Phase detector and method providing rapid locking of delay-lock loops

This invention relates to clock generating systems and methods, and, more particularly, to a phase detector that may be used in a delay-lock loop that operates in a manner that allows the loop to rapidly achieve a locked condition.

Periodic digital signals are commonly used in a variety of electronic devices. Probably the most common type of periodic digital signals are clock signals that are typically used to establish the timing of a digital signal or the timing at which an operation is performed on a digital signal. For example, data signals are typically coupled to and from memory devices, such as synchronous dynamic random access memory (“SDRAM”) devices, in synchronism with a clock or data strobe signal. More specifically, read data signals are typically coupled from a memory device in synchronism with a read data strobe signal. The read data strobe signal typically has the same phase as the read data signals, and it is normally generated by the same memory device that is outputting the read data signals. However, by the time the read data strobe signal has propagated to circuitry such as latches that are used to capture the read data signals, the timing of the read strobe signal may have varied from the desired timing. It is therefore sometimes necessary to adjust the timing of the read data strobe signal.

Write data signals are typically latched into a memory device in synchronism with a write data strobe signal. The write data strobe signal has a phase that is the quadrature of the write data signals so that a transition of the write data strobe signal occurs during a “data eye” occurring at the center of the period in which the write data signals are valid. The write strobe signal is typically generated by the memory controller from an internal clock signal that is derived from the system clock signal, and it is coupled to the memory device into which the data are being written. Unfortunately, the phase of the system clock signal is normally substantially the same as the phase of the write data signals. Therefore, it is necessary for the memory controller to generate the write data strobe signal as a quadrature signal having a phase that is 90-degrees relative to the phase of the internal clock signal. In other cases, a quadrature clock signal used for latching write data is generated in the memory device to which the data are being written. The quadrature clock signal is typically generated in the memory device from an internal clock signal that is also derived from the system clock signal. Other periodic signals are also sometimes used in integrated circuits, such as memory devices, for a variety of purposes.

Various techniques can be used and have been used by integrated circuits, such as memory controllers and memory devices, to control the timing of periodic signals, such as clock signals and strobe signals. For example, a delay-lock loop, such as the delay-lock loop 10 shown in FIG. 1 , can be used for this purpose. The delay-lock loop 10 includes a phase detector 14 to which a reference signal S R and a feedback signal S FB are coupled. The phase detector 14 compares the phase of the S R signal to the phase of the S FB signal, and generates either an “Up” output signal or a “Down” signal depending on the polarity of the phase comparison.

The “Up” and “Down” signals are applied to a charge pump 16 , which applies either a positive or negative current to a capacitor 18 to either increase or decrease a control voltage V C , which is generated across the capacitor 18 , respectively. More specifically, the charge pump 16 outputs a positive current to increase the control voltage V ctrl responsive to the “Up” signal, and it outputs a negative current to decrease the control voltage V ctrl responsive to the “Dn” signal.

The control voltage V ctrl is applied to a control input of a voltage controlled delay line “VCDL” 20 , which has a clock input receiving the S R signal. The VCDL 20 delays the S R signal to provide an output signal S OUT , which is coupled to the phase detector 14 as the S FB signal. The magnitude of the delay provided by the VCDL 20 is determined by the magnitude of the control voltage V ctrl . More specifically, the magnitude of the delay is inversely proportional to the magnitude of the control voltage V ctrl . As a result, the delay of the VCDL 20 decreases responsive to an increase in the magnitude of the control voltage V ctrl , and the delay of the VCDL 20 increases responsive to a decrease in the magnitude of the control voltage V ctrl .

In operation, the phase detector 14 and charge pump 16 adjust the magnitude of the control voltage V ctrl so that the phase of the S FB signal is substantially equal to the phase of the S R signal. For example, if the S R signal leads the S FB signal (i.e., the delay of the VCDL 20 is too large), the phase detector 14 detects the difference in phase and generates an “Up” signal, which causes the charge pump 16 to increase the magnitude of the control voltage V ctrl . This increase in the magnitude of the control voltage V ctrl causes the VCDL 20 to reduce the delay of the S R signal to reduce the phase of the S FB signal. The phase detector 14 and charge pump 16 continue to increase the magnitude of the control voltage V ctrl until the phase of the S FB signal is substantially equal to the phase of the S R signal. The delay-lock loop 10 operates in an opposite manner if the S R signal lags the S FB signal (i.e., the delay of the VCDL 20 is too small).

