cost of
ownership, environmental offset contributions, and other aspects
important to the actual and economically-ascribed value of
embodiments of the invention as a green technology. For example,
reduction for the need of vast-volume air handlers reduces energy
consumption but Joule heating, control processor power consumption,
and performance limitations of thermoelectric devices reduce
contributions of the technology. FIG. 13, adapted from [12] depicts
an example representation of the total energy-flow budget relevant
for actual and economically-ascribed value of embodiments of the
invention as a green technology.
3.1 Electrical Arrangements for Pluralities of Thermoelectric
Devices
[0324] FIG. 14a depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so That each cell is electrically connected
to at least two switching transistors, the first switching
transistor connecting to a power source and the second switching
transistor connecting to a load, or load interface circuit, to
which the thermoelectric cell provides power.
[0325] FIG. 14b depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so that each cell is electrically connected
to at least two switching transistors, the first switching
transistor connecting to a power source and the second switching
transistor connecting to a measurement circuit.
[0326] FIG. 14c depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so that each cell is electrically connected
to at least two switching transistors, the first switching
transistor connecting to a load, or load interface circuit, to
which the thermoelectric cell provides power, and the second
switching transistor connecting to a measurement circuit.
[0327] FIG. 14d depicts a plurality of thermoelectric cells, such
as those examples depicted in FIGS. 7c-7f, as can be physically
arranged in various ways, such as those examples depicted in FIGS.
11b-11e, to be arranged so that each cell is electrically connected
to at least three switching transistors, the first switching
transistor connecting to a power source, the second switching
transistor connecting to a load, or load interface circuit, to
which the thermoelectric cell provides power, and the third
switching transistor connecting to a measurement circuit.
[0328] Each of FIGS. 15a-15j depicts a collection of
switching-transistor configuration mode states that can be
periodically attained in a mutually exclusive fashion according,
for example, to a periodic state transition map with periodic
behavior determined by parameters such as at least one of
frequency, duty cycle, period, or duration in each state.
[0329] Note that idle and measurement modes provide a `safe`
intermediate state (for "break before make" action) between the
power source mode and the power load mode.
3.2 Control Systems for Multimode Thermoelectric Devices
[0330] The aforedescribed arrangements can be operated by a control
system. The control system can be configured to create a wide range
of operational capabilities for the adaptive optimized reactions to
needs for heat removal, opportunities for energy harvesting,
prevention of condensation (endemic to Peltier devices), prevention
of thermal runaway, backup safety provisions, and many additional
other functions. In some instances or implementations, individual
control systems can be provided to arrays of multimode
thermoelectric devices. In other instances or implementations, a
plurality of arrays of multimode thermoelectric devices can be
controlled by a single control system.
[0331] In an embodiment, as facilitated further in subsequent
sections, a control system can be integrated together with
switching transistors and an array of multimode thermoelectric
devices to form a physically self-contained system.
[0332] In an embodiment, the control system can switch among any
two or more of: [0333] Any of the ten periodical mode state
configurations depicted in FIGS. 15a-15j; [0334] Any of the four
individual modes such as idle, power source, power load, and
measurement.
[0335] In an embodiment, the invention provides pulse-width
modulation and other duty-cycle control interleave modes of
operation.
[0336] In an embodiment, the invention provides pulse-width
modulation and other duty-cycle control to prevent Peltier cooling
induced condensation.
[0337] FIG. 16a depicts an example arrangement wherein an array of
thermoelectric cells, physically arranged for example in ways such
as the examples depicted in FIGS. 11b-11e, are interfaced with
arrangements of switching transistors, for example in ways such as
the examples depicted in FIGS. 14a-14d, such that each arrangement
of switching transistors selects modes of operation for at least
one thermoelectric cell, the modes including at least measurement,
heat transfer, and heat-to-electricity energy harvesting. The
switching transistors can be controlled by a control system. The
control system can be provided measurement information and
externally provided control parameters or commands. Such an
arrangement can be configured to select the mode of each cell
thermoelectric independently or collectively, for example with a
control system comprising one or more of logic circuit,
clock-driven binary counters, a computer bus interface, algorithmic
microprocessor control, field-programmable logic array (FPLA)
control, analog circuitry, and driving transistors.
[0338] FIG. 16b depicts a variation of the example arrangement of
FIG. 1a wherein the control system is provided input signals from
one or more additional sensors, for example a remote temperature
sensor.
[0339] FIG. 16c depicts a variation of the example arrangement of
FIG. 1a wherein there is no measurement mode and the control system
is provided input signals from one or more additional sensors, for
example a remote temperature sensor rather than measurement signals
from the switching transistor array.
[0340] In an embodiment, control systems as described above can
interact with control systems in other parts of a heat management
and energy harvesting hierarchical chain.
[0341] The invention also provides for control systems as described
above can interact with peer control systems of a heat management
and energy harvesting arrangement.
3.3 Consideration of and Compensation for Dynamic Behavior of
Thermoelectric Devices
[0342] Thermoelectric devices actually have complex dynamic
behavior. For example, FIG. 17 (adapted from [14]) depicts a
representation of the graphical determination of pole locations for
a representative thermoelectric device.
[0343] FIG. 18a (adapted from [14]) depicts a representation of the
unit-step change in current for a representative thermoelectric
device.
[0344] FIG. 18b (adapted from [14]) depicts a representation of the
unit-step change in heat load for a representative thermoelectric
device.
[0345] FIG. 18c (adapted from [14]) depicts a representation of the
unit-step change in ambient temperature for a representative
thermoelectric device.
[0346] FIGS. 19a-19c (adapted from [14]) depicts a representation
of the normalization of time in relation to the cold junction
temperature for various values of normalized quiescent electrical
current.
[0347] FIG. 20a (adapted from [14]) depicts a representation of the
unit-step in input power for a representative thermoelectric
device.
[0348] FIG. 20b (adapted from [14]) depicts a representation of the
unit-step in ambient temperature for a representative
thermoelectric device.
[0349] FIG. 21a (adapted from [14]) depicts a representation of the
unit-step in load resistance for a representative thermoelectric
device.
[0350] FIG. 21b (adapted from [14]) depicts a representation of the
unit step in load back-emf for a representative thermoelectric
device.
[0351] In an embodiment, the invention provides for a control
system to include consideration of the dynamic behavior of at least
one type of thermoelectric device.
[0352] In an embodiment, the invention provides for a control
system to include compensation for the dynamic behavior of at least
one type of thermoelectric device.
