A Batteryless 19μW MICS/ISM- Band Energy Harvesting Body Area Sensor Node SoC Fan Zhang, Yanqing Zhang, Jason Silver, Yousef Shakhsheer, Manohar Nagaraju, Alicia Klinefelter, Jagdish Pandey, James Boley, Eric Carlson, Aatmesh Shrivastava, Brian Otis, Benton Calhoun University of Washington, Seattle, WA University of Virginia, Charlottesville, VA 1
41
Embed
A Batteryless 19µW MICS/ISM-Band Energy Harvesting Body Area
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A Batteryless 19µW MICS/ISM-
Band Energy Harvesting Body
Area Sensor Node SoC
Fan Zhang, Yanqing Zhang, Jason Silver, Yousef
Shakhsheer, Manohar Nagaraju, Alicia Klinefelter,
Jagdish Pandey, James Boley, Eric Carlson,
Aatmesh Shrivastava, Brian Otis, Benton Calhoun
University of Washington, Seattle, WA
University of Virginia, Charlottesville, VA
1
• Wireless body area sensor nodes (BASN) enable
inexpensive continuous monitoring of patients
• Battery replacement for body-worn devices may not be
feasible or desirable
Information
Assessment,
Treatment
Introduction
Commercial-off-the-shelf (COTS) wireless BASN
Other Components: Amplifiers, ADC, Power, Interface
• RF Transceiver
• Matching network
• RF crystal
• Antenna connector
(bottom) Radio
• PAVG > 20 mW (continuous transmission)
< 2 hours of battery life (33mAh)
• Memory
• Microcontroller
• Watch Crystal
(top)
Digital Logic Memory
3
Custom low-power BASN IC
Rai, ISSCC 2009. Morrison, DAC 2010
• Digitize & transmit ECG, EMG, neural data
• No signal processing
• Battery life: 3 days (100% duty cycle (DC)) 4
Custom IC Battery
(33 mAh)
Opportunity: energy harvesting
COLD
HOTn
p
LOAD
HEAT FLOW
0
0.01
0.02
0.03
0.04
0.05
5 10 50 100 1 k
Room Temp
Ou
tput
Vo
ltage (
V)
Load Resistance (Ω) 5
On-body measurement of
COTS thermoelectric generator (TEG) (4 x 4 cm2)
0
0.02
0.04
0.06
0.08
0.1
5 Ω 10 Ω 50 Ω 100 Ω 1 kΩ
6°C
ForearmBack of NeckChestShoulderAbdomenLower ThighLower BackUpper Calf
Opportunity: energy harvesting
COLD
HOTn
p
LOAD
HEAT FLOW
0
0.01
0.02
0.03
0.04
0.05
5 10 50 100 1 k
Room Temp
Ou
tput
Vo
ltage (
V)
Load Resistance (Ω)
~60µW ~200µW
6
On-body measurement of
COTS thermoelectric generator (TEG) (4 x 4 cm2)
0
0.02
0.04
0.06
0.08
0.1
5 Ω 10 Ω 50 Ω 100 Ω 1 kΩ
6°C
ForearmBack of NeckChestShoulderAbdomenLower ThighLower BackUpper Calf
PROPOSED:
• NO battery harvest power
• Extract information
• Selective transmission
Our proposed solution
7
CONVENTIONAL:
• Battery
• No signal processing
• Transmission at 100% DC
ADC RFAFE
Boost
Converter
Voltage
Regulation
µController
Memory
DSP
Power Mgmt.
