Synopsys, Inc.

Skye Wolfer

David Andrew Yokoyama-Martin

John Theodore Stonick

Jianping Wen

Daniel Keith Weinlader

James P Flynn

Junqi Hua

Richard H Steeves

Robert Beverly Lefferts

Dino Anthony Toffolon

Choon Haw Leong

Jasjeet Singh

Zhenchang Du

A CDR method/circuit utilizes a closed-loop clock alignment circuit and a duplicate clock to align a sampling point clock to both mid-interval and optimal sample point phases during data receiving processes. An initial clock is generated having the mid-interval sampling point phase, then the closed-loop clock alignment circuit generates a phase correction signal based on a phase difference between the data sampling clock and the initial clock, and then the phase correction signal is fed back to a high-speed phase mixer to adjust/align the sampling point clock to the initial clock. Subsequently, the duplicate clock is generated and utilized to determine an optimal sampling point phase while the data sampling clock is utilized to read the received data signal, and then the closed-loop clock alignment circuit is re-used to re-align the data sampling clock to the duplicate clock when the optimal sampling point phase is identified.

Clock and data recovery using closed-loop clock alignment and duplicate sampling clock

A digital circuit that divides a high speed digital clock by a fractional value is described. The circuit utilizes a divider circuit and shifts the divider clock by a fraction of a phase to achieve the desired fractional division. A clock mux is used to perform the clock shift, and a masking mux is used to eliminate glitches during the clock shift.

Low power digital fractional divider with glitchless output

Circuitry to facilitate testing of serial interfaces is described. Specifically, some embodiments of the present invention facilitate testing the clock and data recovery functionality of a receiver. A serial interface can include a multiplying phase locked loop (MPLL) clock generator, a transmitter, and a receiver. The MPLL clock generator can generate a first clock signal and a second clock signal, and can vary a phase and/or frequency difference between the first clock signal and the second clock signal. During test, the transmitter and the receiver can be directly or capacitively coupled to each another. Specifically, during test, the serial interface can be configured so that the transmitter transmits data using the first clock signal, and the receiver receives data using the second clock signal. The clock and data recovery functionality of the receiver can be tested by comparing the transmitted data with the received data.

Circuitry to facilitate testing of serial interfaces

A digital phase selector circuit that switches an output clock between N input clock phases is described. The phase selector utilizes a special output mux and switches clock phases during a safe zone to avoid glitches. The phase selector is used in the feedback path of a PLL to implement functions such as spread spectrum or fractional reference clocks. An example with N=4 and an optimized latch mux is shown.

High speed phase selector with a glitchless output used in phase locked loop applications

Some embodiments provide a clock and data recovery (CDR) system to recover clock and data information from an analog signal. The CDR system may include an integral path and a proportional path that are part of an integral-proportional control loop. The integral path may be used to track frequency changes in a clock signal that is embedded in the analog signal, while the proportional path may be used to track phase changes in the clock signal that is embedded in the analog signal. The proportional path may be executed at a first clock frequency, while the integral path may be executed at a second clock frequency that is lower than the first clock frequency to reduce the power consumption of the CDR system.

Reducing power consumption in clock and data recovery systems

An integrated circuit that includes a receive data path is described. The receive data path: equalizes a received analog signal, converts the resulting equalized analog signal to digital data values based on a clock signal, and recovers the clock signal in the digital data values. The integrated circuit also includes an on-chip oscilloscope. The oscilloscope includes: two comparators, a phase rotator that outputs an oscilloscope clock signal whose phase can be varied relative to that of the recovered clock signal, and an offset circuit that outputs a voltage offset. Based on the voltage offset and the oscilloscope clock signal, the comparators output digital values which can be used to determine eye patterns that correspond to the analog signal before and after equalization. The eye patterns can then be correlated with an error rate associated with the received data.

Multiple-input, on-chip oscilloscope

One embodiment of the present invention provides a phase-locked loop (PLL) for synthesizing a fractional frequency. The PLL can include a 1/N frequency divider, a voltage-controlled oscillator (VCO), a programmable phase mixer, and a phase detector. The programmable phase mixer can be coupled between an output of the VCO and an input of the frequency divider, wherein the programmable phase mixer is configured to receive the output clock signal from the VCO and generate a first clock signal of frequency fby varying a phase of the output clock signal. The frequency divider is configured to receive the first clock signal from the programmable phase mixer and generate a second clock signal of frequency f=f/N. The phase detector can receive a reference clock signal and the second clock signal as inputs, and the phase detector's output can be used to generate the control voltage for the VCO.

[object Object]

Systems and techniques for adapting and/or optimizing an equalizer of a receiver are described. The equalizer's behavior can be adjusted by modifying one or more equalization parameters. At the beginning of the adaptation and/or optimization process, the system can determine robust initial values for the one or more equalization parameters. The system can then adapt and/or optimize the equalizer by iteratively adjusting the one or more equalization parameters. Specifically, in each iteration, the system can use the receiver's clock and data recovery (CDR) circuitry to determine the number of early and late data transitions associated with one or more data patterns. Next, the system can adjust the one or more equalization parameters so that, for each data pattern in the one or more data patterns, the ratio between the number of early data transitions and the number of late data transitions is substantially equal to a desired value.

Method and apparatus for performing adaptive equalization

An on-die scope is described. The on-die scope can include one or more scope slicers, phase sweeping circuitry, voltage sweeping circuitry, and eye-diagram data collection circuitry. The clock and data recovery circuitry can receive an input signal, and output a recovered clock signal and a recovered bit-stream. The phase sweeping circuitry can receive the recovered clock signal, and output the scope clock signal by adding a phase offset to the recovered clock signal. A scope slicer can receive the voltage threshold, the scope clock signal, and the input signal, and output a scope bit-stream. The eye-diagram data collection circuitry can detect one or more bit-patterns in the recovered bit-stream, and modify values of one or more scope counters based solely or partly on the scope bit-stream and the recovered bit-stream.

Pattern agnostic on-die scope

Some embodiments of the present invention provide a voltage-mode transmitter. The transmitter can include configuration circuitry, bias circuitry, and a set of driver slices. Each driver slice can include driver transistors which drive an output value. The outputs of each driver slice can be directly or capacitively coupled with the transmitter's outputs. Each driver slice can also include one or more impedance-matching transistors which are serially coupled to at least some of the driver transistors. The configuration circuitry can configure a subset of driver slices so that the down (or up) impedance of the transmitter is within a first tolerance of a desired impedance value. The bias circuitry can bias the one or more impedance-matching transistors in each driver slice in the subset of driver slices so that the up (or down) impedance is within a second tolerance of the down (or up) impedance.

Circuitry for matching the up and down impedances of a voltage-mode transmitter

Some embodiments of the present invention provide a programmable transmitter which includes a set of drivers and one or more chains of configuration registers. Each driver is capable of being configured to perform a transmission function from a predetermined set of transmission functions. Each configuration register can correspond to a driver, and can store configuration data which is used to configure the corresponding driver. The programmable transmitter can include configuration circuitry which serially shifts configuration data into the one or more chains of configuration registers. The programmable transmitter can also include programming circuitry which can determine configuration data for each driver based partly or solely on a desired transmitter behavior.

Growing community of inventors

Hillsboro, OR, United States of America

Skye Wolfer