Switching noise reduction in a multi-clock domain transceiver

Abstract

A method for reducing system performance degradation caused by switching noise in a system which includes a set of subsystems. Each of the subsystems includes an analog section and a digital section. Each of the analog sections operates in accordance with a corresponding one of a set of sampling clock signals which are synchronous in frequency. The digital sections operate in accordance with a receive clock signal. The receive clock signal is generated such that it is synchronous in frequency with the sampling clock signals and has a phase offset with respect to one of the sampling clock signals. This phase offset is adjusted such that system performance degradation due to coupling of switching noise from the digital sections to the analog sections is substantially minimized.

What is claimed is:  
     
         1 . A method for reducing system performance degradation due to switching noise in a system, the system comprising a set of subsystems, each of the subsystems comprising an analog section and a digital section, each of the analog sections operating in accordance with a corresponding one of a set of sampling clock signals, the sampling clock signals being synchronous in frequency, the digital sections operating in accordance with a receive clock signal, the method comprising the operations of: 
 generating the receive clock signal such that the receive clock signal is synchronous in frequency with the sampling clock signals and having a phase offset with respect to one of the sampling clock signals; and    adjusting the phase offset such that system performance degradation due to coupling of switching noise from the digital sections to the analog sections is substantially minimized.    
     
     
         2 . The method of  claim 1  wherein, in the operation of adjusting the phase offset, the phase offset is adjusted such that a time difference between a transition occurrence of the receive clock signal and transition occurrences of sampling clock signals, that are adjacent in time to the transition occurrence of the receive clock signal, is substantially maximized.  
     
     
         3 . The method of  claim 1  wherein the operation of adjusting the phase offset of the receive clock comprises the operations of: 
 (1) determining a set of phase offset values for the phase offset;  
 (2) computing a set of system performance errors corresponding one-to-one to the phase offset values; and  
 (3) selecting one of the phase offset values, said one phase offset value corresponding to a minimum of the system performance errors.  
 
     
     
         4 . The method of  claim 3  wherein the set of phase offset values comprises 64 phase offset values.  
     
     
         5 . The method of  claim 3  wherein operation (2) comprises the operations of: 
 computing a subsystem performance error, corresponding to one of the phase offset values, for each of the subsystems;  
 combining the subsystem performance errors to generate the corresponding system performance error.  
 
     
     
         6 . The method of  claim 5  wherein the operation of computing a subsystem performance error for a corresponding subsystem comprises: 
 squaring a slicer error associated with the subsystem;  
 accumulating a number of associated squared slicer errors via a filter for a period of time; and  
 outputting an accumulated squared error as the subsystem performance error after the period of time.  
 
     
     
         7 . The method of  claim 1  further comprising the operation of: 
 adjusting a sampling phase of at least one of the sampling clock signals such that a subsystem performance error of the subsystem which corresponds to said one of the sampling clock signals is substantially minimized.  
 
     
     
         8 . The method of  claim 7  wherein the operation of adjusting the sampling phase of at least one of the sampling clock signals comprises the operations of: 
 (1) determining a set of sampling phase values for the sampling phase;  
 (2) computing a set of subsystem performance errors corresponding one-to-one to the sampling phase values; and  
 (3) selecting one of the sampling phase values, said one sampling phase value corresponding to a minimum of the subsystem performance errors.  
 
     
     
         9 . The method of  claim 8  wherein the set of sampling phase values comprises 16 sampling phase values.  
     
     
         10 . The method of  claim 8  wherein the operation of computing a subsystem performance error for the corresponding subsystem comprises: 
 squaring a slicer error associated with the subsystem;  
 accumulating a number of associated squared slicer errors via a filter for a period of time; and  
 outputting an accumulated squared error as the subsystem performance error after the period of time.  
 
     
     
         11 . The method of  claim 1  further comprising the operation of: 
 adjusting a sampling phase of each of the sampling clock signals such that a subsystem performance error of a corresponding subsystem is substantially minimized.  
 