[0353] In an embodiment, the invention provides for a control
system that selects the mode of operation of at least one
reciprocal thermoelectric device to include consideration of and/or
compensation for the dynamic behavior of the reciprocal
thermoelectric device.
3.4 Interfacing Multimode Thermoelectric Devices with Chip
Packaging
[0354] FIG. 22a depicts an example "top cover" or "top socket"
arrangement for making thermal connections to an integrated circuit
package. The example "top cover" or "top socket" arrangement can be
made to be mechanically compatible with a traditional electrical
socket for the integrated circuit package, or can be designed
together with an associated form of electrical socket for the
integrated circuit package.
[0355] FIG. 22b depicts an example arrangement wherein the example
"top cover" or "top socket" of FIG. 22a comprises a thermoelectric
array. The "top cover" or "top socket" arrangement can include
switching transistors and one or more of a control system and
additional sensors.
[0356] In an embodiment, the invention exploits reciprocity
properties of thermoelectric materials in contact with a chip
package.
[0357] FIG. 23a depicts an example arrangement wherein a first side
of an active thermoelectric device is attached atop an integrated
circuit package and in thermal-transfer contact with one or more of
the integrated circuit wafer and heat conducting elements of the
integrated circuit package. The active thermoelectric device can
include switching transistors and one or more of a control system
and additional sensors.
[0358] FIG. 23b depicts the example arrangement of FIG. 23a wherein
the second side of the active thermoelectric device is attached and
in thermal-transfer contact with a thermal interface to a cooling
fin arrangement.
[0359] FIG. 23c depicts the example arrangement of FIG. 23a wherein
the second side of the active thermoelectric device is attached and
in thermal-transfer contact with a thermal interface which is in
turn in thermal-transfer contact with a fluidic cooling
arrangement.
[0360] In an embodiment, the invention provides for at least one
thermal interface to be in thermal contact with a circulating
cooling fluid, for example possessing a high heat-carrying
capacity. In an embodiment, the invention provides for thermal
interfaces with a circulating cooling fluid to be designed to
easily connect to integrated circuit packaging.
[0361] FIG. 23d depicts the example arrangement of FIG. 23a wherein
the second side of the active thermoelectric device is attached and
in thermal-transfer contact with a thermal interface to a fluidic
cooling arrangement.
[0362] FIG. 24a depicts a thermal interface provided by an article
of integrated circuit packaging wherein the thermal interface is
located on the top of the integrated circuit packaging.
[0363] FIG. 24b depicts an alternative or supplemental thermal
interface provided by an article of integrated circuit packaging
wherein the thermal interface is located on the bottom of the
integrated circuit packaging.
[0364] FIG. 24c depicts an alternative or supplemental thermal
interface provided by an article of integrated circuit packaging
wherein the thermal interface is located on one or more sides of
the integrated circuit packaging.
[0365] FIG. 24d depicts an alternative or supplemental thermal
interface provided by an article of integrated circuit packaging
wherein the thermal interface is located within the integrated
circuit packaging. In some arrangements, the internal thermal
interface can be accessed via one or more openings in the
integrated circuit packaging. The one or more openings can be in
the form of a window, a fastener (for example a threaded hole, a
mating thermal connector, etc.).
[0366] In an embodiment, the invention provides for the placement
of thermoelectric material inside, on top of, on the bottom of or
around a chip package in order to retrofit existing computer chips
in lieu of using chip packages adapted with internal features
provided for by the invention.
[0367] FIG. 25a represents a thermal interface that is built upon
the top of a computer chip.
[0368] FIG. 25b shows a thermal interface that is built on the
bottom of a computer chip.
[0369] FIG. 25c illustrates a thermal interface that is placed
inside a computer chip packaging and makes an appearance at the
edge of the chip packaging.
[0370] FIG. 25d depicts a thermal interface that is built around a
computer chip (i.e. the computer chip package is largely contained
within the thermal interface).
4. Micro-Droplet (a.k.a. "Digital") Microfluidic IC Chip Cooling
and Heat Transport
[0371] Micro-droplet microfluidic cooling is also currently under
research and development, some employing some minor interworking
with thermoelectric devices. Treatment of such approaches employing
planar (two-dimensional) micro-droplet transport can be found in
[1] and the references therein, and approaches employing
three-dimensional and multiple-layer micro-droplet transport are
taught in co-pending U.S. Patent Application 61/599,643.
[0372] In [1] the authors describe first other approaches and
general aspects of controlled electrowetting micro-droplet
transport via microfluidic device structures. For example, FIG.
26a, adapted from the afore-cited text, depicts a side view
representation of a microfluidic electrowetting micro-droplet
transport "chip" implementation fitted over an integrated circuit
package and in turn in thermal contact with an active cooling
element such as a thermoelectric cooler. Additionally, FIG. 26b,
adapted from the afore-cited text, depicts a top view
representation of a number of micro-droplets being transported (via
electrowetted transport) through various straight and
right-angle-turn paths over a planar array of microelectrodes
comprised by such a microfluidic electrowetting micro-droplet
"chip." The micro-droplets are transported over the planar array of
microelectrodes in tightly-controlled fashion by temporally
sequencing the electric potential applied to individual
microelectrodes. The micro-droplets are moved into areas of thermal
contact with portions of a heat-producing integrated circuit dye,
housing, packaging, heat-sink, etc., where they absorb heat and
then are moved to other areas, volumes, or reservoirs where the
absorbed heat can be discharged, for example by means of an active
cooling element such as a thermoelectric cooler. In addition to the
transport of micro-droplets, those authors describe various means
of controlling the surface-area and temporal duration of
micro-droplets exposure to heat sources, droplet routing
strategies, and other innovations. Also useful experimental data
resulting from prototypes are reported, including the fact that
larger droplets with longer exposure times to heat sources perform
cooling functions better than smaller droplets with shorter
exposure times to heat sources.
[0373] FIG. 26c (adapted from [1]) depicts a general planar
microfluidic micro-droplet transport arrangement.
[0374] FIG. 26d (adapted from [1]) depicts a "side view" of
electrowetting transport of a micro-droplet through the
microfluidic transport arrangement via sequencing of potential
applied to microelectrodes 1, 2, 3, and 4.
[0375] FIG. 26e (adapted from [1]) depicts a "top view" motion of
the micro-droplet via electrowetting transport through the
microfluidic transport arrangement of FIG. 26d.
