energy harvesting for wireless …eh-network.org/events/dissemination2011/presentations/rob van...©...

© IMEC 2011 / CONFIDENTIAL

ENERGY HARVESTING FOR WIRELESS

AUTONOMOUS SENSOR SYSTEMS

Rob van Schaijk

© IMEC 2011 / CONFIDENTIAL < 2

“Blood pressure too high”

“Traffic jam ahead”

“I’m sensing corrosion”

“We’re ripe”

“I’m here Mummy”

“I’m all out of milk”

“Time for walkies”

“We’re 98% full”

“You left me here”

“Send me energy”

SMART SYSTEMS EVERYWHERE


THE WIRELESS AUTONOMOUS SENSOR NODE

3

S10µW

A10µW

FrontEnd

20µW DSP20µW

Radio20µW

Micropower System -100µW

µP20µW

Thermal, Vibrational, RF, Light, Bio-chemical

NonElectrical

World


CAN WE STICK WITH BATTERIES?

Few specifications:- Smart systems everywhere Large numbers Accessibility to devices Battery replacement is not always an option Device autonomy exceeds lifetime Size and Weight important factor

4

Thin-film-flexible Lithium-coin PrintableLithium-flexible SupercapacitorThin-film-flexible Lithium-flexible


ONLY BATTERY

5

Primary (Li-ion)

Rechargeable (Li-ion)Thin film

capacitor

Harvester + energy storage

1cm3 size

Small battery alone do not offer autonomy


ONLY HARVESTER

Energy buffer in sensor system is essential.- NO buffer: harvested energy = Peak energy- With buffer: harvested energy = Average Energy

A small battery or Super Capacitor is therefore needed

6

25

20

15

10

5

0

-5

Curre

nt (m

A)

0.250.200.150.100.050.00-0.05

Time (s)

SLEEP SLEEP

RECEIVE

TRANSMIT

SENSING and

POWER CONTROL

SLEEP

Peak Energy

Average Energy

A typical energy consumption scenario


Man4 µW/cm2

Machine100 µW/cm2

WiFi0.01 µW/cm2

GSM0.1 µW/cm2

Machine1-10 mW/cm2

Man20 µW/cm2

Indoor10 µW/cm2

Outdoor10 mW/cm2

Power levels “MEMS” based harvesters*

ThermalPhotovoltaic RFVibration

*Vullers et al, Micropower Energy Harvesting, Solid-State Electronics 53 (7) Pgs 684-693, DOI: 10.1016/j.sse.2008.12.011


Smart package (Perishables)

BAN

Smart buildings

APPLICATION FIELDS ENERGY HARVESTING

8

Predictive MaintenanceTPMS

RF

ThermalVibrational

ADVANCED ENERGY HARVESTING TECHNIQUES

© IMEC 2011 / CONFIDENTIAL 9

“NOT ENOUGH POWER IS

GENERATED”

you consume

too much !

you don’t generate enough !

And this largely explains the absence of killer applications

using energy harvesting


THE ENERGY BALANCE

For a successful introduction of MEMS based Energy Harvester: The Power usage needs to be reduced

- Of the shelf components use ‘too’ much power- Power optimization needed towards ultra low power

Energy harvesters have to increase power output- Increase of harvesting efficiency- Increase of conversion efficiency -> Power Management is key!

Sensor

Read-out

DSP

Radio

Energy Harvester


COST REDUCTION THROUGH MICRO-SYSTEM APPROACH

< 11

Size (or Cost)

Pow

er/u

nit

μW

mW

μm mm

Fine-machining

Micromachining

http://www.micropelt.com/applications/power_generation.php


SOMETHING ELSE IS (REALLY) WRONG WITH ENERGY HARVESTING TODAY: SCALING !

