Jitter Fundamentals:

Agilent 81250 ParBERT Jitter Injection and

Analysis Capabilities

Application Note

Introduction

In digital communications, a Jitter Fundamentals

sequence of 0’s and 1’s flows
from a transmitter to a receiv- • What is Jitter?
er. The transmission media is • Jitter Components
copper, fiber or air. • Jitter Challenges
Communication systems are
• Jitter in SONET and Ethernet
maximized for bandwidth and
minimized for errors. These • Stressed Eye
errors are counted as Bit • Jitter Measurement in Time/Frequency Domain
Error Ratio (BER).
81250 ParBERT Jitter Capabilities
Digital communication systems
transmit the pure bit stream • Delay Control / Clock Modulation
in terms of a NRZ data stream • Bandwidth Control
only, and then regenerate the
• Bath Tub Measurement with RJ / DJ Separation
bit clock at the receiver
through the use of a clock and • Spectral Decomposition of Jitter
data recovery (CDR) circuit. Figure 1: Jitter Fundamentals
Timing aberrations of the
incoming signal causes mal-
function of the CDR circuit
resulting in bad sampling of
the data causing bit errors.
These timing aberrations are
called jitter.
 Why is jitter an important issue?

Jitter is significant because it Data Cycle, Bit Period

is one of the major potential
causes for data being received
in error. For example, if a long
string of bits is on the short
side, eventually a receiver will
make a decision at the edge of
the bit rather than the center It causes bit
(if the clock rate is held con- errors!
stant). This will result in
errored bits. Another perspec-
tive is to view the eye dia-
gram. As the jitter increases,
eventually the eye will close
horizontally. In the eye open-
ing display, one can see how
the eye closes the higher the
BER figure gets. The represen- Figure 2: Why Jitter is an important issue
tation here defines so called
ISO-BERs. In the original
Graphical User Interface (GUI),
this diagram would be in
color, and the color coding on
the right defines the BER fig-
ure for each line. If the clock
is derived from the data
(CDR), the sampling point can
then follow the jitter and allow
the system to tolerate jitter. A
clock recovery process is limit-
ed by the loop bandwidth of
the clock recovery circuitry.

There are many challenges to - Design goal for maturing

today’s jitter measurements: high speed interfaces: Generate
robustness with circuits that
- Huge problems on complex show more margin against
ASICs to avoid ground bounce process influence on DJ and
and clock crosstalk induced RJ.
jitter.
- Increasing difficulties for
- New high speed IO standards design validation and debug-
require DJ, RJ specified sepa- ging.
rately.
- A better analysis of jitter
- SONET based jitter transfer mechanisms is required.
specs only make sense when
there is no other (intrinsic) - P robabilistic view (histogram)
jitter source than the artificial- is no longer sufficient.
ly injected jitter.

- New defect types in produc-

tion: process variation affects
gross yield through different
influences on DJ and RJ.
2
By examining the edges of a
• Significant instants can
digital communications bit usually be defined as
stream, we can better illustrate edges
our definition. Figure 3 shows
• The system clock can be
an oscilloscope display of a used to define what the
data stream with the system ideal positions in time are
clock waveform. If the timing
• Edge position of the data
of this bit stream is jitter free, should consistently align
the period for all of the bits with the same relative
will always be precisely identi- points on the reference
cal. Thus the time between any clock waveform
two rising or two falling edges
will always be a precise inte-
ger multiple of the nominal bit
period. Figure 3: Ideal signal = constant bit period

Another way to look at this is

to look at the data stream rel-
ative to an ideal clock source.
The time between a data edge
and the closest clock edge
should always be the same. If
the data signal is jitter free,
then the 50% amplitude points
on the data waveform should
consistently align with points
on the clock waveform .
However, if the bit period fluc-
tuates for any reason, the bit
stream will no longer be jitter
free.

Although there are differences A more sophisticated definition

in the many definitions of jit- from the viewpoint of the
ter, the fundamental similarity SONET standard [1] is: “Jitter
is that jitter has to do with is the short term phase vari-
the time difference between ation of the significant
the ideal and actual occur- instants of a digital signal
rence of an event. A simple from their ideal positions in
definition is: time. It is primarily con-
cerned with non-cumulative
“Jitter is the time difference variations above 10 Hz.
between when a pre-defined Cumulative phase variations
event should have occurred below 10 Hz are referred to
and when it actually did as wander.”
occur. The time difference is
expressed in unit interval The Fibre Channel community
(UI), 1 UI is the value of the defined it thus: “Jitter is the
bit period of the ideal clock deviation from the ideal tim-
signal. This time difference ing of an event. The reference
can be treated as phase mod- is the differential zero cross-
ulation, there are one (or ing for electrical signals and
more) signals modulating the the nominal receiver threshold
ideal position of the data sig- power level for optical sys-
nal.” tems.” [2]
3
 Jitter Components (1)

Jitter consists of two funda-

mental components called
Total Jitter (pp) :
Random Jitter (RJ) and Data jitter
Deterministic Jitter (DJ). plus
n(BER) xs
Figure 4 illustrates an example
of a signal with an extremely
large amount of DJ. Random s :sigma

Jitter is unbounded and is BER n

usually best described by a -------------------
10 -6 9.8
Gaussian probability density 10-9 12.2
function. Deterministic jitter is 10-10 12.7
bounded, ie it has definite 10-12 14.1
10-14 15.3
amplitude limits from an earli-
est to a latest trace. The
Random Jitter (RJ) is defined Random Jitter (RJ): Deterministic Jitter (DJ):
by an rms value which in this Defined by RMS value which spacing between mean values of
case equals the ‘s’ (sigma) of equals s (sigma) of the “earliest” and “latest” trace
Gaussian distribution
the Gaussian distribution. The
total jitter is a function of ‘n’ Figure 4: A real signal
times the rms value, depending
on what BER limit a system is
specified, plus the
Deterministic Jitter (DJ) which
is by nature a peak- to-peak
value.