[0376] <<FIG. 27a and FIG. 27b (adapted from [1]) depict
"top" and "side" views of a first heat transfer contact (graded)
modulation scheme explained in [1].
[0377] FIG. 27c (adapted from [1]) depicts a second heat transfer
contact (on-off) modulation scheme explained in [1].
[0378] The arrangements described above can also be applied to
printed circuit boards as taught, for example, in [1] Chapter 6.
Extensions and improvements of these techniques are also possible,
for example as taught in co-pending U.S. Patent Application
61/599,643.
[0379] In [1] Chapter 6 the authors describe adapting microfluidic
electrowetting micro-droplet planar microelectrode array and
micro-droplet transport to implementations using Printed Circuit
Boards ("PCBs"). Two approaches are considered in some detail,
these being the "confined system" represented in FIG. 28b and the
"open system" represented in FIG. 28b. In each of these systems,
micro-droplets are moved into areas of thermal contact with
portions of a heat-producing integrated circuit dye, housing,
packaging, heat-sink, etc., where they absorb heat and then are
moved (via sequencing the electric potential applied to the
microelectrodes) to other areas, volumes, or reservoirs where the
absorbed heat can be discharged. FIGS. 29a and 29b (each adapted
from [1]) depict example routing paths of micro-droplets over the
planar microelectrode array.
[0380] FIG. 30 depicts an representation of the "top" or "bottom"
view an example array of microelectrodes, each microelectrode
rendered as conductor area on a Printed Circuit Board (PCB) and
provided with an associated electrically-conducting "trace" for
electrically connecting the microelectrode to voltage potential
control circuitry, and interspersed between some pairs of
electrodes a physical open hole suitable for a micro-droplet to
travel through. (The subsequent remaining Figures in this section,
namely FIGS. 31-38, depict a side-view representation incorporating
a side-view of the arrangement in FIG. 30.)
[0381] FIG. 31 depicts a side-view representation of an example
two-layer micro-droplet transport arrangement with conduits linking
the two micro-droplet transport region. This Figure incorporates a
side-view of the arrangement like that depicted in FIG. 30. The
view shown in FIG. 30 would herein lie in the center facing
downwards and comprises additional microelectrodes; two of the
physical open holes suitable for a micro-droplet to travel through
depicted in FIG. 5 appear (in side-view) in FIG. 31 as the conduits
linking the two micro-droplet transport region. In the depiction of
FIG. 31, above the upper micro-droplet transport region is a solid
layer of PCB material punctuated with thermally-conducting segments
that conduct heat from the item to be cooled into the upper
micro-droplet transport region. In this example embodiment, the
punctuating thermally-conducting segments are also electrical
conductors configured to serve as an electrical ground plane that
provides both electrical shielding and serves as the ground plane
for forming electric fields for micro-droplet transport via
electrowetting. Also in the depiction of FIG. 31, below the lower
micro-droplet transport region is a solid layer of material (for
example, PCB material) whose upper area comprises an electrical
conductor layer configured to serve as an electrical ground plane
that provides both electrical shielding and serves as the ground
plane for forming electric fields for micro-droplet transport via
electrowetting.
[0382] The arrangements for at least Printed Circuit Boards
described above have many shortcomings, and many of these are
addressed (along with additional advantages) by the
three-dimensional micro-droplet routing arrangements taught in
co-pending U.S. Patent Application 61/599,643.
[0383] However, in the afore-cited text, those authors limit
themselves to planar microelectrode arrays and accordingly planar
micro-droplet transport paths. For a micro-droplet exposed to heat
in central areas of a microelectrode array and which must then be
transported to the edges of the microelectrode array to dispense
the absorbed heat, the micro-droplets can unfortunate radiate heat
back into other portions of the heat-producing integrated circuits.
Those authors allude to methods for minimizing the time over which
unintended heat-radiation can occur by heated microdroplets.
[0384] Further, the afore-cited text does not provide consideration
to avoiding undesired electromagnetic field and electrical field
effects that can interfere with adjacent high-performance
electronic circuitry.
[0385] In addition to these issues and problems, the afore-cited
text only considers the cooling of heat-producing integrated
circuits. Energy harvesting is not considered.
[0386] Accordingly, the reciprocal properties of heat transfer and
energy harvesting (via classical Peltier and Seebeck processes) are
not considered, nor therefore arrangements to implement adaptive
selection between cooling and energy harvesting modalities.
[0387] Additionally, the afore-cited text only considers
traditional semiconductor thermoelectric elements and does not cite
nor anticipate the far higher-efficiency quantum-based
thermoelectric materials such as quantum well and Atvo metals.
These transform classical Peltier and Seebeck processes to vastly
different effects with not only radically improved performance
crossing (for the first time) important application-feasibility
thresholds but also, in many areas, entirely different engineering
and economic tradeoffs.
[0388] The present invention addresses each of these, namely:
[0389] Implementation of 3D micro-droplet transit structures
suitable for thermal cooling and/or energy harvesting applications,
and further doing so in a manner suitable for implementation in
inexpensive multilayer Printed Circuit Boards ("PCBs"); [0390]
Incorporating electrical-field shielding in the above 3D
micro-droplet transit structures and PCB implementations to avoid
undesired electromagnetic field and electrical field effects that
can interfere with adjacent high-performance electronic circuitry;
[0391] Using the above 3D micro-droplet transit structures and PCB
implementations to avoid undesired heat radiation by heated
micro-droplets as they are transported in areas with thermal
contact to the electronic component or other heat-producing
element; [0392] Using the above 3D micro-droplet transit structures
to facilitate arrangements to implement adaptive selection between
cooling and energy harvesting modalities. [0393] Employing
higher-efficiency quantum-based thermoelectric materials, such as
quantum well and Atvo metals, so as to radically improved
performance beyond important application-feasibility thresholds and
access entirely different engineering and economic tradeoffs. The
material below is adapted from U.S. Patent Application
61/599,643.
[0394] FIG. 32 depicts an example a situation where wherein the
momentum of the micro-droplet is not suppressed (i.e., the
micro-droplet is not locked into position under the activated
microelectrode for an interval of time) and the micro-droplet
continues moving a bit beyond the immediate region crowned by the
activated microelectrode. Here the micro-droplet moves towards the
opening of the conduit joining the lower micro-droplet transport
region and the upper micro-droplet transport region.
