Injection Locking EECS 242 Lecture 26 Prof. Ali M. Niknejad Outline • Injection Locking - Adler’s Equation (locking range) - Extension to large signals • Examples: - GSM CMOS PA - Low Power Transmitter - Dual Mode Oscillators - Clock distribution • Quadrature Locked Oscillators • Injection locked dividers Ali M. Niknejad University of California, Berkeley Slide: 2 Injection Locking http://www.youtube.com/watch?v=IBgq-_NJCl0 locking is also known as frequency entrainment or • Injection synchronization • Many natural examples including clocks on the same wall observed to synchronize - pendulum over time - fireflies put on a good light show • Injection locking can be deliberate or unwanted Ali M. Niknejad University of California, Berkeley Slide: 3 Injection Locking Video Demonstration http://www.youtube.com/watch?v=W1TMZASCR-I metronomes (similar to pendulums) are initially excited • Several in random phases. The oscillation frequencies are presumably • • Ali M. Niknejad very close but vary slightly due to manufacturing imperfections When placed on a rigid surface, the metronomes oscillate independently. When placed on flexible table with “springs” (coke cans), they couple to one another and injection lock. University of California, Berkeley Slide: 4 A Study of Injection Locking and Pulling Unwanted Injection Pulling/Locking in Oscillators Behzad Razavi, Fellow, IEEE of the difficulties in designing a • One fully integrated transceiver is • • • Abstract—Injection locking characteristics of oscillators are derived and a graphical analysis is presented that describes injection pulling in time and frequency domains. An identity obtained from phase and envelope equations is used to express the requisite oscillator nonlinearity and interpret phase noise reduction. The behavior of phase-locked oscillators under injection pulling is also formulated. exactly due to pulling / pushing If the injection signal is strong Index Terms—Adler’s equation, injection locking, injection enough, it will lock the source. pulling, oscillator nonlinearity, oscillator pulling, quadrature oscillators.it will “pull” the source Otherwise and produce unwanted modulation NJECTION of a periodic signal into an oscillator leads to interesting locking orthe pullingtransmitter phenomena. Studied by In the I first example, Adler [1], Kurokawa [2], and others [3]–[5], these effects have is locked a XTAL whereas found to increasingly greater importance for theythe manifest themselvesisin locked many of today’s transceivers and frequency synthesis receiver to the data clock. techniques. Unwanted coupling This paper describes new(package, insights into injection locking and pulling Vdd/Gnd) and formulates the behavior of phase-locked oscillators substrate, can cause under injection. A graphical interpretation of Adler’s equation pulling.illustrates pulling in both time and frequency domains while derived from the phase and envelope equations A PA isanexpresses aidentity classic source of trouble the required oscillator nonlinearity across the lock range. in a direct-conversion transmitter Section II of the paper places this work in context and Section III deals with injection locking. Sections IV and V respectively consider injection pulling and the required oscillator nonlinearity. Section VI quantifies the effect of pulling on phase-locked loops (PLLs) and Section VII summarizes the experimental results. Ali M. Niknejad Fig. 1. Source: [Razavi] Oscillator pulling in (a) broadband transceiver and (b) RF transceiver. Injection locking becomes useful in a number of applications, including frequency division [8], [9], quadrature generation [10], [11], and oscillators with finer phase separations [12]. Injection pulling, on the other hand, typically proves undesirable. For example, in the broadband transceiver of Fig. 1(a), , is lockedSlide: to University of California, Berkeley the transmit voltage-controlled oscillator, 5 I. GENERAL CONSIDERATIONS Injection Locking is Non-Linear */0&1)2(/$3(142/5 !"# !'$()* !(' -!$%&.+/- -!'$()*.+/!$%& , +$ ! +$ 0!+ !+ !+ !"#$%&'("$%&#)& *+%,-%)'.&/"& 012&+%$")3)/& *+%,-%)'.&'3)& '3-$%&012& *+%,-%)'.&/"& $4#*/&54%)&#/$& 367(#/-8%& 9%'"6%$&4#:4& %)"-:4 -!$%&.+/- -!'$()*.+/, +$ +$ Ali M. Niknejad , 0!+ !+ !+ -!$%&.+/- -!'$()*.+/, +$ +$ , 0!+ !+ !+ -!$%&.+/- -!'$()*.+/, +$ !+ "#"$%&''()) , +$ 0!+ !+ , Source: M. Perrott MIT OCW 6.976 !*+$,-. • For weak injection, you get a response at both side-bands • As the injection is increased, it begins to “pull” the oscillator for large enough injection, the oscillation locks to the • Eventually, injection signal University of California, Berkeley Slide: 6 Injection Locking in LC Tanks 1416 IE (a) a free-running • Consider oscillator consisting of an ideal • positive feedback amplifier and an LC tank. Now suppose (b) we insert a phase shift in the loop. We know this will cause the oscillation frequency to (c) shift since the loop gain has to have exactly 2π phase shift (or multiples) Gm ZT (ω0 ) = gm R = 1 Gm ejφ0 ZT (ω1 ) = 1 Gm ejφ0 |ZT (ω1 )|e−jφ0 = 1 ∠ZT (ω1 ) = −φ0 Ali M. Niknejad Fig Source: Fig. 2. (a) Conceptual oscillator. (b) Frequency[Razavi] shift due to additional phase shift. (c) Open-loop characteristics. (d) Frequency shift by injection. wh shift), and the ideal inverting buffer follows the tank to create a total phase shift of 360 around the feedback loop. What hap- if University of California, pens if Berkeley an additional phase shift is inserted in the loop, e.g.,Slide: as 7 Injection Locking in LC Tanks [cont] phase shift in the tank will cause the oscillation frequency to • Achange in order to compensate for the phase shift through the • • • tank impedance. The oscillation frequency is no longer at the resonant frequency of the tank. Note that the oscillation amplitude must also change since the loop gain is now different (tank impedance is lower) Maximum phase shift that the tank can provide is ± 90° In a high Q tank, the frequency shift is relatively small since 1.5 1 ω0 dφ Q= 2 dω 0.5 0 ω0 ∆φ ∆ω ≈ 2 Q -0.5 -1 -1.5 8 9.25 · 10 Ali M. Niknejad 8 9.5 · 10 University of California, Berkeley 8 9.75 · 10 9 1 · 10 9 1.025 · 10 9 1.05 · 10 9 1.075 · 10 9 1.1 · 10 Slide: 8 Phase Shift for Injected Signal interesting to observe that if a signal is • It’s injected into the circuit, then the tank current is • • • • Ali M. Niknejad a sum of the injected and transistor current. Assume the oscillator “locks” onto the injected current and oscillates at the same frequency. Since the locking signal is not in general at the resonant center frequency, the tank introduces Source: [Razavi] a phase shift Fig. 2. (a) Conceptual oscillator. (b) Frequency shift due to additional phase shift. (c) Open-loop characteristics. (d) Frequency shift by injection. In order for the oscillator loop gain to be equal to unity with zero phase shift, the sum of the shift), andinjected the ideal inverting buffer follows the tank to create a current of the transistor and the total phase shift of 360 around the feedback loop. What hapcurrents must have the proper phase shift to pens if an additional phase shift is inserted in the loop, e.g., as compensate for the tank phase shift. depicted in Fig. 2(b)? The circuit can no longer oscillate at because the total phase shift at this frequency deviates from 360 We see that the oscillator current, tank by all. Thus, as different illustrated in Fig. 2(c), the oscillation frequency current, and injected current have such that the tank contributes must change to a new value phases enough phase shift to cancel the effect of . Note that, if the Fig wh if De be en contribute no phase shift, then the drain current m buffer and must remain in phase with under all condi- re of tions. University of California, Berkeley Slide: 9tan by adding a sinuNow suppose we attempt to produce Injection Locked Oscillator Phasors 0 Itank Vtank = Itank ZT (!1 ) Itank = Iosc + Iinj Itank j Iosc e Iosc 0 Iinj 6 Iosc = 6 Vtank = 6 Itank + 0 Iinj that the frequency of the injection signal determines the • Note extra phase shift Φ of the tank. This is fixed by the frequency 0 • • Ali M. Niknejad offset. The current from the transistor is fed by the tank voltage, which by definition the tank current times the tank impedance, which introduces Φ0 between the tank current/voltage. The angle between the injected current and the oscillator current θ must be such that their sum aligns with the tank current. University of California, Berkeley Slide: 10 Injection Geometry sin φ0 = B A B Itank Itank B cos(π/2 − θ) = sin(θ) = Iinj Iosc Iinj sin φ0 = sin(θ) Itank Iinj sin(θ) Iinj sin(θ) ! sin φ0 = = jθ |Iosc e + Iinj | 2 + I 2 + 2 cos θI Iosc osc Iinj inj 0 Iinj geometry of the problem implies the following constraints • The on the injected current amplitude relative to the oscillation • amplitude. The maximum value of the rhs occurs at: −Iinj cos θ = Iosc Ali M. Niknejad sin θ = ! 