4
Jitter Components (2)

This is the whole picture of

the jitter components: Total Jitter
The Deterministic Jitter (DJ) pk-pk
has several faces, e.g. it is
caused by bandwidth limita- Random Deterministic
tions and component interac- Jitter Jitter
tion (crosstalk). The diagram Bounded, pk-
shown here separates the indi- Unbounded, rms
pk
vidual components and pins
the nature of the origin. Periodic Data Dependent Bounded
The first layer separates the Jitter Jitter Uncorrelated
DJ into: Sinusoidal Data smearing Crosstalk

- Periodic Jitter (PJ), which Duty Cycle Inter Symbol

displaces the timing of rising Distortion Interference
and falling edges with a peri- Lead/trail Long/Short bits
odic pattern (or as the origin
edge
of this type of jitter is sinu-
soidal modulation, it is also Figure 5: Jitter Component Segmentation
called SJ).

- Data Dependent Jitter (DDJ)

(which is a function of bit
patterns).

- Bounded Uncorrelated Jitter This can also be explaind from

(BUJ) which is caused by Bandwidth limitations which
interference with asynchronous occur from AC coupling (low
signals: cross-talk between frequency cut-off) or from high
sub-circuits, power supply frequency roll-off. However,
noise and electro-magnetic this is from filters with non-
interference (EMI). linear phase characteristic
only. Linear filters, such as
In the next layer the Data Bessel filters, do not cause jit-
Dependent Jitter can be sepa- ter. DCD and ISI jitter are a
rated into: function of the pattern. The
jitter appears when changing
- Duty Cycle Distortion (DCD), the pattern from a clock-like
which is caused by voltage off- pattern to real data. PRBS
sets between differential inputs type patterns are good for
and differences between transi- testing as they contain many
tion times within a system. variants of frequency compo-
nent.
- Inter-Symbol Interference
(ISI), which is caused by the
different symbols (long and
short bit cycles).

5
 Random Jitter (RJ)

RJ is caused by thermal and

noise effects. These effects are
statistical by nature. So the
Random Jitter (RJ) is
unbounded and is modeled by General: #
events = n x s (sigma)
a probability density function
Random Jitter: n(BER) xs
and is quantified by the rms
value of the density function. mean value N # events BER
In most cases the Gaussian Normalized
Events sigma sigma sigma sigma ---------------------------
distribution is used for the
2 67% 0.33
characterization of Random
4 97% 0.03
Jitter. In this case the rms
6 99.7% 0.003
value equals the ‘s’ (sigma) of
9.8 106 10-6
the Gaussian distribution.
12.2 109 10-9
There is a fixed relation
14.1 1012 10-12
between ‘s’ (sigma) and the
number of events. While a
time
range of 6 sigma already
defines a high number of Figure 6: Random Jitter: The Gaussian Distribution
events (99.7%), this is a small
number for BER. Typically a
BER is required to be as low
as 10-12, this needs a range of
14.1 sigma to be included for
the total jitter budget.

In the past the common model

was to fit the actual jitter his-
togram into the Gaussian curve
and the rms value was
extracted. For a signal with
significant DJ this resulted in
a value way too high for the
rms figure. This led to the
RJ/DJ separation model. So
first any DJ jitter component
is isolated and only the
‘remaining’ jitter is then treat-
ed as RJ. DJ would broaden
the Gaussian curve, while the
tail portions of the histogram
represent the RJ jitter compo-
nent [3]. However, it is not
necessary to perform his-
togram measurements to iso-
late RJ/DJ components. Later
in this Note it will be shown
that the BERT Scan
Measurement allows
Random/Deterministic Jitter to
be separated directly. This
eliminates the need for specific
jitter test sets, because the jit-
ter measurements can be
obtained from BER test equip-
ment.
6
Inter-Symbol Interference

Inter-Symbol Interference caused by bandwidth limitation / loss

describes the amount of jitter
occuring through data content. • low/(high) pass filter
Here the issue is the transmis-
sion of long and short bits. • cable droop
Within the data stream the
data cycles most affected are R 1
1 2 2
those which contain a single
C
bit state which is opposite to
the surrounding bits. In Figure
7 this is represented by the 3
3
trace which shows a data bit Zl
being one cycle a ‘1’ while it
is surrounded by ‘0’s. 1: 1-> 0 transition
2: 0 -> 1 transition
This type of jitter results from 3: 1 UI pulse
bandwidth limitations which
are low/high pass filters with
non-linear phase characteristic, Figure 7: ISI: Inter-Symbol Interference
or from loss within transmis-
sion lines. Loss in
cables/microstrips causes droop
which again limits signal set-
tling before the next transition
occurs. Both bandwidth limita-
tion and cable loss lead to a
shortening of this bit as this is
insufficient time to settle the
signal to 100% before the
opposite state starts. This
leads to an early start of the
transition with the disadvan-
tage that 50% is too early and
within random data, the eye
will start to close.

A lot of jitter tests use specif-

ic filters which incorporate jit-
ter as inter-symbol interfer-
ence. These filters are called
JIMs (Jitter Injection Modules),
and are built as L-C chains,
which resulting in a higher
order. Here, Phase Shift occurs
not only for a single bit but
also for 2,3 or more consecu-
tive bits. This is applied in the
so called “Stressed Eye Tests”,
which will be covered later in
this Note. [4]

7
 Periodic Jitter (PJ)

Periodic Jitter provides sinu-

soidal varying jitter, where a
sinusoidal signal modulates the
Ideal clock: sin(2p f c t )
phase of the ideal clock or
data signal. The sinusoidal sig-
nal may be asynchronous or
synchronous to the clock / sin (2p f c t + 43 p sin( 101 2p f c t ) )
Jittered clock:
data signal. Synchronous clock
sub-rates cause every ‘n’th
cycle to be shorter/longer than
the others. Jitter: 4
p sin( 101 2p f c t )
3

Often this type of jitter is 2

3 UI
used in jitter testing. In this
case a sinusoidal signal is
used to displace the data
edges. This helps to character-
ize the bandwidth of CDR cir-
cuits. SONET Jitter testing is
Figure 8: Periodic Jitter
standardized for all jitter com-
ponents to be sinusoidal.