A variety of designs for the phase detector 14 can be used in the delay-lock loop 10 . The operation of one commonly used phase detector can be explained with reference to FIG. 2 , which shows the reference signal S R leading the feedback signal S FB derived from the output S OUT signal. In response to the rising edge of the S R signal, the phase detector 14 generates the “Up” signal, which terminates after a predetermined time delay. Similarly, the phase detector 14 generates the “Down” signal in response to the rising edge of the S FB signal, which terminates at the same time the “Up” signal terminates. The period of time that the “Up” signal is generated but the “Down” signal is not being generated causes a positive current to be applied to the capacitor 18 for the period represented by the shaded area shown in FIG. 2 . The positive current increases the control voltage V ctrl , which results in the phase of the S FB signal being reduced.

One limitation of the delay-lock loop 10 is the time required for the loop 10 to become “locked,” i.e., the phases of the S R and S FB signals substantially equaling each other. The loop 10 could become locked faster if the magnitude of the control signal V ctrl could be changed either more quickly or to a greater degree. The operation of a phase detector (not shown) that is able to cause the delay-lock loop 10 to lock earlier will be explained with reference to FIG. 3 . Like the phase detector explained with reference to FIG. 2 , the phase detector generates the “Up” signal responsive to the rising edge of the S R signal, and it generates the “Down” signal responsive to the rising edge of the S FB signal. However, rather than generating the “Up” and “Down” signals for a fixed period following the first of these signals to be generated, the phase detector generates the “Up” signal until the falling edge of the S FB signal, and it generates the “Down” signal until the falling edge of the S R signal. As a result, the period during which the “Up” signal but not the “Down” signal is generated is greatly increased thereby greatly increasing the magnitude of the V ctrl signal, as indicated by the shaded bar in FIG. 3 . The increased magnitude of the V ctrl signal causes the delay-lock loop 10 containing the phase detector to achieve lock more rapidly. In the event the S R signal lags the S FB signal, the phase detector operates in the opposite manner.

Although a phase detector operating as shown in FIG. 3 has the advantage of allowing a locked loop to achieve faster lock time, it has the disadvantage of allowing a fixed phase error between the phase of the S R signal and the phase of the S FB signal in the event the duty cycle of either the S R signal or the S FB signal, as will now be explained with reference to FIG. 4 . As shown in FIG. 4 , the S R signal lags the S FB signal. As with the example shown in FIG. 3 , the “Down” signal is generated on the rising edge of the S FB signal, and the “Up” signal is generated on the rising edge of the S R signal. The “Down” signal is generated prior to the “Up” signal since the S R signal lags the S FB signal. The period that the “Down” signal is generated but not the “Up” signal is not generated results in a negative current being applied to the capacitor during the period indicated by the shaded area shown in FIG. 4 . As a result, the magnitude of the V ctrl signal is decreased.

The duty cycle of the S FB signal is greater than the 50% duty cycle of the S R signal. As a result, the S R signal transitions low before the S FB signal transitions low even though the S R signal lags the S FB signal. As in the example shown in FIG. 3 , the falling edge of the S R signal terminates the “Down” signal and the falling edge of the S FB signal terminates the “Up” signal. The “Down” signal therefore terminates before the “Up” signal terminates. The period that the “Up” signal continues beyond the termination of the “Down” signal results in a positive current being applied to the capacitor 18 during the period indicated by the shaded area shown in FIG. 4 , thereby increasing the magnitude of the V ctrl signal. In fact, the period during which the positive current is applied to the capacitor 18 is equal to the period during which the negative current is applied to the capacitor 18 . The net change in the magnitude of the V ctrl signal is therefore zero. Under these circumstances the phase of the S FB signal is not corrected to match the phase of the S R signal, and the S R signal continues to lag the S FB signal. The difference in the duty cycle of the S FB signal from the duty cycle of the S R signal thus results in the maintenance of a permanent phase error between the S FB and the S R signals. As a result, the S OUT signal generated by the locked loop may not have the proper phase value.

There is therefore a need for a locked loop that is capable of achieving a rapid lock condition without allowing a phase error in the event the duty cycle of a reference signal is not equal to the duty cycle of a feedback signal generated by the locked loop.