2 − I2 Iosc inj Iosc University of California, Berkeley sin φ0,max Iinj = Iosc Slide: 11 Locking Range the edge of the lock range, the injected • At current is orthogonal to the tank current. phase angle between the injected current • The and the oscillator is 90° + Φ lock range can be computed by noting that • The the tank phase shift is given by Itank 0,max 2Q tan φ0 = (ω0 − ωinj ) ω0 Iosc 2Q Iinj Iinj tan φ0 = (ω0 − ωinj ) = =! ω0 Itank 2 − I2 Iosc inj ω0 Iinj 1 ! (ω0 − ωinj ) = 2Q Iosc 1− Ali M. Niknejad 2 Iinj 2 Iosc 0 Maximum Locking Range Iinj University of California, Berkeley Slide: 12 Weak Injection Locking Range first derived these results in a celebrate paper [Adler] • Adler under the conditions of a weak injection signal. results for a large injection signal were first derived by • The [Paciorek] and then re-derived (as shown here) by [Razavi] • Under weak injection: I ! I inj osc Iinj sin φ0 ≈ sin(θ) Iosc Iinj 2Q sin φ0 = sin(θ) ≈ tan φ0 = (ω0 − ωinj ) Itank ω0 2Q Itank sin(θ) = (ω0 − ωinj ) ω0 Iinj the edge of the locking range, the angle reaches 90°, and the • At injected signal is occurring at the peaks of the output, which cannot produce locking (think back to the Hajimiri phase noise model) ω0 Iinj (ω0 − ωinj ) = 2Q Itank Ali M. Niknejad University of California, Berkeley Slide: 13 IEEE JOURNAL OF SOLID-STATE CIRCUITS, VOL. 39, NO. 9, SEPTEMBER 2004 Injection Pulling Dynamics , and (15) Source: [Razavi] ains phase modula- Fig. 6. LC oscillator under injection. now that the oscillator is under injection phase shift can Suppose be oscillator signal feeding the tank can be written as antaneous input freEquating this result to , we obtain • so that VX = Vinj,p cos ωinj t + Vosc,p cos(ωinj t + θ) VX = (Vinj,p + Vosc,p cos θ) cos ωinj t − Vosc,p sin θ sin ωinj t (16) (23) This We canalso benote written as athat cosine with a phase shift • from (19) slowly-varying ), ection phenomena. V = Vinj,p + Vosc,p cos θ cos(ω t + ψ) X inj cos ψ • Which allows us to write! the output as Vosc,p sin θ tan ψ = Vinj,p + Vosc,p cos θ " ! #$# m shown in Fig. 6, (24) Vinj,p + Vosc,p cos θ dψ −1 2Q e input. The output Vout = cos ωinj t + ψ + tan ω0 − ωinj − cos ψ ω0 dt having a carrier fre(25) words, the output is Ali M. Niknejad University of California, Berkeley sibly time-varying) Slide: 14 Injection Pulling Dynamics [cont] have assumed that the phase shift through the tank is given • We by the simple expression derived earlier where the instantaneous frequency is dψ ω+ dt ! " 2Q dψ tan α ≈ ω0 − ω − ω0 dt • Ifto:the injection signal is weak, then the previous result simplifies ! Vout = Vosc,p cos ωinj t + ψ + tan −1 " 2Q ω0 ! ω0 − ωinj dψ − dt #$# must equal to the output voltage, or the phase shifts • Which must equal ψ + tan −1 Ali M. Niknejad ! 2Q ω0 " ω0 − ωinj dψ − dt University of California, Berkeley #$ =θ Slide: 15 Pulling [cont] • These equations can be manipulated into a form of Adler’s equation: dθ ω0 Vinj,p = ω0 − ωinj − sin θ dt 2Q Vosc,p dθ = ω0 − ωinj − ωL sin θ dt ω0 Vinj,p ωL ! 2Q Vosc,p equation describes the dynamics of the phase change of the • This oscillator under injecting pulling. If we set the derivative to zero, • • we obtain the injection locking conditions (same as before). Notice that maximum value of the rhs is quite small, which means that the rate of change of phase is This equation can be used to study the behavior of locking signals outside the lock range. Note that this equation agrees with our graphical analysis: dθ =0 dt Ali M. Niknejad University of California, Berkeley Slide: 16 Pull-In Process θ(t) = 2 tan −1 ! 1 − cot θ0 tanh sin θ0 θ0 = sin −1 " #$ ωL cos θ0 (t − t0 ) 2 ω0 − ω1 ωL the steady-state phase shift between the injected signal • θand0 isthe active device signal and t is an integration constant that 0 • depends on the initial phase shift. From this equation the lock-in time can be computed (it’s approximately an exponential process): ! 2 −1 1 − sin θ0 tan tL = tanh ωL cos θ0 cos θ0 Ali M. Niknejad University of California, Berkeley "θ #$ L 2 + t0 Slide: 17 Phase Noise in Injection Locked Systems IEEE JOURNAL OF SOLID-STATE CIRCUITS, VOL. 39, NO. 9, SEPTEMBER 2004 input frequency deviates uction becomes less proeither edge of the lock g the impedance seen by o rely on the phase noise g. Since the lock range is atural frequency of oscilocess variations and poor noise due [Razavi] to injection locking. he edge of the lock range, Fig. 14. Reduction of phaseSource: slightly. For example, if Under % and the natural fre- a lock, the phase of the oscillator follows the phase of the % with process and tem-injection signal. If a “clean” signal is used to lock a VCO, thenat the phase noise would improve up to the locking range. njection locking occurs 37) and the above obserAt the edge of the lock, the injected signal cannot correct for e noise falls from infinity the phase noise since it injects energy at a 90° phase offset, degradation in the phase where the signal has a peak amplitude . • • LOCKED OSCILLATORS with pulling in nominally Ali M. Niknejad Fig. 15. PLL under injection pulling. University of California, Berkeley Slide: 18 Fig. 8. Microstrip balun. Example: GSM Class E PA 964 IEEE JOURNAL OF SOLID-STATE CIRCUITS, VOL. 34, NO Fig. 10. Output power and PAE versus frequency for Fig. 9. Output power, efficiency, and PAE versus supply voltage. 10 and 5 mil, respectively. A single-ended signal at port 1 can be converted to differential signals at ports 2 and 3 or vice versa. The edge-coupled microstrip lines and the discrete tuning capacitors are shown in the enlarged portion of Fig. 8. and V. Fig. 5. Illustration of the mode-locking concept. B. Mode-Locking Technique Fig. 11. Amplified GMSK modulated signal ( spectral emission Modemask. locking refers to the ) and the GSM Source: [Tsai]condition in which an self-oscillating circuit is coupled and forced to of the same PAE infrequency this regionasis an dueinput to thesignal, progressively morein a resulting VI. MEASUREMENT RESULTS prominent input power in driving the PAE requirement. definition. At This 2-V is reduction in theterm input supply, the PAE was measured to be 48%. Fig. A. 4. Power-Amplifier Schematic of thePrototype complete power amplifier. each stage of the amplifier by a pair of cross-coupl Fig. 10 shows the output power and PAE at two different devices, as shown in Fig. The twowhere input the volta Large single-ended output stage is hard to drive capacitance). voltages(large across the range of 5. frequencies A 10-dBm input device was converted to supply of is phase, are the two output The isload mode as locked successfully. Thisvoltages. locking range differential signals, using a commercial balun (Murata amplifier Useequivalent the and PAoutput toapplied drive an oscillator, right? a at the nodes isYes. designed suchGHz. that measured to beoutput 490 MHz, centering at about 1.9 LDB20C500A1900), then to itself! the PA. capacitors RFThat’s power and inductors and load, and larger verify its potential in practical communication output ports. Fig. 9 maximized shows a was measured at the 50- transistors in phase to control the composite switch.applicaAs far a switches. In fact,injection the switch areto often inTo modulation. Use locking inject phase Need “off” mode-locking class-Ethe PAoperation was tested is with a Gaussian , measured at tions, the typical plot of the output power versus circuit is concerned, similar to the si size to reduce the on-resistance loss. The input capacitance of switch to power turnincreases off transmission. 