[0395] FIG. 33 (adapted from U.S. Patent Application 61/599,643)
depicts transmission through a first conduit joining two
droplet-transport layers from a non-heat-gathering-layer to a
heat-gathering-layer wherein the transmission through the conduits
employs a component of capillary forces and electric fields from
distant microelectrodes.
[0396] FIG. 34 (adapted from U.S. Patent Application 61/599,643)
depicts depict transmission through the first conduit joining the
two droplet-transport layers wherein the transmission through the
conduits employs essentially only proximate microelectrodes.
[0397] Microelectrodes can be implemented within the conduits
through a variety of ways, including insertion of prefabricated
cylindrical structures within the conduits. Further, the voltage
potential applied to microelectrodes within the conduit in various
implementations and transport schemes take on different values over
time, for example sometime the electrowetted transport voltage
potential and sometimes the ground plane voltage potential. In some
implementations and transport schemes, other voltage potentials can
also or alternatively be used so as to manipulate the path and
shape of the micro-droplet as advantageous. Other approaches
differing in various ways from that depicted in this series of
figures can also be used and are anticipated by the invention.
[0398] While in the upper transport region the micro-droplet
absorbs heat generated by the item to be cooled through the thermal
conducting layer segment and electrical ground plane, or via other
arrangements in alternate implementations. The absorbed heat in the
resulting heated micro-droplets can then be transported to other
regions where the heat can be processed in various ways (as in the
examples to be described as well as other ways applicable to
various applications and/or alternate embodiments of the
invention).
[0399] FIG. 35 (adapted from U.S. Patent Application 61/599,643)
depicts example transmission through the first conduit joining the
two droplet-transport layers.
[0400] FIG. 36 (adapted from U.S. Patent Application 61/599,643)
depicts the attraction of the micro-droplet to a region immediately
to the right of the second conduit joining the two
droplet-transport regions via activation of the microelectrode
immediately to the right of the second conduit joining the two
droplet-transport regions.
[0401] FIG. 37 (adapted from U.S. Patent Application 61/599,643)
depicts a representation of heat transfer from the previously
heated micro-droplet to the electrical ground plane and further
into the material joined to the electrical ground plane wherein the
material joined to the electrical ground plane comprises a "global"
(large area) thermoelectric structure.
[0402] In various embodiments the thermoelectric structure can be a
thermoelectric cooler, a thermoelectric electric current generator,
or a reciprocal thermoelectric device capable of operating in
either a thermoelectric cooler or a thermoelectric electric current
generator as determined by imposed thermal conditions and
electrical connections to the reciprocal thermoelectric device. In
an embodiment, the role of electrical ground plane (used for
micro-droplet transport) can be served by the electrical conditions
and physical location of a portion of the thermoelectric device
itself (such as electrically conducting material joining two legs
of the thermoelectric device). In some embodiments, the role of
electrical shielding (from electrical field and electromagnetic
generation noise) can also be served by the electrical conditions
and physical location of the same portion of the thermoelectric
device itself. In other embodiments, the role of electrical
shielding can also be served by the electrical conditions and
physical location of another portion of the thermoelectric device
itself. In yet other embodiments, the role of electrical shielding
can also be served by another electrical shielding element.
[0403] Additionally, in some embodiments, the thermoelectric device
can serve as a temperature sensor.
[0404] In some embodiments, the mode of the thermoelectric device
is switched over time. As one example, the thermoelectric device
can be a thermoelectric cooler one moment and a temperature sensor
at another moment. As another example, the thermoelectric device
can be a thermoelectric electric current generator one moment and a
temperature sensor at another moment. As yet another example, the
thermoelectric device can be a thermoelectric cooler one moment and
a thermoelectric electric current generator at another moment. As
still another example, the thermoelectric device can be a
thermoelectric cooler one moment, a temperature sensor at another
moment, and a thermoelectric electric current generator at yet
another moment.
[0405] FIG. 38 (adapted from U.S. Patent Application 61/599,643)
depicts an expanding variation on the arrangement of FIG. 38
wherein the electrical ground plane depicted throughout earlier
figures is replaced by an extended array of local thermoelectric
structures.
[0406] In an embodiment, the role of electrical ground plane can be
served by the electrical conditions and physical location of a
portion of the thermoelectric device itself (such as electrically
conducting material joining two legs of the thermoelectric device),
and the individual portion of each of the plurality of
thermoelectric devices collectively serve as an electrical
equivalent to an electrical ground plane used for micro-droplet
transport. In some embodiments, the role of electrical shielding
(from electrical field and electromagnetic generation noise) can
also be served by the electrical conditions and physical location
of the same portion of the thermoelectric device itself. In other
embodiments, the role of electrical shielding can also be served by
the electrical conditions and physical location of another portion
of the thermoelectric device itself. In yet other embodiments, the
role of electrical shielding can also be served by another
electrical shielding element. In various embodiments, each of the
local thermoelectric structures can be a thermoelectric cooler, a
thermoelectric electric current generator, or a reciprocal
thermoelectric device capable of operating in either a
thermoelectric cooler or a thermoelectric electric current
generator as determined by imposed thermal conditions and
electrical connections to the reciprocal thermoelectric device.
[0407] In some embodiments, all of the local thermoelectric
structures are thermoelectric coolers. In other embodiments, all of
the local thermoelectric structures are thermoelectric electric
current generators.
[0408] In yet other embodiments, each of the local thermoelectric
structures are reciprocal thermoelectric devices capable of
operating in either a thermoelectric cooler or a thermoelectric
electric current generator as determined by imposed thermal
conditions and electrical connections to the reciprocal
thermoelectric device. In some implementations of such (i.e., all
reciprocal thermoelectric device) embodiments, all local
thermoelectric structures are used in the same mode at the same
time. In other implementations of such (i.e., all reciprocal
thermoelectric device) embodiments, a first plurality of local
thermoelectric structures are used in thermoelectric cooler mode at
the same time that a second non-overlapping plurality of local
thermoelectric structures are used in thermoelectric electric
current generator mode. In yet other implementations of such (i.e.,
all reciprocal thermoelectric device) embodiments, each of the
local thermoelectric structures are reciprocal thermoelectric
devices is configured to be independently operable in either a
thermoelectric cooler or a thermoelectric electric current
generator as determined by imposed thermal conditions and
electrical connections to the reciprocal thermoelectric device.
[0409] Additionally, in some embodiments, at least one of the
thermoelectric devices can serve as a temperature sensor.
