Design and Testing of a Self-Powered Wireless Hydrogen Sensing Platform

Jerry Chun-Pai Jun,

Jenshan Lin, Hung-Tan Wang Fan Ren, Stephen Pearton and Toshikazu Nishida University of Florida

Motivation Behind a Self-Powered Wireless Hydrogen Sensing Platform

• Popular topic due to need of inexpensive sensor devices requiring minimal maintenance to monitor harsh and dangerous environs.

• Growing interest in hydrogen as a fuel cell, which is dangerous if not properly contained.

• Combustion gas detection in Spacecrafts and Proton Exchange Membrane (PEM) Fuel Cells • Greater than 4% of hydrogen concentrations are explosive.

Limitations Of Sensor Development

• Limitations of Energy Harvesting Devices • Limitations of Low-Power and Low Voltage Commercial Components • Limitations of a Wireless System – Wireless Channel Estimation – FCC Regulations

Energy Harvesting Techniques

Solar Energy Harvesting

• Solar Cells are a mature commercial Product • Dependent upon real time lighting and temperature conditions • Pulse Resonant Power Converter – Self-powered and self controlled – Convert input voltage of 0.8-1.2V to steady 2V output

Vibration Energy Harvesting

• Collection of energy proportional to volume of device • Limited to magnitude and frequency of vibrations • For Proof of Concept – PSI D220-A4-203YB Double Quick Mounted Y-Pole PZT Device – Direct Charging Circuit

Energy Harvesting Techniques cont.

Solar Energy Harvesting

IXOLAR XOD17 04B Solar Cell

Vibration Energy Harvesting

Pulse Resonant Power Converter Functional Block Diagram (a) Bare die photo (b) Four mounted PSI D220-A4-203YB Double Quick Mounted Y-Pole Bender (a) Direct Charging Circuit (b)

ZnO Nano-Rods as a Sensing Mechanism

Al 2 O 3 Substrate

Schematic of Multiple ZnO Nano-Rods Close-Up of Packaged ZnO Nano Rod Sensor • ZnO currently used for detection of humidity, UV light and gas detection • Easy to synthesize on a plethora of substrates • Bio-safe characteristics • Large chemically sensitive surface to volume ratio • If coated with Pt or Pd, can increase device’s sensitivity to hydrogen • High compatibility to microelectronic devices

Pt-ZnO Nano-Rod Sensors

Pt-coated ZnO Nano-Rod - Relative Resistance Change for Various Hydrogen Concentrations

• Sputtered with Pt coatings of approximately 10 Å in thickness • Show no response to the presence of O 2 and N 2 at room temperature • Pt increases conductivity of Nano-Rods • Up to 8% change in resistance after 10 min. exposure to 500 PPM of hydrogen • Greater than 2% change in resistance after 10 min exposure to 10 PPM of hydrogen • 90% recovery within 20 seconds upon removal of hydrogen from the ambient

Comparison of ZnO Nano-Rods Coated with Different Metals

4 2 8 500ppm H 2 6 Air Pt Pd Au Ag Ti Ni 0 0 5 10 15 20 Time(min) 25 30

Relative Resistance Change for Various Metal-coated ZnO Nano-Rods

Differential Measurement

V s

R 2 R 2

V 1 R 1 V g R 3 V 2 R 2 R 4

R 3 R 3

• Wheatstone Resistive Bridge – Can limit current consumption of resistive bridge – Best way to detect changes in resistance • Difference Amplifier – Using differential architecture of operational amplifier to subtract difference at input, and apply gain – Form of differential measurement

V 1

Instrumentation Amplifier

R g R 1 R 2 V 3 R 3 R 2 R 3 V 2 V 4

V OUT

 (

V

2 

V

1 ) 1  2 

R

1

R g

 

R R

2 3

V OUT

• Provides High Impedance Input Buffers isolate V 1 V 2 from resistive and network of difference amplifier • Buffers and provides gain before difference amplifier • Gain can be easily adjusted by varying a single resistor, R g .

Differential Detection Circuit

R 2 R 2 R 3 R 3 V OUT • Since Pt-ZnO Nano-Rod devices react to both hydrogen and temperature, the use of a passivated ZnO as a reference resistor can mitigate the temperature dependency of the differential Detection Circuit.

• Rbias used to limit current flowing into both legs of resistive bridge • Maintains concept of a differential measurement • Instrumentation Amplifier helps balance input offset voltages, while providing gain, and conditioning signal for ADC

Fabricated Pt-ZnO Nano-Rod for Use in Differential Detection Circuit

ZnO with increase Pt catalyst

1580 1560 1540 1520 1500 1480 1460 1440 1420 1400 0.

00 1.

75 3.

50 5.

25 7.

00 8.

75 10 .5

0 12 .2

5 14 .0

0 15 .7

5 17 .5

0 19 .3

3 21 .0

0 22 .8

3 24 .5

0

time(min)

Fabricated Differential Detection Circuit

Fabricated Differential Detection Circuit

400 300 200 100 0 1460

Output voltage vs sweep of exposed Pt-ZnO Nominal Resistance

1480 1500 1520

Nominal Resistance (Ohms)

1540 1560

Microcontroller Selection

Type of Program Memory Program Memory RAM I/O Pins ADC Interface

Flash 8 kB 256 Bytes 22 pins 10-bit SAR ( successive approximation register ) 1 Hardware SPI or UART, Timer UART

Supply Voltage Range

1.8 V – 3.6 V

Active Mode Standby Mode # of Power Saving Modes

200uA @ 1 MHz, 2.2 Vsupply 0.7 uA 5

Features of Texas Instruments’ MSP430F1232IPW

REQUIREMENTS

• • • • • • •

Low-Voltage Low-Active Current Low-Sleep Current Onboard Memory Onboard ADC Serial Output Reprogrammable

Microcontroller Operation

• • •

Level Monitoring State Machine

Runs through state until a discernable presence of hydrogen is detected.

