Digitally Assisted Wireless Transceivers and · PDF fileDigitally Assisted Wireless...

90
Symposia on VLSI Technology and Circuits Kenichi Okada Tokyo Institute of Technology Digitally Assisted Wireless Transceivers and Synthesizers

Transcript of Digitally Assisted Wireless Transceivers and · PDF fileDigitally Assisted Wireless...

Symposia on VLSI Technology and Circuits

Kenichi Okada

Tokyo Institute of Technology

Digitally Assisted Wireless

Transceivers and Synthesizers

Outline

Slide 1

•Analog to Digital

•Digitization of Wireless TRX

•Digital Assistance

•Wireless Transceiver

• Frequency Synthesizer

• Future Analog Design

•Synthesizable Analog

Analog to Digital

Slide 2

AnalogDigitalAnalog

Digitization

Digital

Digital Assistance

• Scalability

• Portability

• Robustness

• Less redundancy

Further Digitization

Slide 3

DigitalAnalog

Digitization

Massive Digital Assistance

• Scalability

• Portability

• Robustness

• Less redundancy

Digital

Digital

Case Study

Slide 4

Partially replaced by pure digital-domain “calculation”

(NOT time-domain analog processing)

• Filter: LPF in AD-PLL, LPF in wireless TX

• Equalizer: FIR in wireless, OFDM

• PLL: carrier and timing recovery in wireless RX

•Mixer: Low-IF transceiver

Only analog-domain

•Oscillator: Clock generation

•Data converter: V-to-D, D-to-V, D-to-I, C-to-D,..

•Analog amplifier: voltage-to-voltage

Diffic

ult fo

r dig

itizatio

n

History of Timing Recovery

*F. Gardner, IEEE Trans. Comm. 1993

Analog recovery

Digitally-assisted

analog recovery

Digital recovery

sampling clocktiming recovery

analog

processing

digital

processing

timing recovery

timing recovery

Slide 5

Analog Digital

analog

processing

digital

processing

analog

processing

digital

processing

Analog Demodulator

Slide 6

Costas-loop for BPSK

*H. Suzuki, et al., IEEE Trans. VT 1985

timing recovery loop (carrier & phase)

0/1

90o

0o

loop filter

Digital Carrier and Timing Recovery

Slide 7

NCO: Number-Controlled OscillatorLF: Loop FilterPD: Phase Detector

timing recovery (phase)

timing-recoveredsignal

PDLFNCO

carrier recovery (freq.)

rotationfilter

PDLFNCO

interpolationfilter

timing- & carrier-recovered signal

ADC

Everything is implemented in digital domain.

*F. Gardner, IEEE Trans. Comm. 1993

Recent Digital Transceiver

Slide 8

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

Analog Digital

RX

filter

TX

filter

Carrier

& Timing

Recovery

Dem

ap

per

FE

C

Deco

der

Map

per

FE

C

En

co

der

Zero-IF

Digital LPF in Wireless

Slide 9

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

Analog Digital

RX

filter

TX

filter

Carrier

& Timing

Recovery

Dem

ap

per

FE

C

Deco

der

Map

per

FE

C

En

co

der

LPF for transmitting bandwidth restriction

Digital Equalizers in Wireless

Slide 10

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

Analog Digital

RX

filter

TX

filter

Carrier

& Timing

Recovery

Dem

ap

per

FE

C

Deco

der

Map

per

FE

C

En

co

der

Adaptive RF&BB equalizer

RF&BB equalizer

Digitization of IF Mixer

Slide 11

Analog Digital

ADC

ADC

Analog Digital

ADC

Hetero-dyne RX

Low-IF RX

Very common for BT

1/f noise

Overhead for ADC

RF

RF IF

IF

Aim of This Talk

Slide 12

• Digitization

• Wireless transceiver is a good example of

digitized analog circuit. (for hinting)

• Toward “Synthesizable Analog Circuit”

• Digital assistance

• Digitally-designed analog

• Digital calibration/compensation is implemented

in a system level to satisfy complicated

requirements for wireless system.