A dual mode phase detector and method includes both a first phase detector and a second phase detector. The first phase detector generates at least one output signal indicative of the phase relationship between a first predetermined transition of a first input signal, such as its rising edge, and the first predetermined transition of a second input signal. The second phase detector generates at least one output signal indicative of both the phase relationship between the first predetermined transition of the first input signal and the first predetermined transition of the second input signal, and the phase relationship between a second predetermined transition of the first input signal, such as its falling edge, and the second predetermined transition of the second input signal. The dual mode phase detector also includes a selector that selects the at least one output signal from the first phase detector if the first and second input signals have a first phase relationship, and it selects the at least one output signal from the second phase detector if the first and second input signals have a second phase relationship that is different from the first phase relationship. The dual-mode phase detector may be used in a locked loop, such as a delay-locked loop. In such case, the dual edge phase detector is used to allow the loop to quickly achieve a locked condition. Thereafter, the single edge phase detector is used so that differences in the duty cycles of the first and second signals do not create fixed phase errors.

FIG. 1 is a block diagram of a typical delay-lock loop of conventional design.

FIG. 2 is a timing diagram illustrating the operation of one example of a phase detector used in the delay-lock loop of FIG. 1 .

FIG. 3 is a timing diagram illustrating the operation of another example of a phase detector used in the delay-lock loop of FIG. 1 , which has the advantage of achieving a locked condition more rapidly.

FIG. 4 is a timing diagram illustrating the manner in which the phase detector of FIG. 3 can allow a fixed phase error to exist in the event the duty cycles of the signals applied to the phase detector differ from each other.

FIG. 5 is a block diagram of a delay-lock loop using a dual mode phase detector according to one example of the present invention.

FIG. 6 is a block diagram of an example of a single edge phase detector that may be used in the dual mode phase detector of FIG. 5 .

FIG. 7 is a logic diagram of an example of a single edge phase detector that may be used as the single edge phase detector of FIG. 6 .

FIG. 8 is a logic diagram of an example of a dual edge phase detector that may be used in the dual mode phase detector of FIG. 5 .

FIG. 9 is a block diagram of a memory device using an output signal generated by a delay-lock loop using the dual mode phase detector of FIG. 5 or some other embodiment of the invention.

FIG. 10 is a block diagram of a processor-based system using the memory device of FIG. 9 .

A delay-lock loop 40 according to one example of the present invention is shown in FIG. 5 . The delay-lock loop 40 uses the same charge pump 16 , capacitor 18 and VCDL 20 that are used in the delay-lock loop 10 of FIG. 1 . These components operate in essentially the same manner in both loops 10 , 40 . Therefore, in the interests of brevity and clarity, an explanation of their function and operation will not be repeated.

The delay-lock loop 40 differs from the delay-lock loop 10 of FIG. 1 in the use of a dual mode phase detector 50 . The phase detector 50 includes both a single edge phase detector 54 , which operates as shown in FIG. 2 , and a dual edge phase detector 56 , which operates as shown in FIGS. 3 and 4 . Both of the phase detectors 54 , 56 receive the reference signal S R and the feedback signal S FB , and they generate the “Up” and “Down” signals in response. The “Up” and “Down” signals are applied to respective inputs of a switch 58 . The switch 58 selects the “Up” and “Down” signals produced by one of the phase detectors 54 , 56 and couples them to the charge pump 16 . The switch 58 is operated by a lock detector 60 , which receives the S R signal and the S FB signal.

In operation, before the lock detector 60 detects a locked condition in which the phase of the S R signal is within a predetermined range of the S FB signal, the lock detector 60 causes the switch 58 to couple the “Up” and “Down” signals from the dual edge phase detector 56 to the charge pump 16 . As explained with reference to FIG. 3 , the increased magnitude of the control signal V ctrl resulting from the use of the dual edge phase detector 56 causes the loop 40 to achieve a locked condition relatively quickly. However, as explained with reference to FIG. 4 , the dual edge phase detector 56 has the disadvantage of allowing a fixed phase error to exist in the loop 40 . This disadvantage is avoided in the phase detector 50 of FIG. 5 by the lock detector 60 detecting the locked condition of the loop 40 and causing the switch 58 to couple the “Up” and “Down” signals from the single edge phase detector 54 to the charge pump 16 . As explained above, the use of a signal edge phase detector 54 does not permit the loop 40 to establish a fixed phase error. Therefore, the dual mode phase detector 40 is able to achieve a rapid lock condition without allowing a fixed phase error to exist.