1.98 GHz. The output from 49 mW to 1.0 minimum shift keying (GMSK) [21] modulated input signal. version as shown in Fig. 1, except for two feat these is asusually tuned out by inductors. W transistors monotonically the supply is swept from 0.6 to 2 VHowever, and GMSK and its close relative, Gaussian frequency shift keying thearecurrent circulating at each tuned lo at gigahertz frequencies, beyond transistor size the (GFSK), constantoriginally envelope modulation schemes in which . The maximum is approximately proportional to a certain to assist switching of the other circu is carried in the phase variation of the half signal, used in this caserequired is determined by the peak drain too voltage inductance values for tuning may become smallinformation to utilized capacitance input can now be significan makingthe them well suitedattoeach switching power amplification. on chip, which estimatedthe to be slightly above 5 V capacitance at 2-VBerkeley be realizable. In is addition, large gate-to-drain Ali M. Niknejad University of California, Slide: 19 • • • supply, approaching the drain-gate oxide breakdown limit of These modulation schemes are used by popular cellular and )+!J57()*+!K.L!'+=!25+75!*+6?!(25!=51);+!*>!(25!D*354!*17)66'(*4!)1!56'M*4'(5=!2545.!!! Example: Low Power Transmitter ! !"#"1$ 233-,-(40$%&'()$*+,-../0&)$5(+-64$ B25!1725:'()7!*>!(25!D*354!*17)66'(*4!)1!12*3+!)+!E);.!-.-.!!! .! G74!TA.! NF! N-D72!TA.! :>>$ ! ?1$ ! ! ! ! !V!N-D72!W-:82-,! ! ! 3-09,18%-:!821:A3%8872! U181!V!N-D72!W-:82-,! ! 5=6!?:>7.8%-:!,-./70!821:A3%8872! :-4;=$ ! -./!0%1&213!-4!516!0%27.8!3-09,18%-:!;<!1:0!5=6!%:>7.8%-:!,-./70!;<! ! ! :! ,-./70! 821:A3%8872B! 8@7! C-D72! 13C,%4%72! %A! !27C,1.70! =E! 1:! 744%.%7:8! '! ! F! C-D72! -A.%,,18-2! %A! :7.7AA12E! A%:.7! 8@7!! $+FG! -A.%,,18-2! .1::-8! ! ! 8@7! 1:87::1! D%8@-98! 07&210%:&! %8A! JK41.8-2! A9=A81:8%1,,E'! ! ;@7! C-D72! E&)$0F-+$>(+-64D$ 4$G$1$0&$!$ 78$ 7@4<1$"$"$"$71 =$ :A$ <$ ?@$ B?$ C1$ 7@$"$"$"$7@4$ 79 ! ! ! ! ! 7C$ ! :,0).D-4;$ :1$ ! "$"$"$ : ! ! ! ! E);.!-.-N!J725:'()7!*>!(25!)+C57()*+!6*785=!*17)66'(*4.! :-4;< Source: [Chee] $%&'!(')*!+,-.,-!/.01-2,3!4506!.7402!7/1%889-72!%/!:80;-<!;200!2,66%6&!:2%&5-<!871=0>'! 7:!%:8-!8@7!H-,81&7K,%3%870!27&%37!8-!1,,-D!8@7!-98C98!H-,81&7!8-!AD%:&! a low-power radio, the overall transmitter efficiency is very • Inimportant. The PA output power is modest (~1mW), but since ! $%&'!('?@!/574/!-50!/%6&80!/%>0>!871=A%6!296&0! !B!9/!9!;,61-%76!7;!-50!C%9/!1,2206-!7;!-50! C,E'!!;@%A!2709.7A!8@7!07H%.7!,-AA!5? 0ALM0A6!1:0!%3C2-H7A!%8A!744%.%7:.E'!! "# %6D01-%76!871=%6&!-296/%/-72!EF'!!F!5%&502!C%9/!1,2206-!%61209/0/!-50!-296/176>,1-9610!7;! ,,18-2! %A! A7,4K02%H7:! 1:0! @7:.7! 0-7A! :-8! ,-10! 8@7! 274727:.7! -A.%,,18-2! the overall efficiency should be large, the entire transmitter -5%/!98/7!%61209/0/!-50!.7402!176/,3.-%76!7;!-296/%/-72!E !96>!E !45%15!>0&29>0/!-50! should not consume more than 2-3mW. ! 7J02988! 0;;%1%061K'! ! L7! 3%6%3%M0! 0;;%1%061K! >0&29>9-%76H! -50! 871=A%6! 296&0! ! ! 196! C0! By using injection locking one can reduce the power 157/06!-7!C0!N"!EIM!96>!-50!.09=!0;;%1%061K!%/!20>,10>!CK!768K!CK!N?O'! ::1! ,-10A! 8@7! C-D72! -A.%,,18-2RA! -98C98! 81:/! 1:0! 07&2107A! %8A! JK41.8-2'!! consumption of the driver stages and end up with a minimal -A.%,,18-2!A94472A!42-3!1!C--2!C@1A7!:-%A7!C724-231:.7!1:0!1:!%3C27.%A7! transmitter architecture. -296/%/-72!EF!96>!EGH!45%15!%61209/0/!-50!%6D01-0>!/%&698!96>!871=A%6!296&0'!!I740J02H! @9A! 8@7! C27KNF! C-D72B! .-:A%A8%:&! -4! 8@7! 274727:.7! -A.%,,18-2! C-D72B! %A! 8,@3-7A)1,B*1?0,!",;9CD= • F G " '# &' '! &% &! G/*1+<)220/,H44)-)01- 7:.E'!!;-!-=81%:!1!A81=,7!G$!.122%72B!8@7!C-D72!-A.%,,18-2!%A!,-./70!8-!1:! &# &! %E 274727:.7!-A.%,,18-2B!D@-A7!-A.%,,18%-:!427S97:.E!%A!A81=%,%O70!=E!1!@%&@! -2'!!!!! Ali M. Niknejad %# %F %! University of California, Berkeley %' Slide: 20 Consider interconnection of two from the locked half mode 4.8GHz reference. the locked half the mode 4.8GHz reference. A A/0 tanks as conventional mode VCO fabricated as hown in Fig.single 1. Each tank isisa also parallel /0ascircuit, butFig.1 they ntional single mode VCO is also fabricated Fig.1 Capacitive-coupledparallel parallel/0 /0tank tank Capacitive-coupled reference to show the power saving advantage of the nce to show the power saving advantage of the re coupled together through a coupling capacitor 01. By structure. The most important feature for this coupled /0 tank ure. Example: Dual Mode VCO The most important feature for this coupled /0 tank arying the value of coupling capacitor, differentis dynamic that it inherently has two resonantfrequencies. frequencies.AAsimple simple is that it inherently has two resonant Index Term – Coupled LC tank, dual-mode, henomenon be observed. calculation yields the following expression for the two ndex Term – can Coupled LC tank, dual-mode, calculation yields the following expression for the two resonant frequencies e-controlled oscillator, injection locking. resonant frequencies 22 % ! 22 " ! I. INTRODUCTION Eq. 1 2 " ! 2 $ $ # # 2 % ! 2 / 3 I. INTRODUCTION Eq. 1 $/ # $3 # 22 3 22 3 22 3 22 3 Generation of a low phase noise spectrally pure tone where Generation of a low phase noise spectrally pure tone where 2 has been a major challenge for communication circuits 2 + 03 03 01 022( 0 3 Eq. 2a en a majorinchallenge for communication !+#0)) 0" 0 " 0 % ( && % 40 2 fabricated CMOS technology. The lack circuits of a high Q Eq. 2a ! # )) *3 /"1 3 /"2 1 /%2 2/&&1 '% 4 /31/2 ated in resonator CMOS technology. of consumption a high Q stable necessitates The high lack power in * /1 0 3/2 0 3/2 01/1 '0 2 /1/2 resonator highloop power consumption in As the form ofnecessitates a phase-locked frequency synthesizer. Eq. 2b 22 # % % % 0 0 0 0 m of a phase-locked loop frequency synthesizer. As Eq. 2b we move to higher frequencies, the problem is exacerbated 2 2 # 3 /%1 3 /%2 1/%2 2/1 Capacitive-coupled parallel /0 tank ve totohigher frequencies, problem is of exacerbated due theFig.1 increasing powerthe consumption the prescalar Eq. 2c 23 / #10301/%2 030/22 % 01/01 2 frequency dividers in the frequency synthesizer. Since the increasing power consumption of the prescalar Eq. 2c 2 3 shows # 0301 the % 0differential Fig.2 302 % 0102 dual-mode /0 tank used digital logicmost isinoften employed synthesizer. tofeature perform for thisSince task, ncy dividers theimportant frequency The thisthe coupledinFig.2 /0 tank this design. /0 tank coredual-mode is simply the shows The the differential /0 differential tank used power consumption increases as we operate at a closer logic is often employed to perform this task, the version of Fig. values carefully in this design. The 1. /0The tankcomponents core is simply theare differential sfraction that it inherently has two resonant frequencies. A simple of the process . . Injection locked dividers consumption increasesT as we operate at a closer calculated and1.selected for desired resonant /1 version of two Fig. The components values frequencies. are carefully alculation