• Mutual re-use of TX and RX for calibration

Outline

Slide 13

•Analog to Digital

•Digitization of Wireless TRX

•Digital Assistance

•Wireless Transceiver

• Frequency Synthesizer

• Future Analog Design

•Synthesizable Analog

Impairments in Wireless Transceiver

Slide 14

Mismatch in differential block

• DC offset in RX

• IIP2 in RX

• LO leakage in TX

Mismatch btw I and Q blocks

• Image signal

• Analog filter BW (LPF)

Non-linearity

• IMD in PA

*A. Jerng, “Digital Calibration for RF Transceivers,” ISSCC 2012, Tutorial 9

PVT variation

• Gain control

• Power control

• VCO LC tank

• ILFD/ILO

Environmental variation

• TX-to-RX distance

• Fading

• Antenna reflection

Digital Equalizers

Slide 15

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

Analog Digital

RX

filter

TX

filter

Carrier

& Timing

Recovery

Dem

ap

per

FE

C

Deco

der

Map

per

FE

C

En

co

der

Adaptive RF&BB equalizer

RF&BB equalizer

Environmental Variation

Fading by multi-path

flat or frequency selective

Automatic Gain Control (AGC)

Slide 16

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

Analog Digital

RX

filter

TX

filter

Carrier

& Timing

Recovery

Dem

ap

per

FE

C

Deco

der

Map

per

FE

C

En

co

der

Automatic Gain Control

DC offset cancel

Environmental Variation

Received signal strength

Antenna reflection

I/Q Mismatch

Slide 17

RX

ADC

ADC

TX

DAC

DAC

Digital

BB

LPF

LPF

VGA

VGA

LPF

LPF

Analog Digital

gain&phase

mismatch

phase mismatch

I

Q

I

Q

Image Rejection Ratio (IMRR)

Slide 18

LO leak

Desired

𝝎𝐋𝐎 −𝝎𝐁𝐁

[dB

m]

freq. [GHz]𝝎𝐋𝐎 +𝝎𝐁𝐁𝝎𝐋𝐎

IMRR [dB]

e.g. 𝝎𝐁𝐁 = 𝟏𝟎𝐌𝐇𝐳

Image

Image for Modulated Sig.

Slide 19

Image

Desired

𝝎𝐋𝐎 −𝝎𝐁𝐁

[dB

m]

freq. [GHz]𝝎𝐋𝐎 +𝝎𝐁𝐁𝝎𝐋𝐎

IMRR [dB]

I/Q mismatch degrades SNR.

Image Rejection Ratio (IMRR)

Slide 20

0

1

2

3

4

5

6

7

0 0.2 0.4 0.6 0.8 1

Ph

as

e E

rro

r [d

eg

]

Amplitude Error [dB]

Δg: Amplitude error ratio

Δθ: Phase error

Target:

0.2dB, 1.0degree

for IMRR of 35dB

𝑰𝑴𝑹𝑹 ≈𝜟𝒈𝟐 + 𝜟𝜽𝟐

𝟒

TX IMRR Calibration

Slide 21

I: 0o

Q: -90o

test signal

coupler

TX

Down-conversion is

required.

Detector can be used.

(2nd-order distortion)

IMLO

Desired

freq. [GHz]

𝝎𝐋𝐎 −𝝎𝐁𝐁 𝝎𝐋𝐎 +𝝎𝐁𝐁

𝝎𝐋𝐎

TX IMRR Calibration

Slide 22

I: 0o

Q: -90o

test signal

coupler

Detector

TX

IMLO

Desired

IMLO

Desired

𝝎𝐁𝐁

freq. [GHz]

𝟐𝝎𝐁𝐁0𝝎𝐋𝐎 −𝝎𝐁𝐁 𝝎𝐋𝐎 +𝝎𝐁𝐁

𝝎𝐋𝐎

freq. [GHz]

Detector

Slide 23

RF (e.g. 5GHz)

bias

LPF BB (e.g. 10MHz)

IMLO

Desired

IMLO

Desired

𝝎𝐁𝐁 𝟐𝝎𝐁𝐁0𝝎𝐋𝐎 −𝝎𝐁𝐁 𝝎𝐋𝐎 +𝝎𝐁𝐁

𝝎𝐋𝐎

𝒗𝐨𝐮𝐭 = 𝐋𝐏𝐅(𝑨 𝒗𝐢𝐧𝟐)

I/Q Mismatch Calibration by Loop-back

Slide 24

RX

ADC

ADC

TX

DAC

DAC

Digital

BB

LPF

LPF

VGA

VGA

LPF

LPF

coupler

Detector

TX

filter

• I/Q Amplitude offset

• I/Q Phase offset

I/Q Mismatch Calibration by Loop-back

Slide 25

RX

ADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

coupler

Detector

TX

filter

• I/Q Amplitude offset

• I/Q Phase offset*Iason Vassiliou, et al., IEEE JSSC 2003.