Figure 9 shows the traditional

eye measurement taken from a
sampling scope. This allows
the jitter to be viewed as rms What you see today...
and peak- to-peak values and a
histogram is generated. The
histogram can tell us some
details about the jitter distri-
bution. But the histogram is
not able to get insight on what
lies beneath this. In this case
there is a mix of Random and
Sinusoidal jitter. As mentioned ...What is behind
earlier, with help of some DSP (RJ=0.2UIrms 80MHz bandwidth
(Digital Signal Processing) tools , DJ=0.05UIpkpk, 10MHz)
it would be possible to do a
RJ/DJ separation from the Figure 9: Jitter Measurement Situation
Histogram for gaining the
rms/peak- to-peak value of the
jitter components. But it is
impossible from the histogram
to gain the bandwidth of the
noise or the frequency of the
sinusoidal interference. This
requires another approach.

8
Jitter in the Frequency Domain

Figure 10 introduces jitter in Clock with 1MHz Sinusoidal Jitter

the frequency domain. This
shows a diagram with frequen-
cy on the x-axis and power
Jitter Spectrum, Peak at
(or power factor) on the y- 1MHz
axis. The graphs show the two
jitter components separated for
sinusoidal and random. In the
histogram there is only a Jitter 25ps (0.08UI pp)
slight difference between the
Clk with 80MHz Broadband Gaussian Jitter
Sinusoidal (DJ) and the
Gaussian (RJ) jitter. In the fre-
quency decomposition the
sinusoidal jitter is represented
as a single line occurring at a
single frequency. The random
noise shows a wide frequency
spectrum up to the bandwidth
Jitter 13ps rms (0.04UI rms)
of the random spectrum (in
this case 80 MHz). Above this Figure 10: Spectral Decomposition of Jitter
value, the power decreases
down to the noise floor.
This kind of representation is
possible using a Spectrum
Analyzer with a Phase
Discriminator. The Phase
Discriminator (or Demodulator)
de-modulates the signal to
obtain the phase modulation
signal which is what we see in
the figures above. This type of
measurement is very common
to jitter measurements on
SONET devices. However, one
has to be aware that this is
practical for jitter on a clock
signal, but cannot be used for
true data signals.

There is a new tool usable on

BER test equipment, which can
perform these Spectral
Decomposition Measurements
eliminating the need of
Spectrum Analyzer equipment
and the capability to do these
measurements on true data
signals. This will be explained
later in this Note.

9
 The BERT Scan Measurement

The Gigabit Ethernet approach

to measuring something that is Measurements:
Data OUT Vector n
unbounded is through a Bit- -> Optimum Sampling
Error-Ratio measurement, Range for Point
specifically through a “Bath
Analyzer sampling delay -> Skew
Tub plot”: -> Phase Margin
Bit Error
-> Setup/Hold Time
Data Jitter will result in the Ratio -> Jitter
signal edges moving toward the
Timing Measurements:
center of the eye diagram.
-> Bath Tube
Extreme excursions will occur
less frequently than minor Bit Error Eye Opening
excursions. If the transmit sig- Threshold at BER 10-12
nal is fed to an error detector Time
Jitter Measurements:
and the sampling point is opti-
mized in both time and ampli-
tude, the error rate should be
well below 10E-12 (as close to
zero as can be measured). Figure 11: BERT Scan
As the sampling point is con-
tinually moved into the edges
of the eye, the BER will get
steadily worse.

The important element for The IEEE 802.3ae standard

doing characterization is the sets the allowable jitter magni-
ability to move timing edges tude at the 10E-12 BER level.
around. On an analyzer this Thus the “eye” must have a
allows to move the sampling specified opening at this BER.
point around.
Bath Tub measurements are
It is important to state: this the basis for all measurements
measurement can be performed listed above including jitter
on differential signals! measurements. For the jitter
Moving the sampling point over measurement (especially the
one cycle and plotting the RJ/DJ separation) the transi-
error rate, results in a graph tional behavior of the left and
called ‘Bath Tub’. The x-axis the right slope between two
of this graph is time, the y- points specified as BER figures
axis is the BER figure. The will be extrapolated.
name ‘Bath Tub’ results from
the specific shape of the curve.
This specific shape results
from the fact that the lower
the BER figure gets, the more
test vectors have to be
processed. The longer the test
runs, the wider the jitter band
of a real world signal will be.
So the lower the BER thresh-
old is specified, the less the
resulting phase margin will be.

10
Jitter within the Standards

Jitter is described differently SONET/SDH – ITU- Gigabit Ethernet and 10 Gigabit

T0.172, GR-253-CORE Ethernet – IEEE 802.3z and 802.3ae
within the various standards
within the communication Asynchronous Architecture
industry. The SONET as a syn- Synchronous Architecture
Transceiver timing not derived from received
chronous architecture deals Transceiver timing derived from received
signal therefore jitter propagation
signal therefore jitter propagation less critical

with Jitter Generation, Jitter critical • 1 Gb En

Tolerance and Jitter Transfer. • Jitter Generation Bathtub plots to measure jitter
Jitter Generation is the jitter Characterizes transmitter jitter generation
performance
coming from the device, Jitter Transfer
• 10 Gb En: Serial TDP
• Jitter Transfer