yields the following expression for the consume less power, but have narrow locking range and n of the process .T. Injection locked dividers and /2 and areselected on chipforspiral inductors. center tap calculated desired resonant The frequencies. /1 require an additional oscillator. me less power, but have narrow locking range and esonant frequencies connection of the two inductors is used as the dc current and /2 are on chip spiral inductors. The center tap While multimode oscillator have been proposed e an additional oscillator. path which allows the active circuits presented at current the two connection of the two inductors is used as the dc (a) before [1] [2], in this paper we present a coupled /0 2 2 % ! ports to 2 2 " !have been proposed While multimode oscillator share the same dc current. By terminating each Eq.allows 1 the active circuits presented at the two path which oscillator with dual frequency outputs that naturally $ $ # # / 3 port with a negative resistance, generated for example by a [1] [2], in this paper we present a coupled /0 ports to share the same dc current. By terminating each 22 provides a technique 22 for combining the VCO and a 3 3 cross-coupled active device, both oscillation modes can be tor with dual frequency outputs that naturally frequency divider in a phase-locked loop. Typically, we port with a negative resistance, generated for example by a es technique for combining the VCO andlower a mode to activated at the same time. This is the basic idea behind the where useathe higher mode to drive a mixer and the cross-coupled active device, both oscillation modes can be dual-mode oscillator. ncy divider in a phase-locked loop. Typically, we 2 2 drive the PLL dividers. activatedThe at the same /0 time. This iscan thealso basic idea phase behindnoise the + ( coupled network benefit 0 0 0 0 0 e higher mode to drive mode to 3 a mixer 3 and1 the lower 2 3 Eq.oscillator. 2a dual-mode ! # )) " " % && % 4 performance. Analytical analysis shows that this higher he II. PLL dividers. / / / / / / DUAL-MODE LOCK The resonant coupled /0 network also benefit phase noise 1 ' 1 2 * 1VCO2WITH2 INJECTION order network can can enhance the equivalent tank performance. Analytical shows that this higher DESIGN quality factor thereby analysis improving the phase noise. An 0 0 0 0 3 3 Eq. 2b 1 2 DUAL-MODE VCO WITH I NJECTION L OCK order resonant can enhance equivalent tanka 22 # % % % intuitive waynetwork of understanding thisthe approach is that /ESIGN / 2 /of2 Electrical /1 Engineeringquality The authors are withDthe factor thereby improving phase noise. An (b) 1 Department “zero” is designed at the vicinity of the the resonant frequency, Fig. 2 roll-off (a) Differential dual-mode !" tank (b) and Computer Science, University of California at Berkeley,intuitive way of understanding this approach is that a shown in Fig. 2(b), so that it increases the rate from Eq. 2c 2the 0301Email: % 0of Berkeley, CA 94720 USA. dengzm@eecs.berkeley.edu impedance over frequency 3 #Department 30 2 % 0102Engineering thors are with Electrical the peak resulting in avicinity sharper of transition and a frequency, narrower “zero” is designed at the the resonant Ali M. Niknejad University of California, Berkeley Slide: 21 voltage-controlled oscillator, injection locking. coupled tank has two resonant modes. • AFrom each port of the oscillator, one mode is • • dominant. In a fully differential version shown to the right, the impedance variation with frequency is shown. Can we build an oscillator that sustains both modes? omputer Science, University of California at Berkeley, Dual Mode VCO (cont) Source: [Deng] cross-coupled pairs are used to sustain each mode by • Two providing sufficient negative resistance. The PMOS side is • Ali M. Niknejad running at a lower frequency whereas the NMOS side runs at the higher frequency. Transistors M3a/M3b are used for injection locking. University of California, Berkeley Slide: 22 Dual Mode VCO Spectrum • • Ali M. Niknejad • Without injection locking, each mode runs independently. The frequencies are not integrally related, so the unlocked spectrum contains not only the two modes and harmonics, but also every intermodulation component. When the modes are locked, the intermodulation components disappear. The high frequency VCO is divided by two in this configuration. More on injection locking dividers to come... University of California, Berkeley Slide: 23 ter 2: Overview of Electrical Clock Distribution 14 Clock Distribution Chapter 2: Overview of Electrical Clock Distribution 16 PLL Figure 2.7: igure 2.6: A buffered H-tree. A clock grid driven by buffers. Source: [Mahony] 2.3.1.3 Local distribution is completely symmetric to simplify the design and provide nominally • source of power consumption. Figures of merit include skew, The final level topology of distribution network is the local which is the portion o a clock in acommonly large chip isa clock a major challenge andlevel, a big kew (assuming no Delivering loading variations). The most used is a buff- the network that follows the clock pin. This network drives the final loads of the clock dis H-tree [7] which is shown in Figure 2.6. H-trees are an extremely popular choice at tribution and hence consumes the most power. As a rule of thumb, the power at the loca jitter, and power. evel of the clock distribution because they are a simple pattern that efficiently and level is about one order of magnitude larger than the power in the global and regional lev A buffered H-tree or a grid are common approaches. The grid metrically covers large areas. H-trees minimize the els distance fromwith thethe PLL to notable the fur-exceptions being clock networks that use a combined, only has skew and larger capacitance. • low-impedance grid atthe theinterconnect regional level [2]. clock pins (without using diagonal traces) and thereby minimize The layout of the local is generally included in the design of the macro block and cy. The number of buffer levels within the global network − which alsogrid determines is not the responsibility of global and regional clock network designer. Because of its rela tency − depends on the signal dispersion, loss and on the required power fanout. Ali M. Niknejad tively limited span, it is sufficient to use automatic layout for this portion of the clock University of California, Berkeley Slide: 24 ould be expected based on the delay between points on the clock network. Despite vantages, however, an analysis of the relationship between skew and interconnect Clock Distribution: Distributed ndicates that wire loss will limit the use of standing-waves in an on-chip, global distribution. This type of clock distribution also requires limiting amplifiers to con2.5: Directions in global clock distribution 27 he sinusoidal standing waves to digital levels, which could add additional skew and 2.5: Directions in global clock distribution l << λ 2 VCO and loop filter Phase detector Regional Clock Distribution ÷N Clock load gure 2.12: Standing-wave clock distribution [41]. Figure 2.13: Clock distribution using a coupled array of PLLs [44].System Clk PD, LPF & CP Source: [Mahony] Vc (to VCOs) algorithm is used to avoid instability during the phase locking process. There are two main advantages to this topology. First, the PLLs filter the noise of the global H-tree and Coupled oscillator arrays replace it with the noise of the local VCOs, thereby reducing the jitter that accumulates in wave oscillators tune out the • Standing capacitance of the line and form a resonant VCO Coupling wires the H-tree. The second advantage is that, like in [44], the skew is a function of the error in he research that has been described to this point generally concentrates on minimiz- the phase detectors rather than the accumulated skew between branches of the