Digital

BB

DFT

-20

-15

-10

-5

0

0.01 0.1 1 10 100

Frequency-Dependent I/Q Mismatch

Slide 26

TX DAC

DAC

LPF

LPF

Frequency [MHz]

Gain

[d

B]

I

Q

5% cut-off mismatch

causes a serious

frequency-dependent

I/Q mismatch.

-180

-150

-120

-90

-60

-30

0

0.01 0.1 1 10 100

-20

-15

-10

-5

0

0.01 0.1 1 10 100

Frequency-Dependent IMRR

Slide 27

Gain/phase mismatch can be frequency-dependent.

5% cut-off frequency mismatch

Freq. [MHz]

Freq. [MHz]

Gain

[d

B]

Ph

ase [

deg

.]

IMR

R [

dB

]

Freq. [MHz]

BW

-50

-40

-30

-20

0.01 0.1 1 10 100

1.5% (-42.5dB)

2.6% (-38.5dB)

-37dB@1MHz, -27dB@10MHz

Key Idea of Wireless Calibration

Slide 28

Self-calibration with less additional blocks

Reuse of TRX each other

TX = Signal Generator for RX calibration

RX = Spectrum Analyzer for TX calibration

Overall Procedure of TRX Calibration

Slide 29

1. RX BB LPF Calibration (using TX BB)• I/Q gain mismatch

• LPF cut-off mismatch (including VGA and ADC)

2. TX BB LPF Calibration (using RX BB)• I/Q gain mismatch

• LPF cut-off mismatch

3. TX I/Q Calibration (using detector and RX BB)• Impairments of mixer, LO, RF I/Q amps., etc

• compensated by digital BB

4. RX I/Q Calibration (using TX)• Impairments of mixer, LO, RF I/Q amps., etc

• compensated by digital BB

RX Filter Calibration (BB loopback)

Slide 30

RXADC

ADC

TX

DAC

DAC

Digital

BB

LPF

LPF

VGA LPF

LPF

Test-tone

generation

Gain/Phase

calculation

VGA and ADC are also included in RX BB calibration.

VGA

target

TX Filter Calibration (BB loopback)

Slide 31

RXADC

ADC

TX

DAC

DAC

Digital

BB

LPF

LPF

VGA LPF

LPF gain/cut-off mismatch between I/Q paths are calibrated.

VGA LPF

target

Test-tone

generation

Gain/Phase

calculation

target

RF Loop-Back Calibration for TX

Slide 32

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

Detector

*Iason Vassiliou, et al., IEEE JSSC 2003.

IM/LO

Calculation

by DFT

Test-to

ne

gen

era

tion

TX

filtercompensate

ADC is re-used for IM/LO calculation with DFT in BB.

target

RF Loop-Back Calibration for TX

Slide 33

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

Detector

*Iason Vassiliou, et al., IEEE JSSC 2003.

IM/LO

Calculation

by DFT

Test-to

ne

gen

era

tion

TX

filtercompensate

IMLO

Desired IMLO

Desired

freq. [GHz]

0

HPF

RF Loop-Back Calibration for RX

Slide 34

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

*Iason Vassiliou, et al., IEEE JSSC 2003.

I/Q mismatch

calculation

Test-to

ne

gen

era

tion

TX

filter

ATT

TX is used for a test-tone generator.

Desired

target

I/Q Gain/Phase Mismatch Calculation

Slide 35

t

t

*S. Lerstaveesin, et al., IEEE JSSC 2006.

𝐋𝐏𝐅 𝑰𝟐 −𝑸𝟐 ≅ 𝜟𝒈

𝐋𝐏𝐅 𝑰 ∗ 𝑸 ≅ −𝜟𝜽/𝟐

𝑰

𝑸

At least, a 10-bit ADC is required for a IMRR of 40dB.

𝐋𝐏𝐅 𝑰𝟐 +𝑸𝟐

RSSI AGC

Modulated signal can be used.

Background calibration

target

RF Loop-Back Calibration for RX

Slide 36

RXADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

*Iason Vassiliou, et al., IEEE JSSC 2003.

