摘要:
A high speed phase detector circuit operating at a clock speed equal to one-half an input data rate (i.e. a half-rate clock) provides phase information and transition information from incoming serial data. The high speed phase detector circuit samples the incoming serial data on both the rising edge and falling edge of the half-rate clock to provide equivalent full high speed data rate sampling. The high speed phase detector circuit generates a delay between the phase information and the transition information. The phase information is produced in a first bit period and the transition information is produced in a second bit period relative to the first bit period.
摘要:
A transceiver includes a receiver section and a transmitter section. The receiver section includes a clocking circuit, a serial-to-parallel module, and compensation. The transmitter section includes a clocking circuit, parallel-to-serial module, and compensation. The compensation within the receiver section and transmitter section compensates for integrated circuit (IC) processing limits and/or integrated circuit (IC) fabrication limits within the clocking circuits, serial-to-parallel module, and parallel-to-serial module that would otherwise limit the speed at which the transceiver could transport data.
摘要:
A high-speed, wide bandwidth data detection circuit includes a phase detection module, a data detection module, a loop filter, and a voltage controlled oscillator. The phase detection module is operably coupled to produce a controlled current based on a current mode mathematical manipulation of differences between an incoming data stream and a recovered clock. The phase detection module performs the current mode mathematical manipulations and produces the controlled current in the current domain. The data detection module is operably coupled to produce the detected data based on the incoming data stream and the recovered clock. The loop filter is operably coupled to convert the controlled current into a controlled voltage. The voltage controlled oscillator is operably coupled to convert the control voltage into the recovered clock.
摘要:
A clock recovery circuit that operates at a clock speed equal to one-half the input data rate is presented. The clock recovery circuit uses dual input latches to sample the incoming serial data on both the rising edge and falling edge of a half-rate clock signal to provide equivalent full data rate clock recovery. The clock recovery circuit functions to maintain the half-rate clock transitions in the center of the incoming serial data bits. The clock recovery circuit includes a phase detector, charge pump, controlled oscillation module and a feedback module. The phase detector produces information on the phase and data transitions in the incoming data signal to the charge pump. Generally, the circuit is delay insensitive and receives phase and transition information staggered relative to each other.
摘要:
Described are fast synchronous counters with reduced combinatorial logic. In one embodiment, a four-bit shift register is configured in a ring and preset with a data pattern (e.g., 1000). The register is then rapidly shifted into any of four unique states. Combinatorial logic connected to the shift register converts the four unique states into a two-bit binary signal representative of the four states. In the general case, counters in accordance with this embodiment represent N-bit binary numbers using 2N synchronous storage elements. Two or more counters can be combined to produce larger synchronous counters. An up/down counter in accordance with yet another embodiment is connected to a multi-path delay line to create a variable delay circuit. The switching speed of the delay circuit is independent of the number of delay settings. Also advantageous, the delay circuit scales linearly, in terms of power consumption and area, with changes in delay granularity.
摘要:
A data density independent clock and data recovery system includes a lock phase adjust charge pump operably coupled to receive phase information and transition information from a phase detector and to produce a current signal, responsive to the phase information and transition information, to a loop filter that converts the current signal to a control voltage signal operably coupled to a voltage controlled oscillator that produces a clock signal to the phase detector based on the control voltage signal. The lock phase adjust charge pump includes a phase charge pump, a transition charge pump, a programmable DC bias current sink, and two programmable offset bias current sinks. The transition charge pump includes a programmable transition current sink. The control logic operates under external control to adjust the currents conducted by the transition charge pump, the programmable DC bias current sink, and the two programmable offset bias current sinks.
摘要:
In many electronic systems, it is common to communicate data from a transmitter in one device to a receiver in another. Accurate communications requires use of several matched clock signals. Mismatches in these clock signals cause transmitters to add “jitter” to transmitted data or receivers to be more intolerant of jitter in received signals, increasing the chances of mis-interpreting the data. Accordingly, the inventors devised an exemplary clock-distribution method which entails generating a base set of matched clock signals, deriving at least two separate sets of matched clock signals from the base set, and distributing one of the sets of clock signals to a set of matched components in a circuit and the other set of matched clock signals to a different set of components in the same circuit. The clock signals driving the matched components are isolated from mismatched aspects of the other components, and thus exhibit better matching. This method improves the accuracy of high-speed receivers, transmitters, transceivers, and other communications circuits that use it.
摘要:
Described are fast synchronous counters with reduced combinatorial logic. In one embodiment, a four-bit shift register is configured in a ring and preset with a data pattern (e.g., 1000). The register is then rapidly shifted into any of four unique states. Combinatorial logic connected to the shift register converts the four unique states into a two-bit binary signal representative of the four states. In the general case, counters in accordance with this embodiment represent N-bit binary numbers using 2N synchronous storage elements. Two or more counters can be combined to produce larger synchronous counters. An up/down counter in accordance with yet another embodiment is connected to a multi-path delay line to create a variable delay circuit. The switching speed of the delay circuit is independent of the number of delay settings. Also advantageous, the delay circuit scales linearly, in terms of power consumption and area, with changes in delay granularity.
摘要:
Described are fast synchronous counters with reduced combinatorial logic. In one embodiment, a four-bit shift register is configured in a ring and preset with a data pattern (e.g., 1000). The register is then rapidly shifted into any of four unique states. Combinatorial logic connected to the shift register converts the four unique states into a two-bit binary signal representative of the four states. In the general case, counters in accordance with this embodiment represent N-bit binary numbers using 2N synchronous storage elements. Two or more counters can be combined to produce larger synchronous counters. An up/down counter in accordance with yet another embodiment is connected to a multi-path delay line to create a variable delay circuit. The switching speed of the delay circuit is independent of the number of delay settings. Also advantageous, the delay circuit scales linearly, in terms of power consumption and area, with changes in delay granularity.
摘要:
A method and circuit for determining variation between an input clock signal (CLK0) and a reference clock signal (REFCLK) is provided. A plurality of time shifted input clock signals (CLK0, CLK1, . . . , CLK09) can be generated from a single input clock signal (CLK0). The plurality of time shifted input clock signals (CLK0, CLK1, . . . , CLK09) can be sampled at successive periodic intervals occurring relative to the reference clock signal (REFCLK). For each of the time shifted input clock signals (CLK0, CLK1, . . . , CLK09), a sampled value for a succeeding and a preceding periodic interval can be compared to determine whether there is a variation between an input clock signal (CLK0) and a reference clock signal (REFCLK).