H-tree. This e skew of a global clock distribution. However, based on the scaling analysis in means that the scaling of this type of distribution does not depend on the latency of the Figure 2.15: Clock distribution using a coupled array of VCOs [48]. H-tree but rather on the scaling of the VCO and phase detector. The presence of the global on 2.4, it appears that jitter may be more difficult to reduce for future high-perfor- network. been proposed that have the potential to reduce both jitter and skew by eliminating A distributed approach locks an array of amounts of latency from the clock network. Some also reduce skew and jitter by totheit’s ying a phase-averaging effectVCO’s that tends to reduce impactneighbors. of uncorrelated static The neighbors can also be injection locked to one another to eliminate the phase detectors. e microprocessors. Several clock distributions based on coupled oscillator arrays H-tree also allows for low-frequency testing by bypassing the PLLs. The only drawbacks • • Ali M. Niknejad Coupled distributed oscillators to this approach are the additional resources necessary for multiple2.5.3.3 PLLs and the com- plexity of synchronization. This approach has not yet been prototyped so no measured The last type of coupled-oscillator clock distributions uses coupled arrays of distrib data is available. uted oscillators. The oscillators are distributed in the sense that the coupling wires are pa 2.5.3.2 of the oscillators themselves, so their free-running frequency depends to some degree o VCO arrays the geometry of the interconnects. The cooperative ring oscillator (CRO) clock distribu The global clock distributions described in [47] and [48] are similar to the PLL array tion described in [49] uses overlapping ring oscillators. The frequency of the oscillato designs, except that they use a single control loop with multiple VCOs. The clock distribudepends on both the buffer and interconnect delays around the ring. Figure 2.16 shows tion shown in Figure 2.15 and described in detail here is based on [48], but the same basic three-phase CRO grid and its electrical equivalent. The oscillators are allowed to run ope concepts are used in [47]. A control voltage is distributed to all of the VCOs around the loop, and they phase lock to each other due to mutual coupling between rings, similar die to set their frequency. A clock signal is fed back from a regional clock tap (driven by a [47] and [48]. University of California, Berkeley A similar approach is proposed in [50], which implements coupled arrays rotar Slide: of25 of the“keep-out” PMOS andregions, areas where the signal amplitude is too low to be tapped b the line, so they are optimized for maximum gd/cd. The relative sizes ered Thetovoltage standing-wave pattern for a portion of the grid is shown in Figu NMOS devices can also be optimized, but a good rule of thumb is tobuffers. size them be Standing Wave Oscillators (SWO) roughly equal. Note that the largest magnitude occurs around the square portion of the grid and t nodes of the standing-wave are pushed onto the stubs. 4.2.3 Design example Single SWO (folded) 63µm:0.18 µm Clkinj Differential interconnect Cross-coupled pair Injection-locked cross-coupled pair Clock buffer 63µm:0.18 µm 14µm 14µm Chapter 5: Standing-Wave Clock Distribution Network 3.5mA 3.5µm 4µm Figure 5.3: 5.1.1 SWO coupling A resonant clock grid of coupled SWOs. Figure 4.8: SWO and transmission line cross-section for design example. ccp ccpof theccp ccp ccp This clock grid of coupled oscillators solves two fundamental problem negative resistance to sustain • The oscillation is distributed along the line. disadvantage is that the amplitude of • One the clock varies along the line. can be injection locked together • SWO’s to form larger clock trees. accumulation of timing uncertainty and sensitivity to buffer delay. A A SWO is straightforward to design from (4.11) and (4.2). Table 4.1H-trees: lists the the parame- in spaced Chapter 2, H-trees accumulate timing uncertainty because the clock si ters that will be used to design a SWO with five equally sized andcussed equally cross-coupled pairs. The transmission-line cross-section (Figure 4.8) is optimized generatedfor bymina single source and then regenerated by each ccp buffer along the tree. I ccp ccp ccp ccp the ground imum loss given a total track width of 32µm and a distance of 3.5µm totrast, each plane. SWO in the grid locally generates the clock signal and only uses the A unit cross-coupled pair (ccp) is defined to have NMOS and PMOS devices that are 18µm wide and 0.18µm long and is optimized for maximum gd/cd with 1.0mA of bias cur- Figure 5.1: Two coupled SWOs. Source: [Mahony] Multiple SWOs can be coupled together by simply connecting their transmission li Two coupled SWOs are shown in Figure 5.1. Ideally, the two oscillators are matche frequency and a transient current between the two oscillators brings them into per phase lock. If the oscillators are detuned, they will still lock if they are within their mu Ali M. Niknejad University of California, Berkeley locking range. A steady-state coupling current will keep them frequencySlide: locked,26 altho ROUFOUGARAN et al.: SINGLE-CHIP 900-MHz SPREAD-SPECTRUM WIRELESS TRANSCEIVER IN Quadrature Locked VCOs (QVCO) 270° 90° 0° 180° 0° 180° Fig. 20. (a) Q showing the re across the reso Source: [Rofoug] Fig. 19.VCO’s Quadrature oscillator comprises a pair of cross-coupled LC oscil-as Two are coupled together using extra transistors •lators synchronized in frequency. are identical, we expect that the shown. If the oscillators amplitude and frequency of oscillation should be identical. Becausebalanced of the phase of the outputs coupling, can be shown that they recently, quadrature areit obtained from two lock in quadrature... • Ali M. Niknejad cross-coupled relaxation oscillators [30]. Another oscillator consists of two LC biquad inBerkeley feedback [31]. Four-stage Universityfilters of California, Slide: 27 UM WIRELESS TRANSCEIVER IN 1- m CMOS—PART I QVCO: Quadrature Lock 525 Source: [Rofoug] that the tank “A” (a) is driven by the tank voltage VA plus (b) • Note the tank voltage of “B”. Fig. 20. (a) Quadrature oscillator redrawn, emphasizing its symmetry and rightthe tank, though, is driven by the voltage V phases but voltage -V . • The showing resonator currents; and (b) relative of the voltages across the resonators required forand symmetric operation. that the LC tanks FETs are identical, then the • Assuming phasor currents flowing into the tanks must have equal B pled LC oscil- magnitude: A |vA + vB | = | − vA + VB | |1 + ejθ | = |1 − ejθ | d from two Ali M. Niknejad cos θ = 0 → θ = ±90◦ University of California, Berkeley Slide: 28 (a) (b) QVCO: Lead/Lag Ambiguity Fig. 20. (a) Quadrature oscillator redrawn, emphasizing its symmetry and showing the resonator currents; and (b) the relative phases of the voltages across the resonators required for symmetric operation. r of cross-coupled LC oscil- are obtained from two 30]. Another oscillator edback [31]. Four-stage at taps two stages apart lthough apparently not ite low phase noise in ture oscillator. It comSource: [Rofoug] d above, labeled A and ted in parallel with the Oscillator A is direct21. Magnitude and phase of a realistic resonator’s impedance, show- relation, or the NoteFig.that the tank current/voltage has a 90° llator B, and the output ing three possible oscillation frequencies. The oscillator selects the highest of the impedance Three frequencies satisfy this because the feedback loop is gain±45°. is largest here. Oscillator A (Fig. phase 19). frequency constraint, but f3 has the highest loop gain. ctly the same frequency y the coupling topology. circuit impedance is andpossible its phase is solutions: either 45 It’s clear that there are two lead or lag in plains how, owing