[0410] In some embodiments, the mode of a given thermoelectric
device is switched over time. As one example, a given
thermoelectric device can be a thermoelectric cooler one moment and
a temperature sensor at another moment. As another example, a given
thermoelectric device can be a thermoelectric electric current
generator one moment and a temperature sensor at another moment. As
yet another example, a given thermoelectric device can be a
thermoelectric cooler one moment and a thermoelectric electric
current generator at another moment. As still another example, a
given thermoelectric device can be a thermoelectric cooler one
moment, a temperature sensor at another moment, and a
thermoelectric electric current generator at yet another
moment.
[0411] Arrangements such as those depicted in FIG. 38 provide a
wide range of capabilities. As one example, local thermoelectric
elements on the left side of the figure could remove heat from
previously-heated micro-droplets and then send the cooled
micro-droplets to the upper level for another cycle of heat
gathering. As another example, the duration of a micro-droplets
exposure to heat in the upper region can be modulated by the
measured temperature of previous heated micro-droplets returning
from that particular area of the item to be cooled. As yet another
example, local thermoelectric elements on the left side of the
figure could pre-cool micro-droplets to below-ambient temperatures
and then send the extra-cool micro-droplets to the upper level for
a cycle of additional heat gathering. Many other capabilities are
made possible by the invention.
5. Modular Hierarchical Structure
[0412] FIG. 39 illustrates the concept that an arbitrary number of
thermal interface stages can be cascaded and arranged
hierarchically so as to remove heat from any number of computer
chips or other heat sources, transferring the heat via an
associated number of heat exchanges and subject to an arbitrary
number of energy harvesting operations. Energy harvesting can occur
in heat exchange and/or heat transfer steps as advantageous.
[0413] FIG. 40 illustrates the concept that heat from an arbitrary
number of computer chips or other heat sources can be transferred
through an arbitrary number of heat exchanges and heat transfers as
proves advantageous.
[0414] FIG. 41a represents a heat-aggregating subsystem wherein
heat generated by an integrated circuit chip is transferred via
circulating cooling fluid or heat pipes through a plurality of
levels of heat handling and aggregation. At one or more of these
levels heat can be either converted to electricity, eliminated via
a fan, or transferred via a heat exchange for further transfer of
the heat for harvest and reuse.
[0415] FIG. 41b represents a combination of any number of
subsystems to form a heat-aggregating system wherein each component
subsystem is comprised as described in FIG. 41a. Such
heat-aggregating systems could be formed by any number of computer
cages; a heat-aggregating system could be comprised on any number
of computer racks; a heat-aggregating system could be made up of
any number of rack clusters; a heat-aggregating system could be
formed by any number of data center floors; a heat-aggregating
system could be created within a data center building; or a
heat-aggregating system could be comprised of any number of data
center buildings.
[0416] FIG. 41c shows multiple heat-aggregating systems wherein any
number of subsystems make up each of the heat-aggregating systems
and each subsystem is comprised as described in FIG. 41a. Multiple
heat-aggregating systems can be formed by combining any number of
computer cages, computer racks, rack clusters, data center floors,
data center buildings or data center complexes.
[0417] In an embodiment, the invention provides for the adaptive
collection of heat or conversion of heat to electricity from at
least one integrated circuit.
[0418] In an embodiment, the invention provides for the adaptive
collection of heat or conversion of heat to electricity from at
least a plurality of integrated circuits forming a central
computing system.
[0419] In an embodiment, the invention provides for the collection
of adaptive heat or conversion of heat to electricity at the scale
of a data center.
5.1 Heat-Transfer Interconnection
[0420] FIG. 42a depicts an abstract thermodynamic representation of
a passive heat transfer arrangement for the transfer heat from a
hotter heat transport system to a cooler heat transport system.
[0421] FIG. 42b depicts an abstract thermodynamic representation of
a heat transfer arrangement transfer heat from a hotter heat
transport system to a cooler heat transport system further
comprising a heat engine, generating energy or work such a system
can be used for opportunistic energy harvesting, for example under
the careful operation of a control system.
[0422] FIG. 42c depicts an abstract thermodynamic representation of
a heat transfer arrangement comprising a heat pump, using energy or
work to unidirectionally transfer heat from a hotter heat transport
system to a cooler heat transport system. Such an arrangement can
be used to improve the rate of heat transfer and prevent a "stall"
in heat transfer should both depicted heat transport systems
operate at roughly the same temperature (for example, in heavy heat
situations).
[0423] FIG. 43a and FIG. 43b depict a two-stage heat transfer
arrangement that can be used to illustrate how the arrangement
represented by FIG. 42b can further be used for opportunistic
energy harvesting, for example under the careful operation of a
control system. At times when there is extra heat build up, one of
the thermoelectric heat engines can be mode switched to act as a
thermoelectric heat engine.
[0424] FIG. 44a illustrates an example thermal interconnection
arrangement by which an arbitrary system (or subsystem) can be
thermally connected to another system (or subsystem) by pressure
contact. FIG. 44b illustrates an example thermal interconnection
arrangement by which an arbitrary system (or subsystem) can be
thermally connected to another system (or subsystem) by
fastener-facilitated contact (for example using threaded fastener
arrangements). FIG. 44c illustrates an example thermal
interconnection arrangement by which an arbitrary system (or
subsystem) can be thermally connected to another system (or
subsystem) by a mating arrangement (for example, spring-spread
pins, friction pins, twist-lock, etc.).
[0425] In embodiments of any of these, or other, example thermal
interconnection arrangements, heat can be transferred from one
system/subsystem to the next via metal or heat-conducting polymers,
ceramics, composites, etc.
[0426] In embodiments of any of these, or other, example thermal
interconnection arrangements, the invention provides for one or
both ends of the thermal interface to be constructed from
thermoelectric materials.
[0427] In embodiments of any of these, or other, example thermal
interconnection arrangements, the invention provides for one or
both ends of the thermal interface to be constructed from arrays of
thermoelectric devices. In an embodiment, the operational modes of
these thermoelectric devices are controlled by a control
system.
[0428] 5.2 Modular Hierarchical Heat Transfer, Energy Harvesting,
and Control Systems
[0429] Various arrangements for modular hierarchical heat transfer,
energy harvesting, and control systems are provided for by the
invention.
[0430] For example, FIG. 45 (adapted from [15]) depicts an
arrangement wherein a thermoelectric device is introduced at the
thermal interface between two closed loop fluid cooling
systems.
