28
Sam Palermo Analog & Mixed-Signal Center Texas A&M University ECEN720: High-Speed Links Circuits and Systems Spring 2021 Lecture 4: Channel Pulse Model & Modulation Schemes

ECEN720: High-Speed Links Circuits and Systems Spring 2021...(symbols/s) without ISI is R S/2 (Hz) • Thus, a system with bandwidth W=1/2T=R S/2 (Hz) can support a maximum transmission

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

  • Sam PalermoAnalog & Mixed-Signal Center

    Texas A&M University

    ECEN720: High-Speed Links Circuits and Systems

    Spring 2021

    Lecture 4: Channel Pulse Model & Modulation Schemes

  • Announcements• Lab 1 Report and Prelab 2 due Feb. 3

    • For Prelab 2 Question 2, look ahead to Lecture 5 Slide 8

    • Lab 2 Report and Prelab 3 due Feb. 10

    • Reference material• Peak distortion analysis paper by Casper• Notes from H. Song, Arizona State• Papers posted on PAM-2/4 transceivers

    2

  • Agenda• ISI and channel pulse model

    • Peak distortion analysis

    • Compare NRZ (PAM-2) and PAM-4 modulation

    3

  • Inter-Symbol Interference (ISI)• Previous bits residual state can distort the current bit,

    resulting in inter-symbol interference (ISI)• ISI is caused by

    • Reflections, Channel resonances, Channel loss (dispersion)• Pulse Response

    4

    tc 1 th tc 1 ty 1

    thtcty 11

  • NRZ Data Modeling• An NRZ data stream can be modeled as a

    superposition of isolated “1”s and “0”s

    5

    Data = “1000101”

    “1” Symbol

    “0” Symbol

    TktukTtutck 11

    tctc kk 10

    0001

    wherett

    tu[Song]

  • NRZ Data Modeling• An NRZ data stream can be modeled as a

    superposition of isolated “1”s and “0”s

    6

    k

    dki tctV k

    [Song]

  • Channel Response to NRZ Data• Channel response to NRZ data stream is

    equivalent to superposition of isolated pulse responses

    7

    k

    d

    k

    dkio kTtytcHtVHtV kk

    [Song]

  • Channel Pulse Response

    8

    thtcty kk dd

    y(1)(t) sampled relative to pulse peak:[… 0.003 0.036 0.540 0.165 0.065 0.033 0.020 0.012 0.009 …]

    k =[ … -2 1 0 1 2 3 4 5 6 …]By Linearity: y(0)(t) =-1*y(1)(t)

    cursor

    post-cursor ISI

    pre-cursor ISI

  • Channel Data Stream Response

    9

    Input Data Stream

    Pulse Responses

    Channel Response

  • a-1

    a0

    a1a2 a3

    Channel “FIR” Model

    10

    tc 10

    tc 10

    tytcH 1010

    tytcH 1010

    y(1)(t) sampled relative to pulse peak:[… 0.003 0.036 0.540 0.165 0.065 0.033 0.020 0.012 0.009 …]

    a =[ … a-2 a-1 a0 a1 a2 a3 a4 a5 a6 …]

    D is the delay from the channel input to the output pulse peak

  • Agenda• ISI and channel pulse model

    • Peak distortion analysis

    • Compare NRZ (PAM-2) and PAM-4 modulation

    11

  • Peak Distortion Analysis• Can estimate worst-case eye

    height and data pattern from pulse response

    • Worst-case “1” is summation of a “1” pulse with all negative non k=0 pulse responses

    12

    0

    0

    )1(01

    kk kTty

    d kTtytyts k

    • Worst-case “0” is summation of a “0” pulse with all positive non k=0 pulse responses

    0

    0

    )0(00

    kk kTty

    d kTtytyts k

  • Peak Distortion Analysis• Worst-case eye height is s1(t)-s0(t)

    13

    0

    0

    0

    0

    )0(0

    )1(001

    kk kTty

    d

    kk kTty

    d kTtykTtytytytststs kk

    • If symmetric “1” and “0” pulses (linearity), then only positive pulse response is needed

    0

    0

    1

    0

    0

    1)1(02

    kk kTty

    kk kTty

    kTtykTtytyts

    tyty )1(0)0(0 1 Because

    “1” pulse worst-case “1” edge

    “1” pulse worst-case “0” edge

  • Peak Distortion Analysis Example 1

    14

    288.0389.0007.0540.02

    389.0

    007.0

    540.0

    0

    0

    1

    0

    0

    1

    )1(0

    ts

    kTty

    kTty

    ty

    kk kTty

    kk kTty

  • Worst-Case Bit Pattern• Pulse response can be used to find the worst-case bit pattern

