Lecture-Set3
Lecture-Set3
Thas A Nirmalathas
nirmalat @ unimelb . edu . Au
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems :
University of Melbourne LectureSet3_slide2
Optical Modulation
T0= 1 / f0
Modulation Bias
Signal Light
Current
Driver Driver
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems
University of Melbourne :
Laser Rate Equations
• Describe complex nonlinear device behaviour (but not perfect)
• Relationships between photon densities and carrier densities
• What can the rate equations tell us?
– threshold current, carrier density, photon density
– small-signal amplitude and frequency modulation response
– turn-on and turn-off times for digital modulation
– trade-offs in device performance through choice of device parameters
• Important device parameters
– injected current (density)
– material gain and gain suppression
– losses due to device material and facet reflections
– other electron and photon losses
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems
University of Melbourne :
Complete Laser Rate Equations
• Coupled differential equations
Contribution from Depletion due to spontaneous
Injected current recombination
rate of change in
dN J N
the carrier density = - - vg g (N ) S Depletion due to
(N- carriers per dt qd te stimulated
unit volume) recombination
Fraction of photons having same
energy contributed by spontaneous
rate of change in
dS S Gb N emission
the photon = - + + G vg g (N ) S
density
(S- photons per
dt tp te
unit volume) Contribution
Loss of photons due to from stimulated
Intrinsic and mirror losses emission
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems
University of Melbourne :
Laser Dynamics
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems :
University of Melbourne LectureSet3_slide8
dN
Carrier Density (I)
dt
• Rate of change of carrier density over time influenced by
– rate of injected electrons (applied current)
– rate of random recombination (spontaneous emission)
– rate of light-induced recombination (stimulated emission)
• Injected current density
I - Injected current,
Carrier density
I J J - Transverse current density,
injection rate =
qV qd V - Cavity volume,
d - Thickness of active layer
• Spontaneous recombination
– determined by lifetime of carriers at thigher
e energy state,
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems :
University of Melbourne LectureSet3_slide9
dN
dt N - Carrier Density (II)
• Stimulated recombination
– determined by photon density, carrier density and material gain
– optical gain - pseudo-linear relationship between g and N
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems
University of Melbourne :
dS
dt Photon Density (II)
• Rate of change of photon density over time influenced by
– rate of coupled spontaneous emission
– rate of photon loss through cavity losses
– rate of stimulated emission
• Coupled spontaneous emission Spontaneous emission
– determined by rate
Gb N
• spontaneous carrier recombination te
G
• optical confinement factor, b
• spontaneous emission factor (fraction of spontaneous
emission entering into the laser mode) 10-5<b<10-3
tp S
• Cavity loss-induced photon loss Photon loss rate tp
– determined by photon lifetime
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems :
University of Melbourne LectureSet3_slide11
dS
dt Photon Density (II)
Stimulated emission
– determined by
• Gain of the material (g(N))
• Photon density at the lasing frequency (S)
• Optical confinement factor (G),
Stimulated Emission Rate
G v g g ( N ) S » G v g a0 ( N - N 0 )(1 - eS ) S
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems :
University of Melbourne LectureSet3_slide12
Steady State (I)
• Steady-state calculations dN dS
= 0 =
require dt dt
Gb N S J N
G v g g( N ) S + - =0 - - v g g( N ) S = 0
t e (N ) t p q d t e (N )
Gb N
S ss =
(
t e 1 / t p - Gv g a ( N - N 0 ) )
• Since S ³ 0, N is limited by upper
bound 1
N £ No +
• Which is called threshold carrier density Gv g at p
1
N th = N o +
Gv g at p
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems
University of Melbourne :
Rate Equation at Steady-State:
dN J N dS GbN S There is an upper bound for the carrier density and this upper bound
= - - vg g (N )S = Gvg g (N )S + - is called the threshold carrier density, Nth and it is given by:
dt qd t e dt te t p 1
N th = N o + (3)
Gv g at p
dN dS dS
At steady-state: = =0 , so looking at =0
dt dt Looking at the rate equation for the
dt
Gb N S dN
0 = Gv g g ( N )S + - carrier density at steady-state, =0
te tp dt
I N
Gb N S = th + v g g ( N th ) S SS (4)
= - Gv g g ( N )S qV te
te tp
And from (2)
Gb N æ 1 ö 1
= S ç - Gv g g ( N )÷ - Gv g g ( N th ) = 0
te çt ÷ tp
è p ø (5)
Gb N 1
Gv g g ( N th ) =
te tp
S= (1) Substitute (5) into (4):
1
- Gv g g ( N ) I N S
tp = th + SS
At steady-state, we expect SSS >= 0, therefore from (1) qV t e Gt p
1
- Gv g g ( N ) ³ 0 S SS I N
tp = - th
Gt p qV t e
1
Gv g g ( N ) £
tp æ I N ö
S SS = Gt p ç - th ÷ (6)
Gv g a ( N - N o ) £
1 è qV t e ø
tp
æ I I ö
1 = Gt p ç - th ÷
N - No £
Gv g at p è qV qV ø
1 (2) Gt p
N £ No +
Gv g at p
= ( I - I th )
qV A Nirmalathas 2016
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas Optical Communications systems
University of Melbourne LectureSet3_slide14
Therefore the optical output power at steady-state is given by: Given that we know that:
1 vg æ 1 ö
Pout = S SS ´ hu ´ V ´ ´ ln ç ÷ (7)
2 L è Rø 1
= v g ( a i + a mirror ) and
Substituting SSS from Equation (6) into (7) then: tp
1 vg æ 1 ö Photon escape rate from laser cavity
Pout = S SS ´ hu ´ V ´ ´ ln ç ÷ h D = hi
2 L è Rø Photon generation rate
Gt p 1 vg æ 1 ö v ga mirror
= ´ hu ´ V ´ ´ ln ç ÷ ´ ( I - I th ) (8) = hi
qV 2 L è Rø v g ( a i + a mirror )
Gt p vg æ 1 ö
= ´ hu ´ ln ç ÷ ´ ( I - I th )
qL 2 è Rø then (9) can be simplified to the following:
Rearranging Equation (8), we are able to get:
1 æ 1ö
ln
Gt p vg
æ 1ö hu L çè R ÷ø
Pout = ln ç ÷ ´ ( I - I th )
´ hu ´ Pout = ´ ´ ( I - I th )
qL 2 è Rø 2q 1
1 æ 1ö Gt p v g
ln (9)
hu L çè R ÷ø hu
= ´ ´ ( I - I th ) = h ( I - I th )
2q 1 2q D
Gt p v g
assuming hi = G = 1.