[0431] FIG. 46 depicts the component layout of a daisy-chain heat
transfer arrangement employing closed systems of circulating fluids
for use in a hierarchical or peer arrangement. There can be any
number of circulating fluid systems from which heat can be pulled
from any number of heat-generating sources for energy harvest or
transfer; alternatively, any number of cooling fans can be utilized
so that heat transferred from any number of heat-generating sources
can be dispelled into the air.
[0432] In an embodiment, the invention provides for the thermal
interface to send the heat to a fan where air convection can be
utilized to remove the heat, to feed the heat into a heat-transport
interface that relies upon a circulating coolant to remove the
heat, or to both.
[0433] In an embodiment, the invention provides for repeated
hierarchical steps of heat transfer from thermal sources, conducted
through a heat exchange or other thermal interface, and transferred
to a thermal sink. In an embodiment, the hierarchy can comprise use
of the heat gathered at a thermal sink at one hierarchy level to
serve as the heat of the heat source in an adjacent level in the
hierarchy.
[0434] In an embodiment, at any one or more places in the
hierarchy, energy harvesting operations can be introduced.
[0435] In an embodiment, energy harvesting operations convert heat
into electricity.
[0436] In an embodiment, electricity created by energy harvesting
operations is used to provide power for current or future heat
transfer operations.
[0437] In an embodiment, an energy harvesting operation improves
the efficiency of the inventive cooling system.
[0438] In an embodiment, each energy harvesting operations improve
the effectiveness of the inventive cooling system.
[0439] In an embodiment, the invention provides for heat that can
not be efficiently or effectively harvested for energy to be
dispersed via fan(s) at one or more suitable location(s) within the
system.
[0440] In an embodiment, the invention comprises one or more
heat-aggregating system and/or one or more heat-aggregating
subsystems.
[0441] FIG. 47 illustrates two layers of a hierarchical system with
a tree architecture that provides increasing degrees of
hierarchical aggregation at each sequentially lower level. External
control and reporting signal flows are not shown in this figure but
are provided for by the invention.
[0442] In an embodiment, the invention provides for one portion of
the chip-generated heat to be spatially transferred and another
portion of the heat to be energy harvested. The greater the amount
of energy that can be harvested at the heat exchange level, the
better the resultant overall cooling effect and the greater the
overall efficiency of the composite system.
[0443] FIG. 48 illustrates a hierarchical system for heat gathering
compatible with the hierarchical arrangements depicted in FIGS.
2a-2j. At any level aggregated heat can be extracted for various
purposes (such as heat-to-electricity conversions, dissipation by
fans, etc.) as suggested by the dashed lines.
[0444] FIG. 49 illustrates a hierarchical system for
heat-to-electricity conversion compatible with the hierarchical
arrangements depicted in FIGS. 2a-2j. At any level generated
electricity can be extracted for various useful purposes (such as
charging backup batteries, operating cooling fans, supplementing
local power to one or more associated integrated circuits,
circuits, subsystems, boards, etc.) as suggested by the dashed
lines. In an embodiment, the invention provides for concerted
effort to convert as much heat to electricity at the local chip
level as possible.
[0445] FIG. 50 illustrates a hierarchical control system compatible
with the hierarchical arrangements depicted in FIGS. 2a-2j. At any
level control parameters or commands can be provided to the control
system as suggested by the inward dashed lines. At any level
measurements, state variable, or control status can be provided
from the control system as suggested by the outward dashed
lines.
6. Modular Adaptive Multi-Level Control for
Variable-Hierarchy-Structure Hierarchical Systems
[0446] in an embodiment, the invention can comprise various types
of modular adaptive multi-level control for
variable-hierarchy-structure hierarchical systems. Various types of
modular adaptive multi-level control for
variable-hierarchy-structure hierarchical systems applicable for
use in the present invention are taught in co-pending U.S. Patent
Application 61/599,403. Selected material from U.S. Patent
Application 61/599,403 is provided below.
[0447] In a further aspect of the invention, the hierarchical
multiple-level control system comprises a plurality of subsystems,
each with their own control system, that can operate in isolation,
but when interconnected or networked with additional subsystems
associated with other hierarchical levels, each subsystem will
assume their respective role in the hierarchy with respect to
(those) additional subsystems.
6.1 General Topological, Communications, and Hierarchical
Framework
[0448] FIG. 51 depicts a representation of an example hierarchical
multiple-level control system comprising N levels, each level in
the hierarchy comprising a single subsystem.
[0449] FIG. 52 depicts a representation of an example
strictly-layer parent-to-child and child-to-parent communications
between pairs of consecutive subsystem levels in the example
hierarchy depicted in FIG. A.
[0450] FIG. 53 depicts a representation wherein more general
communications between pairs of subsystems in levels in the example
hierarchy is provided for. In one extreme, all subsystems can be
interconnected in a full-mesh topology. In another
[0451] The invention pertains to the area of hierarchical
multiple-level control systems, and more specifically to the design
of subsystems, each with their own control system, that can operate
in isolation but--when interconnected or networked with additional
subsystems associated with other hierarchical levels--will assume
their respective role in the hierarchy with respect to those other
additional subsystems.
[0452] FIG. 54 depicts a variation on the representation of FIG. 53
wherein additionally only some of the subsystems associated with
some of levels in the example hierarchy are present.
[0453] FIG. 55 depicts a variation on the representation of FIG. 52
wherein there are a plurality of subsystems associated with each
level in the example hierarchy. Here strictly-layer parent-to-child
and child-to-parent communications between pairs of consecutive
subsystem levels in the example hierarchy is shown. However, the
invention also provides for more general communications between
pairs of subsystems in levels in the example hierarchy, for example
such as in the arrangements provided in FIG. 53 and FIG. 54.
[0454] FIG. 56 depicts a variation on the representation of FIG. 55
wherein there is at least one subsystem associated with each level
in the example hierarchy. Here again, strictly-layer
parent-to-child and child-to-parent communications between pairs of
consecutive subsystem levels in the example hierarchy is shown.
However, the invention also provides for more general
communications between pairs of subsystems in levels in the example
hierarchy, for example such as in the arrangements provided in FIG.
53 and FIG. 54.
[0455] The invention provides for inclusion of communications among
the subsystems assembled in an aggregate system. In one approach, a
common network can be used. Such a network can be an IP network
(such as cabled or wireless Ethernet.RTM.), a tapped buss (such as
I.sup.2C, Dallas One-Wire.RTM., etc.), USB, fiber, radio, infrared,
power-line carrier (as in X10.RTM.), etc. If cables are used, such
a network can be implemented in a daisy-chain among subsystems,
implemented via connection hubs or switches (Ethernet, USB,
etc.)
