University of Central FloridaSchool of Electrical Engineering and Computer Science
1
Passive, Wireless Surface Passive, Wireless Surface Acoustic Wave Technology Acoustic Wave Technology for Identification and Multi-for Identification and Multi-
Sensor SystemsSensor SystemsD. C. Malocha
School of Electrical Engineering & Computer Science, University of Central Florida,
Orlando, Fl., [email protected]
Piezoelectric Substrate
f1 f4 f6 f0f2 f5 f3
0 1 2 3 4 5 6 71
0.5
0
0.5
1
Normalized Time (Chip Lengths)
University of Central FloridaSchool of Electrical Engineering and Computer Science
2
What is a typical SAW Device?What is a typical SAW Device?• A solid state device
– Converts electrical energy into a mechanical wave on a single crystal substrate
– Provides very complex signal processing in a very small volume
• It is estimated that approximately 4 billion SAW devices are produced each year
Applications:Cellular phones and TV (largest market)Military (Radar, filters, advanced systemsCurrently emerging – sensors, RFID
University of Central FloridaSchool of Electrical Engineering and Computer Science
3
SAW IntroductionSAW Introduction• Operates from 10MHz to 3 GHz
• Fabricated using IC technology
• Manufactured on piezoelectric substrates
• Operate from cryogenic to 1000 oC
• Small, cheap, rugged, high performance
2mm
10mm
Quartz Filter SAW packaged filter showing 2 transducers, bus bars, bonding, etc.
University of Central FloridaSchool of Electrical Engineering and Computer Science
4
General SAW BackgroundGeneral SAW Background
2.5mm
10mm
LiNbO3 Filter
From: Siemens
University of Central FloridaSchool of Electrical Engineering and Computer Science
5
SAW AdvantageSAW Advantage
University of Central FloridaSchool of Electrical Engineering and Computer Science
6
Basic Operation of a SAW Basic Operation of a SAW Electromechanical TransductionElectromechanical Transduction
Velocity*time=distance
Velocity=distance/time= T
A SAW transducer is a mapping of time into spatial distance on the substrate
University of Central FloridaSchool of Electrical Engineering and Computer Science
7
SAW Reflector ArraySAW Reflector Array
With ¼ wavelength electrodes, all reflections add in phase (synchronous) which makes a distributed reflector. This is an acoustic mirror. Perturbation at each electrode is small which minimizes losses and mode conversion (BAW generation)
SAW Input
SAW Output
University of Central FloridaSchool of Electrical Engineering and Computer Science
8
SAW Tag & Sensor AdvantagesSAW Tag & Sensor Advantages• Extremely robust
• Operating temperature range: cryogenic to ~1000 oC• Radiation hard, solid state
• Wireless and passive• Coding and spread spectrum embodiments
• Security in coding• Multi-sensors or tags can be interrogated
• Wide range of sensors in a single platform• Temperature, pressure, liquid, gas, etc.
University of Central FloridaSchool of Electrical Engineering and Computer Science
9
Why Use SAW Sensors and Tags?Why Use SAW Sensors and Tags?
• External stimuli affects device parameters (frequency, phase, amplitude, delay)
• Operate from cryogenic to >1000oC
• Ability to both measure a stimuli and to wirelessly, passively transmit information
• Frequency range ~10 MHz – 4 GHz
• Monolithic structure fabricated with current IC photolithography techniques, small, rugged
University of Central FloridaSchool of Electrical Engineering and Computer Science
10
Response of SAW Reflector Test StructureResponse of SAW Reflector Test Structure
Measurement of S21 using a swept frequency provides the required data.
