Semiconductor Optical Communication

Components and Devices

Questions and Problems

 Review Questions Lec 4:

1. Take a one dimensional periodic structure of a=5nm, b=20nm, and a

Vo=100meV. Take an effective mass of the electron to be me*=0.07mo.
mo is the free electron mass. Write a computer program to find the

/
Energy at k=0 and k=p (a+b) to an accuracy of 0.1meV. This is the first
quantization energy.

2. (i) Increase ‘b’ till you reach where the two energies are same to the
accuracy that you are working with. This is the energy for a single
quantum well.
(ii) Now increase Vo in steps till you reach a few eV. Compare this result
with that of an infinite potential well En=n2h2 (8me*a2) for n=1.
/
 Review Questions Lec 5:

1. Take a=5nm and b=4nm with me*=0.07mo. Check for the k=0 and

/
k=p (a+b) energy difference again. This would be the first mini-band of
a superlattice. What is your comment with respect to that found in P2
above.
Review Questions Lec 6:
1. Find the composition of InGaAsP alloy lattice matched to InP for
which the band gap is 0.85eV.

2. Check the band shape of an indirect semiconductor and that of a

direct band gap semiconductor. Justify whose electron effective
mass would be larger.

3. Look at the shapes of the conduction band and the valence band of
a direct band gap semiconductor. For electrons in the conduction
band and holes in the valence band, which should have a larger
effective mass?

4. Take me*=0.07mo and mh*=0.7mo and plot the density of states for
both. How similar are they?

5. Note that there are two valence bands at the G-point (k=0). Their
curvatures are very different. Which should be called Heavy Hole
(hh) and which should be called Light Hole (lh)?

6. Do a similar exercise as that in the previous lecture and find the

quantization energies for a=5nm, b=200nm, me*=0.07mo, and
mh*=0.7mo
Review Questions Lec 7:
1. Of the two different kind of Phonons studied which would probably
most interact with an inter-band transition in an indirect
semiconductor? Justify your answer.

2. Write a mathematical expression of the temperature dependence of

the electron distribution in a conduction band.

3. Out of the three different transitions, absorption of a photon,

spontaneous emission of a photon, and stimulated emission of a
photon, which are resonant processes ?

4. What happens when an electron in the conduction band is

accelerated to an energy above the bottom of the conduction band
which is larger than the separation in energy between the bottom
of the direct conduction band and the bottom of the indirect
conduction band in a direct band gap semiconductor?

5. Why is it less probable to have an optical transition in an indirect

semiconductor as compared to that of a direct semiconductor?
 Review Questions Lec 8:
1. Which system should one use for the growth of InxGa(1-
x)AsyP(1-y) ? What are the disadvantages of this growth
system?

2. Why is VPE system not popular for the growth of InxGa(1-

x)AsyP(1-y) communication device applications?

3. Which growth system is suitable for obtaining highest

quality material ? What are the disadvantages of this
system?

4. How should one select the growth temperature for a

particular semiconductor?

5. Between MBE and MOCVD, which is more suitable in

obtaining better electronic devices (Should have less non-
compensated unintentional doping)

6. Which growth system should one choose if both optical

 Review Questions Lec 9:

1. A graded Al0.36Ga0.64As is to be grown on GaAs for a thickness of

1mm at a growth rate of 1mm/hr. in an MOCVD system using TMG,
TEG, and AsH3. Find the flow rates of TMG, TEG, and AsH3 with time
o
if growth is done at 750 C. The carrier gas (H2) flow rate is 10 SLM
(Standard Litres per minute).

2. An InxGa(1-x)AsyP(1-y) is to be grown lattice matched to InP at a

band gap of 0.8eV. Find the flow rates of TMG, TMI, AsH3, and PH3, if
o
growth is to be done at 700 C. The carrier gas (H2) flow rate is 15
SLM.
 Review Questions Lec 10:

1. A graded Al0.36Ga0.64As is to be grown on GaAs for a thickness of

1mm at a growth rate of 1mm/hr. in an MOCVD system using TMG,
TEG, and AsH3. Find the flow rates of TMG, TEG, and AsH3 with time
o
if growth is done at 750 C. The carrier gas (H2) flow rate is 10 SLM
(Standard Litres per minute).