[0456] The invention provides for the communications among the
subsystems to include at least one or more of: [0457] Subsystem
presence messages or indications, [0458] Subsystem identification
messages or indications, [0459] Status messages or indications,
[0460] Measurement information to be shared with one or more other
subsystems, [0461] Control information directed to one or more
other subsystems, [0462] Configuration information directed to one
or more other subsystems, [0463] Diagnostics control and
measurement information, [0464] Logging information, [0465] Timing
and/or clock information. 2. Linear Controllers, Bilinear
Controllers, and their Variations
[0466] In an embodiment, the invention provides for hierarchical
multiple-level control system to include linear control systems,
therein permitting the additive control of at least one controller
state variables of one subsystem by control signals generated by or
associated with at least one other subsystem.
[0467] FIG. 56a depicts a representation of an example linear
control system accepting outside control and measurement inputs and
internal feedback paths. The scalar or (more typically) vector
state-variable x of the control system is directed, at least in
some form and/or part, to the control of at least the internals of
the subsystem to which the controller is associated. Typically the
controller is internally comprised within the subsystem to which
the controller is associated, but this is not required. The
controller can be implemented in software, firmware, digital
hardware, analog hardware, or various combinations of these.
[0468] FIG. 57b depicts a representation of an example variation on
the arrangement of FIG. 57a wherein additional inputs are provided
by other subsystems and additional outputs are provided to other
subsystems. Each dashed oval represent operations such as scaling,
offset, dynamical filtering, state-variable selection/suppression,
etc. that can be relevant in various designs, implementations, and
embodiments. The additional input and additional output information
can be exchanged between and/or among subsystems employing one or
more types of communication arrangements described earlier in
Section 1.
[0469] Additionally, the representation depicted in FIG. 57b
provides for changes to parameters and/or configuration of the
controller responsive to the presence or existence of other
subsystems (in other layers of the hierarchy, same layer of the
hierarchy, etc.) as advantageous in various implementations and
embodiments.
[0470] In an embodiment, the invention provides for hierarchical
multiple-level control system to include bilinear control systems,
therein permitting the multiplicative control of at least one
controller state variables of one subsystem by control signals
generated by or associated with at least one other subsystem.
[0471] In an embodiment, the invention provides for hierarchical
multiple-level control system to include bilinear control systems,
therein permitting both (1) additive control of at least one
controller state variables of one subsystem by control signals
generated by or associated with at least one other subsystem and
(2) multiplicative control of at least one controller state
variables of one subsystem by control signals generated by or
associated with at least one other subsystem.
[0472] FIG. 58a depicts a representation of an example bilinear
control system accepting outside control and measurement inputs and
internal feedback paths. The scalar or (more typically) vector
state-variable x of the control system is directed, at least in
some form and/or part, to the control of at least the internals of
the subsystem to which the controller is associated. Typically the
controller is internally comprised within the subsystem to which
the controller is associated, but this is not required. The
controller can be implemented in software, firmware, digital
hardware, analog hardware, or various combinations of these.
[0473] FIG. 58b depicts a representation of an example variation on
the arrangement of FIG. 58a wherein additional inputs are provided
by other subsystems and additional outputs are provided to other
subsystems. Each dashed oval represent operations such as scaling,
offset, dynamical filtering, state-variable selection/suppression,
etc. that can be relevant in various designs, implementations, and
embodiments. The additional input and additional output information
can be exchanged between and/or among subsystems employing one or
more types of communication arrangements described earlier in
Section 1.
[0474] Additionally, the representation depicted in FIG. 58b
provides for changes to parameters and/or configuration of the
controller responsive to the presence or existence of other
subsystems (in other layers of the hierarchy, same layer of the
hierarchy, etc.) as advantageous in various implementations and
embodiments.
3. Nonlinear Controllers
[0475] FIG. 59a depicts a representation of an example nonlinear
control system accepting outside control and measurement inputs and
internal feedback paths. The scalar or (more typically) vector
state-variable x of the control system is directed, at least in
some form and/or part, to the control of at least the internals of
the subsystem to which the controller is associated. Typically the
controller is internally comprised within the subsystem to which
the controller is associated, but this is not required. The
controller can be implemented in software, firmware, digital
hardware, analog hardware, or various combinations of these.
[0476] FIG. 59b depicts a representation of an example variation on
the arrangement of FIG. 59a wherein additional inputs are provided
by other subsystems and additional outputs are provided to other
subsystems. Each dashed oval represent operations such as scaling,
offset, dynamical filtering, state-variable selection/suppression,
etc. that can be relevant in various designs, implementations, and
embodiments. The additional input and additional output information
can be exchanged between and/or among subsystems employing one or
more types of communication arrangements described earlier in
Section 1.
[0477] Additionally, the representation depicted in FIG. 59b
provides for changes to parameters and/or configuration of the
controller responsive to the presence or existence of other
subsystems (in other layers of the hierarchy, same layer of the
hierarchy, etc.) as advantageous in various implementations and
embodiments.
4. Addition of Synthesized Hysteresis to Open-Loop and Closed-Loop
Controllers
[0478] It is common for many control systems, for example those
controlling temperature, motor-controlled position, etc. to
incorporate hysteresis. Additionally, many systems (such motor gear
chains, mechanical thermostats, etc.) inherently comprise
hysteresis processes. Accordingly the invention provides for at
least one of: [0479] Introduction of synthesized hysteresis into
controllers so as to obtain better performance, [0480] Introduction
of synthesized hysteresis into controllers so as to obtain better
stability, [0481] Introduction of synthesized hysteresis into
controllers so as to allow for settling times during parameter or
configuration changes, [0482] Inclusion of synthesized hysteresis
in closed loop controller to compensate for inherently comprise
hysteresis processes within controlled elements, [0483] Other
uses.
[0484] Systems and methods for synthesized hysteresis for use in
control and other systems is taught in, for example U.S. Pat. No.
7,309,828 and pending U.S. patent application Ser. No. 13/186,459.
The synthesized hysteresis can be implemented in software,
firmware, digital hardware, analog hardware, or various
combinations of these.
6. Product Evolution and Phased Deployment
[0485] Recall the hierarchy of environments involved in heat
transfer illustrated in FIGS. 2a-2j.
[0486] In an embodiment, the invention provides for a modular
product hierarchy that can be designed to meet market need and
demand.
