Glitch-free digital phase detector circuits and methods with...

Miscellaneous active electrical nonlinear devices – circuits – and – Specific signal discriminating without subsequent control – By phase

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C327S012000

Reexamination Certificate

active

06809555

ABSTRACT:

FIELD OF THE INVENTION
The invention relates to phase detector circuits and methods. More particularly, the invention relates to digital phase detector circuits and methods having optional lock window offset and lock window extension features.
BACKGROUND OF THE INVENTION
Phase detector circuits are commonly used in today's digital integrated circuits. For example, a phase detector circuit is included in virtually every phase lock loop (PLL) and delay lock loop (DLL). A phase detector circuit is a circuit that compares the corresponding edges (e.g., rising edges) of two input clock signals, and provides two output signals indicating the phase relationship between the two input clock signals. The two output signals (sometimes called “Up” and “Down”, or “Add” and “Subtract”) indicate whether a feedback clock signal needs to be delayed or moved forward in order to be in phase with a reference clock signal. Some phase detector circuits also supply a third output signal (e.g., “Lock”) that indicates when the two input clock signals are in phase and no adjustment is needed.
FIG. 1
is a block diagram of an exemplary system that includes a phase detector circuit. System
100
includes a clock adjust circuit
101
(e.g., a DLL in the pictured embodiment), a clock network
102
, circuit elements
103
, and phase detector circuit
104
. Clock adjust circuit
101
includes a finite state machine
112
, a delay chain including delay elements D
0
-Dn, and a multiplexer
111
. The input clock signal CLKIN is provided to state machine
112
, to the first element D
0
in the delay chain, and to multiplexer
111
. State machine
112
controls multiplexer
111
via select signal SEL to select one of the delayed clock signals from the delay chain. Thus, the amount of delay through clock adjust circuit
101
from input terminal CLKIN to output terminal CLKOUT depends on the state of state machine
112
, which in turn depends on the signals ADD and SUB from phase detector
104
.
Output signal CLKOUT from clock adjust circuit
101
is provided to a clock network
102
, which drives the various circuit elements
103
included in the system. Clock network
102
also provides feedback clock signal FBCLK to phase detector circuit
104
. Phase detector circuit
104
compares the phase of feedback clock signal FBCLK to the phase of input clock signal CLKIN, providing signals ADD and SUB. Signals ADD and SUB, in turn, control state machine
112
to either add more delay or shorten the delay through clock adjust circuit
101
, as necessary to align corresponding edges of the two clock signals FBCLK and CLKIN. Phase detector circuit
104
is clocked by input clock signal CLKIN, and can optionally be reset by signal RST, which can also be used to reset state machine
112
.
Optional output signal LOCK from phase detector circuit
104
can be used for various purposes, e.g., to enable and disable additional circuitry. For example, a system can include two clock adjust circuits coupled in series. The first clock adjust circuit can perform a coarse phase adjustment, and the second clock adjust circuit can perform a fine phase adjustment. The LOCK signal from the first clock adjust circuit can be used to enable the second clock adjust circuit only after a coarse lock has been achieved.
FIGS. 2-4
illustrate three value combinations for signals LOCK, ADD, and SUB. In
FIG. 2
, input clock signal CLKIN and feedback clock signal FBCLK are in phase. In this example, being “in phase” means the rising edges of the two clocks occur at the same time. Thus, the relative phase of the two signals is “locked”, and the delay through the clock adjust circuit need not be changed. Signal LOCK is high (
1
), and signals ADD and SUB are low (
0
).
In
FIG. 3
, the rising edge of the feedback clock precedes the rising edge of the input clock by a time −Delta. Thus, additional delay needs to be added to the path through the clock delay circuit to bring the feedback clock signal FBCLK into phase with the input clock signal CLKIN. Thus, signal LOCK is low, signal ADD is high, and signal SUB is low. Because the clock signals are periodic signals, the clocks could also be brought into phase by reducing the delay through the clock delay circuit. However, it is preferable to make the smallest adjustment that will have the desired effect.
In
FIG. 4
, the rising edge of the feedback clock follows the rising edge of the input clock by a time +Delta. Thus, some of the delay needs to be removed from the path through the clock delay circuit to bring the feedback clock signal FBCLK into phase with the input clock signal CLKIN. Thus, signal LOCK is low, signal ADD is low, and signal SUB is high. Because the clock signals are periodic signals, the clocks could also be brought into phase by increasing the delay through the clock delay circuit. However, as noted above, it is preferable to make the smallest adjustment that will have the desired effect.
Returning now to
FIG. 1
, in the ideal case clock adjust circuit
101
operates as follows. When reset signal RST is high, signal SEL selects the clock input signal CLKIN as output clock signal CLKOUT. When reset signal RST goes low, signal ADD goes high. Clock adjust circuit
101
continues inserting additional delay into the clock path between the CLKIN and CLKOUT terminals until feedback clock signal FBCLK is precisely in phase with input clock signal CLKIN. At that point, signal LOCK goes high and signal ADD goes low. State machine
112
holds the current selection in multiplexer
111
. However, if variations occur in temperature, power high voltage level, or the frequency of input clock signal CLKIN, the two edges can lose synchronization. In this case, one of signals ADD and SUB goes high, and signal LOCK goes low. Finite state machine
112
selects a differently delayed clock signal to provide to the CLKOUT terminal of clock adjust circuit
101
, until the feedback clock signal FBCLK is again in phase with the input clock signal CLKIN. At that point, signal LOCK again goes high, while signals ADD and SUB are both low.
The previous paragraph describes the behavior of the clock adjust circuit in the ideal case. However, in the ideal case the delay through the clock adjust circuit has infinitely small gradations and the two clock signals can be precisely synchronized. Thus, exactly one of signals LOCK, ADD, and SUB can be high at any given time. However, in reality, it is often true that the two clock signals are rarely precisely synchronized. Nevertheless, the LOCK signal must be allowed to signal a lock on the clock signal. Therefore, a “lock window”, is frequently provided, wherein the LOCK signal is active (e.g., high) whenever the feedback clock edge occurs within a predetermined time of the input clock edge. Hence, the LOCK signal can be high at the same time as one or the other of the ADD and SUB signals.
FIG. 5
illustrates the concept of a lock window. The lock window ranges from a time −Delta_syn preceding the input clock edge to a time +Delta_syn following the input clock edge. Whenever the corresponding edge (e.g., the rising edge) of the feedback clock signal FBCLK falls within this window, signal LOCK is high.
A small lock window is desirable when high accuracy is required. However, a small lock window makes the phase detector circuit sensitive to clock jitter. Therefore, some known systems include phase detector circuits having programmable lock windows. For example, one known DLL has a lock window that supports two predetermined values of Delta_syn. When the DLL first becomes active, Delta_syn has a first and lower value. Thus, the clock signals are precisely locked. However, as soon as the two clock signals are locked (e.g., signal LOCK goes high), the lock window is extended by increasing the value of Delta_syn. This feature reduces noise and power consumption by causing the DLL to go inactive again until the new extended window is exceeded.
Clearly, phase detector circuits are common and useful circuits. Therefore, it is desirable to provide improved

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Glitch-free digital phase detector circuits and methods with... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Glitch-free digital phase detector circuits and methods with..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Glitch-free digital phase detector circuits and methods with... will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-3281118

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.