2. An InxGa(1-x)AsyP(1-y) is to be grown lattice matched to InP at a

band gap of 0.8eV. Find the flow rates of TMG, TMI, AsH3, and PH3, if
o
growth is to be done at 700 C. The carrier gas (H2) flow rate is 15
SLM.
 Review Problems Lec 12:

1. For a TM wave incident from the medium n1 to the interface of n1 and n2 where
n1 > n2, derive the phase change due to reflection for p/2 > qi > qc.

2. A plane wave ‘A’ is incident in a planar waveguide as shown in the figure above,
where n1 > n2. Show that for qi > qc the propagation in the x-direction in the
medium n2 is exponentially decaying whereas in the z-direction in the same
medium the propagation is identical to that of medium n1.

n2
A qi n1 z

n2
x
 Review Problems Lec 13:

1. For a planar GaAs waveguide of thickness 0.5mm and refractive indices

nf = 3.5, ns = 3.45, and nc= no =1.0, find a value of the thickness tg for
which there would be propagation of only the lowest order mode for a
wavelength of 1.5µm but with the maximum propagation constant
possible. Calculate this propagation constant. Find the cutoff
wavelength for the lowest order mode in this waveguide.

2. A five layer slab waveguide structure composed of layers no=1.000, tgo=

semi-infinite; n1=3.498, tg1=1.5mm; nf =3.500, tgf=2.5mm; n2 = 3.495,
tg2=1.5mm; and ns = 3.5, tgs=semi-infinite is supposed to guide TE light
at a wavelength ‘l’. Find the range of wavelengths for which there would
be single mode propagation. [Hint: Write the wave equation, i.e. the field
expressions in the five layers for the field components Ez and Hz. Match
the boundary conditions at the four interfaces. Then use Matlab or any
other suitable software to solve the equations and find the relationship
of b vs V. Define b and V with respect to nf and tgf. Get your answer from
the plot.]
 Review Problems Lec 14:

1. A rib waveguide is formed with nc=1.0, nf=1.55, and ns=1.63. If the width
of the stripe for the rib is 4 µm and the thickness of the film is 1.0µm and
the rib height is 0.2µm. Find the propagation constant of the guided
mode for a wavelength of 1.55µm in air.

2. For a planar waveguide of thickness ‘2d’ and refractive indices nf=1.501,

ns=1.500, and nc=1.0 find the thickness of the guide for which there
would be propagation of only the lowest order mode for a wavelength of
1.65µm but with the maximum propagation constant. If a rib waveguide is
to be made from this slab waveguide, given that the rib height ‘h’ needs
to be only 5%-20% of the slab waveguide thickness, choose a slab
waveguide thickness and plot the maximum rib width ‘a’ as a function of
‘h’ for a single transverse mode of propagation in the rib waveguide.
 Review Problems Lec 15:
1. Photons of wavelength l=813 nm are absorbed in InP at room
temperature (Eg=1.344eV, me*=0.08mo, mh*=0.60mo) and excites
electron-hole pairs (EHP). Calculate the average kinetic energy of the
electrons and holes before they relax to the bottom of the bands. [Hint:
they will not be the same]
2. Find the peak emission wavelength (lo) for an Al0.2Ga0.8As LED operating
at 400oK, given the band gaps AlAsG=3.03eV, AlAsX=2.15eV, AlAsL=2.36eV,
GaAsG=1.43eV, GaAsX=1.73eV, and GaAsL=1.89eV. (assume linear
interpolation to be valid)
3. Derive the fraction of radiation escaping from an LED (point source)
imbedded in a medium of refractive index n1 into the upper medium of
refractive index n2, assuming there is no absorption in the medium. Also
assume the transmission coefficient is that of normal incidence at the
interface.
4. An LED at room temperature under 0.8V forward bias conducts a current of
I =12mA and emits light at a peak wavelength of 1.0µm. The radiative and
the non-radiative time constants are 0.1ns and 1ns, respectively. Assuming
that the unity extraction efficiency and the injection efficiency is 0.9,
calculate the power conversion efficiency of the diode. Is this calculated
efficiency greater than or less than unity? Does it surprise you, explain.
What is the new conversion efficiency, if for the same set of conditions, the
emission wavelength is now 1.3µm ?
 Review Questions Lec 16:
1. How does a double - heterostructure LED help in increasing the
efficiency ?