only or 45 (Fig. 20). In an LC resonator with series loss in the phaseinductor, between “A” and “B”. n quadrature. these phases appear at three different frequencies, , ctively, are the steady, real and oscillator, (Fig. 21). However, amongoscillators them only is are stable,not perfectly In any the two with reference polarities because at this frequency the impedance is largest, and thus the symmetric, and it can be shown that there is only one unique rrent into the tuned load the phase feedback factor in the oscillator is strongest. At solution (where the loop gain is highest and the net phase shift is into of the resonator’s impedance is 45 , which means that the differential average around 0°) must the lead loop by 90is, as is shown in Fig. 20. Note also that the ET’s. Arguing purely by oscillation frequency is higher than the resonance of the tuned Niknejad University of California, Berkeley s Ali inM.the two circuits are • • • Slide: 29 performances of the S-QVCO compared to the P-QVCO. possible oscillation The frequencies and , both differing f measurement results for a prototype of thethe S-QVCO fabricated Further, the same LC: noise also natural tank resonance frequency Fig. 6.show Fair phase-noise comparison between P-QVCO and S-QVCO. , whic an oscillation in a standard 0.35- m CMOS process which modulates YSIS AND DESIGN OF A 1.8-GHz CMOS LC QUADRATURE VCO frequency of 1.8 GHz, a tuning range of 18%, a phase noise of transconductance e P-QVCO, in the presence of different values for . 140 dBc/Hz or less at a 3-MHz offset frequency across the we obtain frequencies and , both differing f , tuning range, and an error ofoscillation at most 0.25 Linear model ofequivalent QVCO isphasepossible ption, there will be noconsumption doubt that theofS-QVCO LC: the natural tank resonance for ashown. current mA from a 2-V power supply. frequency The loop gain 25 is easily QVCO Loop Gain • m the P-QVCO.5 computed and the frequency at which the loop gain is zero is II. A SIMPLIFIED QVCO MODEL APPENDIX A To estimate the effect on th the oscillation frequency. These equations can be simplified noting that appears squared in (17) duces a linear model for the P-QVCO, where that the is offoscillation frequencies to define InNotice this appendix, weoscillation find the possible in the expres the transconductance of the negative-resistance resonance. results lower for the linearized This QVCO circuit in in Fig. 7. The loop gain is easily d the transconductance of the coupling pair phase noise. calculated as ng to this figure and also in the following, we • Fig. 7. Linear model for a QVCO. onverted baseband signals and LO leakage at 1.75 GHz carrier URE VCO 1745 ge and current signals to be fully differential: 56 dB). Source: [Andreani] wing into the tank is the difference between the These equations can be simplified noting that As shown in Appendix A, the oscillation frequency retwo branches of the differential stage. As a LC from the and therefore possible oscillation frequencies and , both differing from displaced sults slightly tank resonance . Using (20), we can approxim and assuming that (17) isthe thenatural loss resistance of one half-tank. LC: tank resonance frequency by anthe offset , whose magnitude . This can to ( inner square root in (18) asdepends on . According According to Barkausen’s criteria, thealso circuit oscillates when be explained in thelarger following, way. Referring to , which can be neglec this term is much than intuitive is carrier satisfied; further, there are in two thesignals condition baseband and LO leakage at 1.75 GHz in Fig. 7,using the losses tankare balanced by a current in Finally, (19) and the approximation B). matters even further, an additional variable is the value of the phase , which is provided by withfor , , we arrive at valid or widths, . The results shown in . The tank is now lossless, and the current from acts on d when both P-QVCO and S-QVCO shared the same value LC-parallel. This second current, , is thus in quadrar, can be largely reduced for the P-QVCO, in which an ideal (18) . Using andwith assuming that in turn implies e decreases by approximately 2 dB in the region, but , which that (20), andwe can areapproxim phase ture to ( the inner in (18) l decibels in the region. It should be noted that a lower shifted . Fig.root 8 shows the as phasors of the. According voltage across by square ows the P-QVCO to achieve maximum oscillation These equations can abehigher simplified noting that whichTocan this termand is of much larger than the currents entering the, tank. findbetheneglec red to the S-QVCO, or, when the oscillation frequency and the the tank, whichusing is the same as the (2). approximation Finally, and , we consider the loop in Fig.Slide: 7 as e Ali same for both P-QVCO and S-QVCO, the capacitanceUniversity in lation between M. Niknejad of California, Berkeley (19)and 30 both transistors h Fig. 2. • • • Ali M. Niknejad QVCO in the Literature Schematic view of the quadrature VCO presented in [7]. To see how the (SSB) upconvers so that the overal very difficult to m into the ratio of th image band [to b In the case of the of 0.1% between IBR of 70 dB for and to 49 dB for Source: [Andreani] quickly larger wh is weak Several toquadrature the QVCO have inbeen proposed P-QVCO for Fig. 3. modifications Schematic view of the VCO proposed this work. improved performance easy to check that In particular, it has been shown that if a weaker coupling isdecreasing . Th The present analyzes alternative waycircuit [16] oflocks cross closer noise performanc introduced, thepaper phase noise an improves (the differential VCOs obtain a QVCO, in which quadrature the error performanc tocoupling the tanktworesonance), but attothe cost of imperfect is placed in series with , rather P-QVCO present coupling transistor generation in parallel (Fig. quadrature 3). This choice is motivated by the connoise FoM, the h A than series connected generation scheme proposed in the phase P-QVCO is responsible for a large choosing that has better bysideration [Andreani] noise performance. contribution to the overall phase noise, and connecting to unity. Slide: 31 University of California, Berkeley in series with , in a cascode-like fashion, should greatly CK out,I Static Frequency Dividers /0102&34536$-0'780)$9:(;<=(<$-(8<)&'> CK out,Q tD Q 23456738 !"#$%&' !"#$%&1 ( ( ) *+, -. -. ( ) *+, ) /0# ) /0# Fig. 5.22 (a) Typical static divider, (b) its waveform #-. -. dition. As frequency goes up, the idle time d /0# VDD ! !"#$%&%'()*%+,%-".(/$&$'$0-(1.("20"3$-4(560(275"#%'( R R This is a simple divider structure uses a • master/slave 8$9%9:(7(*%4$'5%*;($-(-%475$&%()%%/17"3 topology. ! <75"#%'(=7.(1%($=>2%=%-5%/($-(&7*$0,'(67.'( Two latches are cascaded into a • negative 7""0*/$-4(50('>%%/?>06%*(*%+,$*%=%-5' feedback loop (the output will D • therefore toggle). Since two clock cycles are required to pass the data from one latch to the next, it naturally divides by 2. D in M 1 D M2 Dout M3 D in M4 CKin CKin M6 M5 !"#"$%&''()) Ali M. Niknejad RC CKin !