Although the switch 58 used in the dual mode phase detector 50 is shown in FIG. 5 as being a separate component, it may alternatively be incorporated into the charge pump 16 . Further, rather than using a switch or similar device to selecting the “Up” and “Down” signals from either the single edge phase detector 54 or the dual edge phase detector 56 , the single edge phase detector 54 and the dual edge phase detector 56 may both be coupled to the charge pump 16 at all times. Either the single edge phase detector 54 or the dual edge phase detector 56 , but not both, would then be enabled by the lock detector 60 or other device. By way of further example, rather than generating “Up” and “Down” signals, the phase detectors 54 , 56 could each generate a signal indicative of the phase relationship between the S R and S FB signals, thus making the charge pump 16 unnecessary. The switch 58 would then simply select the signal from one of the phase detectors. Other modifications will be apparent to one skilled in the art.

One example of a single edge phase detector 70 that may be used as the single edge phase detector 54 of FIG. 5 is shown in FIG. 6 . With reference to FIG. 6 , the phase detector 70 includes a first flip-flop 72 having its data input biased high and its clock input coupled to receive the S R signal. Therefore, the Q output of the flip-flop 72 transitions high responsive to the rising edge of the S R signal to generate the “Up” signal. The phase detector 70 also includes a second flip-flop 74 , which has its data input biased high and its clock input coupled to receive the S FB signal. Therefore, the Q output of the flip-flop 74 transitions high responsive to the rising edge of the S FB signal to generate the “Down” signal. However, the Q outputs of the flip-flops 72 , 74 are also applied to respective inputs of an AND gate 76 so that the output of the AND gate transitions high when both the S R and the S FB signal have transitioned high. The high output of the AND gate 76 resets the flip-flops 72 , 74 when both the S R and the S FB signal have transitioned high. Therefore, if the S R signal leads the S FB signal, the “Up” signal will be generated until the S FB signal transitions high, but no “Down” signal will be generated. Conversely, if the S FB signal leads the S R signal, the “Down” signal will be generated until the S R signal transitions high, but no “Up” signal will be generated.

In practice, the output of the AND gate 76 is coupled to the reset inputs of the flip-flops 72 , 74 through a delay circuit 78 . The delay circuit delays the termination of the “Up” and “Down” signals for a short period after the later generated of the “Up” and “Down” signals is generated. As a result, if the S R signal leads the S FB signal, the “Down” signal will be generated for a short period in addition to the “Up” signal being generated for a longer period. If the S FB signal leads the S R signal, the “Up” signal will be generated for a short period in addition to the “Down” signal being generated for a longer period. If the S R signal has the same phase as the S FB signal, i.e., the loop 40 is locked, a short “Up” signal and a short “Down” signal of equal duration will be generated. By generating short “Up” and “Down” signals even when the loop 40 is locked or nearly locked, any “dead band” in the operation of the loop 40 when the loop 40 is locked is avoided. Without the use of the delay circuit 78 , a small difference between the phases of the S R and S FB signals might not result in a control signal V ctrl being generated so that this phase difference would continue to exist.

One example of the single edge phase detector 70 ′ is shown in FIG. 7 . The first and second flip-flops 72 , 74 of the phase detector 70 ′ are implemented by respective circuits 72 ′, 74 ′ that are identical to each other. The first circuit 72 ′ includes a flip-flop 82 formed by a pair of cross-coupled NAND gates 80 , 82 . The S R signal is applied to the NAND gate 80 through an inverter 86 and a NAND gate 88 . Prior to the rising edge of the S R signal, the NAND gate 88 is enabled so that the low S R signal causes the NAND gate 88 to output a low A1 signal. The low A1 signal resets the flip-flop 82 so that the B1 output of the NAND gate 80 is high. The high at the output of the NAND gate 80 is applied to a NAND gate 90 , which implements a second flip-flop 84 in combination with the NAND gate 88 . The high at the output of the NAND gate 90 is also applied to an input of a NAND gate 92 , which then outputs a low since the other input of the NAND gate 92 normally receives an inactive high FUP signal. This low is applied through a pair of inverters 93 , 94 to generate an inactive low “Up” signal and it is applied through a pass gate 95 and an inverter 96 to generate an inactive high “Up_” signal. Thus, prior to the rising edge of the S R signal, the A1 signal is low and the B1 signal is high.