Test-to

ne

gen

era

tion

TX

filter

ATT

I/Q mismatch

calculation

Desired

t

t

Calibration vs Compensation

Slide 37

Frequency independent (RF)

TX I/Q mismatch(RF) Digital compensation (BB TX filter)

RX I/Q mismatch(RF) Digital compensation (BB RX filter)

FDE/OFDM

Frequency independent (BB)

TX I/Q mismatch(BB) Digital compensation

RX I/Q mismatch(BB) Digitally-calibrated analog (AGC)

/ Digital compensation

Frequency dependent (BB)

TX I/Q mismatch(BB) Digitally-calibrated analog

RX I/Q mismatch(BB) Digitally-calibrated analog

/ Digital compensationas a typical case

60GHz Transceiver Calibration

Slide 38

RX

ADC

ADC

TX

DAC

DAC

LPF

LPF

VGA

VGA

LPF

LPF

DetectorDigital

BB

*T. Tsukizawa, et al., ISSCC 2013

5b 3.5GS/s

8b 1MS/s ADC

One additional ADC is used for a fine resolution.

Outline

Slide 39

•Analog to Digital

•Digitization of Wireless TRX

•Digital Assistance

•Wireless Transceiver

• Frequency Synthesizer

• Future Analog Design

•Synthesizable Analog

Calibration in Frequency Synthesizer

Slide 40

• AFC for capacitor-bank in LC-VCO

• ILFD/ILO Calibration

• Linearity calibration/compensation

− Loop-BW, Quantization noise, FM/Polar-TX

AFC: Automatic Frequency Calibration

ILFD: Injection-Locked Frequency Divider

ILO: Injection-Locked Oscillator (Multiplier)

VCO: frequency ← voltage (varactor, C-bank)

DCO: frequency ← code (C-bank, I-control)

TDC: code ← delay (PVT, noise, layout, etc)

(ADC: code ← voltage)

(DAC: voltage ← code)

(Amp: voltage ← voltage)

ILFD Calibration

Slide 41

Locked*/Free-run** frequency is used.

*S. Pellerano, et al., ISSCC 2008 **T. Shima, et al., APMC 2011

***W. Deng, et al., A-SSCC 2012

60GHz PLL

Summary of Transceiver Calibration

Slide 42

• Wireless transceiver is a big system.

• Historically, architecture-level digitization

has been applied with system-level

calibration and compensation for PVT

and environmental variations.

• Re-use of counter-part block for

calibration

Outline

Slide 43

•Analog to Digital

•Digitization of Wireless TRX

•Digital Assistance

•Wireless Transceiver

• Frequency Synthesizer

• Future Analog Design

•Synthesizable Analog

Issues of Analog Circuit Design

Slide 44

PVT

layout non-ideality

• mismatch

• isolation/coupling

Imperfection caused by physical implementation

Compensated by

digital assistance

Why is the simulated performance degraded?

Layout Design Issues beyond 20nm

Slide 45

Double patterning

A A AB B B

STI stress (LOD)

SA SB

LOD

Well proximity effect

• Larger Rd, Rs, Rg

• Fixed fin height (for FinFET)

• Self-heating

• No body effect

Transistor matching cannot be expected any more.

Scaled CMOS Layout

Slide 46

65nm layout style 32nm layout style

• Uni-directional features

• Uniform gate dimension

• Gridded layout*M. Bohr, ISSCC 2009

Massive Digital Assistance

Slide 47

PVT

layout non-ideality

• mismatch

• isolation/coupling

Compensated by

digital assistance

Delay and linearity in delay can be calibrated

easily in time-domain analog circuits,

e.g. AD-PLL.

Further Analog Circuit

Slide 48

DigitalAnalog

Digitization

Massive Digital Assistance

• Scalability

• Portability

• Robustness

• Less redundancy

Digital

Digital

Outline

Slide 49

•Analog to Digital

•Digitization of Wireless TRX

•Digital Assistance

•Wireless Transceiver

• Frequency Synthesizer

• Future Analog Design

•Synthesizable Analog

Synthesizable Analog Circuits

Slide 50

GDSDigital design flow

Commercial P&R tools…

HDLmodule PLL

(CLK, …, OUT)

endmodule

with a standard-cell library

without any custom-designed cells

without manual placement

Analog Synthesis by Digital Tools

Slide 51

Verilog netlist

(gate-level)

Verilog

RTL

DCO

Logic

Logic

Logic Synt. Tool

GDSII

P&R Tool

Netlist

*W. Deng, et al., ISSCC 2014

Layout uncertainty

Pure digital

Digitally-designed Analog

Issue: Layout Uncertainty

Slide 52

Ideal placementActual placement

Unbalanced loading

No layout symmetry

Massive digital assistance can

overcome the layout uncertainty issue.

