Optical Receivers

Theory and Operation

Block diagram of fiber optic
receiver.
 Photo Detectors
• Optical receivers convert optical signal (light)
to electrical signal (current/voltage)
– Hence referred ‘O/E Converter’
• Photodetector is the fundamental element of
optical receiver, followed by amplifiers and
signal conditioning circuitry
• There are several photodetector types:
– Photodiodes, Phototransistors, Photon multipliers,
Photo-resistors etc.
 Requirements
• Compatible physical dimensions (small size)
• Low sensitivity (high responsivity) at the
desired wavelength and low responsivity
elsewhere  wavelength selectivity
• Low noise and high gain
• Fast response time  high bandwidth
• Insensitive to temperature variations
• Long operating life and low cost
 Optical detector materials.

 Si,GaAs, GaAlAs – 850nm

 Ge, InP, InGaAs -1300nm and 1550nm.

 Materials determine the responsivity of the

detector which is the ratio of the output
photocurrent to the incident optical power.

 It’s a function of the wavelength and efficiency of

the device.
 Photodetector Types
• There are several Photodetector types:
– Photomultipler
– Pyroelectric detector
– Photodiodes
– Phototransistors
– Photon multipliers
– Photo-resistors etc.
 Photomultiplier
– Uses photocathode and electron multiplier in a
vaccum tube
– High gain at low noise
But
– Large Size and high voltage requirements
– Unsuitable for optical communication systems
 Pyroelectric detectors
• Involve the conversion of photon to heat
• Photon absorption result in temperature
change of the detector material.
• Dielectric constant change
• Response is flat over a wide range but not
suitable for optical communication systems.
 Photodiodes

• Mostly used
• Small size
• High sensitivity
• Fast response time
 Photodiodes
• Photodiodes meet most the requirements, hence
widely used as photo detectors.
• P-N photodiode
• Positive-Intrinsic-Negative (pin) photodiode
– No internal gain, robust detector
• Avalanche Photo Diode (APD)
– Advanced version with internal gain M due to self
multiplication process
• Photodiodes are sufficiently reverse biased during
normal operation  no current flow without
illumination, the intrinsic region is fully depleted of
carriers
 PN Photodiode
• Operated in reverse-bias.
• A photon striking in the depletion region cause a free
electron and a hole. The carriers move to their respective
junctions creating current flow.
 Physical Principles of
Photodiodes
• As a photon flux Φ penetrates into a semiconductor, it will be
absorbed as it progresses through the material.
• If αs(λ) is the photon absorption coefficient at a wavelength λ,
the power level at a distance x into the material is

Absorbed photons
trigger photocurrent
Ip in the external
circuitry
Optical Absorption Coefficient
Examples of Photon
Absorption
 pin energy-band diagram

Cut off wavelength:

hc 1.24 Cut off wavelength depends on the

c μm
Eg E g (eV ) band gap energy
 Quantum Efficiency
• The quantum efficiency η is the number of the
electron–hole carrier pairs generated per incident–
absorbed photon of energy hν and is given by

Ip is the photocurrent generated by a steady-state

optical power Pin incident on the photodetector.
 Responsivity
• The primary photocurrent resulting from absorption is:

q s( )w
Ip P0 (1 e )(1 R f ) [6-3]
h

• Quantum Efficiency:

# of electron - hole photogenerated pairs

# of incident photons
[6-4]
IP / q
P0 / h
• Responsivity:
IP q [6-5]
[A/W]
P0 h
Responsivity vs. wavelength
Advantage of PIN photodiodes.
The output electrical current is linearly
proportional to the input optical power
making it a highly linear device.
Low bias voltage(<4v).
Low noise
Low dark current
High-speed response
 AVALANCHE Photodiodes.
An APD internally amplifies the photocurrent by
an avalanche process when a large reverse-bias
voltage is applied across the active region.

The gain of the APD can be changed by

changing the reverse-bias voltage.
AVALANCHE Photodiode.
 Avalanche Photodiode (APD)
• APD has an internal gain M, which is obtained by
having a high electric field that energizes photo-
generated electrons.
• These electrons ionize bound electrons in the
valence band upon colliding with them which is
known as impact ionization
• The newly generated electrons and holes are also
accelerated by the high electric field and gain
energy to cause further impact ionization
• This phenomena is the avalanche effect
APD Vs PIN
 Responsivity ( )
Quantum Efficiency ( ) = number of e-h pairs
generated / number of incident photons
Ip / q Ip q
mA/mW
P0 / h P0 h

APD’s have an internal gain M, hence

APD PIN M where, M = IM/Ip

IM : Mean multiplied current

M = 1 for PIN diodes
 Responsivity

When λ<< λc absorption is low hc

c
When λ > λc; no absorption Eg
 Light Absorption Coefficient
• The upper cutoff
wavelength is
determined by the
bandgap energy Eg of
the material.

