This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 0, NO. 0, 2012 1
Modeling and Implementation of an All Digital
Phase-Locked-Loop for Grid-Voltage Phase Detection
Hua Geng, Member, IEEE, Jianbo Sun, Shuai Xiao, Student Member, IEEE, and Geng Yang, Senior Member, IEEE
Abstract—In this paper, a novel all digital phase-locked-loop
(ADPLL) is proposed for the phase detection of power grid
voltage. The proposed ADPLL features wide track-in range and
fast pull-in time, and it can be e asily integrated into the digital
controller with low cost. The nonlinear model of such digital
system is derived from the operation principle of the ADPLL and
the linearized model is then obtained to evaluate the steady and
dynamic performance due to the nonlinear and discrete property.
Compared with the conventional DPLL, the proposed ADPLL has
almost no steady state phase error when the frequency of the input
signal deviates from the center frequency. Moreover, the tracking
speed is highly improved. In contrast with other AD PLLs, the
proposed one employs the loop filter with proportional-Integral
(PI) structure, which can help to eliminate the steady state error
excited by high-order disturbancesornoise.Asaresult,alow
system clock or sampling frequency is enough to get a satisfactory
performance, which is an attractive advantage for the low cost ap-
plications. Simulation and experimental results verify the analysis
and the effectiveness of the ADPLL.
Index Terms—All digital phase-locked-loop (ADPLL), digital
system modeling, phase detection.
I. INTRODUCTION
P
HASE-LOCKED LO OP (PLL) technique has contributed
significantly toward the advancement in power converter
based control s ystem s since the 1970s [1]–[3]. In the motor
servo control system, PLL can be employed to track the rotor
position according to the resolver signals [1]. In the grid-con-
nected power conversion systems, such as wind energy con-
version system (WECS) [4] or flexible ac transmission system
(FACTS) [5], PLL is crucial to extract the frequency and angle
information of the utility voltage for the s ynchronization of the
inverter ou tput voltages [6].
Manuscript received October 27, 2011; revised January 16, 2012, May 01,
2012, and June 20, 2012; accepted July 09, 2012. Date of publication July 20,
2012; date of current version nulldate. This work was sponsored by National
Natural Science Foundation of China (61104046), Tsinghua University Initia-
tive Scientific Research Program and grants from the Power Electronics Science
and Education Development Program of Delta Environmental & Educational
Foundation. Paper no. TII-11-679.
H. Geng, S . Xiao and G. Yang are with Automation Department, Tsinghua
University, Beijing 100081, China (e-mail: geng hua@tsinghua.edu .cn; xi-
aoshuai1000@126.com; yanggeng@tsinghua.edu.cn).
J. Sun was with Automation Department, Tsinghua University, Bei-
jing 100081, Chin a. He is now w ith the Departmen t o f Electrical Engi-
neering, Univer sity of California, Lo s Angeles, CA 90 095 USA (e-m ail:
jsun.thu@gmail.com).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.i eee.org.
Dig
ital Object Identifier 10.1109/TII.2012.2209666
Fig. 1 . Basic str ucture of the PLL.
The basic structure of any PLL is shown in Fig
. 1 [7]– [12].
It consists of three blocks: a phase detect
or (PD ), a loop filter
(LF) and a voltage controlled oscilla
tor (VCO). The input of the
PLL is a continuous-time sinusoidal
signal or the zero crossing
signal
, which is compared w ith
output of the VCO
in the PD. The high-order harmonics
in the phase error
can be filtered out with the LF and th
e output
modifies
the frequency of the VCO. The feed
back action is such that the
phase error is forced to zero. Wh
en zero error is achieved, the
input signal is sampled exac
tlyatzerocrossingsandthePLL
is lo cked. All PLLs have thi
s basic structure and differ mainly
through the various method
s in the impl emen tati on of the PD
[7].
For grid application s, PL L
s are usu ally implemented inside
a higher level control alg
orithm. The y should b e designed with
the goals of having a go
od performa nce and not being exces-
sively resource consu
ming [8]. However, tradeoff alw ays exists
between the performa
nce and the complex ity of the PLLs. In
order to improve the
dynamic resp onse and p recision, different
PDs are proposed to c
onstruct the PLL.
Transformation-ba
sed PDs are common especially in multi-
phase systems, with
the synchronous reference frame (SRF)
[12]–[14], being
the m ost popular. By reducing the band width,
the SRF based PLL
can also operate in grid unbalanced condi-
tions but this is
an inefficient solution [15]. The PDs can also
be designed at t
he fixed reference frame (FRF) [16]. So as to
alleviate the
effect of harmon ic distortion that is present in the
grid voltage,
the adaptive notch filters (ANFs), also referred to
as second-o
rder generalized integrators (SOGIs), are employed
to reconstr
uct the pu re sinusoidal signal and estimate the funda-
mental fre
quency [6], [16]. In single-phase systems, these trans-
forms are m
ore complex du e to the lack of multiple independent
input var
iables [9]. As such, methods to g enerate secondary or-
thogonal
signals range from transport delay filters to compl e x
integra
tive feedback networks [17], [18]. Accuracies in these
method
s all rely on the precision of the generated orthogon al
signa
l [7]. To avoid such disadvantage, product-type or mixer
PD sys
tems, which rely on the pr oduct of two sinusoidal signals
to pr
oduce trigonometric relationships between the summation
and d
ifference of the input and estimated signal, are applied to
the
single-phase systems [7].
1551-3203/$31.00 © 2012 IEEE