In response to the rising edge of the S R signal, the output of the inverter 86 transitions low thereby setting the flip-flop 84 , which causes the A1 signal at the output of the NAND gate 88 to transition high. As explained above, the B1 signal will also be high at this time, and, as will be explained below, the third input of the NAND gate 90 is high whenever the flip-flop 82 is not being reset. As a result, the output of the NAND gate 90 transitions low, thereby causing the NAND gate 92 to output a high. The “Up” and “Up_” signals then become active high and low, respectively. Thus, the “Up” and “Up_” signals become active responsive to the rising edge of the S R signal. The low at the output of the NAND gate 90 also disables the NAND gate 88 . However, the A1 signal at the output of the NAND gate remains high since the S R signal has transitioned high. Thus, after the rising edge of the S R signal, both the A1 signal and the B1 signal are high.

As previously mentioned, the circuit 74 ′ contains the same components as the circuit 72 ′, and they operate in the same manner to generate active “Down” and “Down_” signals responsive to the rising edge of the S FB signal. Therefore, in the interest of brevity, and explanation of the operation of the circuit 74 ′ will not be repeated.

The function of the delay circuit 78 ( FIG. 6 ) is performed by a NAND gate 97 and inverter 98 . When the later of the rising edge of the S R signal or the rising edge of the S FB signal occurs and the output of the NAND gate 76 ′ transitions low, the resetting of the flip-flops 72 , 74 is delayed by the propagation delay of the NAND gate 96 and inverter 98 so that short “Up” and “Down” signals are generated when the phase of the S R signal is equal to the phase of the S FB signal, as previously explained.

The reset function of the AND gate 72 ( FIG. 6 ) is performed by a NAND gate 76 ′ that receives A1 and A2 signals from the outputs of each of the NAND gates 88 in the circuits 72 ′, 74 ′, respectively. As previously explained, these signals transition high responsive to the rising edges of the S R and S FB signals, respectively. The NAND gate 76 also receives B1 and B2 signals from the outputs of the NAND gate 80 in each of the circuits 72 ′, 74 ′, respectively. When the rising edges of both the S R and S FB signals has occurred, the A1, A2, B1 and B2 signals are all high. As a result, the NAND gate 76 ′ outputs a low reset signal.

When the reset signal is generated, the low at the output of the NAND gate 76 ′ causes the output of the inverter 98 to output a low. This low resets the flip flop 82 , thereby causing the NAND gate 82 to output a high. Since the A1 signal is still high at that time, the B1 signal at the output of the NAND gate 80 transitions low. The low B1 signal causes the output of the NAND gate 90 to go high, thereby terminating the active “Up” and “Up_” signals. However, the low signal at the output of the NAND gate 90 does not cause the output of the NAND gate 88 to change since the other input of the NAND gate 88 is still low because the S R signal is still high. As are result, the A1 signal remains high.

When the B1 and B2 signals transition low responsive to the reset signal, the reset signal generated by the NAND gate 76 ′ terminates, thereby causing the output of the inverter 98 to transition high. However, this transition has no effect on the circuit other than to enable the NAND gates 82 , 90 since the low B1 signal continues to make the outputs of the NAND gates 82 , 90 high.

Before the next rising edge of the SR signal, the SR signal transitions low. Since the NAND gate 88 is enabled by the low B1 signal applied to the input of the NAND gate 90 , the A1 signal at the output of the NAND gate 88 transitions low. The low A1 signal resets the flip-flop 84 , and it sets the flip-flop 82 thereby making the B1 signal high. However, the output of the NAND gate 90 remains high because of the low A1 signal. As a result, the “Up” and “Up_” signals remain inactive. Thus, after the S R and S FB signals have transitioned low, the A1 signal is low and the B1 signal is high, which are the initial conditions used in the previous explanation of the operation of the circuit 72 ′ responsive to the rising edge of the S R signal.