P.D. Mitcheson et al.; Proceedings of the IEEE, 96, No. 9, pp. 1 (2008)

Hmeas

norm EPPP ==

max

3161 3

4

ωρ VolY

PFOMAuo

measV =

± PRESENT

NEEDED (AT LEAST)

DISPROPORTIONAL SCALING WITH MINIATURIZATION (CASE OF VIBRATIONAL ENERGY HARVESTING)


Advanced

energy harvest

ing

Man4 µW/cm2

Machine100 µW/cm2

WiFi0.01 µW/cm2

GSM0.1 µW/cm2

Machine1-10 mW/cm2

Man20 µW/cm2

Indoor10 µW/cm2

Outdoor10 mW/cm2

Power levels “MEMS” based harvesters


< 13

© IMEC 2011 / CONFIDENTIAL < 14

VIBRATION HARVESTING: THREE MECHANISMS

Permanent magnet

coil

springmass

Vout

ElectrostaticElectro-

magnetic Piezoelectric

14

http://en.wikipedia.org/wiki/Image:SchemaPiezo.gif


10-7

10-6

10-5

10-4

Out

put p

ower

(W)

3 4 5 6 7 8 91000

2

Frequency (Hz)

damp=0.016 damp=0.075 damp=0.133

Mechanical power generation: General principle

Possible range:10 -100 µW/cm2

Velocity damped resonator Maximum power conditions

Optimal transformation of mechanical energy into electricity occurs at the resonance frequency of the harvester!

Power

LX

m

LXmP

0

0

030

0

2

4)(

==

=

ωγς

ωω

X0: vibration amplitude L: system dimensionm: system massw0: resonance frequency

15


INTELLIGENT TIRE

16

Radial mode

80km/h

• Intelligent tire: measurement of forces for improvement of active safety systems

• Resonant excitation ↔ high acceleration shock

• Shock → self resonance → ring down mode

80km/h → ~20 shocks/sec

300G shock → comparable to resonance at 1G for Q=300

Source: Pirelli tire company Ring down signal after shock (16G for 0.5ms) in matched load resistor

Road

Tyre

Forces

Friction coefficient between tire and road → improvement of active safety systems


MEMS BASED PIEZOELECTRIC ENERGY HARVESTERS

17

Acceleration force

mass

beam

Piezo capacitor

Motion changes polarization

PZT: 40µ[email protected] (2.5g)

AlN: 60µW@572Hz (2.0g)

World Record output power

2007

2008

AlN power vs frequency

2009

Aiming for 500µW

AlN: 100µW@572Hz (1.0g)

2010 AlN: 225µW@929Hz (2.5g)

0

50

100

150

200

250

920 925 930 935 940

Pow

er [µ

W]

frequency [Hz]IEDM 2009: R. Elfrink et al., “First autonomous wireless sensor node powered by a vacuum-packaged piezoelectric MEMS energy harvester”

http://en.wikipedia.org/wiki/Image:SchemaPiezo.gif


PIEZOELECTRIC MATERIALS

SolGel

PZT • not IC fab compatible• SolGel

- limited thickness- low reproducibility- no topography allowed

• Sputtering- low deposition rate- composition control

sputtering

AlN • IC fab compatible • Sputtering

- high deposition rate- stoichiometric

Aluminum Nitride is good candidate for energy harvester

r

eFOMε

231=

e31 ~ -8 -12 C/m2

er ~ 1000tan d ~ 0.03 high lossZ ~ few kΩ < 0.1 Volt

e31 ~ -1.1 C/m2

er ~ 10tan d ~ 0.003 low lossZ ~ few MΩ few Volt

18


WAFER LEVEL PACKAGE

• SU-8 on rolling wheel

• wafer at elevated temperature (~80C)

• topography is no issue

• vacuum inside package

0.0001

0.001

0.01

0.1

1

10

100

985 990 995 1000 1005 1010Frequency [Hz]

Pow

er [µ

W]