[0487] Rather than necessitate replacement of existing system
hardware structures all at once, the innovation will enable phased
replacement as required due to the end of operating life, adequate
degradation, or functional obsolescence.
[0488] Until such replacement or upgrade is enacted, the innovation
can readily be incrementally implemented via incremental retrofit
of computers, chip(s) within individual computers, cages, racks,
etc.
[0489] The modular features used to implement scalability of the
innovation can be implemented in such a way that each modular level
can operate in a stand-alone mode, for example, relying on backup
fans to expel excess heat. This can also provide a failsafe backup
for heat dispersion should some part of a hierarchical deployment
fail.
[0490] In an implementation or a deployment, aspects of the
invention can be implemented at any one or more levels as
determined appropriate in a given situation.
7. Other forms of Operation
[0491] In an embodiment, the invention, the resulting system could
be operated in a way that results in higher operating costs but
which will provide active, controllable heat removal and heat
aggregation for computer systems and data centers.
CLOSING
[0492] The terms "certain embodiments", "an embodiment",
"embodiment", "embodiments", "the embodiment", "the embodiments",
"one or more embodiments", "some embodiments", and "one embodiment"
mean one or more (but not all) embodiments unless expressly
specified otherwise. The terms "including", "comprising", "having"
and variations thereof mean "including but not limited to", unless
expressly specified otherwise. The enumerated listing of items does
not imply that any or all of the items are mutually exclusive,
unless expressly specified otherwise. The terms "a", "an" and "the"
mean "one or more", unless expressly specified otherwise.
[0493] While the invention has been described in detail with
reference to disclosed embodiments, various modifications within
the scope of the invention will be apparent to those of ordinary
skill in this technological field. It is to be appreciated that
features described with respect to one embodiment typically can be
applied to other embodiments.
[0494] The invention can be embodied in other specific forms
without departing from the spirit or essential characteristics
thereof. The present embodiments are therefore to be considered in
all respects as illustrative and not restrictive, the scope of the
invention being indicated by the appended claims rather than by the
foregoing description, and all changes which come within the
meaning and range of equivalency of the claims are therefore
intended to be embraced therein.
[0495] Although exemplary embodiments have been provided in detail,
various changes, substitutions and alternations could be made
thereto without departing from spirit and scope of the disclosed
subject matter as defined by the appended claims. Variations
described for the embodiments may be realized in any combination
desirable for each particular application. Thus particular
limitations and embodiment enhancements described herein, which may
have particular advantages to a particular application, need not be
used for all applications. Also, not all limitations need be
implemented in methods, systems, and apparatuses including one or
more concepts described with relation to the provided embodiments.
Therefore, the invention properly is to be construed with reference
to the claims.
REFERENCES
[0496] [1] Paik, P.; Chakrabarty, K.; Pamula, V. Adaptive Cooling
of Integrated Circuits Using Digital Microfluidics. Artech House,
Inc. Norwood, Me. 2007. ISBN 13: 978-1-59693-138-1. [0497] [2]
Aung, Win. Cooling Techniques for Computers. Hemisphere Publishing
Corporation. New York. 1991. IBSN 0-89116-756-0. [0498] [3]
Rasmussen, Neil. APC by Schneider Electric. White Paper 138,
Revision 7. "Cooling Strategies for Ultra-High Density Racks and
Blade Servers". 2010. [0499] [4] Moss, David, Dell Inc.; Bean, Jr.,
John H., APC by Schneider Electric. APC by Schneider Electric.
White Paper 46, Revision 1. "Energy Impact of Increased Server
Inlet Temperature". 2010. [0500] [5] Bell, Lon E. "Cooling,
Heating, Generating Power, and Recovering Waste Heat with
Thermoelectric Systems". Science magazine. Volume 321. Sep. 12,
2008. [0501] [6] Rowe, D. M. Thermoelectrics Handbook: Macro to
Nano. Taylor & Francis. Boca Raton, Fla. 2006. ISBN
0-8493-2264-2. Chapter 57: Quantum Well Thermoelectric Devices and
Applications. [0502] [7] "The Science Behind Cool Chips.TM., An
Introduction". http://www.coolchips.gi/technology/overview.shtml
(visited Feb. 16, 2012). Cool Chips PLC. 2012. [0503] [8] "Cool
Chips.TM., Technical Overview".
http://www.coolchips.gi/technology/Coolchipstech3Jan06.pdf (visited
Feb. 16, 2012). Cool Chips PLC. 2012. [0504] [9] "Power Chips.TM.,
A Quantum Leap in Power Generation." http://www.powerchips.gi/
(visited Feb. 16, 2012). Power Chips PLC. 2012. [0505] [10] "Avto
Metals--A World of Possibilities". http://www.avtometals.gi/
(visited Feb. 16, 2012). Avto Metals PLC. 2012. [0506] [11] Li, Kam
W. Applied Thermodynamics: Availability Method and Energy
Conversion. Taylor & Francis. Washington, D.C. 1996. ISBN
1-56032-349-3. [0507] [12] Maraquand, C.; Croft, D. Thermofluids:
An Integrated Approach to Thermodynamics and Fluid Mechanics
Principles. John Wiley & Sons. Chichester, West Sussex,
England. 1994. ISBN 0-471-94357-6. [0508] [13] Ioffe, A. F.
Semiconductor Thermoelements and Thermoelectric Cooling. Infosearch
Limited. London. 1957. [0509] [14] Gray. The Dynamic Behavior of
Thermoelectric Devices. The Technology Press of The Massachusetts
Institute of Technology and John Wiley & Sons, Inc. New York.
1960. LCN 60-12981. [0510] [15] Rowe, D. M. Thermoelectrics
Handbook: Macro to Nano. Taylor & Francis. Boca Raton, Fla.
2006. ISBN 0-8493-2264-2. Chapter 59, Section 3: Heat Transfer
Systems. [0511] [16] Burshteyn, A. I. Semiconductor Thermoelectric
Devices. Temple Press Books Ltd. London. 1964. [0512] [17]
Benedicks, Carl Axel Fredrik. The Homogeneous Electro-thermic
Effect (including the Thomson effect as a special case). A. B.
Svenska Teknologforeningens Forlag. Stockholm. 1921. [0513] [18]
MacDonald, D. K. C. Thermoelectricity: An Introduction to the
Principles. John Wiley & Sons. Inc. New York. 1962. LCN
62-10927.
* * * * *
References