USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Photonic time-stretching of 102 GHz millimeter waves

using 1.55 µm nonlinear optic polymer EO modulators

H. Erlig

Pacific Wave Industries

H. R. Fetterman and D. Chang

University of California Los Angeles

M. Oh, C. Zhang, W. H. Steier, and L. R. Dalton

University of Southern California

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Outline

• Time-stretch concept

• Experimental description

• Polymer modulator

• Data and data analysis

• Dispersion penalty & SSB modulation

• Conclusions

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Time-Stretching Concept

BroadbandFemtosecond

SourceL1, Dλ

Time

Optical PulseTime Domain

Mach-ZehnderModulator

Applied RFSignal, 50 GHz

Time

Optical PulseAfter Modulator

L2, Dλ

Time

StretchedModulated

Optical Pulse

OpticalAmplifier

Photodetector

Detected RFSignal, 10 GHz

BroadbandFemtosecond

SourceL1, Dλ

Time

Optical PulseTime Domain

Mach-ZehnderModulator

Applied RFSignal, 50 GHz

TimeTime

Optical PulseAfter Modulator

L2, Dλ

TimeTime

StretchedModulated

Optical Pulse

OpticalAmplifier

Photodetector

Detected RFSignal, 10 GHz

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Time-Stretching - Intensity

( ) ∝tI ( )tf

π

tMf2

cos m

βπ 2m

2''

2 fML

2cos

PulseEnvelope Time-Stretched

WaveformDispersion

Penalty,Leads to amplitude

attenuation

1

2

LL

1M += Stretch Ratio

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Experimental Arrangement

optic

FiberSpool L2

optic

FiberSpool L1

MM-WaveSource

8-12GHz32dB Amp

Photo-detector

ND-Filter +Adj. Attenuator

Modelocked1.55um Laser

optic

EDFA

SamplingScope

SpectrumAnalyzer

PolymerModulator

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Polymer Modulator

• PC/CLD polymer traveling wave modulators

• Optical network analyzer: 1 dB down at 20 GHz compared to 2 GHz

• Modeled effects: velocity mismatch and electrical loss

• Vπ ~ 7 V, 1.3 cm interaction length

• W-band response relatively flat

Measured and Calculated Optical Response

-12

-9

-6

-3

0

3

2 4 6 8 10 12 14 16 18 20

Frequency (GHz)

Op

tica

l Re

spo

nse

(d

B)

Measured Optical Response

Calculated Optical Response 0

5

10

15

20

25

30

70 80 90 100 110 120

Frequency (GHz)

Rel

ativ

e R

esp

on

se (

dB

)

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Transmission Through Modulator

Influence Of Modulator 5 On Optical Spectrum

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600

Wavelength (nm)

Norm

aliz

ed In

tensi

ty

After Modulator

Before Modulator

Influence Of Modulator 3 On Optical Spectrum

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600

Wavelength (nm)

Norm

aliz

ed In

tensi

ty

After Modulator

Before Modulator

• Mach-Zehnders used with broadband femtosecond source

• Modulators reshape 50 nm spectrum• Different modulators on chip introduce

different amount of spectral reshaping• Slight optical path mismatch• Highly chirp pulses key• Effect also observed in LiNbO3

modulators

Light

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Pulse Shape After System

• Two interfering highly chirped optical pulses

( )( )

δ

⋅

β

−∝

tcz

b21

cos

TL

texptI

2

20

''

2

For system parameters and period of 1.5 ns

• Calculated effective optical path mismatch of 100 µm

• Calculated effective index mismatch 0.005

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

33 To 50 GHz Time-Stretched Signals

• Sweep oscillator

• L1=1.5 Km

• L2=4.5 Km

• Measured Meff 3.86

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

61.8 GHz Time-Stretched Signal

• Source GUNN diode at 61.8 GHz

• L1=1.5 Km

• L2=6.5 Km

• Measured Meff 5.13

• PSD shape not significantly changed

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

101.7 GHz Time-Stretched Signal

• Source Klystron at 101.7 GHz

• L1=0.5 Km

• L2=5.0 Km

• Measured Meff 9.8

• Change in input pulse chirp

• Drop in signal level

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Stretch Ratios

• Discrepancies between calculated M and measured Meff

• Need to account for dispersive elements ahead of first fiber spool such as 50m fiber patch cord

• Then:

• where δ has length units and represents dispersion equivalent to that length of fiber

• Solutions for δ are correct in magnitude and self-consistent:

f (GHz) L1 (Km) L2 (Km) M Meff δ δ (m)33-50 1.5 4.5 4.00 3.86 7361.8 1.5 6.5 5.33 5.1 74

101.7 0.5 5 11.00 9.8 68

δ++=

1

2

LL

1Meff

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Dispersion Penalty

• Sidebands slip out of phase in L2due to group velocity dispersion

• Result: dispersion penalty (Coppinger et. al.)

• Tradeoff: M vs. aperture time vs. bandwidth.

• Practical limit for A/D preprocessing: “must stay in 1st lobe”

• Modulator operation exceeds this limit in our experiment

βπ 2m

2''

2 fML

2cos

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

SSB Modulation

Optical Field,Frequency Ω

tsinmV)t(V ω=

Γ+ω= VtcosmV)t(V

Mach-ZehnderDriven To ProduceSingle Side Band

Frequency

Amplitude

Ω Ω+ω

Optical SpectrumAt Output Of Modulator

VΓ chosen to produce additional π/2 opticalphase shift

( )( )

( )( )

π+ω

β+

ω∆∆

⋅

τ−⋅∝

4M2L

tM

cosJJ

M

t2exp

M1

tI 2m

2''

m

0

12

2

• Amplitude limitation imposed by dispersion penalty removed

USCUNIVERSITY

OF SOUTHERNCALIFORNIA

Conclusions

• Demonstrated time-stretching of 102 GHz signal

• Enabled by broadband 1.55 µm polymer modulator

• Modulator performance spans useful bandwidth range determined by dispersion penalty

• Importance of optical path length mismatch observed

• Single-Sideband Modulator removes high-frequency attenuation