Low Noise Microwave Generation by Regenerative ModeLocking of a Nd:YVO4MgO:LiNbO3 Microchip Laser

1
Low Noise Microwave Generation by Regenerative
Mode-Locking of a NdYVO4/MgOLiNbO3 Microchip
Laser

• David Yoo
• Center for Microwave and Lightwave Engineering
• Director Peter Herczfeld
• Drexel University

2
Overview

• Oscillators
• What are they?
• Problem Statement
• Whats the need?
• Why is this important?
• Optoelectronic Oscillation Using Regenerative
  Mode-Locking
• (This is what I do.)
• Conclusions and Future Work
• What have I done this year?
• What do I want to do next?

3
Oscillators

• Oscillators electronic devices used for
  generating signals
• Ideal oscillator output
• In time domain - a sinusoidal wave
• In frequency domain a delta function at a
  particular operating frequency
• Traditional oscillator structure
• An amplifier whose output is fed back into itself
  in phase
• This is what causes the oscillation signal to
  regenerate and sustain itself
• Real world analogy loud hum you sometimes
  hear when a speaker gets too close to a microphone

4
Important Parameters of Oscillators

• Output Power
• Determined by how much power the amplifier can
  supply
• Operating Frequency
• Determined by what signal you feed back into the
  amplifier (and how)
• Noise
• Determined by ratio of energy stored vs. energy
  lost

5
Impact of Oscillator Noise

• Real oscillators have a width to their peaks
• phase noise
• Large phase noise makes it impossible to
  distinguish 2 signals that are close in frequency

6
Problem Statement

• Whats the need and why?
• High frequency oscillators are needed in every
  area of electronics referred to as broadband
• Higher frequency means higher information content
• BUT oscillators get noisier as the frequency
  goes up

7
Project Goals

• What did I want to do?
• Develop an oscillator
• Can be scaled to high frequencies
• 10s to 100 GHz
• Low phase noise
• Measured in dBc/Hz at an offset frequency of 10
  kHz
• Good low frequency signals can be -100 dBc/Hz
• At high frequencies (e.g., 94 GHz), -50 to -70
  dBc/Hz is not uncommon
• Cost effective
• Compact

8
Optoelectronic Oscillators (OEOs)

• Convert continuous light energy from a laser to
  microwave signals
• Laser signal is converted into electrical signal
  by photodetector, and electrical signal is fed
  back into laser, regenerating the laser signal
• Laser is mode-locked it pulses its output at
  the inserted frequency
• Use of optical fiber as an energy storage device
  results in signals with extremely low phase noise
• An electrical signal might travel 10 feet through
  wires before half the power is lost
• An optical signal can easily travel 3 miles
  through fiber before half the power is lost

9
Regeneratively Mode-Locked OEO Block Diagram
Modulated Laser
Optical Fiber Spool
Optical Line Stretcher
Photodetector
RF Amplifier
RF Coupler
RF Bandpass Filter
Microwave Output
10
The Results

• Oscillation at 20 GHz has been demonstrated
• 300 m of fiber used
• Phase noise of -99 dBc/Hz at 10 kHz offset

• This OEO is
• Frequency scalable
• Same type of laser can just as easily generate up
  to 200 GHz signals by changing its length
• Low noise
• First prototype already about as good as the best
  synthesizer we have in the lab
• Cost effective
• Optical fiber is pennies per kilometer
• Compact
• Microchip laser is 1 x 0.4 x 0.3 cm

11
Future Work

• Lower Phase noise
• Theory supports possibility of phase noise as low
  as -140 dBc/Hz
• Higher Frequency
• Shorter laser could generate 100 GHz
• Change lasing wavelength
• Different materials (e.g., Er-doped glasses) lase
  at frequencies where fiber is even less lossy
  (and cheaper)
• Interesting possibilities along the way
• Can we exploit the optical signal as well as the
  microwaves? (Remember, mode-locked laser signal
  is pulsed.)
• Can we find a replacement for the fiber
  (OEO-in-a-chip)?