e.g., DCO &TDC linearity

Synthesizable Analog Circuits

Slide 53

• Synthesizable PLL*

• Synthesizable DCO

• Synthesizable DAC

• Synthesizable TDC

• Synthesizable ADC**

*W. Deng, et al., ISSCC 2014 **S. Weaver, et al., IEEE TCAS-I 204

only by standard cells

Synthesizable ADC

Slide 54

*S. Weaver, et al., IEEE TCAS-I 2014

Comparator by NAND3

Gaussian offset distribution

ADC architecture

SNDR of 35.9dB, 210MS/s

Linearity compensation by inverse Gaussian

Synthesizable DCO

Slide 55

*D. Sheng, et al., IEEE TCAS-II 2007

**A. Matsumoto, et al., JSSC 2008 ***W. Deng, et al., ISSCC 2014

MUX and varactor* Phase-Interpolator**, I-DAC***,

and fine varactor***

Synthesizable DAC

Slide 56

Only standard cell

D0D1= 11 Vout= 0V

D0D1= 10 Vout= 0.5V

D0D1= 01 Vout= 0.5V

D0D1= 00 Vout= 1V

D0= 0 Vout= 1V

D0= 1 Vout= 0V

Synthesizable I-linear DAC

Slide 57

Only standard cell

D0

D1

D2

D3

1

1

Iout

×1

×2

×4

×8

*W. Deng, et al., ISSCC 2014

Current-starving RO

by NAND2

V-linear DAC vs I-linear DAC

Slide 58

0 4 8 12

Control Code

Cu

rren

t[m

A]

0.24

0.20

0.16

0.12

0 4 8 12

Control Code

Fre

qu

en

cy[G

Hz]

1.6

1.2

0.8

Current-starving RO

by NAND2

Stdcell Varactor

Slide 59

Vin

1.55ps@200MHz

*P. L. Chen, et al., TCAS II 2005

Performance Trade-off

Slide 60

Performance A

Pe

rfo

rman

ce B

Custom design

Custom design but easy

Synthesizable design

Performance Trade-off

Slide 61

Performance A

Pe

rfo

rman

ce B

Custom design

Custom design but easy

Synthesizable design

Synthesis-friendly architecture

A synthesis-friendly architecture

can improve performances.

Injection-Locked PLL (IL-PLL)

Slide 62

Conventional CP-PLL and TDC-PLL (AD-PLL)

• Phase lock: feedback

• Frequency lock: feedback

Injection-Locked PLL (IL-PLL)

• Phase lock: feed-forward

• Frequency lock: feedback

Injection-lock

Counter

*W. Deng, et al., ISSCC 2014

The fine timing feedback is not required.

Synthesis-friendly

Injection-Locked PLL

Slide 63

fref fout = N * fref

0.4*fref

Ph

as

e n

ois

e [

dB

c/H

z]

Offset Freq. [Hz]

PNVCO

PNref

+20log10(N)

Reference injection

*S. Ye, et al., IEEE JSSC 2002 **N. D. Dalt, IEEE TCAS-II 2014

Flicker of PNVCO

& Ref noise

Noise folding

& phase lock

Synthesizable IL-PLL

Slide 64

Frequency tracking with very narrow BW

*W. Deng, et al., ISSCC 2013 **A. Musa, et al., JSSC 2014 ***W. Deng, et al., ISSCC 2014

Phase lock

Gated Edge Injection

Slide 65

Vx

gating

window

Inj. edge

Vy

Severe timing design is NOT required.

(1)

(2)

(3)

*W. Deng, et al., ISSCC 2014 **D. Park, et al., ISSCC 2012

gating

window

Layout

Slide 66

CMOS 65nm

Area: 0.0066mm2

Jitter: 1.7ps

PDC: 780mW

FOM: -236.5 dB

60mm

110mm

Comparison of the state-of-the-art PLLs

Slide 67

*W. Deng, et al., ISSCC 2014

Comparison of Synthesizable PLL

Slide 68

This work

65nm

[22]

28nm

[23]

65nm

[24]

65nm

Power

[mW]0.78

@900MHz

13.7 @2.5GHz

3.1@250MHz

2.1@403MHz

Area

[mm2]0.0066 0.042 0.032 0.1

Integ.