One example of a dual edge phase detector 100 that may be used as the dual edge phase detector 56 ( FIG. 5 ) is shown in FIG. 8 . The phase detector 100 includes two identical detector circuits 102 , 104 . The detector circuit 102 generates the “Up” and “Up_” signals responsive to the rising edge of the S R signal for a period that is terminated by the falling edge of the S FB signal. The detector circuit 102 also receives an active low reset signal RSTA_and an active low set signal SETA_, which are normally high. The S R signal is applied to an inverter 110 and one input of a NAND gate 112 . The other input of the NAND gate 112 receives the S R signal through an enabled NAND gate 116 and three inverters 118 , 120 , 122 . Prior to the rising edge of the S R signal, the inverter 122 applies a high to the input of the NAND gate 112 . Therefore, when the S R signal transitions high, both inputs of the NAND gate 112 are initially high thereby causing the NAND gate 112 to output a low. However, after a period corresponding to the propagation delays of the NAND gate 116 and the inverters 118 , 120 , 122 , the output of the inverter 122 transitions low thereby causing the output of the NAND gate 112 to transition high. As a result, the NAND gate 112 outputs a short negative-going pulse responsive to the rising edge of the S R signal.

The negative-going pulse occurring on the rising edge of the S R signal sets a flip-flop 130 formed by a pair of NAND gates 132 , 134 . The NAND gate 132 then outputs a high, which is applied to a NAND gate 138 . As will be explained below, the NAND gate 132 is enabled at all times except for a short period after the falling edge of the S FB signal. Therefore, an active low “Up_” signal is generated at the output of the NAND gate 138 . Similarly, the NAND gate 134 outputs low, which is applied to a NAND gate 140 . The NAND gate 140 is enabled through a pair of inverters 144 , 146 by the high at the output of the NAND gate 112 at all times other than for a short period after the rising edge of the S R signal, as previously explained. As a result, the NAND gate 140 outputs an active high “Up” signal.

The flip-flop 130 is reset by a circuit that is similar to the circuit that is used to set the flip-flop 130 . An inverter 150 and a NAND gate 152 receive a S FB signal (which is the compliment of the S FB signal). The output of the inverter 150 is coupled to one input of a NAND gate 156 , and the output of the NAND gate 152 is coupled to the other input of the NAND gate 156 through a series of inverters 158 , 160 , 162 . As a result, in response to the falling edge of the S FB signal, the NAND gate 156 outputs a short negative-going pulse, which resets the flip-flop 130 . The output of the NAND gate 156 is high at all other times to enable the NAND gate 138 through a pair of inverters 166 , 168 . When the flip-flop 130 is reset responsive to the falling edge of the S FB signal, the active “Up” and Up_” signals terminate. In summary, the “Up” and Up_” signals become active a short time after the rising edge of the S R signal, and they become inactive a short time after the falling edge of the S FB signal. The flip-flop 130 may also be set to make the “Up” and “Up_” signals active by driving the SETA_signal active low, and it may be reset to make the “Up” and “Up_” signals inactive by driving the RSTA_ signal active low.

The detector circuit 104 operates in the same manner as the detector circuit 102 to make the “Down” and Down_” signals active a short time after the rising edge of the S FB signal, and to make the “Down” and Down_” signals inactive a short time after the falling edge of the S R signal. Again, the “Down” and Down_” may also be made active by driving the SETB_ signal active low, and they may be made inactive by driving the RSTB_ signal active low.

Delay-lock loops using dual mode phase detectors according to various embodiments of the present invention can be used for a variety of purposes in electronic devices, such as in memory devices. For example, with reference to FIG. 9 , a synchronous dynamic random access memory (“SDRAM”) 200 includes a command decoder 204 that controls the operation of the SDRAM 200 responsive to high-level command signals received on a control bus 206 and coupled through input receivers 208 . These high level command signals, which are typically generated by a memory controller (not shown in FIG. 9 ), are a clock enable signal CKE*, a clock signal CLK, a chip select signal CS*, a write enable signal WE*, a row address strobe signal RAS*, a column address strobe signal CAS*, and a data mask signal DQM, in which the “*” designates the signal as active low. The command decoder 204 generates a sequence of command signals responsive to the high level command signals to carry out the function (e.g., a read or a write) designated by each of the high level command signals. These command signals, and the manner in which they accomplish their respective functions, are conventional. Therefore, in the interest of brevity, a further explanation of these command signals will be omitted.