62 64 66 68 70 72 74 76 78 80-90
-80
-70
-60
-50
-40
-30
-20
Frequency (MHz)
dB
(S21
)
Transducer response
Reflector response is a time echo which produces a frequency ripple
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5-80
-70
-60
-50
-40
-30
-20
-10
Time (s)
dB
(s
21)
Direct SAW response
Reflector response
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5-80
-70
-60
-50
-40
-30
-20
-10
Time (s)
dB
(s
21)
Direct SAW response
Reflector response
Time ResponseFrequency Response
University of Central FloridaSchool of Electrical Engineering and Computer Science
11
Current SAW RF ID Tag SchematicCurrent SAW RF ID Tag Schematic
•Good for ID tags in close proximity
•All reflectors are at the same frequency
•Typical insertion loss is from 40 to 60 dB
University of Central FloridaSchool of Electrical Engineering and Computer Science
12
Orthogonal Frequency Coded Orthogonal Frequency Coded (OFC) SAW Sensors – (OFC) SAW Sensors –
a New Embodimenta New Embodiment• Simultaneous sensing and tagging possible
using multiple frequencies• Interrogation using RF chirp is possible• Reduced time ambiguity of compressed pulses• Improved security using spread spectrum• Ultra-wide band (UWB) possible
Piezoelectric Substrate
f1 f4 f6 f0f2 f5 f3f1f4f6f0 f2f5f3
University of Central FloridaSchool of Electrical Engineering and Computer Science
13
Orthogonal Frequency Coded Orthogonal Frequency Coded (OFC) SAW Device Concept(OFC) SAW Device Concept
Piezoelectric Substrate
f1 f4 f6 f0f2 f5 f3
Wideband input transducer- coded or uncoded, connected to antenna.
A chip in time is represented by a reflector at a given Bragg frequency.
Each reflector approximates an ideal Rect (t/T)*cos (wot) time response with a specified carrier frequency.
Multiple chips (reflectors) constitutes a bit ( entire reflector bank).
Coding is contained in chip frequency, phase and delay.
Example OFC Tag
University of Central FloridaSchool of Electrical Engineering and Computer Science
14
UCF OFC Sensor Successful UCF OFC Sensor Successful DemonstrationsDemonstrations
• Temperature sensing– Cryogenic: liquid nitrogen – Room temperature to 250oC– Currently working on sensor for operation to
750oC
• Cryogenic liquid level sensor: liquid nitrogen
• Pressure sensor• Hydrogen gas sensor
University of Central FloridaSchool of Electrical Engineering and Computer Science
15
Schematic of OFC SAW ID TagSchematic of OFC SAW ID Tag
0 1 2 3 4 5 6 71
0.5
0
0.5
1
Normalized Time (Chip Lengths)
Piezoelectric Substrate
f1 f4 f6 f0f2 f5 f3
University of Central FloridaSchool of Electrical Engineering and Computer Science
16
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.80
0.2
0.4
0.6
0.8
Normalized Frequency
Mag
nit
ud
e (L
inea
r)
0 1 2 3 4 5 6 71
0.5
0
0.5
1
Normalized Time (Chip Lengths)
Chip length
Bit Length
c/11 cc ff
τB = N·τC
constantlength, Chip c
cc N* cf
The peak of one chip is at the null of all others
University of Central FloridaSchool of Electrical Engineering and Computer Science
17
Bit, PN, OFC Signal ComparisonBit, PN, OFC Signal Comparison
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2-40
-35
-30
-25
-20
-15
-10
-5
0
Normalized Frequency
No
rma
lize
d to
Pe
ak
of S
ing
le C
arr
ier
(dB
)
7 chips/bit PN-OFC7 chips/bit PN Single CarrierBPSK
0 1 2 3 4 5 6 70
0.2
0.4
0.6
0.8
1
Time Normalized to a Chip Length
No
rma
lize
d A
mp
litu
de
7 chips/bit PN-OFC7 chips/bit PN Single CarrierBPSK
Matched Filter Correlated Response
OFC format: 7 chips & 7 frequencies, PG=49
Bit Frequency Response
Processing Gain ~ time-bandwidth product
University of Central FloridaSchool of Electrical Engineering and Computer Science
18
OFC Sensor Platform for Many OFC Sensor Platform for Many Sensor ApplicationsSensor Applications
• OFC reflectors repeated on both sides of transducer• Transceiver yields two compressed pulses• Pulse separation proportional to sensed information
• Different free space delays (τ1 ≠ τ2) yield temperature