n ce
(Transmitter dispersion Penalty) for
Tolerance is the jitter which

er a
transmitters Stressed Eye receiver test

n
Characterizes clock recovery

atio
the device can handle, and

Tol
performance against jitter – similar to Jitter Tolerance

n er
Jitter Transfer is the jitter XAUI Bathtub,

Ge
• Jitter Tolerance
Stressed Eye
which moves through the Characterizes receiver’s tolerance to
jitter
device from input to output.
The jitter used for the meas-
urements is sinusoidal only.
Figure 12: Jitter: the Standards
The more modern standards
are dominated by the Ethernet
community dealing with asyn- SONET/SDH = Synchronous clocking

chronous architecture. Here the • Important spec is Jitter Transfer

• Outgoing signal is retimed with recovered clock
jitter is a mixture of random Signal of incoming signal so that clock propagates
and deterministic content. And processing through the system
CDR • Disadvantage: imparities can accumulate until
the 10GbE community has clean-up PLL is used
established new concepts of • Advantage: as data rate is identical everywhere
number of data bits on both sides of DUT is
measuring: Transmitter identical, too
Dispersion Penalty (TDP) and
Stressed Eye Test. FIF (10Gb)Ethernet
Signal O • No such spec as Jitter Transfer
processing
• Outgoing signal is retimed with independent
The difference between SONET CDR DUT clock which may have a slightly different
and Ethernet is the clocking indep. rate (+/- 100ppm)
clk • Advantage: no accumulation of clock disparities
system: / jitter
• Disadvantage: as number of data bits on both
sides of DUT is not identical, bit-
SONET/SDH uses a common stuffing/discarding scheme is required
system frequency which is
recovered by CDR within each Figure 13: The difference in clocking of SONET/SDH and Ethernet
receiver, and then used within
the whole circuity. Therefore it
is called synchronous or a- each sides of the DUT is iden- advantage is that no accumula-
synchronous clocking. The out- tical, too. tion of clock disparities/jitter
going signal is retimed with occur throughout the device,
the recovered clock of the For (10Gb) Ethernet there is so no jitter transfer happens
incoming signal so that the no such specification as Jitter and therefore no such specifi-
clock propagates through the Transfer. The reason for this is cation or measurement is nec-
system. An important specifi- that the outgoing signal is essary.
cation therefore is Jitter retimed with an independent
Transfer. A disadvantage is Reference Clock within the
that the imparities can accu- Receiver. This new Reference
mulate, as from sub-system to Clock is allowed for a slightly
sub-system jitter can add up different rate (typically +/-100
until a very specific clean-up ppm). The resulting disadvan-
PLL is used. An advantage tage is that the number of
(especially for testing) is that data bits on each side of DUT
the data rate is identical is not identical. So an Idle
everywhere in the system, and scheme with bit-stuffing/dis-
the number of data bits on carding is required. The major
11
 Stressed Eye

Stressed Eye for Ethernet [5]

is similar to Jitter Tolerance
• Sinusoidal jitter modulates the
for SONET/SDH. While clock timing the pattern
SONET/SDH deals with sinu- generator
soidal jitter only, here the
stressed eye is a well defined
• Sinusoidal interference signal
mix of periodic jitter in time
summed with jittered data
and amplitude. pattern
0 to 40 10.3125 Pattern
An arbitrary Waveform GHz
Generator is used to modulate MHz Clock Generator • Summed signal low-pass
filtered
a Signal Generator which
sends the jitter modulated
clock into the Pattern ~1-2 GHz 7.5 GHz
Generator. The data stream is sine Laser
LPF
added with a sinousoidal
interference signal and finally Figure 14: Stressed Eye Example
is low-pass filtered for a spe-
cific shape supplied to optical
converters. Often there are
additional filters to the Pattern
Generator output to include
Data Dependent Jitter (ISI)
with the data pattern.

12
ParBERT 81250 Jitter Capabilities

The 81250 ParBERT can emu-

late (1) and measure (2+3) jit- BER
ter. For Jitter Tolerance type 1st
testing, the ParBERT allows 2nd
either a modulated clock to be Delay
Bath Tub RJ
worked with and/or the delay Control
/ DJ
of the generator output to be
controlled (by 3.3 Gb/s genera- Modulated
tor E4862B and 13.5 Gb/s gen- Clock (<10 3rd
kHz)
erator N4872A).

Clock
Jitter Decomposition
The modulation of the external Data In Data Out
clock can be done to achieve
multiple Unit Intervals (UI’s) DUT
as jitter, but this is limited to
a certain bandwidth as there Generators 3.3G only

are PLLs within the internal

clock distribution path. The
Figure 15: 81250 ParBERT Jitter Capabilities
PLLs will filter the higher fre-
quency contents of the modu-
“Clean Eye” Jitter modulated with Sine - Wave
lation signal. With the delay
control input each generator
can be controlled individually
up to a modulation frequency
of 200 MHz and a peak- to -
peak modulation of 500 ps. At
3 Gb/s speed this allows a jit-
ter budget exceeding the eye
totally. The modulation signal
type controls the distribution Jitter modulated with Rectangle -Wave
of the jitter. With a mix of Jitter modulated with Noise
random and square wave sig-
nal one can emulate a mix of
RJ and DJ.