The SDRAM 200 includes an address register 212 that receives row addresses and column addresses through an address bus 214 . The address bus 214 is generally coupled through input receiver 210 and then applied to a memory controller (not shown in FIG. 9 ). A row address is generally first received by the address register 212 and applied to a row address multiplexer 218 . The row address multiplexer 218 couples the row address to a number of components associated with either of two memory banks 220 , 222 depending upon the state of a bank address bit forming part of the row address. Associated with each of the memory banks 220 , 222 is a respective row address latch 226 , which stores the row address, and a row decoder 228 , which decodes the row address and applies corresponding signals to one of the arrays 220 or 222 . The row address multiplexer 218 also couples row addresses to the row address latches 226 for the purpose of refreshing the memory cells in the arrays 220 , 222 . The row addresses are generated for refresh purposes by a refresh counter 230 , which is controlled by a refresh controller 232 . The refresh controller 232 is, in turn, controlled by the command decoder 204 .

After the row address has been applied to the address register 212 and stored in one of the row address latches 226 , a column address is applied to the address register 212 . The address register 212 couples the column address to a column address latch 240 . Depending on the operating mode of the SDRAM 200 , the column address is either coupled through a burst counter 242 to a column address buffer 244 , or to the burst counter 242 which applies a sequence of column addresses to the column address buffer 244 starting at the column address output by the address register 212 . In either case, the column address buffer 244 applies a column address to a column decoder 248 .

Data to be read from one of the arrays 220 , 222 is coupled to the column circuitry 254 , 255 for one of the arrays 220 , 222 , respectively. The data is then coupled through a data output register 256 and data output drivers 257 to a data bus 258 . The data output drivers 257 apply the read data to the data bus 258 responsive to a read data strobe signal S R generated by a a delay-lock loop using a dual mode phase detector in accordance with various examples of the present invention. The delay-lock loop receives a periodic iCLK signal and generates the read data strobe S R responsive thereto. As a result, the read data are coupled to the data bus 258 substantially in phase with the iCLK signal.

Data to be written to one of the arrays 220 , 222 are coupled from the data bus 258 through data input receivers 261 to a data input register 260 . The data input receivers 261 couple the write data from the data bus 258 responsive to a write data strobe signal S W . The write data strobe signal S W is also generated by a delay-lock loop using a dual mode phase detector in accordance with various embodiments of the present invention. The delay-lock loop receives the periodic iCLK signal and generates the write data strobe signal S W responsive thereto. Using known techniques, the delay-lock loop may generate the write data strobe signal S W in quadrature with the iCLK signal. As a result, the write data are coupled to the data bus 258 at a predetermined phase relative to the phase of the iCLK signal. The write data are coupled to the column circuitry 254 , 255 where they are transferred to one of the arrays 220 , 222 , respectively. A mask register 264 responds to a data mask DM signal to selectively alter the flow of data into and out of the column circuitry 254 , 255 , such as by selectively masking data to be read from the arrays 220 , 222 .

The SDRAM 200 shown in FIG. 9 can be used in various electronic systems. For example, it may be used in a processor-based system, such as a computer system 300 shown in FIG. 10 . The computer system 300 includes a processor 302 for performing various computing functions, such as executing specific software to perform specific calculations or tasks. The processor 302 includes a processor bus 304 that normally includes an address bus, a control bus, and a data bus. In addition, the computer system 300 includes one or more input devices 314 , such as a keyboard or a mouse, coupled to the processor 302 to allow an operator to interface with the computer system 300 . Typically, the computer system 300 also includes one or more output devices 316 coupled to the processor 302 , such output devices typically being a printer or a video terminal. One or more data storage devices 318 are also typically coupled to the processor 302 to allow the processor 302 to store data in or retrieve data from internal or external storage media (not shown). Examples of typical storage devices 318 include hard and floppy disks, tape cassettes, and compact disk read-only memories (CD-ROMs). The processor 302 is also typically coupled to cache memory 326 , which is usually static random access memory (“SRAM”), and to the SDRAM 200 through a memory controller 330 . The memory controller 330 normally includes a control bus 336 and an address bus 338 that are coupled to the SDRAM 200 . A data bus 340 is coupled from the SDRAM 200 to the processor bus 304 either directly (as shown), through the memory controller 330 , or by some other means.

From the foregoing it will be appreciated that, although specific embodiments of the invention have been described herein for purposes of illustration, various modifications may be made without deviating from the spirit and scope of the invention. Accordingly, the invention is not limited except as by the appended claims.