RF Kick-StartTEG
Signal Path
Power Path
VBOOST
ECG
EMG
EEG
Functional diagram
POWER
MANAGEMENT
Regulate VBOOST
Monitor Stoplight
SIGNAL PATH
AFE Processing
RF
DPM
TEG
Boost
ROM
Boot
EN
ER
GY
HA
RV
ES
TIN
G
8
DPM: Digital Power Management
AFE: Analog Front-End
Startup and energy harvesting
RF-kick
POWER
MANAGEMENT
Regulate VBOOST
Monitor Stoplight
SIGNAL PATH
AFE Processing
RF
DPM
TEG
Boost
ROM
Boot
EN
ER
GY
HA
RV
ES
TIN
G
Startup and energy harvesting
RF-kick
POWER
MANAGEMENT
Regulate VBOOST
Monitor Stoplight
SIGNAL PATH
AFE Processing
RF
DPM
TEG
Boost
ROM
Boot
EN
ER
GY
HA
RV
ES
TIN
G
POR
Signal processing path
11
TEG
Boost
ROM
Boot
POWER
MANAGEMENT
Regulate VBOOST
Monitor Stoplight
EN
ER
GY
HA
RV
ES
TIN
G
SIGNAL PATH
AFE Processing
RF
DPM
Signal processing path
12
TEG
Boost
ROM
Boot
POWER
MANAGEMENT
Regulate VBOOST
Monitor Stoplight
EN
ER
GY
HA
RV
ES
TIN
G
SIGNAL PATH
AFE Processing
RF
DPM
Flexible signal path
Power management
13
TEG
Boost
ROM
Boot
POWER
MANAGEMENT
Regulate VBOOST
Monitor Stoplight
EN
ER
GY
HA
RV
ES
TIN
G
SIGNAL PATH
AFE Processing
RF
DPM
Power management
14
TEG
Boost
ROM
Boot
POWER
MANAGEMENT
Regulate VBOOST
Monitor Stoplight
EN
ER
GY
HA
RV
ES
TIN
G
SIGNAL PATH
AFE Processing
RF
DPM
DVS
System block diagram
15
Energy harvesting
16
ROM
Hybrid energy harvesting
VBOOST: 1.35V
6 stages
Rectifier
Over-Voltage Clamp
N1
Boost Converter
RF:
-10 dBm
VTEG:
30 mV
Gate
Control
TEG
Power Mgmt.
Storage CapN1
Power-On Reset (POR)
Off-chip
Off-chip
17
Capable of thermal, photovoltaic, and/or RF energy harvesting
1cm
• Measured efficiency of 38% for Vin=30mV, Vout=1.35V
• Requires minimum voltage (VKILL) to sustain conversion
18
Boost converter
Carlson, Strunz, Otis, JSSC 2010
Measured RF kick-start
0 2 4 6 8 100
0.5
1
1.5
time (s)
Vo
ltag
e (
V)
RF Pulse, -10dBm
Boost Converter
turns on
VBOOST
VTEG
• Wireless RF pulse provides one-time kick-start
• The node runs indefinitely thereafter 19
Boot-up sequence
• POR issues reset at 1.0V
• Upon boot-up, the chip fetches instructions from the ROM
(default) or the RAM
• If VBOOST < VKILL , RF burst can ‘revive’ chip to default
algorithm in ROM
POR
Programmed?
132B ROM
(AFib Detection)
1404B RAM
(Reprogrammable)
Yes No
1.2 V
1 V
VBOOST POR out
20
Instruction memory (IMEM)
Signal path
21
Analog front-end (AFE)
22
AmpOut+
AmpOut-
LNA
Cs
Cf
In+
In-
AmpIn+
AmpIn- GND
Cfilter
Out+
Out-
Gm Gm
Cfilter
Gm Gm
Chopper
modulatorPseudo-resistor
Cs
Cf
Cfilter
Cfilter
VgaOutVGA
CsVgaIn+
VgaIn-Cs
Cf
GmVCM
AmpIn+
Vbias1
Vbias2
VDD
GND
AmpIn-
Vbias3
Am
pO
ut-
Am
pO
ut+
Vcs
CMFB
VgaIn+
Vbias4
Vbias5
VDD
GND
VgaIn-
Vbias6
Vg
aO
ut
Vcs
Cf
Gm
• Chopper-stabilized
low-noise amp
• Variable-gain amp
• 6 programmable
gain 40 – 78 dB
• 3 µW / channel
1 of 4 channels
Flexible signal path
23
Example of Mixed Path
ENV
DetectFIR
Example Custom Path
RR+
AFibFIR
Generic Path
MCU
Flexible Architecture for Data Processing
MCU
Event-Based Burst
Store and Burst
Stream
If event
Processed
Data
Flexible Architecture for Data Transmission
Data for TX
4kB
DMem
4kB
DMem
ECG
EMG
EEG
AFE
Data processing Data transmission
• Data processing: max flexibility (generic path) or max efficiency (biosignal accelerators)
• Data transmission: supports modes from streaming (100% DC) to rare event detection (~0% DC)
MCU: microcontroller
TX
MCU
RR+AFib Accel.