0.1 g atmospheric0.1 g vacuum1.0 g atmospheric1.0 g vacuum

Piezocapacitor

5 mm

19


HARVESTER EFFECTIVENESS OF REPORTED DEVICES

P.D. Mitcheson et al.; Proceedings of the IEEE, 96, No. 9, pp. 1 (2008)

Hmeas

norm EPPP ==

max


LEAKAGE OF WAFER LEVEL PACKAGE Reduction of power due to air leak into package ~50% power loss in half year time Hermetic wafer level package (long term vacuum to

minimize air damping)- Improvements of polymer based bonding- extra (metal) barriers- extra cavity - molding layer- Eutectic bonding

0.0

0.5

1.0

1.5

2.0

2.5

3.0

518 519 520 521 522 523 524 525Frequency [Hz]

Pow

er [µ

W]

t=0t=10 dayst=60 dayst=140 dayst=220 days

Piezocapacitor

SU-8 polymeric bond

air in leak

21


Man4 µW/cm2

Machine100 µW/cm2

WiFi0.01 µW/cm2

GSM0.1 µW/cm2

Machine1-10 mW/cm2

Man20 µW/cm2

Indoor10 µW/cm2

Outdoor10 mW/cm2




Seebeck effect (mechanism) Thermocouple structure

THERMO ELECTRIC ENERGY HARVESTING

Seebeck effect: Voc = Seebeck * DTHuman body DT ~ 1-3 K

Voltage output typically few mV


MEMS BASED TEG ENERGY HARVESTERS; RESULTS

Planar 250 mV/K

In-plane 92 mV/K

World Record output voltage

2007

2008

Experimental result

2009

Aiming for 1V/K

High-topography 258 mV/K

High-topography thermocouple

Planar High-topographyIn-plane


HISTORY OF MICROPOWERDEMONSTRATORS

• show feasibility of WATS systems for body area networks

• use of off-the-shelf components

• supporting research by investigate limits

• Includes new technology and circuit designs developed by IMEC

•Has resulted in Frost & Sullivan Award for ECG shirt


Man4 µW/cm2

Machine100 µW/cm2

WiFi0.01 µW/cm2

GSM0.1 µW/cm2

Machine1-10 mW/cm2

Man20 µW/cm2

Indoor10 µW/cm2

Outdoor10 mW/cm2




RF ENERGY TRANSFER INSTEAD : THREE MECHANISMS

•Inductive coupling•Close contact

27

Receiver coil

Vout

Transmit coil

Vin

Vin Vout

Transmit coil

Receiver coil

•Non-radiative coupling•Intermediate distances

•Radiative Coupling•Large distances

( ) 22

2

4 rGGPP RTT

inc πλ

=Power transfer: Friis equationλ: Wavelength usedGr, GT: Antenna Gain

PTPR; Power transmitted, received


DCDC

boostconverter

battery

Rectenna Charger

For use in WSN, battery needs to be recharged:Minumum voltage is needed for IC = 0.2V→Vout determines distance→ Pinc determines charging time

RF BATTERY CHARGING

Vout, PoutPinc

28


SMALL-SIZE RECTENNAAPPLICATIONSRF Battery charging

29

Charging 4.2V Li-Ion battery at 60 cm with EIRP=1.2W at 2.45GHz

Charging 2.9V Li-Ion battery at 166 cm with EIRP=1.2W at 2.45GHz


< 30

Buck-boost converterVin-min=0.21V, Vout=3.0-4.2V

Theoretical distances for single-cell rectennas

More cells is larger distance


SUMMARY

For autonomous wireless sensor system one needs:- Small low cost energy harvester- Power optimization of complete sensor system- Harvester + power management + energy storage

31

MEMS technology- Capable of 100µW/cm2

- Key for mass application

Still in research phase- Higher power output needed- Reliability and robustness


THANK YOU FOR YOUR ATTENTION.

energy harvesting for wireless …eh-network.org/events/dissemination2011/presentations/rob van...©...

Documents