• For gas, chemo or bio sensing a sensitive film, such as palladium for hydrogen gas, is placed in one delay path and a change in differential delay senses the gas (τ1 = τ2)
Piezoelectric Substrate
f1 f0f2 f3f1f0 f2f3
University of Central FloridaSchool of Electrical Engineering and Computer Science
19
Schematic and Actual OFC Gas SensorSchematic and Actual OFC Gas Sensor
Piezoelectric Substrate
f1 f0f2 f3f1f0 f2f3
•For palladium hydrogen gas sensor, Pd film is in only in one delay path, a change in differential delay senses the gas (τ1 = τ2)
OFC Sensor Schematic
Actual device with RF probe
University of Central FloridaSchool of Electrical Engineering and Computer Science
20
Orthogonal Frequency Coded Orthogonal Frequency Coded (OFC) SAWs(OFC) SAWs
• Multiple access operation using Spread Spectrum Coding
• Improved range due to enhanced processing gain and low loss (due to OFC reflectors)
• One platform for diverse sensors• Inherent security using spread spectrum• Fractional bandwidth can meet ultra-
wideband (UWB) specifications
University of Central FloridaSchool of Electrical Engineering and Computer Science
21
COM predictions
Experimental Measurement
Piezoelectric Substrate
f1 f4 f6 f0f2 f5 f3f1f4f6f0 f2f5f3
Dual delay 250 MHz, BW~28%, 7 chips/bank, YZ LiNbO3
•Ng*r = .72 @fo
Chip reflector loss~4dB
COM Simulation versus COM Simulation versus Experimental Results – Example 1 Experimental Results – Example 1
Ng*r ~ .72 Ng*r ~ .72
University of Central FloridaSchool of Electrical Engineering and Computer Science
22
COM Simulation versus COM Simulation versus Experimental Results Experimental Results
Example 2 - Example 2 - Ng*r~2.38Ng*r~2.38
Piezoelectric Substrate
f1 f4 f6 f0f2 f5 f3f1f4f6f0 f2f5f3
2 3 4 5 6 7 8-70
-60
-50
-40
-30
-20
-10
Time (s)
Imp
uls
e R
esp
on
se (
dB
)
SimulationExperiment
First reflectorbank
response
Second reflectorbank response
250 MHz, YZ LiNbO3,
8 chips %BW=11.5
Al shorted-electrode reflectivity was ~3.4% Ng=70 @f0
Ng*r~2.38
Chip reflector loss<.5dB
University of Central FloridaSchool of Electrical Engineering and Computer Science
23
Chirp Interrogation Transceiver Chirp Interrogation Transceiver SchematicSchematic
SAWsensor
RF Oscillator
Digital control and analysis circuitry
SAW up-chirp filter
SAW down-chirp filter
IF Oscillator
A / D
IF Filter
Picture of RF Section Transceiver, A/D and Post Processing is Accomplished in Computer
Transceiver Block Diagram
University of Central FloridaSchool of Electrical Engineering and Computer Science
24
Compressed Pulse ResponsesCompressed Pulse Responses
Piezoelectric Substrate
f1 f4 f6 f0f2 f5 f3f1f4f6f0 f2f5f3
Temperature Sensor Example
University of Central FloridaSchool of Electrical Engineering and Computer Science
25
Example of Current Hardware Example of Current Hardware Simulator ResultsSimulator Results
• A simple RF front end and wired SAW device with digital oscilloscope captures trace and simulates A/D and processor
• Picture scale: – Vertical: 5mV / div– Horizontal: 50ns / div
Auto-Correlation
University of Central FloridaSchool of Electrical Engineering and Computer Science
26
Temperature Sensor ResultsTemperature Sensor Results
• 250 MHz LiNbO3 OFC SAW sensor tested using temperature controlled RF probe station
• Temp range: 25-200oC• Results applied to simulated
transceiver and compared with thermocouple measurements
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200Temperature Sensor Results
Time (min)
Te
mp
era
ture
( C)
LiNbO3 SAW Sensor
Thermocouple
University of Central FloridaSchool of Electrical Engineering and Computer Science
27
OFC Cryogenic Sensor ResultsOFC Cryogenic Sensor Results
0 5 10 15 20 25-200
-150
-100
-50
0
50
Time (min)
Tem
pera
ture
( C
)
ThermocoupleLiNbO
3 SAW Sensor
ScaleVertical: +50 to -200 oC
Horizontal: Relative time (min)
Measurement system with liquid nitrogen Dewar and vacuum chamber fro DUT
OFC SAW temperature sensor results and comparison with thermocouple measurements at cryogenic temperatures. Temperature scale is between +50 oC and -200 oC and horizontal scale is relative time in minutes.