• At lower-wavelength
end, the photo response
diminishes due to low
absorption (very large
values of αs).
Comparison of photodetectors
 Photodetector Noise
• In fiber optic communication systems, the photodiode is
generally required to detect very weak optical signals.
• Detection of weak optical signals requires that the
photodetector and its amplification circuitry be optimized to
maintain a given signal-to-noise ratio.
• The power signal-to-noise ratio S/N (also designated by SNR)
at the output of an optical receiver is defined by

SNR Can NOT be improved by amplification

 Notation: Detector Current
• The direct current value is denoted by, IP (capitol main
entry and capital suffix).
• The time varying (either randomly or periodically) current
with a zero mean is denoted by, ip (small main entry and
small suffix).
• Therefore, the total current Ip is the sum of the DC
component IP and the AC component ip .

IP Ip ip
T /2
2 1 2
i p Lim T i (t )dt
p
T T /2
 Quantum (Shot Noise)
Quantum noise arises due optical power fluctuation
because light is made up of discrete number of photons

2 2
iQ 2qI p BM F (M )

F(M): APD Noise Figure F(M) ~= Mx (0 ≤ x ≤ 1)

Ip: Mean Detected Current
B = Bandwidth
q: Charge of an electron
 Dark/Leakage Current Noise
There will be some (dark and leakage ) current without any
incident light. This current generates two types of noise

2 2
Bulk Dark Current Noise i
DB 2qI D BM F (M )
ID: Dark Current

2
Surface Leakage iDS 2qI L B
Current Noise
(not multiplied by M) IL: Leakage Current
 Thermal Noise
The photodetector load resistor RL contributes to
thermal (Johnson) noise current

iT2 4K BTB / RL

KB: Boltzmann’s constant = 1.38054 X 10(-23) J/K

T is the absolute Temperature
• Quantum and Thermal are the significant noise
mechanisms in all optical receivers
• RIN (Relative Intensity Noise) will also appear in
analog links
 Signal to Noise Ratio
Detected current = AC (ip) + DC (Ip)
Signal Power = <ip2>M2
i p2 M 2
SNR
2q ( I p I D ) M 2 F ( M ) B 2qI L B 4k BTB / RL

Typically not all the noise terms will have equal weight.
Often thermal and quantum noise are the most significant.
Noise Calculation Example
 Limiting Cases for SNR
• When the optical signal power is relatively high, then the shot
noise power is much greater than the thermal noise power. In
this case the SNR is called shot-noise or quantum noise
limited.
• When the optical signal power is low, then thermal noise
usually dominates over the shot noise. In this case the SNR is
referred to as being thermal-noise limited.
 Limiting Cases of SNR
In the shot current limited case the SNR is:
i p2
SNR
2q ( I p ) F ( M ) B

For analog links, there will be RIN (Relative

Intensity Noise) as well
i p2 M 2
SNR
2q ( I p I D ) M 2 F ( M ) 4k BT / RL ( RIN ) I p2 B
Typical SNR vs. Received Power

• Note, APD
has an
advantage
only at low
received
power levels
 Noise-Equivalent Power
• The sensitivity of a photodetector is describable in terms of the
minimum detectable optical power to have SNR = 1.
• This optical power is the noise equivalent power or NEP.
• Example: Consider the thermal-noise limited case for a pin
photodiode. Then

To find the NEP, set the SNR = 1 and solve for P:

Response Time in pin photodiode

Transit time, td and carrier drift velocity vd are related by

td w / vd For a high speed Si PD, td = 0.1 ns
 Rise and fall times

Photodiode has uneven rise and fall times depending on:

1. Absorption coefficient s( ) and
2. Junction Capacitance Cj Cj o rA
w
 Junction Capacitance
A
Cj o r
w
εo = 8.8542 x 10(-12) F/m; free space permittivity
εr = the semiconductor dielectric constant
A = the diffusion layer (photo sensitive) area
w = width of the depletion layer

Large area photo detectors have large junction

capacitance hence small bandwidth (low speed)
 A concern in free space optical receivers
 Various pulse responses

Pulse response is a complex function of absorption coefficient

and junction capacitance
Comparisons of pin Photodiodes

NOTE: The values were derived from various vendor data

sheets and from performance numbers reported in the
literature. They are guidelines for comparison purposes.
 Comparisons of APDs

NOTE: The values were derived from various vendor data sheets
and from performance numbers reported in the literature. They
are guidelines for comparison purposes only.
Part B

OPTICAL RECEIVER
Signal Path through an Optical Link
 Fundamental Receiver Operation
• The first receiver element is a pin or an avalanche photodiode, which
produces an electric current proportional to the received power level.
• Since this electric current typically is very weak, a front-end amplifier
boosts it to a level that can be used by the following electronics.
• After being amplified, the signal passes through a low-pass filter to reduce
the noise that is outside of the signal bandwidth.
• The also filter can reshape (equalize) the pulses that have become distorted
as they traveled through the fiber.
• Together with a clock (timing) recovery circuit, a decision circuit decides
whether a 1 or 0 pulse was received,
 Optical receiver schematic

Bandwidth of the front end:

CT: Total Capacitance = Cd+Ca

B 1 2 RT CT
RT: Total Resistance = Rb // Ra
Try Example 6.7 in Keiser
 Noise Sources in a Receiver
The term noise describes unwanted components of an electric signal that tend to
disturb the transmission and processing of the signal
• The random arrival rate of signal photons produces quantum (shot) noise
• Dark current comes from thermally generated eh pairs in the pn junction
• Additional shot noise arises from the statistical nature of the APD process
• Thermal noises arise from the random motion of electrons in the detector
load resistor and in the amplifier electronics
 Probability of Error (BER)
• BER is the ratio of erroneous bits to correct bits
• A simple way to measure the error rate in a digital data stream
is to divide the number Ne of errors occurring over a certain
time interval t by the number Nt of pulses (ones and zeros)
transmitted during this interval.
• This is the bit-error rate (BER)
• Here B is the bit rate.
• Typical error rates for optical fiber telecom systems range
from 10–9 to 10–12 (compared to 10-6 for wireless systems)
• The error rate depends on the signal-to-noise ratio at the
receiver (the ratio of signal power to noise power).
 Logic 0 and 1 probability distributions

1
P0 (Vth ) p( y / 0)dy
Pe 2 P1 (Vth ) P0 (Vth ) Vth

Asymmetric distributions Vth

P1 (Vth ) p ( y /1)dy
Select Vth to minimize Pe
 Deciding Threshold Voltage
Probability of error assuming P 1
P1 (Vth ) P0 (Vth )
e 2
Equal ones and zeros

Where, P0 (Vth ) p( y / 0)dy

Vth

Vth
P1 (Vth ) p ( y /1)dy

Depends on the noise variance at on/off levels and the

Threshold voltage Vth that is decided to minimize the Pe

Question: Do you think Vth = ½ [Von + Voff] ?

 Derived BER Expression
• A simple estimation of the BER can be calculated by assuming the
equalizer output is a gaussian random variable.
• Let the mean and variance of the gaussian output for a 1 pulse be bon
and σ2on, respectively, and boff and σ2off for a 0 pulse.
• If the probabilities of 0 and 1 pulses are equally likely, the bit error
rate or the error probability Pe becomes
 Probability of Error Calculation
• The factor Q is widely used to specify receiver performance, since it
is related to the SNR required to achieve a specific BER.
• There exists a narrow range of SNR above which the error rate is
tolerable and below which a highly unacceptable number of errors
occur. The SNR at which this transition occurs is called the
threshold level.
 BER as a Function of SNR
BER as a function of SNR when the standard deviations are equal
(σon = σoff) and when boff = 0
 Receiver Sensitivity
• A specific minimum average optical power level must arrive at
the photodetector to achieve a desired BER at a given data rate.
The value of this minimum power level is called the receiver
sensitivity.
• Assuming there is no optical power in a received zero pulse,
then the receiver sensitivity is

Where, including an amplifier noise figure Fn, the

thermal noise current variance is
 Receiver Sensitivity Calculation
The receiver sensitivity as a function of bit rate will change for a given
photodiode depending on values of parameters such as wavelength,
APD gain, and noise figure.
 The Quantum Limit
• The minimum received optical power required for a specific bit-error rate
performance in a digital system.
• This power level is called the quantum limit, since all system parameters
are assumed ideal and the performance is limited only by the detection
statistics.
 Eye Diagrams
• Eye pattern measurements are made in the time domain and
immediately show the effects of waveform distortion on the
display screen of standard BER test equipment.
– The eye opening width defines the time interval over which signals can be
sampled without interference from adjacent pulses (ISI).
– The best sampling time is at the height of the largest eye opening.
– The eye opening height shows the noise margin or immunity to noise.
– The rate at which the eye closes gives the sensitivity to timing errors.
– The rise time is the interval between the 10 and 90% rising-edge points
 Stressed Eye Tests
• The IEEE 802.3ae spec for testing 10-Gigabit Ethernet (10-GbE) devices
describes performance measures using a degraded signal.
• This stressed eye test examines the worst-case condition of a poor
extinction ratio plus multiple stresses, ISI or vertical eye closure, sinusoidal
interference, and sinusoidal jitter.
• The test assumes that all different possible signal impairments will close
the eye down to a diamond shaped area (0.10 and 0.25 of the full pattern
height).
• If the eye opening is greater than this area, the receiver being tested is
expected to operate properly in an actual fielded system.

The inclusion of all possible signal

distortion effects results in a
stressed eye with only a small
diamond-shaped opening

62
 Architecture of a Typical PON
• A passive optical network (PON) connects switching equipment in a
central office (CO) with N service subscribers
• Digitized voice and data are sent downstream from the CO to customers
over an optical link by using a 1490-nm wavelength.
• The upstream (customer to central office) return path for the data and voice
uses a 1310-nm wavelength.
 Burst-Mode Receivers
• The amplitude and phase of packets received in successive time slots from
different user locations can vary widely from packet to packet.
• If the fiber attenuation is 0.5 dB/km, there is a 10-dB difference in the
signal amplitudes from the closest and farthest users.
• If there are additional optical components in one of the transmission paths,
then the signal levels arriving at the OLT could vary up to 20 dB.
• A fast-responding burst-mode receiver with high sensitivity is needed

The guard time

provides a sufficient
delay time to
prevent collisions
between
successive packets
that may come from
different ONTs.