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications systems
University of Melbourne LectureSet3_slide15
Steady State (II)
• Total power at the output of the laser can be expressed as
vg
Pout = S ss ´ hn ´ V ´ ln(1 / R)
2L
Account for
photons which
Photon volume, L- manage to
Energy per length
density escape from
photon w-width, d- laser cavity
depth
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems :
University of Melbourne LectureSet3_slide16
Steady State - L-I Curve
Normalised
due to photon density
• N “clamped” at Nth as stimulated emission
current increases
• S increases linearly as
current increases from
threshold
Threshold
• Simplified equations do current
not accurately model N
&S Nth
just below threshold
Normalised
due to carrier density
spontaneous emission Normalised current density
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems
University of Melbourne :
Typical Response to a Turn-On
Photon generation • N < Nth, few photons – spontaneous emission
• N > Nth, sudden increase of photons – stim. emission
• Stim. Emission depletes carrier density; N < Nth
Carriers/Photons
• dS/dt only
B C significantly
increased when net
A gain is positive,
D 1
G v g g(N ) - > 0
tp
• time delay for N to
reach threshold is td,
turn-on delay
Optical Communications Systems
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 :
University of Melbourne LectureSet3_slide19
Wide Pulse Input Response
Region B
• N does not stop at Nth
since S is still low
Region D
when current drops,
B C carrier density also
drops, and net gain is
A highly negative
D
photon density reduces
to zero rapidly, so turn-
off period is very short
period independent of
signal levels
Optical Communications Systems
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016
:
University of Melbourne
LectureSet3_slide21
Small-Signal Analysis
• Similarly, apply small perturbation to injected current, and study
perturbations induced in N and S, defined by the rate equations
• Bias laser by applying current density of Jss then add small signal,
j’, at angular frequency w
J (t ) = J ss + j ¢ e j w t
• Resonance occurs at w= wn
• Small signal modulation response has asymmetrical peak
values vs modulation frequency
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016
University of Melbourne
Laser Parasitics
• Modulation bandwidth of laser can be severely
degraded by internal (chip) and external (packaging)
parasitics
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems:
University of Melbourne Lecture1_slide24
Improve Modulation Bandwidth
• Aim - increase the electrical bandwidth of the laser, in order to transmit signals
at higher frequencies (i.e. greater capacity)
• Reduce parasitics
- Laser chip parasitics
- Driving circuits parasitics
• Parameter manipulation for increase in wn
– Increase optical gain, g
• decrease transparency doping level in active region or decrease laser
temperature
• use quantum well structure - better define states
– Increase photon density Sss
• narrow active region or tight optical field confinement
• bias laser well above threshold
– Reduce tp
• reduce cavity length or increase mirror losses
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems:
University of Melbourne Lecture1_slide25
Frequency Chirp
• When the laser current is directly modulated, the laser may experience
dynamic line broadening
• Changes in the carrier density in the active region result in changes in the
refractive index, which leads to changes in the instantaneous optical
frequency - frequency chirp
BLUE
SHIFT
RED
SHIFT
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems:
University of Melbourne Lecture1_slide27
Laser Driver Circuits
• Current source to determine the bias point (operating point) Idc
• Voltage to current amplifier to generate sufficiently large
modulation current Iac
• Low noise design so that output power is not subjected to the
influence of noise
• Stability of current sources against temperature changes
• Two modes of biasing possible – constant output power or
constant current modes
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems:
University of Melbourne Lecture1_slide28
LD Driver Building Blocks
Input
amplifier
Laser Driver
Bias
Control
Detector
Amplifier
Guest Lecturer, Dept. of Electrical and Electronic Engineering, The ©Thas A Nirmalathas 2016 Optical Communications Systems:
University of Melbourne Lecture1_slide29