2. Why is it necessary to take precautions that in an SLD standing

modes do not develop? How is the facet handled for that purpose?

3. Qualitatively compare the same for the SLED and SLD when the
junction temperature rises to 200oC above room temperature.

4. Explain why the modulation bandwidth of an SLD is larger than that

of a SLED.
 Review Questions Lec 17:
1. A InGaAs LED with minority carrier lifetimes (for both electrons and
holes) of t = 1ns is excited by a modulating current of I=[10+Cos(2pft+q)] mA.
If the steady state optical output is Po=5mW, derive expressions for P(t) for
modulating frequencies of (a) f=1.0MHz and (b)f = 10GHz.

2. An edge-emitting LED has stripe width of 10 µm. The active layer thickness
is 0.5 µm and has a refractive index of 3.5. The length of the LED is 0.5 mm
and the radiative recombination lifetime of the carriers is 0.1 ns. The optical
output is taken out from one end of the diode only. When the bias current is
switched off, find the time required for the output intensity to decrease
from the steady state value to 1% of the steady state intensity.

3. A GaAs LED fabricated from fairly lightly doped materials has an effective
recombination region of width 0.1 µm. If it is operated at a current density
of 2 x 107 A/m2, estimate the modulation bandwidth that can be expected.
Assume the recombination constant B(recomb)=7x10-16 m-3. [Hint: Get
relation between recombination time constant with B(recomb)]

4. Explain why the edge emitting LED (ELED) has a narrower spectral width
than the surface emitting LED, (SLED). How does it differ from the
Superluminescent diode (SLD) ?
 Review Questions Lec 18:
1. Which of the semiconductors Si, Ge, GaAs, and GaP are suitable for
the fabrication of diode lasers? Justify your conclusion.

2. An In0.53Ga0.47As semiconductor at 300K has parabolic conduction

and valence bands [Ec,v a k2]. The effective masses of the electrons
and holes in this material are me*=0.06mo and mhh*=0.15mo,
respectively. If the electron concentration peak is 0.5kBT above the
bottom of the conduction band, then find the hole energy (eV) below
the top of the valence band for efficient photon emission, assuming
the density of states for the conduction and the valence bands to be
the same.

3. What are the different varieties of Fabry-Perrot cavity diode lasers?

4. What are the configurations for which one would be able to get
narrow linewidths for diode lasers?

5. Why is it essential that the diode laser output is emitted normal to

the surface of the semiconductor substrate? What are the
disadvantages associated with it?
 Review Questions Lec 19:

1. How should the direction of the optical waveguide oriented with

respect to the crystal axes for the formation of cleaved cavity
mirrors for the diode lasers?

2. What is the effect of the introduction of Dopuble Heterostructure

in the active region on the performance of the Laser diode?
Calculate the optimum thickness of a GaAs/Al0.3Ga0.7As DH
structure from a consideration of the optical mode – carrier
overlap.

3. What is the advantage of introducing QWs in the DH active

region? How does the number of QWs determine the speed of
operation of the device?

4. How does compressive strain in the active region improve the

efficiency of a diode laser?
 Review Questions Lec 20:

1. Explain how grading of a DH-SQW active region of a diode laser

help in high speed modulation of a diode laser.

2. Design a GaAs/AlGaAs GRIN-SCH-SQW laser for maximum

stability of the threshold current with temperature.

3. What are the considerations of choosing the number of QWs in

the active region of a Diode Laser?