*+$,-. University of California, Berkeley (a) Slide: 32 device capacitance of the former can be absorbed as part of the latter. This topology thus becomes attractive and popular at moderate to high frequencies. Miller Divider (Regenerative) ! in , 3 ! in 2 x (t ) 2 LPF y (t ) ! in 2 output signal of the feedback loop is mixed with the input • The signal. The output of the (ideal) mixer has the sum and Fig. 5.25 Regenerative divider. components. Let usdifference first examine the division behavior and estimate the operation range. For the difference component (due tounity. the LPF) the circuitOnly to divide properly, the loop gain atis!amplified Redrawing in /2 must exceed and up. with simple RC filter and input amplitude A, we have the divider in gained Fig. 5.26(a) • the loop gain is greater! than one and the loop phase is zero • Ifdegrees, ! "A ! ! !the input. the system regenerates )! ≥ 1, (5.36) !H( j • The output frequency2in steady2 state must therefore satisfy: in where " denotes the conversion gain= of f the It can be further derived that fout − fout inmixer. " # 1 $2 2fout = !infin 2 A≥ 1+ 2 ≥ . (5.37) " 2 !c " Ali M. Niknejad University of California, Berkeley Slide: 33 " xy L C R ! in t = B cos M5 2/ " y (t ) ! in t M6 M ! in4 M3 Miller Divider Circuit Details ! in 2 2 Vin 2! n Op. Range A B 190 (b) (a) VDD L1 C1 M2 M1 L2 Vout L Y X C2 (c) VDD Vout L L M 6 with bandpass filter; (b) its input sensitivity, (c) typical CMOS M 5 Regenerative divider Fig. 5.28 (a) M3 M4 M4 M3 realization. Vin M M5 A M6 A B 5 B M4 !I inj M1 It is interesting to note input M 2that a mixer has two V M1 M 2 to two posVin in M 1ports, that leads sible configurations of Miller dividers. As illustrated in Fig. 5.29, the output could (c) ( (a) 8 (a) Regenerative divider with bandpass filter; (b) its input sensitivity, (c) typical CMOS ion. Fig. 5.30 (a) Type II regenerative divider, (b) redrawn to shown injecti LO Port LO Port s interesting to note that ports, that leads V ina mixer has two input be posredrawn as that inFilter Fig. 5.30(b). V It out in fact resembles an i Filter V outto two onfigurations of Miller dividers. As illustrated in Fig. 5.29, the output(which could will be discussed in the next section): M3 and M4 for RF Port LO Port Filter in RF Port Ali M. Niknejad V out LO Port Filter (a) and M5 and M6 appear as diode-connected transistors to lo and RFincrease Port the locking range. The differential injection in lieved to help V inenlarge the range of operation to some extent self-resonance frequency of the circuit if (W /L)3,4 > (W /L) V out (b) • Use a double-balanced Gilbert cell mixer to realize divider Fig. 5.29 Regenerative divider with the output fed back to (a) RF port, (b) LO port. RF Port V in 5.8 Injection-Locked Dividers University of California, Berkeley Slide: 34 Injection Locked Dividers [IL-Div] Fig. 2. Model for a free-running LC oscillator. . terms with freq an th-order s intermodulation o equal to which is the co written as injection locked divider is nothing but an injection locked • An system where the input frequency is at a harmonic of the freeFig. 3. Model for an injection-locked oscillator. Using a comp running frequency of the oscillator. Measurements on a single-ended ILFD (SILFD) are compared and applying th Model the system a non-linearity and a bandpass with simulations. The as simulation results off(e) a differential ILFD be written as transferare function H(ω). (DILFD) reported as well. • that the free-running loop has a stable oscillation • Assume frequency. The system is injection locked to a super-harmonic • Ali M. Niknejad II. MODEL FOR INJECTION-LOCKED OSCILLATORS of the free-running frequency. An non-linearity LC oscillator in can modeled a nonlinear block The thebeloop must as create intermodulation , followed selective block (e.g., an RLC or products thatby falla infrequency the passband of the loop. , in a positive feedback loop as shown in Fig. 2. tank) The nonlinear block models all the nonlinearities in the oscillator, including any amplitude-limiting mechanism. To University of California, Berkeley have a steady-state oscillation, a loop gain of unity should be Slide: 35 IL-Div Analysis the injection signal is a sinusoidal signal which is • Suppose added to the oscillator’s signal (at a sub-harmonic of the injection). The non-linearity acts on both signals and is filtered by the RLC circuit: vi (t) = Vi cos(ωi t + φ) vo (t) = Vo cos(ωo t) u(t) = f (e(t)) = f (vo (t) + vi (t)) H0 H(ω) = r 1 + j2Q ω−ω ωr can be shown that if the output signal contains various • Itharmonics and intermodulation terms which can be written as: u(t) = ∞ ! ∞ ! Km,n cos(mωi t + mφ) cos(nω0 t) m=0 n=0 • where K m,n Ali M. Niknejad is the intermodulation component of f(vi+vo) University of California, Berkeley Slide: 36 IL-Div Fourier Analysis • To see this, write the signals in the following form: vi = Vi cos(β) vo = Vo cos(α) f (e) = f (vi + vo ) means that f is periodic in α and β. For every β define a • which periodic function g(α) as follows g(α) = f (vo + Vi cos(α)) • Note that: g(α + 2π) = g(α) g(−α) = g(α) • So that we can write: g(α) = ∞ ! m=0 Ali M. Niknejad Lm (β) cos(mα) ! 2π 1 L0 (β) = f (Vo cos(β) + Vi cos(α))dα 0 ! 2π 2π 1 Lm (β) = f (Vo cos(β) + Vi cos(α)) cos(mα)dα π 0 University of California, Berkeley Slide: 37 IL-Div [Fourier Analysis cont.] • But since L m Lm (β) = is a periodic and even function of β we can write: ∞ ! Km,n cos(nβ) n=0 • which results in: f (vi + vo ) = ∞ ! ∞ ! ! 2π 1 Km,0 = Lm (β)dβ 2π 0 ! 2π 1 Km,n = Lm (β) cos(nβ)dβ π 0 Km,n cos(mα) cos(nβ) m=0 n=0 that the tank filters out all frequencies except the • Assume ones around ω . That means that only intermodulation terms o that fall at ωo are relevant: |mωi − nω0 | = ω0 Ali M. Niknejad University of California, Berkeley Slide: 38 IL-Div [Fourier cont] the injection signal is an N’th superharmonic, then only the • Ifintermodulation terms n = Nm ± 1 a frequency equal to 1/N the incident frequency. This • possess means that our summation can be written in terms of m as ∞ ! 1 uω0 (t) = K0,1 cos(ω0 t) + Km,N m±1 cos(ωo t + mφ) 2 m=1 complex notation and applying the oscillation condition, • Using the output signal can be written as ! ∞ " jωo t H e 1 0 jω0 t jmφ vo = Vo e = K + K e 0,1 m,N m±1 ∆ω 2 1 + j2Q ωr m=1 # % ! " ∞ 1 $ ∆ω Vo 1 + j2Q = H0 K0,1 + Km,N m±1 ejmφ ωr 2 m=1 Real Part: Imag Part: Ali M. Niknejad ! ∞ " # 1 Vo = H0 K0,1 + Km,N m±1 cos(mφ) 2 m=1 ∞ ∆ω H0 ! 2Vo Q = Km,N m±1 sin(mφ) ωr 2 m=1 University of California, Berkeley # Slide: 39 IL-Div Locking Range two equations can be solved for the unknown oscillation • The amplitude and phase for any incident amplitude and incident frequency, or any offset frequency: ∆ω = (ωi /N ) − ωr be re-written as: • The second equation can ! ∆ω = ∆ωA H0 2Vi ∞ " # Km,N m±1 sin(mφ) m=1 • where Adler’s locking range has been identified: ω r Vi ∆ωA = 2Q Vo • Unlike static dividers, the locking range is limited. Ali M. Niknejad University of California, Berkeley Slide: 40 IL Divide by 2 Circuit equations can be solved analytically for a divide by 2 with • The cubic non-linearity ( )= 0+ 2 3 + + 1 2 3 2Q ∆ω sin(φ) = H0 a2 Vi ωr ! ! ! ! ! ∆ω ! ! H0 a2 Vi ! !<! ! | sin(φ)| < 1 → !! ωr ! ! 2Q ! ! " # $% 4 1 3 Vo = 1 − H0 a1 + a3 Vi2 + a2 Vi cos(φ) 3 a3 H0 2 range is improved by using a large H0/Q or a larger • Locking injection amplitude. For an LC oscillator, this is equivalent to using a larger inductor: H0 = ωL Q impedance node is a convenient place to inject the signal • Ato high limit the injection power. Ali M. Niknejad University of California, Berkeley Slide: 41 Fig. 7. Schematic of the single-ended injection-locked frequency divider. IL-Div Circuit Details a first-order differential (23) to the output phase (23) ig. 6, it is clear that an ion as a first-order PLL. shaped by the low-pass unction, and the output he incident signal within . However, unlike a of an ILO is a function ger for a larger incident Fig. 8. Schematic of the differential injection-locked frequency divider. injection signal is applied as a current to the tail of a cross• The coupled differential oscillator. ity, the biasing circuitry is not shown in this figure. A Colpitts nsfer function is a little oscillator currents forms the core of the SILFD. The incident signal is The transistor of M1/M2 are a non-linear function of e ILO itself. Within the • injected into (feedback) the gate of M1. Transistors M1 and current M2 are used output signal and the injected of M3. LO is suppressed the by the in cascode, mainly to provide more isolation between the input dent power. Outside the