The measure capability of the

81250 ParBERT is BER, Bath
Tub, Eye Opening and Fast
Eye Mask. With the Bath Tub
Figure 16: Jitter Injection
it is possible to read the jitter
separated for RJ and DJ.
As a source for these modula- The control inputs of the dif-
The ParBERT generators (3.3 tion voltages Agilent offers the ferent ParBERT generators
Gb/s E4862B & 13.5 Gb/s 3325A Function/Arbitrary work similarly but differ in
N4872A) provide a Control Waveform Generator. This type parameters: the 3.3 Gb/s gen-
Input for modulating the Delay of instrument can generate any erator (E4862B) allows a delay
with help of an external sig- mentioned type of signal, but modulation of +/- 250 ps up
nal. This modulation can be one at a time. By combining to 200 MHz modulation band-
used to emulate jitter. Figure two of them with help of a width, the 13.5 Gb/s generator
16 shows this jitter emulation power splitter (11667B), a mix (N4872A) allows a range of +/-
as a scope view for three dif- of two signals is possible. So 100 ps up to 1 GHz band-
ferent types of control voltage: mixing random and Sine/rec- width.
sinusoidal, rectangle and ran- tangle emulates jitter with RJ
dom. This Jitter Modulation and DJ components.
can be used to test a DUT for
Jitter Tolerance or to built a 13
 The ParBERT can run on an
external Clock. For Jitter Transfer
Emulation where multiple UI UI Clock
2 Distribution to
jitter is required (what the Modules
delay control cannot offer) an 1
externally modulated Clock can
be fed to the ext. Clk Input of
3
the clock module. But there
are PLLs in the clock path Cutoff~ Frequency Ext.
between the ext. Clk input and 5kHz Clock
the channels. There are some
1: linear transfer Input
specific restriction depending
on the specific speed class of 2: max. bandwidth Clock &
ParBERT data clock and data
3: suppression Data
modules.
Modules
- 675 MHz data modules used Figure 17: ParBERT 81250 Jitter Emulation with a modulated Ext. Clock
with E4805 /E4808A clock
module: In regular operation
there is a PLL inside the clock
module with a cut-off frequen-
cy of around 10 kHz. So one
can generate a modulated jitter
covering multiple UIs (linear
transfer range) only up to this
cut-off frequency. There is a
specific procedure to bypass
the PLL, in this case the cut-
off can be avoided. For details
of operation contact ParBERT
technical support.

- 1.65 Gb/s, 2.7 Gb/s, 3.3 Gb/s A modulated clock is achieved

and 10.8 Gb/s with any clock by modulation of a signal gen-
module: There are always mul- erator. Agilent offers various
tiplying PLL‘s in the clock signal generators. A signal
path inside the data modules generator provides a sinusoidal
which cannot be bypassed. signal, which is fine for the
These have a bandwidth limit ext. Clk input of the ParBERT
of 10 kHz. clock module. The power level
is also sufficient. To establish
- 13.5 Gb/s data modules a modulation, the signal gener-
together with E4809A clock: ator again needs a function
This module offer a clock path generator or another signal
at speed, so there are no mul- generator as its modulation
tiplying PLL‘s. This achieves an input.
unlimited bandwidth of the
clock path, so a modulation of
multiple UIs is possible up to
clock speed.

14
Figure 18 shows three meas-
urements of a Bath Tub curve
with three signals of different
jitter content. As a stimulus to BER Threshold
the ParBERT analyzer a
RJ, Random Jitter, rms
ParBERT generator is used:
Range for
Extra- DJ, Deterministic Jitter
Low RJ: This is the most ideal polation
BER Extraplolated Total Jitter (for
signal (clean data signal, no BER 10^-6 to 10^-15)
modulation) from a ParBERT Min. BER Time Quality of Fit
3.3 Gb/s generator.

High RJ
High RJ: This is the modulated
Low RJ
signal using noise applied to
RJ + DJ
the control input. As the noise
closes the eye, it reduces the
phase margin. The slopes of Figure 18: 81250 Bath Tub with RJ / DJ
the Bath Tub curves get less
steep as in the case of no
modulation.

RJ + DJ: This is a modulated The RJ extrapolation is done

signal using noise and sinu- within the Bath Tub from a
soidal modulation. The amount range specified by two points
of modulation is set for for BER: one is the BER
obtaining the same phase mar- threshold, the other as a Min
gin as the High RJ signalat the BER figure (see Figure 18). In
BER threshold of 10- 3. But the the table below the graph,
shape is different especially for there a 4 colums dealing with
the ‘shoulder‘ at BER = .25, the quality of fit obtained, the
which is characteristic for the R*2 value is the appropriate
DJ jitter component. indicator, this value should be
between .75 and 1 for good
The visualisation of the Bath extrapolation.
Tub Measurement includes a
tabular format under the
graph, which represents the
measured values. Each signal
reads for Phase Margin (out of
the viewable range), Jitter
Mean, Random Jitter (RMS) =
RJ, Deterministic Jitter = DJ,
Estimated Total Jitter (which
is an extrapolation value for
total jitter at a very low BER
threshold, much lower than
measured), and finally there
are 4 colums for representing
the quality of fit for the RJ/DJ
separation.

All the values can be read

either in time (ps) or in UI
(Unit Interval) by configurating
the Window View.

15
 Figure 19 shows the so called
Property Window for the
parameters defining of the
Two values
RJ/DJ separation.
defining the
range for RJ
Two values of the Bath Tub extrapolation
properties define the range for
the RJ extrapolation: the BER This value
Threshold defines the upper defines the BER
value, the Min BER for RJ/DJ value at which
seperation defines the lower the
value. In practice, the values ‚extrapolated‘
should be set to the lower Total Jitter is
region of the Bath Tub curve displayed
where the slope of the curves
is defined by the random jitter
only. However, the range needs
be wide enough for a couple Figure 19: Bath Tub Properties for RJ / DJ
of measurement points to be
included. Otherwise the fit will
be marginal due to insufficient
number of points.

The Residual BER for

Estimated Total Jitter defines •DCD is
the BER value at which the
extrapolated Total Jitter is cal- •Difference of
culated and displayed. This
extrapolation method saves
horizontal eye
measurement time, as it elimi- opening for ‘Errors
nates the need for measure- if 0’s Expected and
ments, which at a BER thresh- Errors if 1’s
old of 10-12 or lower the Expected
measurement time is very
large. So the extrapolation
method is as fast as the meas-
urement for a BER down to
10- 6 which can be done within
a few seconds of measurement
Figure 20: Jitter Measurement, DCD extraction
time.