4. How does cladding layer thickness determine the Far Field

FWHM width of a GRIN-SCH diode laser?
 Review Questions Lec 21:
1. Consider a bump dL on one of the mirrors of a Fabry-Perot cavity, calculate
the mode positions dq for a change of dL in the cavity thickness. Introduce
this value in the expression for Finesse, and show that for a Finesse ‘F’ ,
the mirrors must be flat to l/N. Determine 'N'.
2. Show that the longitudinal mode spacing of a semiconductor laser
resonator cavity of length ‘L’ (considering the presence of dispersion at the
semiconductor band edge) is given by: |dl| = l2.[2nrL{1- (l/nr)(dnr/dl)}]-1
3. Show that DEF > Eph > Eg is the condition for gain in a laser. Where DEF is
(EFn - EFp). What is the condition identified as population inversion in
semiconductors. Derive your answer.
4. A DH diode laser gain profile has gmax=2000m-1 and attenuation as=600m-1
(a) If R1 = R2 = 0.35, what is the minimum value of the length of the cavity
for which lasing action can be obtained ?
(b) If the length is 400 µm, and R1= R2=R, what is the minimum value for R?
(c) How would the laser threshold change if the value of R is increased
from this value by applying reflective coatings to the end facets ?
5. An uncoated Gallium Arsenide (GaAs) injection diode laser with a cavity
length of 500 µm has a loss coefficient of 20cm-1. The measured differential
external quantum efficiency of the device is 45%. Calculate the internal
quantum efficiency of the laser. Assume the refractive index of GaAs is 3.6.
 Review Questions Lec 22:
1. The threshold current density for a stripe-geometry AlGaAs laser is
3kA.cm-2 at a temperature of 15oC. Estimate the required threshold
current at a temperature of 60oC when To for the device is 180oK, and the
contact stripe is 20 x 100 (mm)2.
[Hint: G. H. B. Thompson, IEE proceedings (optoelectronics), 1981,
Vol.128, pp37-43]

2. Explain why in steady state the carrier concentration in the injection laser
active region remains constant even when the current is increased above
threshold.

3. The output power of a junction laser above threshold is proportional to

Rb/tph as Rbias= [Gtph/qw](Jbias- Jth).
Where tph is the photon lifetime in the cavity. Show that, above threshold,
the output power still depends on tph and is proportional to {J-
(qw/ttot)[NT + (g1Gtph)-1]}. Where g1 is the constant of Stimulated
emission, NT is the carrier concentration for which transparency is
obtained, and G is the fraction of injected carrier distribution and mode
overlap.
 Review Questions Lec 23:
1. A diode laser is operated at a current Io=2Ith. If the Carrier and Photon
Lifetimes are 1ns and 10ps, respectively, then estimate the modulation
bandwidth of the laser. Given that the transparency carrier
concentration is 1.0 x 1018 cm-3 and g1=5x10-7cm3 s-1.

2. A 800nm thick DH laser is biased at 0.5Ith and a pulse current step of

1.0Ith is applied to the laser at t=0. If the Carrier and Photon Lifetimes
are 1ns and 10ps, respectively, find the delay time after which the
output power reaches the steady state value. Calculate the time after
which the output power reaches steady sate.

3. Where would you bias a semiconductor laser for high speed direct
modulation? What is the penalty paid for this state of operation?
 Review Questions – I (Lec 24)

1. At what frequencies should a diode laser be directly modulated to

encounter chirping? What is the origin of chirping and how does it
affect the bit rate of optical communication?

2. Does a negative chirp affect the maximum bit rate of transmission

through a dispersive optical fiber in the same way as a positive chirp?
Explain your answer.

3. An InGaAsP/InP DH single mode laser is operating at a power Po in the

steady state with a FWHM linewidth of 0.5nm. Estimate the average
linewidth of the laser due to chirp when it is modulated at a frequency
1.5wo modulated to a power level of 2Po. wo= 5GHz and the Henry factor
is aH=3.5.

4. Does one expect the same aH for all modes of a multimode

semiconductor laser? Explain you amswer.

5. Guess how the linewidth would be affected for a large signal modulated
diode laser.
 Review Questions – II (Lec 24)
5. A direct detection optical fiber PCM
communication system, operating at a
Pout
wavelength of lo=1610nm, is to (DlFWHM)initial
operate over a distance of 50Km (DlFWHM)final
without any repeaters. The link is
formed with a dispersion shifted (zero
dispersion at 1550nm) silica fiber of
0.2dB/Km loss at the operational
wavelength of lo=1610nm and a
residual dispersion of 50 ps.Km-1.nm-
1. The laser output is 200mW at a
line-width of (DlFWHM)initial=0.2nm,
as shown in the adjacent figure. lo l
The chirp in the laser, on modulation, is given as:
(DlFWHM)final=[8.633x10-21].Exp(0.4.fm) nm,
where fm is the modulation frequency in GHz. The modulation delay (exponential) for
the laser, td=12.0ps. The detector Noise equivalent power (NEP) is 12.6 pW/√Hz and
has a transit time limited 3dB bandwidth of 45GHz. (assume the pulses to be Gaussian
in nature). What is the MAXIMUM POSSIBLE BIT RATE of the communication
link? Show calculations to justify your conclusion. (May need iterative solution)
 Review Questions Lec 25:
1. What is the main source of Diode-Laser noise? Enumerate other sources
of this noise.