    15

    ... ...a Pulse 654321012 aaaaaaaaa

    ... )()(1)()()()()(- ... 21123456 asignasignasignasignasignasignasignasign

    • Flip pulse matrix about cursor a0 and the bits are the inverted sign of the pulse ISI

    Worst-Case Bit Pattern Eye 10kbits Eye

  • Peak Distortion Analysis Example 2

    16

    338.0542.0053.0426.02

    542.0

    053.0

    426.0

    0

    0

    1

    0

    0

    1

    )1(0

    ts

    kTty

    kTty

    ty

    kk kTty

    kk kTty

  • Agenda• ISI and channel pulse model

    • Peak distortion analysis

    • Compare NRZ (PAM-2) and PAM-4 modulation

    17

  • PAM-2 (NRZ) vs PAM-4 Modulation

    18

    • Binary, NRZ, PAM-2• Simplest, most common modulation format

    • PAM-4• Transmit 2 bits/symbol• Less channel equalization and circuits run ½ speed

    0

    1

    00

    01

    11

    10

  • Modulation Frequency Spectrum

    19

    Majority of signal power in 1GHz bandwidth

    Majority of signal power in 0.5GHz bandwidth

    0

    1

    00

    01

    11

    10

  • Nyquist Frequency• Nyquist bandwidth constraint:

    • The theoretical minimum required system bandwidth to detect RS (symbols/s) without ISI is RS/2 (Hz)

    • Thus, a system with bandwidth W=1/2T=RS/2 (Hz) can support a maximum transmission rate of 2W=1/T=RS (symbols/s) without ISI

    20

    /Hz)(symbols/s 222

    1

    WRWR

    TSS

    • For ideal Nyquist pulses (sinc), the required bandwidth is only RS/2 to support an RS symbol rate

    Modulation Bits/Symbol Nyquist FrequencyNRZ 1 Rs/2=1/2Tb

    PAM-4 2 Rs/2=1/4Tb

  • NRZ vs PAM-4

    • PAM-4 should be considered when• Slope of channel insertion loss (S21) exceeds reduction in PAM-4

    eye height• Insertion loss over an octave is greater than 20*log10(1/3)=-9.54dB

    • On-chip clock speed limitations21

  • PAM-4 Receiver

    • 3x the comparators of NRZ RX

    22

    [Stojanovic JSSC 2005]

  • NRZ vs PAM-4 – Desktop Channel

    • Eyes are produced with 4-tap TX FIR equalization

    • Loss in the octave between 2.5 and 5GHz is only 2.7dB• NRZ has better voltage margin

    23

    Loss at 2.5GHz = -4.8dB

    Loss at 5GHz = -7.5dB

  • NRZ vs PAM-4 – T20 Server Channel

    • Eyes are produced with 4-tap TX FIR equalization

    • Loss in the octave between 2.5 and 5GHz is 15.8dB• PAM-4 “might” be a better choice

    24

    Loss at 2.5GHz = -11.1dB

    Loss at 5GHz = -26.9dB

  • PAM-4 Peak Distortion Analysis

    25• PAM4 modulation is more sensitive to residual ISI

    PAM4 eye height= 2(𝟏

    𝟑A –

    NRZ eye height= 2(A –

    A c1 c2 c3 c4 c5 c5C-1

    𝑦 𝑡 𝑐 𝑡 ∗ 𝑝 𝑡

    Tb

    Channel pulse response

    Tb

    𝑐 𝑡

    Transmitter symbol

    response

    Channel impulse response

    𝑝 𝑡

    a

    -a

    2a

    PAM2 Eye Diagram a

    a/3

    -a/3

    -a

    2a/3 2a/3

    -2a/3

    0

    PAM4 Eye Diagram

  • Multi-Level PAM Challenges• Receiver complexity increases considerably

    • 3x input comparators (2-bit ADC)• Input signal is no longer self-referenced at 0V differential

    • Need to generate reference threshold levels, which will be dependent on channel loss and TX equalization

    • CDR can display extra jitter due to multiple “zero crossing” times

    • Smaller eyes are more sensitive to cross-talk due to maximum transitions

    • Advanced equalization (DFE) can allow NRZ signaling to have comparable (or better) performance even with >9.5dB loss per octave

    26

  • Modulation Take-Away Points• Loss-slope guidelines are a good place to start in

    consideration of alternate modulation schemes

    • More advanced modulation trades-off receiver complexity versus equalization complexity

    • Advanced modulation challenges• Peak TX power limitations• Setting RX comparator thresholds and controlling offsets• CDR complexity• Crosstalk sensitivity (PAM-4)

    • Need link analysis tools that consider voltage, timing, and crosstalk noise to choose best modulation scheme for a given channel

    27

  • Next Time• Link Circuits

    • Termination structures• Drivers• Receivers

    28