With help of the Bath Tub

measurement and the property Simply perform the measure-
settings, it is also possible to ment once for ‘All Errors’ and
extract the value for the Duty after the measurement, check
Cycle Distortion (DCD) Jitter the values for the phase mar-
Component. The duty cycle gin once for ‘Errors if ‘0‘s
distortion is the difference of Expected’, and once for ‘Errors
the phase margin (horizontal if ‘1‘s Expected’. The difference
eye opening) for the Bath Tub of the two phase margin val-
curves of separating between ues is the DCD value.
Errors on ‘0‘s only and Errors
on ‘1‘s only. The parameter
setup allows the selection for
this settings, indicated in
Figure 20. It is not necessary
to take the measurement twice.

16
Spectral Decomposition of Jitter

Deterministic Jitter can be

analysed for its spectral con- Jitter Power
tents. This is a complimentary
view in the frequency domain, Incoming Data
while the BERT Scan method Deterministic
analyses jitter in the time Error Function, Jitter
domain. DSP
Expected
The Jitter Decomposition is Post-Processing
Data
gained from a specific meas-
urement, which uses an Error
Function obtained from the
real-time compare of incoming Unlimited Frequency
data against expected data.
The error function is the pass- Figure 21: Jitter Decomposition by Spectrum Analysis
fail information over a certain
number of data bits. The con-
tent of the error function is
processed with Digital Signal
clock
Processing Tools (DSP) to visu-
alize the spectral information.
data clock
The DSP tools used are the CDR
Autocorrelation, the Furrier Phase Noise Measurement:
Transformation and the power • Clock only, no Data Jitter
SpectrumA
nalyzer
density calculation. The DSP • CDR, Bandwidth limitted
processing delivers a power Different to ParBERT
factor as a function of fre-
quency which is visualized in data
an x-y graph. The algorithm Real Time Scope Measurement
works in a wide frequency (EZJIT):
range: from DC up to half of • Histogram Scope
the signal’s data rate. • Software CDR
• Jitter Trend/Spectrum
There are two principles to • Spectrum is without Bandwidth
limitting/rating
gain Spectral Jitter informa- Similar to ParBERT
tion: Figure 22: Existing Spectral Jitter Solutions

Firstly, there is the Phase

Noise Measurement. Any CDR has its characteristic data stream. Out of this it
SONET/SDH does all jitter in terms of bandwidth, so part runs a software based CDR,
testing according this principle. of the jitter is removed, espe- and with further help of DSP
This uses a Spectrum Analyzer cially the high frequency part. technology, it gains the spec-
together with a phase discrim- So SONET/SDH looks at the tral information. This is a sim-
inator. A discriminator or de- jitter only within a specified ilar methodology to the
modulator extracts a signal bandwidth range. The phase ParBERT implementation.
which is proportional to the noise measurement is different
phase deviation in the running from the Spectral
signal, and the Spectrum Decomposition offered on the
Analyser visualizes the power ParBERT.
factor versus frequency. This
type of measurement is possi- Another solution is the Jitter
ble on clock signals only. So Measurement Software (EZJIT)
for a data signal a CDR (Clock on the Agilent Infiniium 54850
Data Recovery) is needed to series oscilloscopes. This solu-
convert it into a clock signal. tion samples a portion of the
17
 Use of Jitter Decomposition • DEBUG:
Measurement:

This type of measurement is

Qualitative
embedded within the ParBERT
information on
measurement suite, as a new frequency
1MHz
measurement in the software components
measurement package. As it is within jitter
providing qualitative informa- budget
tion on the frequencies within
the Deterministic Jitter budget, Example:
it is a DEBUG Tool. So for interference of
any kind of design verifica- 100kHz 10MHz 100MHz 1 MHz signal
tion/characterization it can
help the designer to identify
the root cause of Jitter
Injection and sources, which
are impossible to simulate
even with todays sophisticated Figure 23: What ParBERT’s Jitter Decomposition is good for (1)
tools. The final implementation
of high integrated ASICS with
all kind of analog circuitry
(PLL, CDR, ...) within large
digital cores will always give
unexpected results

Another use model is the • Frequency Range of Jitter components not limited by bandwidth -> Inband
charactertization for CDR and Outband frequency spectrum
devices. Assuming the device is • CDR can be characterized for all parameters:
stimulated with a data signal
incorporated with Jitter Data with white
Modulation consisting of wide Jitter Roll
noise
Out Band
band noise (white noise), a InBand Transfer Peaking Off Transfer Signal
processing
CDR will filter this noise CDR
according to the bandwidth of
its feedback loop design. The Outband Jitter is phase difference
feedbacks have a certain band- which CDR cannot compensate due to
Tx PLL Bandwidth cut-off:
with, so a jitter supression will Amplitude
occur beyond this point. Figure
24 gives such an example. Also 1MHz

one can see if there is some Frequency

peaking at the roll-off point.

Figure 24: What ParBERT’s Jitter Decomposition is good for (2)

18
Basics of the Jitter Decomposition Method

The Jitter Modulation on a Jitter

data signal (in this case a BER Modulation
sinusoidal modulation is Strobe
Bit
assumed) causes the data sig- Time
Eye Center
nal to toggle around the ideal
sampling point within the eye
0.5UI
(on the right in Figure 25).
Error
For this measurement the Density
comparator strobe will be dis- Variation
placed by 0.5 UI from the nor-
mal sampling point within the
centre of the eye. So the sam-
pling occurs within the transi-
tion (left in figure). Now when
sampling and comparing
against expected data, there Time
will be errors for those data
cycles where the data is ‘late‘ Figure 25: Basics of Method (1): Modulated Signal at Comparator
due to jitter. If the data is
‘early‘, the sampling and com-
paring deliver no errors. Now
this stream of errors and no
errors is called the ‘Error Sinusoidal Modulation
Density Variation or Error Signal
Function‘. Within the ParBERT
this Error Function is gained
for a configurable block length
and stored in a memory. Modulated data

Figure 26 explains the creation

of the Error Function: At the
top is a sinusoidal jitter mod-
ulation signal applied to a Error Signal
clock signal (middle). The edge
position of the ideal positions
of this clock signal is dotted, Figure 26: Basics of Method (2): Error Density Variation
and the position is straight
due to the modulating signal.
Analyzing the modulated signal
with strobing at the ideal posi-
tion and comparing against ‘1‘,
create the errors as shown in So far it would make no dif-
the bottom row of the table. ference if the modulating sig-
As long as the modulating sig- nal were sinusoidal, triangle or
nal delays the data cycles, rectangle. Within this ideal
there are errors; as long as description the period/frequen-
the modulation accelerates the cy of the modulating signal is
data signal, there are no recorded, but the shape would
errors. be lost.