Jitter [ps]1.7 N.A. 30 N.A.

RMS

Jitter [ps]2.8 3.2 N.A. 13.3

FOM [dB] -236.5 -218.6* -205.5 -214*W/ custom

cells?No No Yes Yes

Topology IL-base TDC-base TDC-base TDC-base*FOM is calculated based on RMS jitter.

Conclusion

Slide 69

• Digitization vs Digitally-Assisted Analog

• Digitally-Assisted Analog

to Digitally-Designed Analog

e.g. Synthesizable Analog Circuit

portability, scalability, robustness,..

References

Slide 70

[1] F. Gardner, "Interpolation in Digital Modems-Part I: Fundamentals," IEEE Trans. on

Communications, Vol. 41, No. 3, pp. 501-507, Mar. 1993.

[2] H. Suzuki, Y. Yamao, and H. Kikuchi, "A Single-Chip MSK Coherent Demodulator for

Mobile Radio Transmission," IEEE Trans. on Vecular Technology, Vol. VT-34, No. 4,

pp. 157-168, Nov. 1985.

[3] A. Jerng, “Digital Calibration for RF Transceivers,” ISSCC 2012, Tutorial 9.

[4] I. Vassiliou, et al., "A Digitally Calibrated 5.15-5.825GHz Transceiver for 802.11a

Wireless LANs in 0.18μm CMOS," IEEE Journal of Solid-State Circuits, Vol. 38, No. 12,

pp. 2221-2231, Dec. 2003.

[5] S. Lerstaveesin, et al., "A Complex Image Rejection Circuit with Sign Detection Only,"

IEEE Journal of Solid-State Circuits, Vol. 41, No. 12, pp. 2693-2702, Dec. 2006.

[6] T. Tsukizawa, et al., "A Fully Integrated 60GHz CMOS Transceiver Chipset Based on

WiGig/IEEE802.11ad with Built-in Self-Calibration for Mobile Applications," ISSCC Dig.

Tech. Papers, pp. 230-231, Feb. 2013.

[7] S. Pellerano, R. Mukhopadhyay, A. Ravi, J. Laskar, and Y. Palaskas, "A 39.1-to-

41.6GHz DS Fractional-N Frequency Synthesizer in 90nm CMOS," ISSCC Dig. Tech.

Papers, pp. 484-485, Feb. 2008.

[8] T. Shima, J. Sato, K. Mizuno, K. Takinami, "A 60 GHz CMOS PLL Synthesizer Using a

Wideband Injection-Locked Frequency Divider with Fast Calibration Technique,"

APMC, pp. 1530-1533, Dec. 2011.

References

Slide 71

[9] W. Deng, A. Musa, K. Okada, and A. Matsuzawa, "A 0.38mm2, 10MHz-6.6GHz

Quadrature Frequency Synthesizer Using Fractional-N Injection-Locked Technique,"

A-SSCC Dig. Tech. Papers, pp. 353-356, Nov. 2012.

[10] M. Bohr, "The New Era of Scaling in an SoC World," ISSCC Dig. Tech. Papers, pp.

23-28, Feb. 2009.

[11] W. Deng, A. Musa, T. Siriburanon, M. Miyahara, K. Okada, and A. Matsuzawa, "A

0.0066-mm2 780-μW Fully Synthesizable PLL with a Current Output DAC and an

Interpolative-Phase Coupled Oscillator using Edge Injection Technique," ISSCC Dig.

Tech. Papers, pp. 266-267, Feb. 2014.

[12] S. Weaver, B. Hershberg, and Un-Ku Moon, "Digitally Synthesized Stochastic Flash

ADC Using Only Standard Digital Cells," IEEE Transactions on Circuits and

Systems-I, Vol. 61, No. 1, pp. 84-91, Jan. 2014.

[13] P. M. Levine, and G. W. Roberts, “A High-Resolution Flash Time-to-Digital Converter

and Calibration Scheme,” IEEE International Test Conference, pp. 1148-1157, Oct.

2004.

[14] D. Sheng, et al., "An Ultra-Low-Power and Portable Digitally Controlled Oscillator for

SoC Applications," IEEE Trans. Circuits and Systems-II, Vol. 54, No. 11, pp. 954-958,

Nov. 2007.

References

Slide 72

[15] A. Matsumoto, S. Sakiyama, Y. Tokunaga, T. Morie, and S. Dosho, "A Design Method

and Developments of a Low-Power and High-Resolution Multiphase Generation

System," IEEE Journal of Solid-State Circuits, Vol. 43, No. 4, pp. 831-843, Apr. 2008.