30-Tap FIR Accel.
0 50 100 150 200 Delay (µs)
Measure
d E
nerg
y/O
p (
pJ)
Energy efficient accelerators
24
Accelerators:
• Programmable FIR
• Heart rate (R-R) extraction
• Atrial Fibrillation (AFib) detection
• Band energy extraction
• Direct memory access (DMA)
• Packetizer
Energy Efficiency / Sample
110x
6800x
4000x
0
1
2
3
4
30 Tap FIR MCU 6.3 nJ
Accel 57.6 pJ
Env. Detect MCU 3.6 nJ
Accel 530 fJ
R-R Extract MCU 12 pJ
Accel 3 fJ
25
DPM: signal path control
ADC
DMA/SRAM
Bio-signal
Accelerators
Packetizer
Power and
Channel control
Sa
mp
ling
ra
te c
on
tro
l
Power/clock gate,
clock rate, and
bus control
Du
ty c
ycle
, d
ata
ra
te c
on
tro
l
Dig
itiz
ed V
BO
OS
T
DPM Chip
program
LNA
VBOOST
VGA
IMEM
MCU vs. DPM
26
DPM
MCU
IMEM
• Execution of instructions toggles automatically
between MCU and DPM.
Operation DPM Energy MCU Energy
NOP 0.7 pJ 1.46 pJ
Control Signals 2.8 pJ 2.92 pJ
Branch Commands 2.9 pJ 4.38 pJ
Generic processing
(e.g. add, multiple)
Control instructions
(e.g. power/clock-gate)
Frequency multiplying TX
Pandey, Otis, JSSC’11
27
• Frequency multiplication:
synthesis at low frequency, transmission at high frequency
• Edge-combiner based frequency multiplier
• ILRO-based edge generator
• PA integrates the edge combiner
BFSK data modulation
FSK Data
CL CL
ΔC
• Quartz reference clock is pulled using DC (~200ppm)
• Δf is multiplied by 9x (~100kHz)
“1” “0”
28
Measured Data Modulation
Power management
29
0.5 V
Bandgap
0.5
V0
.5V
1.0
V
1.2
V
F1
F1
F2
F2
0.25-1.0V
VBOOST: 1.35V Linear Regulators Switched-Cap Regulator
Bias Gen
VREF, IREF
30
Voltage regulation
1.2V PADS, AFE
1.0V TX LO
0.5V TX PA
0.5V DPM, MEM, ACCEL
Variable 0.25-1.0V in 50mV, enabling DVS.
Power management scheme
VREF
MU
XChannel 1
VBOOST
Channel 2
Channel 3Channel 4
ADC
SEL
DPM
• VBOOST is scaled and digitized
• DPM compares VBOOST vs. VTHR (programmable)
• DPM chooses reconfigurable modes (“stoplight”)
Closed-loop power management
VTEG (V)
VBoost VDD, AFE
VGA Out (V)
Supply (V)
TX Duty-
Cycle
Green Yellow Red Yellow Green
Time (s)
32
MODES AFE Process Data Mem. Inst. Mem. Transmit VTHR