19
 Figure 27 is a similar way of
looking at the method of sam-
pling. This assumes that the
incoming datastream runs from
bottom to top. The sampling is Optimal Jitter
strobe histogram
set to strobe within the transi- position Offset Compare
Jitter Error
tional area. The process is not Strobe Modulation Time
position Signal (long
ideal because there is always
term)
some noise within the sam-
pling. Either this noise is
incorporated with the incoming
signal or the there is noise

Bit Time
Jpk-pk

Data-eye
(jitter) on the sampling edge
itself. In reality both will Threshold Same spectral content
occur. Normally a test system
is designed to reduce this
noise to a minimum. Noise can
actually very helpful for this
type of measurement. When
Figure 27: Basics of Method (3): The Noise is important
looking into the Error
Function, the noise will dis-
place the positions of the
errors according to the shape
of the modulating signal. So
over the longer time, a sinu-
soidal signal can be recognised
as sinusoidal, as long as the
Error Function is recorded
long enough. So the full spec-
tral information is embedded
and the later DSP processing
will identify the whole spectral
content. The digitizing does not
change the spectral content.

The ParBERT Waveform Viewer 3125 samples =1us (1MHz)

allows the Error Function to
be visualized. Errors appear in
Figure 28: Error Density Modulation (1MHz Sinusoidal Jitter)
red in the GUI. They appear
as a lighter grey in Figure 28.
In this case a 1 MHz sinu-
soidal modulation generates jit-
ter on a data signal running at
3.125 Gb/s. One can clearly
see that the error distribution
occurs with the period of the
modulating signal. The random
noise involved generates some
displacement.

20
Already mentioned is that the
noise is important for this Sinusoidal signal and noise
type of measurement. As long floor are of similar magnitude
as the noise is of similar mag-
nitude to the Deterministic -> linear Spectrum
Jitter, this will operate as a
linear method So the spectrum
of the Deterministic Jitter is
linear in the error function
Sinusoidal signal is much
and the DSP algorithm oper-
larger than noise floor
ates cleanly. If there is more
Deterministic Jitter than ran- -> non-linear Spectrum,
dom jitter, the error function Distortion, Harmonics
gets clipped and operates in
the ideal way (as described
earlier). Therefore the shape of X(t)=Gauss(m=0, s ) + A*cos(2pft)
the modulation gets lost, which
causes the DSP processing to
Figure 29: Sinusoidal Signal and Noise, Harmonic Distortion
generate Harmonics, which
occur if there is a modulation
with rectangular signals.
Within the spectral view one
would see the Harmonics as a
rectangular waveform.

Some examples

This is the setup for some

ra s
ne bp

aly bps
r
examples: A ParBERT genera-

Ge 25G
to

r
An 25G
ze
3.1
tor (3.3 Gb/s, E4862B or 13.5

3.1
Gb/s, N4872A) is modulated 1 MHz
via the ‘Control Input’. We use Sine
three different signals for jitter
modulation: sine wave (1 MHz), Delay Control
10 MHz Input
Pulse with 10 MHz and 10%
Pulse
Duty Cycle (Width 1 us) and 10% Duty
Noise with 80 MHz bandwidth. 33250A
All these signals can be
achieved with an Agilent 80 MHz
33250A Arbitrary Waveform Noise
Generator.

The generator output connects Figure 30: Simple Loop-back Measurements with help of Generator’s Delay Control
by loop-back cable to an ana-
lyzer. Any ParBERT analyzer
can be used with the Spectral
Decomposition Measurement as
long as the data rate is within
its operating limits.

21
 Jitter Decomposition Examples

Sine Wave: The graph repre-

sents a single line at a single
frequency (top left on Figure
31).

Pulse: This delivers a spectrum

according the sin x/x shape
(top right on Figure 31).

Noise: There is a spectrum of

constant energy level up to the Sinusoidal
(1MHz, 0.08UI pp)
bandwidth of the noise source
(80 MHz). Then the power Pulse with 10% Duty Cycle 0.15UI
pulse jitter injection (Period 10us, Width 1us ->
decreases to the noise floor of sin(x)/x shape
the system (bottom left on
Noise
Figure 31).
(80MHz, 0.04UI rms)

The Spectral Decomposition

offers a couple of parameters
for setup and configuring the Figure 31: Jitter Decomposition Examples
results. Figure 32 shows the
Viewing parameters. First of
all one specifies the Power
Scaling. This allows a setup for
calibrating the jitter power Snapshot of
scale with a pilot tone. The the Parameters
measured jitter spectrum will affecting the
measurement
be referenced power wise to
this reference signal.

Then it allows the Frequency

Ranges to be specified. These
will be highlighted as shown in
the figure before. Further
choices are on scaling of axis,
markers and range of x-axis.