[16] P. L. Chen, et al., "A Portable Digitally Controlled Oscillator Using Novel Varactors,"

IEEE Trans. Circuits and Systems-II, Vol. 52, No. 5, pp. 233-237, May 2007.

[17] S. Ye, L. Jansson, and I. Galton, "A Multiple-Crystal Interface PLL With VCO

Realignment to Reduce Phase Noise," IEEE Journal of Solid-State Circuits, Vol. 37,

No. 12, pp. 1795-1803, Dec. 2002.

[18] N. D. Dalt, "An Analysis of Phase Noise in Realigned VCOs," IEEE Transactions on

Circuits and Systems-II, Vol. 6, No. 3, pp. 143-147, March 2014.

[19] W. Deng, A. Musa, T. Siriburanon, M. Miyahara, K. Okada, and A. Matsuzawa, "A

0.022mm2 970μW Dual-Loop Injection-Locked PLL with -243dB FOM Using

Synthesizable All-Digital PVT Calibration Circuits," ISSCC Dig. Tech. Papers, pp.

248-249, Feb. 2013.

[20] A. Musa, W. Deng, T. Siriburanon, M. Miyahara, K. Okada, and A. Matsuzawa, "A

Compact, Low Power and Low Jitter Dual-Loop Injection Locked PLL Using All-

Digital PVT Calibration," IEEE Journal of Solid-State Circuits, Vol. 49, No. 1, pp. 50-

60, Jan. 2014.

References

Slide 73

[21] D. Park, and S. Cho, "A 14.2mW 2.55-to-3GHz Cascaded PLL with Reference

Injection, 800MHz Delta-Sigma modulator and 255fsrms Integrated Jitter in 0.13um

CMOS," ISSCC Dig. Tech. Papers, pp. 344-345, Feb. 2012.

[22] Y. Park, and D.D. Wentzloff, "An All-Digital PLL Synthesized from a Digital Standard

Cell Library in 65nm CMOS," IEEE Custom Integrated Circuits Conf., Oct. 2011.

[23] W. Kim, et al., "A 0.032mm2 3.1mW Synthesized Pixel Clock Generator with 30psrms

Integrated Jitter and 10-to-630MHz DCO Tuning Range," ISSCC Dig. Tech. Papers,

pp.250-251, Feb. 2013.

[24] M. Faisal, and D.D. Wentzloff, "An Automatically Placed-and-Routed ADPLL for the

MedRadio Band using PWM to Enhance DCO Resolution," IEEE Radio Frequency

Integrated Circuits Symposium, pp.115-118, Jun. 2013.

Other example:

[25] D. Kaczman, et al., "A Single-Chip 10-Band WCDMA/HSDPA 4-Band GSM/EDGE

SAW-less CMOS Receiver With DigRF 3G Interface and +90 dBm IIP2,“ IEEE JSSC,

Vol. 44, No. 3, pp. 718-739, March 2009.

Acknowledgement

Slide 74

This work was partially supported by MIC,

SCOPE, MEXT, STARC, and VDEC in

collaboration with Cadence Design

Systems, Inc., and Agilent Technologies

Japan, Ltd.

Appendix

Slide 75

𝐣

𝟏

𝟎

Up-conversion by cos(𝝎𝐋𝐎𝒕)

Slide 76

𝐜𝐨𝐬(𝝎𝐁𝐁𝒕) ∗ 𝐜𝐨𝐬(𝝎𝐋𝐎𝒕)

𝝎𝐋𝐎

−𝝎𝐋𝐎

∗ 𝐜𝐨𝐬 𝝎𝐋𝐎𝒕 =𝒆𝐣𝝎𝐋𝐎𝒕 + 𝒆−𝐣𝝎𝐋𝐎𝒕

𝟐𝒆−𝐣𝝎𝐋𝐎𝒕

𝟐

𝐜𝐨𝐬(𝝎𝐁𝐁𝒕) =𝒆𝐣𝝎𝐁𝐁𝒕 + 𝒆−𝐣𝝎𝐁𝐁𝒕

𝟐

𝝎𝐋𝐎 𝒆𝐣𝝎𝐋𝐎𝒕

𝟐

𝐣

𝟏

𝟎

Up-conversion by sin(𝝎𝐋𝐎𝒕)