2. Find the relative intensity noise at l=1.5mm for a laser operating at

100mW, with a small signal bandwidth of 30GHz. Given that the active
volume is 250x4x0.01mm3, tph= 2.5ps, tc= 0.5ns, tnr= 1.0ns, N=1019cm-3,
|H(w)|=wr4/[(wr2-w2)+(wg)2], g= (2+0.3wr2)x10-9, and fr=25GHz.

3. If the resonance frequency of a diode laser is 15GHz and the small signal
f3dB=30GHz, what should be the range of the upper limit of the relative
intensity noise given that at low frequency modulation the relative
intensity noise at unit bandwidth is -100dB/Hz and tc= 1.0ns.

4. How is the noise affected by the size of the cavity? Explain your
conclusions.
 Review Questions Lec 26:

1. Should one use an undoped or heavily doped substrate for the fabrication
of diode lasers? Explain your conclusion.

2. What is the necessity of an insulating layer before the final top contact
metallization?

3. If a DH Diode Laser is grown on an n+ substrate, why is it advantageous

to package the laser with its anode connected to the heat-sink? What
precautions need to be taken during die bonding of this laser?

4. What should be the special properties of the carrier material of a fiber

which is used for fiber coupling of a diode laser?

5. Why is it essential to mount both the fiber carrier as well as the diode-
laser on the same base plate when thermoelectric cooling is used?

6. Why are diode laser packages for high speed modulation different from
those of a TO3, TO5, or TO8 packages? Which ones should be used for
this purpose? Explain how these packages are suitable for high
frequency operation.
 Review Questions Lec 27:

1. Why are DFB lasers essentially single mode? How are the modes
determined?

2. What is the need for p/4 phase slip in a DFB Laser?

3. What is the order of linewidth of a DFB laser? Which factors does it

depend on?

4. List the fabrication steps that would be required for the fabrication of a
MQW-DFB laser.

5. A DH DBR laser has a cavity neff =3.4 of length L=100mm and is supposed
to work at lo=1.55mm. The end facets are AR coated with R=0.5 coupled to
two gratings of length LDBR =150mm. Assuming a coupling coefficient of
kLDBR=4, for a 1st order grating find the grating period ‘L’ and the end-loss
of the end facet F-P modes about the selected mode.

6. Why is chirp expected to be much less than that of a Fabry-Perrot Laser

 Review Questions Lec 28:

1. What determines the linewidth of a semiconductor diode laser?

2. Show that for G ≈ Gth, Dw= ћw(1+aH2)/[tph2.2Pout] and hence the
cavity quality factor Q = Df/f = wtph.
3. A F-P cavity of length L=250mm has an effective refractive index
neff=3.7 and aH=2. The F-P cavity has bare end facets at one end
and is coated to R=1 at the other end. Assume cavity losses as
to be negligible. If the laser is operating at an output power of
10mW at a wavelength lo=1.24mm, find the cavity linewidth and
compare it with the intrinsic laser linewidth.
4. Compare “thermal” and “electronic” tuning mechanisms of
DFB lasers.
5. What are the advantages of multi-slot tuning and what are the
constraints of this technique?
6. Check QCSE tuning of diode laser principles: B. Cai, A. J.
Seeds, and J. S. Roberts, IEEE Photon. Technol. Lett. , vol. 6(4),
496 (1994).
 Review Questions Lec 29:

1. What are the usages of VCSEL lasers?

2. What are the reasons for narrow linewidth and low threshold for
VCESLs?
3. If RTop.RBot= 0.95 and a1= 1000cm-1, a2= 10cm-1, and adiff=
20cm-1, find the threshold gain required for d=0.25mm and
L=1mm. Compare this with that of a typical F-P diode laser of
length 250mm.
4. Why is oxide aperture very useful in the fabrication of VCSEL
arrays? Which other technique is used?
5. What are the usages of tunable VCSELs? What method is used
for this tuning?
6. Which techniques are used for controlling the polarization of
VCSELs? Compare the efficiency of these techniques and the
complexities of fabrication of each of these techniques.
 Review Questions Lec 30:

1. A diode laser had been working fine for 200 hrs. at the rated
power, but suddenly failed to produce any power except
that expected from an LED. Can you suggest a reason for
the failure? What would be the possible solutions to avoid
this kind of failure?
2. A similar diode as above, which has been working for 5000
hrs. is found to produce less output power at the same bias
current over a period of a month. Suggest a reason for this
kind of failure. Can this failed laser be restored?
3. What precautions can be taken to avoid soldering and fiber-
couple degredation?
4. If the optical output power of a diode laser is linearly
dependent on the drive current, then why does the failure
rate be dependent exclusively on the mth power of current
and nth power of the output optical power?.
 Review Questions Lec 31:

1. Upon the sudden removal of the external generation stimulus

(rate=Gext) in an n-type semiconductor (no >> po) at time t = 0, where
the excess carrier concentration is Dn(0) >> no, calculate the time
dependence of the excess electron density Dn(t), where no and po are the
equilibrium carrier concentrations. Assume that direct band-to-band
transitions are the only recombination mechanisms. [Let B be the
recombination constant].

2. Is the absorption process a resonant process or a random process?

Justify your answer.

3. Why would one use a direct band gap semiconductor for the fabrication
of a photodiode?

4. For high speed operation in communications, would one choose to

operate the photodiode in the photovoltaic mode or photoconductive
mode? Justify your answer.
 Review Questions Lec 32:
1. A photon of l=1.55mm is absorbed by a lattice matched InGaAs/InP
PIN photodiode at room temperature. If the kinetic energy of the generated hole
is 0.5kBT, then find the kinetic energy of the generated electron. Given that the
electron and hole effective masses are 0.042mo and 0.5mo, respectively.
2. An optical signal Pin=[1.0mW]{1+0.1Coswt} of l=1.5mm is normally incident on
a DH InGaAs/InP PIN photodiode. The two semiconductors having band gaps
Eg(InP)=1.5 eV and Eg(InGaAs)=0.75eV. The absorption coefficients at
l=1.55mm are a(InP)=1.0x102m-1 and a(InGaAs)=1.0x106m-1. Assuming
internal quantum efficiency to be 0.8, calculate the total external quantum
efficiency of the detector for WInGaAs=0.5µm, nr(InP)=3.4, nr(InGaAs)=4.2.
3. An incoherent detection system operates with a PIN photodiode R=0.45A/W at
l=1.55mm at 300oK with a dark current iD=1.0nA and when connected to a load
resistance of 50W the operational bandwidth is B=6.0MHz. The incident power
is Pi=Po+ PmCos(wmt) and that the background radiation is neglected.
Assuming Po=Pm, plot the SNR for 1mW > Po > 1nW.
4. A load resistance RL=100W is connected to a PIN detector of responsivity
0.4A W at a l=1.0µm and dark current of 1.0nA. Assuming the equivalent load
/
resistance (including the diode resistance and the input resistance of the
amplifier) ≈ RL, calculate the NEP of the detector. What is the minimum
detectable power if the operational bandwidth of the photodiode is 250MHz?
 Review Questions Lec 33:

1. Consider the detector of prob. 2, lec. 32. If the photons

are incident from the P+ side and the detector is
operated under saturated velocity of the electrons
(ve=2.5x105 m/s) and holes (vh=5x103 m/s), estimate the
response time of the detector if the junction-area of the
diode is very small.

2. What composition of InGaAsP should be used for the

detection of l=1.3mm from speed considerations.

3. How should the a PIN Photodiode be biased for

maximum speed of operation.
 Review Questions Lec 34:
1. In prob. 1 of lec. 34 if the width of the absorption region is increased to
2.0mm, how is the above estimate in variance with the actual value.
Determine the same by writing a small program.

2. What are the techniques for the speed measurement of fast photodiodes?
If one measures the impulse response by a fs laser, how does one find the
3dB bandwidth of the photodiode from the measured response? How does
one do a photodiode speed measurement when neither a laser can be
modulated at the highest speed that the photodiode responds nor can an
oscilloscope be found to respond to the speed of the detector? Do some
research on it.