Interestingly, evenTransistor in the absence oftoanbeinjection signal, node • and output. M2 is sized smaller than M1 by n increases by 20 M3 dB per is moving at twice ω phase noise region is Ali M. Niknejad o to reduce the parasitic capacitance almost a factor of three at the output node (drain of M2). As a result, a larger University of California, Berkeley Slide: 42 little phase synchronization occurs. The lock range in this case is the mixer conversion gain) at direc thebe zero crossings theorinput at into to injection can obtained fromof(6) (8): fall on the peaks of the output, oscillator. If node and is equivalent switch abruptly and theofca little phase synchronization occurs. The lock range in this case nodeis the conversion gain) at, and (9) can dir is mixer neglected, then be NG IN OSCILLATORS 1417 can be obtained from (6) or (8): oscillator. If and switch abruptly and the c (9) , and (9) can b node is neglected, then IL-Div Intuitive Picture (9) The subtle difference between (6) and (9) plays a critical role in If referred to the input, this range must be doubled: quadrature oscillators (as explained below). The difference betweenphase (6) and (9) playsacross a critical Fig.subtle 4 plots the input-output difference therole lockin If referred to the input, this range must be doubled quadrature oscillators (as explainedinjection below). locking to range. In contrast to phase-locking, Fig. 4 plots the input-output phase difference across the lock mandates operation away from the tank resonance. range. In contrasttotoQuadrature phase-locking, injection With locking 1) Application Oscillators: theto aid of a mandates operation frommodel the tank resonance. feedback model [13] or aaway one-port [14], it can be shown Confirmed by simulations, (13) represents the uppe Application to Quadrature Oscillators: With oscillators the aid of a that1) “antiphase” (unilateral) coupling of two identical the lock range of injection-locked dividers. feedback model [13] or a one-port model [14], it can be shown forces them to operate in quadrature. It can also be shown [14] Confirmed by simulations, (13) represents the up thatthis “antiphase” (unilateral) coupling of twoshifts identical oscillators the lock range of injection-locked dividers. that type ofIntuitively coupling (injection locking) the frequency we canItsee thatbe the injection at the tail III. of Ithe MOS forces them to operate in quadrature. can also shown [14] NJECTION PULLING from resonance so that each tank produces a phase shift of is mixed due to the switching action of M1/ that this type cross-coupled of coupling (injectionpair locking) shifts the frequency the injected signal out of, but n M2. III. frequency INJECTIONlies PULLING Fig. each 5. (a) Injection-locked divider. (b)of Equivalent If circuit. from resonance so that tank produces a phase shift from the lock range, the oscillator is “pulled.” We (10) If theby injected frequency lies out of,osci but For a strong mixing signal, 2/π component ofcomputing thissignal current will behavior the output phase of an from theinjection. lock range, of the ω oscillator flow into the tank, at frequency shift of by harmonics o. In is “pulled.” W low-level thealock range exceeds (9) 3.3%. In other (10) behavior by computing output particular, aforsignal at down-converted intotheω o will be o, phase of an osc where denotes the current injected by 2 oneω oscillator into difference between its input and output, words, a 90 phase low-level injection. the tank has high impedance. This signal therefore A. Phase Shift Through a Tank is an theinjection-locked current produced by the core of each the other and where oscillator need not operate at the edge of the . where Equations (5) denotes the current injected by onefrequency oscillator into oscillator. Use of (5) therefore gives the required shift experiences a large loop gain and can For lock the oscillator. subsequent derivations, we need an express lock range. A. Phase Shift Through a Tank is the current produced by the core of each the other and as phase shift introduced by a tank in the vicinity of re 2) Application to Dividers: shows an injecoscillator. Use of (5) therefore gives the required frequencyFig. shift 5(a) For subsequent derivations, we need an ,expre second-order parallel tank consisting of , and stage [15]. While tion-locked oscillator operating as a phase as by a tank in the vicinity of (11) a phase shift shift introduced of nonlinear function to consisting of , , and (8) previous work has treated the circuit as asecond-order parallel tank derive the lock range [15], it is possible adoptshift a time-variant (11) to a phase of (9) would imply that each oscillator is pushed Ali M.Interestingly, Niknejad University of California, Berkeley Slide: 43 • • view to simplify the analysis. Switching at a rate equal to the Miller Divider / Injection Locked Divider 190 Jri Lee VDD VDD L Vout L L M4 M3 M4 M5 M6 A M5 B Vin M 1 M2 (a) L M3 M6 !I +I inj M1 inj M2 Vin (b) Fig. 5.30 (a) Type II regenerative divider, (b) redrawn to shown injection locking. same circuit topology can be seen to be an injection • The locked divider or a Miller divider. devices M5/M6 are not needed in the normal injection • The locked divider, but here they act to lower the Q of the tank, be redrawn as that in Fig. 5.30(b). It in fact resembles an injection-locked divider (which will be discussed in the next section): M3 and M4 form a cross-coupled pair, and M5 and M6 appear as diode-connected transistors to lower the Q of the tank and increase the locking range. The differential injection in such a manner is believed to help enlarge the range of operation to some extent. It is possible to find a self-resonance frequency of the circuit if (W /L)3,4 > (W /L)5,6 [22]. • Ali M. Niknejad increasing the lock range. A Miller divider can be designed so that it does not oscillate in 5.8 Injection-Locked the absence of an inputDividers signal. The operation speed of dividers can be further boosted up if we simplify the structure at the circuit level. Since a cross-coupled VCO provides ultimate simplicity in generating differential oscillation, may think of injecting a periodic signal (apUniversity ofone California, Berkeley Slide: 44 References [Razavi] B. Razavi, “A Study of Injection Locking and Pulling in Oscillators,” IEEE J. of Solid-State Circuits, vol. 39, no. 9, Sept. 2004, pp. 1415-1424. [Adler] R. Adler, “A Study Locking Phenomena in Oscillators,” Proc. of the I.R.E. and Waves and Electronics, June 1946, pp. 351-357. [Paciorek] [Tsai] K.-C. Tsai and P. R. Gray, “A 1.9-GHz, 1-W CMOS Class-E Power Amplifier for Wireless Communications,” JSSC, vol. 34, July 1999 [Chee] Y. H. Chee, Ultra Low Power Transmitters for Wireless Sensor Networks, Ph.D. Dissertation, U. C. Berkeley, Spring 2006. [Deng] Z. Deng, A.M. Niknejad, “9.6/4.8 GHz dual-mode voltage-controlled oscillator with injection locking,” Electronics Letters, vol. 42, Nov. 9 2006, pp. 1344-1343. [Mahony] Frank P. O’Mahony, 10GHZ GLOBAL CLOCK DISTRIBUTION USING COUPLED STANDING-WAVE OSCILLATORS, Ph.D. Dissertation, 2003, Stanford University [Rofoug] A. Rofougaran et al, “A Single-Chip 900-MHz Spread-Spectrum Wireless Transceiver in 1- m CMOS—Part I: Architecture and Transmitter Design,” IEEE J. of Solid-State Circuits, vol. 33, no. 4, April 1998, pp. 515-533. [Andreani] [Rategh] Ali M. Niknejad L. J. Paciorek, “Injection Locking of Oscillators,” Proc. of the IEEE, vol. 53, no. 11, Nov. 1965, pp.1723-1728. P. Andreani et al, “Analysis and Design of a 1.8-GHz CMOS LC Quadrature VCO”, IEEE J. of Solid-State Circuits, vol. 37, no. 12, Dec. 2002, pp. 1737-1747. H. Ratech and T. H. Lee, “Superharmonic Injection-Locked Frequency Dividers,” IEEE J. of Solid-State Circuits, vol. 34, no. 6, June 1999, pp. 813-821. University of California, Berkeley Slide: 45