Circuits PLL
Perspective Components
Fig. 050-01
Introduction - Continued
Outline:
Operating Principles of PLLs
Classification of PLL Types
Pertinent References:
1. F.M Gardner, Phaselock Techniques, 2nd edition, John-Wiley & Sons, Inc., New York,
1979.
2. B. Razavi (ed.), Monolithic Phase-Locked Loops and Clock Recovery Circuits, IEEE
Press, 1997.
3. R.E. Best, Phase-Locked Loops: Design, Simulation, and Applications, 4th edition,
McGraw-Hill, 1999.
4. Recent publications of the IEEE Journal of Solid-State Circuits.
vin(t) in
vd(t) Loop vc(t)
Phase
osc Filter
Detector
out
o Ko
Voltage 1
vosc(t)
Controlled vc(t)
Oscillator
Fig. 01-02
vin(t) The input or reference signal
in The radian frequency of the input signal
vosc(t) The output of the VCO
osc The radian frequency of the VCO
vd(t) The detector output voltage = Kde
e Phase error between vin(t) and vout(t) = in - osc
vc(t) The output voltage of the loop filter and the control voltage for the VCO
PLL Operation
Locked Operation:
The loop is locked when the frequency of the VCO is exactly equal to the average
frequency of the input signal.
If the input signal has noise, the phase locked loop will remove much of the noise on the
input signal.
To maintain the control voltage needed for locked conditions, it is generally necessary
for the output of the phase/frequency detector to be nonzero.
Unlocked Operation:
The VCO runs at a frequency called the free running frequency, o, which corresponds
to no control voltage.
The capture process is the means by which the loop goes from unlocked, free-running
state to that of the locked state.
t
in
o
t
osc
to t
Fig. 01-03
The digital PLL (DPLL) has been the mainstay of most PLLs and is called the classical
digital PLL.
Input Error
Signal Signal Analog
Analog
Loop
Multiplier
Filter
Oscillator
Output Controlling
Voltage Voltage
Signal
Controlled
Oscillator Fig. 01-04
Input
Signal Error
Analog Controlling
Digital Signal
Loop Voltage
Detector
Filter
Oscillator
Output
N Signal Voltage
Counter Controlled
(optional) Oscillator
Fig. 01-05
Input Digital
Signal Error
Digital Signal Digital Loop
Detector Filter
Oscillator Controlling
Output Digital Signal
Digitally
Signal
Controlled
Oscillator
Fixed
Oscillator
(Clock) Fig. 01-06
Input Output
Signal Analog- Software Digital- Signal
Digital PLL Analog
Converter Converter
Clock
Fig. 01-07
Roadmap
;;;;;
Systems
Perspective
;;;;; Types of PLLs
and PLL Measurements
PLL
Applications and Examples
Circuits
Perspective
;;;;;;;;;
;;;;;;;;;
PLL
Components
Fig. 050-02
PLL COMPONENTS
Building Block of the LPLL
v1(t) 1
Phase
vd(t) Low Pass
2 Detector
Filter
(Multiplier)
Loop Filters
In the PLL, there are many high frequencies including noise that must be removed by
the use of a low pass filter in order to achieve optimum performance.
Types of Loop Filters:
1.) Passive lag filter (lag-lead)
1 + s2
F(s) = 1 + s(1 + 2) where 1 = R1C and 2 = R2C
|F(j)| dB
R1
0dB
+ + -20 dB/decade
R2
Vd Vf
C 2
- - 1+2 dB
1 1 log10()
Fig. 2.1-03 1+2 2
Pole is at 1/(1+2) and the zero at 1/2.
Since the pole is smaller than the zero, the filter is lag-lead
Passive filters should have no amplitude nonlinearity
+ + -20 dB/decade
-
Vd + Vf
2 R2
- - = dB
1 R1
1 log10()
Fig. 2.1-05 2
Has large open loop gain at low frequencies Large hold range
Limited by the linearity and noise of the op amp
Gain limits at the op amp open loop gain
Stability:
To keep the loop stable, it is important to pick the loop filter so that it does not
introduce more than a 90 phase shift in the loop.
ECE 6440 - Frequency Synthesizers P.E. Allen - 2003
Lecture 050 Linear Phase Lock Loops - I (5/14/03) Page 050-19
Phase Signals
It is important to remember that frequency and phase are related as
d
dt = = dt
Transfer functions:
V2(s)
H(s) = V1(s)
where V2(s) and V1(s) are the Laplace transforms of v2(t) and v1(t).
Next, we consider some simple phase signals that are used to excite a PLL.
1(t)
t
Fig. 2.1-06
v1(t)
2.) A step frequency change assuming that 1(t) = o
for t < 0. We may express v1(t) as,
t
v1(t) = V10 sin[ot + t]
= V10 sin[ot + 1(t)]
1(t)
1(t) = t
(the phase becomes a ramp signal) t
1(t)
t
Fig. 2.1-07
ECE 6440 - Frequency Synthesizers P.E. Allen - 2003
Lecture 050 Linear Phase Lock Loops - I (5/14/03) Page 050-21
t 2
1(t) = 2 t
Fig. 2.1-08
2(t) = 2 dt = Kovf(t)dt
1 + s2 KoKd +
1 2
F(s) = 1 + s(1 + 2) H(s) = 1+ KoKd2 KoKd
2
s + s 1+2 + 1+2
10
= 5.0
0
= 3.0
|H(j)| dB
-10
= 0.3
= 0.5
-20
= 0.707
= 1.0
-30
-40
0.1 1 10 100
n Fig. 2.1-10
t2
However, 1(t) = 2 1(s) = s3
Phase error:
s2
e(s) = He(s) s3 = s3(s2 + 2 s + 2) = s(s2 + 2 s + 2)
n n n n
e(t) = L-1[e(s)] = 2 - 2 cos 1- 2nt + sin 1- 2 t e-nt , < 1
n
n n 1+2
= 2 - 2 (1 + nt)e-nt, =1
n n
= 2 - 2 cosh 2-1nt + sinh 1- 2-1 t e-nt , > 1
n
n n 2-1
Steady state error:
t
e() = lim se(s) = 2 (High loop gain) e() = K K F(0)2 + 2 (Low loop gain)
s0 n d o n
First-Order PLL
A firstorder PLL occurs when F(s) = 1. From previous results we have,
2(s) KoKd s
H(s) = (s) = s + K K Also, H e (s) = 1- H(s) = s + KoKd
1 o d
The 3dB bandwidth of H(s) is KoKd.
Comments:
F(s) causes the 3dB bandwidth to be reduced in higher-order systems which means
that the first-order PLL has a wider bandwidth
The hold range of the first-order PLL will be larger than for higher-order PLLs
The first-order PLL will track the signal variations more quickly than higher-order
PLLs
The first-order PLL does not suppress noise superimposed on the input signal to the
extent of higher-order PLLs.
Higher-Order PLLs
Comments:
Generally F(s) has a pole and a zero in order to get better noise rejection without
sacrificing speed.
Open Loop Gain
-20dB/decade
-40dB/decade First-order
Bandwidth
0 dB
Pole Zero
-20dB/decade
Fig. 2.1-11
If the phase shift of the open loop system is more than 90, the stability of the loop may
be poor ( is small).
(To be continued in Lecture 060)