Once hydrogen is detected, microcontroller forces RF front-end to transmit an emergency pulse to the central monitoring station before returning back to an idle mode.

Hydrogen threshold level is at far less than dangerous levels • • • •

Data Transmission State Machine

Runs through states until a discernable presence of hydrogen is detected.

Once threshold is detected, the data from the ADC is queued onto the serial output port of the microcontroller to be transmitted.

Once transmitted, state is reset to sleep For constant tracking of hydrogen levels

Selection of a Modulation Technique

Comparison of Complexity between π/4 DQPSK and OOK • • • • •

MODULATION REQUIREMENTS

RF Power Amplifiers and Oscillators have efficiencies of 50% at best Low parts count Low Duty-Cycle, Low Data Rate.

Expend energy only for transmission of Data Low complexity

Selection of RF Transmitter (1)

Ming TX-99 Transmitter in OOK Mode

300 MHz Ming TX-99

• Onboard antenna • OOK Modulation • Low Part Count • Low Complexity • Tunable Frequency • Colpitts Oscillator Ming TX-99 Transmitter

Selection of RF Receiver (1)

Ming RE-99 Receiver Schematic

300 MHz Ming RE-99

• Onboard antenna • External Antenna Tap • Low Part Count • Low Complexity • Tunable Frequency • Envelope Detection • Little Documentation Ming RE-99 Receiver

Distance Measurements

Receiver Hallway Hallway Transmitter • • • • 0 m 20 m 3.5 m 10 m Layout of Testing Room

Distance (m)

5 10 15 -35 0 -45 -55 -65 -75 20

Received Power vs. Distance With Reference to Room Shape Shape of room resulted in a wave-guide effect at 10 meters Last successful data transfer occurred at 19.4 m Received power at this distance was approximately -70 dBm Can assume Ming RE-99 Receiver sensitivity is approximately -70 dBm

Distance (m) Test Setup Maximum Transmission Distances Received Power at 1m Received Power at 8m

Central Monitoring Station

Moving Average Filter Example Labview Block Diagram Code and Labview Front Panel Gui

• At the time, used Ming RE-99 Receiver • NI USB-6008 DAQ device for power to Receiver, and ADC to capture data • Powered from HP Laptop’s USB Port Running LabVIEW 7.1

• Moving Average Filter to differentiate data “pulse” from noise

Full System Integration and Testing

or Schematic of Hydrogen Chamber Schematic of Hydrogen Chamber

Future Work: New Receiver

System Level Architecture for RXM-315-LR

Linx Technologies RXM-

business

315-LR

• Replacement for Ming RE-99 since Rayming Corp. went out of • OOK Modulation • Low Part Count • Low Complexity • RSSI/PDN • -112 dBm Sensitivity Pin-Out of RXM-315-LR receiver, and receiver test board, shown with SPLATCH antenna

Future Work: Low-Profile Antenna

‘

SPLATCH’ dimensions, matched S parameters Antenna Test Board w/ Matching Circuit

• • • •

Linx Technologies ANT 315 SP ‘SPLATCH’ Style Antenna Grounded Line, Microstrip Monopole Antenna After matching, -9dB gain, trade off for low profile antenna 5 MHz -10 dB BW, Center Frequency = 315 MHz

Future Work: Minimum Redundancy Minimum Energy Coding

CODED – 1 “high” CODED – 1 “high-delay” CODED -2 “high” Proposed Source Coding Technique

• • •

Mapping (n) source bits to message with a maximum of 2, or 3 “high” bits

–

Example: 6 source bits 6 source bits = 64 messages (symbols)Find Codeword of length (m) that allow for 64 symbols, with a maximum of 3 high bits.

–

64 = m C 3 + m C 2 + m C 1 + m C 0 ; m = 7 Power Reduction

–

Assumptions: for now, all source code symbols have equal probability of occurrences, and power is only consumed with the transmission of a high bit.

–

So, Power Consumption Reduction is:

%

Power Reduced

 |  #

avgsourcehighbits

 #   #

avgcodedhighbits avgsourcehighbits

| g 100

By using a minimum energy coding technique, we can expect to reduce the power required to transmit an un-coded message by 20 to 40 percent.

Minimum Redundancy Minimum

70 80 90

Energy Coding (cont.)

Power Consumption Reduction per Additional Redundant Bit

3 high 2 high 1 high 60 1 delay 50 40 30 20 10 0 3 4 5 6 7 8

Original Source Bit Length

9 10

Conclusions

• Successfully designed a low-power sensor interface for the Pt-ZnO Nano-Rod hydrogen sensing mechanism • In conjunction with the microcontroller, RF transmitter, and separate energy harvesting techniques, were successful in detecting and reporting the presence of 500 PPM of H2 in N2. (.05%) using Pt-ZnO Nano-rods as our sensing mechanism • Energy harvesting techniques include solar and vibration energy devices.