Slide 77

𝐬𝐢𝐧(𝝎𝐁𝐁𝒕) ∗ 𝐬𝐢𝐧(𝝎𝐋𝐎𝒕)

𝝎𝐋𝐎

−𝝎𝐋𝐎

∗ 𝐬𝐢𝐧 𝝎𝐋𝐎𝒕 =−𝐣𝒆𝐣𝝎𝐋𝐎𝒕 + 𝐣𝒆−𝐣𝝎𝐋𝐎𝒕

𝟐𝐣𝒆−𝐣𝝎𝐋𝐎𝒕

𝟐

𝐬𝐢𝐧(𝝎𝐁𝐁𝒕) =−𝐣𝒆𝐣𝝎𝐁𝐁𝒕 + 𝐣𝒆−𝐣𝝎𝐁𝐁𝒕

𝟐

−𝐣𝒆𝐣𝝎𝐋𝐎𝒕

𝟐

𝝎𝐋𝐎

𝐣

𝟏

𝟎

𝐣

𝟏

𝟎

0

Ideal I/Q Up-Conversion

Slide 78

TX

-90o

I

0

0

0

0

Q

image

image

desired

desired

Image signals are canceled.

0

I/Q Up-Conversion with Mismatch

Slide 79

TX

-90o

I

00

0

Q

0

image

image

desired

desired

Image signals are NOT canceled.

+f

Calibration of Injection Lock Oscillator

Slide 80

*W. Deng, et al., A-SSCC 2012

PFD CP LPF Gating

Buffer

÷2Prog. Divider

÷32/÷33...÷126/÷127

÷2

÷3

÷4

÷5

÷2

÷3

÷4

4:1

Digital Logic

CLK

IN

4

10

7

3

7

7

16

Div. Chain...

÷2

8:1

÷2

Frac.

ILFD

÷10

10

3 3

4 Div. ChainDiv. ChainDiv. Chain

8:1

Quadrature

Output

4

DAC

MASH

1-1-1

Integ.

÷N

LFSR

Frequency-

to-Digital

Converter

Controller

4:1

4:1

Ref. Clock

Slide 81

start

stop

D Q 0/1

tstart - tstop

Ideal condition (no noise, no PVT)

1

0

t0

Stochastic TDC

Slide 82

start

stop

D Q 0/1 (random)

threshold distribution

tstart - tstopt0

Noise & Process variation

1

0

Stochastic TDC

Slide 83

*P. M. Levine, et al., IEEE ITC 2004

start

stop

+

D Q

D Q

D Q

D Q

tstart - tstopt0

5

10

15

0

𝟎 ~ 𝒏 − 𝟏

potentially synthesizable

Linearity compensation is

not very difficult.

Output code

Stochastic TDC

Pulse Injection

Slide 84

Free-running

Injection Pulse

Injection locked

*B. Helal, et al., JSSC 2009

• Severe timing design is required on the injection pulse width.

Measured Phase Noise

Slide 85

10k 100k 1M 10M

-120

-80

-40

0

Offset Frequency[Hz]

Ph

as

e N

ois

e [

dB

c/H

z] Frequency: 900MHz

Integrating Jitter: 1.7ps

PDC: 780mW

Simulated Cmedium against Vin

Slide 86

0 0.2 0.4 0.6 0.8

4

8

12

16

Cm

ed

ium

[fF

]

Vin [V]

DM=1

DM=0

Miller effect

PMOS+offsetoffset

NMOS

Fine Varactor

Slide 87

t

Miller effect is gain-dependent.

*W. Deng, et al., ISSCC 2014

A transient variation of VOUT can make a fine

capacitance difference in CIN.

Fine Varactor (cont.)

Slide 88

0 10 20 30

Time[ps]0 10 20 30

Time[ps]

Vin

[V]

VDD

Vin

[V]

VDD

4.8ps0.4ps

Vin

DM=1DM=0DF=0DF=1

Vin

(0.066ps×6)

*W. Deng, et al., ISSCC 2014

Robust for Layout Uncertainty

Slide 89

70m

m

DCO DCO

Hierarchical P&R with

synthesized DCOs

(for comparison)

130 mm

60m

m

110 mm

Fully synthesized

(proposed)

Integrating Jitter: 1.7ps

PDC: 780mW

FOM: -236.5 dB

Integrating Jitter: 2.32ps

PDC: 640mW

FOM: -234.6 dB

DCO DCO