3. A communication link is driven by a lo=1.5mm single mode laser of

linewidth 10nm, which has a 3dB modulation bandwidth of 31.8GHz and
negligible chirping. The channel is a single mode dispersion shifted fiber
with a dispersion of 0.5 ps/(Km.nm) at lo. The front end of the receiver is a
PIN photodiode whose response time is transit time limited to 10ps. What
is the maximum length of the fiber for which 10 Gbits/s operation is
possible ?
 Review Questions Lec 35:

1. Why does one expect to have enhanced responsivity in a RCE-PD

even though the absorption region could be quite thin for high
speed operation?

2. How is the transit time limitation overcome without compromising

on the absorption length in a RFPD? What are the disadvantages
of this detector?

3. What are the limitations of a waveguide photodiode, although

speed may be enhanced substantially?

4. Why is it essential to match the velocity of the RF generated from

the photo-response with the optical velocity in the waveguide?

5. How does waveguide photodiodes enable on-chip integration and

for that matter also makes it possible to have a periodic
photodiode structure? (Check last lecture on integration)
 Review Questions Lec 36:

Check review questions on APDs (Lec. 38).

 Review Questions Lec 37:
An intensity modulated optical
signal, PS=Po[1+0.5.Sin(wmt)] is
incident on an InGaAs (lg=1.65mm)/InP
SAM-APD from the p+ end at 300K, as
shown in the figure. The ionization
coefficients of the carriers in the
avalanche region are ae=100bh.
Avalanche width is ‘wa’, aewa=1.5, and
wd=0.7mm. Neglect all capacitances and
inductances. The APD is connected to
an equivalent load of RL=1.0kW.
αw
1. Show that Me e e a and M h 1.
2. If the responsivity R =0.6 A/W at l=1.64mm when this diode is operated as a
PIN-photodiode, write an expression for the RMS signal power of the APD as a
function of Po.
3. Find the Noise-Equivalent-Power (NEP) of the APD for a shot and thermal
noise limited operation neglecting any dark current.
4. The electron and hole velocities are vsat =2x107cm.s-1 & vsath=7x106cm.s-1.
e
Neglecting any delay due to the avalanche process, estimate the bandwidth of
the APD for l=1.64mm.
 Review Questions Lec 38:

1. An avalanche Photodiode has hext= 0.62 at l=1mm, a dark current of 10pA at 100V

bias, a b = 0.85, multiplication M = 30 and a noise equivalent bandwidth of 10

/
GHz. If a signal of 1.0mW is incident on the detector, calculate the current SNR. If
the bias is changed to 250V the dark current, the a b, and the multiplication factor
/
changes to 100pA, 0.02, and 250 respectively. What is the change in the SNR ?

2. A PIN photodiode operating at room temperature, generating a photocurrent

Iph=50mA is connected to a HEMT pre-amplifier through a photodiode load of 50W.
The HEMT operates at ID=1mA, IG=20nA, and gm=10mS to provide an overall
bandwidth (B)=20GHz. What is the r.m.s noise at the input of the pre-amplifier?

3. An APD has responsivity R =0.5, k=0.1, and operated at M=15 receives an optical
signal of 100nW which is intensity modulated with a signal of f(t)=Cos[2pfmt] at a
modulation index of m=0.4. The photodiode sees a total load of RL=1.0kW in
parallel with C=300fF when connected to a transimpedance pre-amplifier having a
feedback resistance of RF=0.5kW to provide an overall bandwidth (B)=1GHz.
Assume the photodiode and the amplifier to work at 300K and fm << B. Also
assume {ia2}=0 and {va2}=2x10-18 V2Hz-1.
Find the SNR at the pre-amplifier output.
 Review Questions Lec 39:

1. What are the disadvantages of direct modulation of diode

lasers?
2. What are the main phenomena on which an external modulator
works?
3. Which devices are suitable for digital and which for analog
external modulation of a laser?
4. What limits the operation frequency of external modulators?
5. What are the advantages and disadvantages of resonant ring
modulators?
6. Why are embeded ring modulators becoming important? What
are its limitations?