University of Central FloridaSchool of Electrical Engineering and Computer Science
28
OFC Bench Marks - CodingOFC Bench Marks - Coding• Number of possible codes: >2N*N! For N chips
– For N=7 chips & 7 frequencies, #Codes = 645,000– For equivalent single frequency tag: #Codes = 128
• PN coding: +/- phase of chip • Time division multiplexing: Extend the possible
number of chips and allow +1, 0, -1 amplitude– # of codes increases dramatically, M>N chips, >2M*N!– Reduced code collisions in multi-device environment
• Frequency division multiplexing: System uses N-frequencies but any device uses M < N frequencies– # of codes decreases– Reduced code collisions in multi-device environment
OFC Bench Marks – OFC Bench Marks – Time & FrequencyTime & Frequency
• System Bandwidth: ~N* -1
– For single frequency: ~ -1
• Processing gain: BW* ~ N2
– For single frequency: ~N– Synchronous time integration using multiple
“pings” can yield increased PG
• OFC and single frequency devices use approximately the same time lengths
University of Central FloridaSchool of Electrical Engineering and Computer Science
29
ccc
c
OFC Bench Marks – OtherOFC Bench Marks – Other
• Device insertion loss: OFC reflector losses can be dramatically reduced yielding 30-60 dB less insertion loss– Ideal OFC devices can have near zero loss
• Size• Number of sensors/codes: Using the OFC
diversity, 25 - 100 devices per interrogator• Interrogation Distance
University of Central FloridaSchool of Electrical Engineering and Computer Science
30
University of Central FloridaSchool of Electrical Engineering and Computer Science
31
DiscussionDiscussion
• Current efforts include OFC SAW liquid level, hydrogen gas, pressure and temperature sensors
• Transceiver is under development for complete wireless, passive SAW OFC sensor system– A/D sampled– Near zero IF– Software radio demodulation – Adaptive filter integration
• Small, efficient antenna design• OFC Code development for multi-sensors• New OFC device embodiments
University of Central FloridaSchool of Electrical Engineering and Computer Science
32
ApplicationsApplications
• Multi-sensor spread spectrum systems
• Cryogenic sensing
• High temperature sensing
• Space applications
• Turbine generators
• Harsh environments
• Ultra Wide band (UWB) Communication – UWB OFC transducers
• Potentially many others
University of Central FloridaSchool of Electrical Engineering and Computer Science
33
Graduate Research Student Graduate Research Student ContributorsContributors
• Daniel Gallagher• Brian Fisher
• Nick Kozlovski• Matt Pavlina
• Bianca Santos• Mike Roller• Rick Puccio
• Nancy Saldanha
University of Central FloridaSchool of Electrical Engineering and Computer Science
34
AcknowledgmentAcknowledgment
Thank you for your attention!
•The authors wish to thank continuing support from NASA, and especially Dr. Robert Youngquist, NASA-KSC. •The foundation of this work was funded through a NASA Graduate Student Research Program Fellowship, the University of Central Florida - Florida Solar Energy Center (FSEC), and a NASA STTR Phase I contract NNK04OA28C. •Continuing research is funded through NASA contracts and industrial collaboration with Applied Sensor Research and Development Corporation, contracts NNK05OB31C, NNK06OM24C, and NNK06OM24C and Mnemonics Corp. Under a new NASA 2007 Phase I STTR.