Figure 32: Calibration and Pilot Tone

22
Windowing

Windowing reduces the effects Windowing reduces the effects of unwanted frequency
of unwanted frequency compo- components by the Fourrier Transformation due to finite length
of acquired bits
nents by the Furrier
Some more literature:
Transformation due to the • The Fundamentals of Signal Analysis 5952-8898E
finite length of acquired bits. • On then Use of Windows for Harmonic Analysis with the Discrete
Some literature references are Fourier Transform, IEEE, Vol. 66, No. 1
given in Figure 33 with the • A Refresher Course on Windowing and Measurements, Real-Time
Update, Hewlett-Packard
primary recommendation for • The Fundamentals of FFT-Based Signal Analysis and Measurement,
‘The Fundamentals of Signal AN041, National Instruments, is Source for:
Analysis (AN243)’ available as:
5952-8898E from Agilent Lit-
station or from the web
(www.agilent.com). [7]

There are three predefined fil-

ters selectable from the config-
uration: Hanning, Hamming
and Blackman. Uniform is not
a filter, but uses the data as Figure 33: FFT Windowing
measured. The filters reduce
the amount of energy in the
unwanted signal areas much
larger than in the areas of
interest.

There is a dependency between

frequencey resolution and
acquisition depth. The longer
the segment of the Error func-
tion, the lower the frequency.
But this is also a matter of
the data rate. This is of course
a function of measurement
time. Using a segment below 1
MBit, let the measurement run
within a few seconds. Larger
Acquisition Depth will slow
down the measurement time.
Frequency Resolution also
defines also the minimum fre-
quency. Therefore if the reso-
lution is 1 kHz, then the first
energy line is available at 1 Figure 34: Frequency Resolution vs. Data
kHz, the next is at 2 kHz and
so on.

23
 References

[1]: NIST Technical Note 1337,

“Characterization of Clocks and
Oscillators,” edited by D.B.
Sullivan, D.W. Allan, D.A.
Howe, F.L. Walls, 1990

[2]: National Committee for

Information Technology
Standardization (NCITS)
T11.2/Project 1230-DT “Fibre
Channel – Methodologies for
Jitter and Signal Quality
Specification” Rev. 5.0,
February 21, 2002

[3]: Jan B. Wilstrup: A new

method for jitter decomposition
through its distribution tail fit-
ting, ITC 1999

[4] Agilent Jitter Injection

Module (JIM), 5988-3583EN

[5] 10GbE Standard IEEE

802.3ae

[6] Infiniium 54850 Series

Oscilloscopes, 5988-7976EN

[7]: The Fundamentals of

Signal Analysis, Agilent AN243,
5952-8898E

24
Related Literature Pub. Number Get assistance with all your
Need to Test BER?, Brochure 5968-9250E
test and measurement needs at:
www.agilent.com/find/assist
Or check your local phone book for
Agilent ParBERT 81250, Mux/Demux Application, 5968-9695E the Agilent office near you.
Application Note
Phone or Fax
Agilent ParBERT 81250 Parallel Bit Error Ratio Tester, 5980-0830E United States:
Photo Card (tel) 800 452 4844

Agilent Productivity Assistance 5980-2160E Canada:

(tel) 877 894 4414
Agilent ParBERT 81250 43.2/45G 5988-3020EN (fax) 905 282 6495
Product Overview
China:
Agilent ParBERT 81250 5988-5901EN (tel) 800 810 0189
Paralel Bit Error Ratio Test Platform (fax) 800 820 2816

Agilent 81250 ParBERT Product Note 5988-5948EN Europe:

(The influence of Generator Transition times (tel) (31 20) 547 2323
on Characterization Measurements) (fax) (31 20) 547 2390

Agilent ParBERT 81250 Automatic Phase 5988-5654EN Japan:

Margin Measurements at 43.2 Gb/s (tel) (81) 426 56 7832
(fax) (81) 426 56 7840
Agilent ParBERT 81250 Product Overview 5968-9188E
Korea:
(tel) (82 2) 2004 5004
Agilent Technologies’ Test and Measurement Support, Services, and Assistance
Agilent Technologies aims to maximize the value you receive, while minimizing
(fax) (82 2) 2004 5115
your risk and problems. We strive to ensure that you get the test and measure-
ment capabilities you paid for and obtain the support you need. Our extensive Latin America:
support resources and services can help you choose the right Agilent products (tel) (305) 269 7500
for your applications and apply them successfully. Every instrument and system (fax) (305) 269 7599
we sell has a global warranty. Support is available for at least five years beyond
the production life of the product. Two concepts underlie Agilent’s overall sup-
port policy: “Our Promise” and “Your Advantage.” Taiwan:
(tel) 0800 047 866
Our Promise
“Our Promise” means your Agilent test and measurement equipment will meet its (fax) 0800 286 331
advertised performance and functionality. When you are choosing new equipment,
we will help you with product information, including realistic performance speci- Other Asia Pacific Countries:
fications and practical recommendations from experienced test engineers. When (tel) (65) 6375 8100
you use Agilent equipment, we can verify that it works properly, help with prod- (fax) (65) 6836 0252
uct operation, and provide basic measurement assistance for the use of specified
capabilities, at no extra cost upon request. Many self-help tools are Email: [email protected]
available.
Your Advantage
“Your Advantage” means that Agilent offers a wide range of additional expert
test and measurement services, which you can purchase according to your unique
Technical data is subject to change
technical and business needs. Solve problems efficiently and gain a competitive
edge by contracting us for calibration, extra- cost upgrades, out-of-warranty repairs, © Agilent Technologies 2003
and on-site education and training, as well as design, system integration, project Printed in the Netherlands July 17th 2003
management, and other professional services. Experienced Agilent engineers and 5988-9756EN
technicians worldwide can help you maximize your productivity, optimize the
return on investment of your Agilent instruments and systems, and obtain
dependable measurement accuracy for the life of those products.
MS Windows NT, Windows 2000 and Microsoft
are U.S. trademarks of Microsoft Corporation.

www.agilent.com/find/emailupdates
Get the lastest information on the products
and applications you select.