     
     
         12 . A method for reducing effect of switching noise in a system, the system comprising a set of subsystems, each of the subsystems comprising an analog section and a digital section, each of the analog sections operating in accordance with a corresponding one of a set of sampling clock signals, the digital sections operating in accordance with a receive clock signal, the method comprising the operations of: 
 generating the sampling clock signals such that the sampling clock signals are synchronous in frequency with each other;    generating the receive clock signal such that the receive clock signal is synchronous in frequency with the sampling clock signals and having a phase offset with respect to one of the sampling clock signals; and    adjusting the phase offset such that effect of switching noise from the digital sections on the analog sections is substantially minimized.    
     
     
         13 . The method of  claim 12  wherein, in the operation of adjusting the phase offset, the phase offset is adjusted such that time difference between a transition occurrence of the receive clock signal and transition occurrences of sampling clock signals that are adjacent in time to the transition occurrence of the receive clock signal is substantially maximized.  
     
     
         14 . The method of  claim 12  further comprising the operation of: 
 adjusting a phase of at least one of the sampling clock signals such that a subsystem performance error of the subsystem which corresponds to said one of the sampling clock signals is substantially minimized.  
 
     
     
         15 . The method of  claim 12  wherein the operation of generating the sampling clock signals comprises the operations of: 
 (a) generating a phase error for each of the sampling clock signals from a corresponding phase detector;  
 (b) inputting each of the phase errors to a corresponding loop filter;  
 (c) generating filtered phase errors from the corresponding loop filters;  
 (d) inputting each of the filtered phase errors to a corresponding oscillator;  
 (e) generating phase control signals from the corresponding oscillators;  
 (f) inputting each of the phase control signals to a corresponding phase selector; and  
 (g) generating the sampling clock signals from the corresponding phase selectors.  
 
     
     
         16 . The method of  claim 15  wherein the operation of generating the receive clock signal comprises the operations of: 
 (1) combining one of the phase control signals with the phase offset to produce a phase shift value;  
 (2) inputting the phase shift value to a receive clock phase selector; and  
 (3) generating the receive clock signal from the receive clock phase selector.  
 
     
     
         17 . The method of  claim 16  wherein the phase shift value comprises a set of phase steps and wherein operation (2) comprises the operation of inputting the phase steps consecutively to the receive clock phase selector.  
     
     
         18 . The method of  claim 16  wherein the operation of adjusting the phase offset of the receive clock comprises the operations of: 
 (4) determining a set of phase offset values for the phase offset;  
 (5) computing a set of system performance errors corresponding one-to-one to the phase offset values; and  
 (6) selecting one of the phase offset values, said one phase offset value corresponding to a minimum of the system performance errors.  
 
     
     
         19 . The method of  claim 18  wherein the set of phase offset values comprises 64 phase offset values.  
     
     
         20 . The method of  claim 18  wherein operation (5) comprises the operations of: 
 computing a subsystem performance error for each of the subsystems for one of the phase offset values;  
 combining the subsystem performance errors to generate the corresponding system performance error.  
 
     
     
         21 . The method of  claim 20  wherein the operation of computing a subsystem performance error for a corresponding subsystem comprises: 
 squaring a slicer error associated with the subsystem;  
 accumulating a number of associated squared slicer errors via a filter for a period of time; and  
 outputting an accumulated squared error as the subsystem performance error after the period of time.  
 
     
     
         22 . The method of  claim 15  wherein, in operation (a), each of the phase detectors receives a corresponding slicer error and a corresponding tentative decision from a decoding system.  
     
     
         23 . The method of  claim 22  wherein operation (a) comprises: 
 (1) generating a pre-cursor phase error by multiplying the corresponding tentative decision by a delayed version of the corresponding slicer error;  
 (2) generating a post-cursor phase error by multiplying the corresponding slicer error by a delayed version of the corresponding tentative decision; and  
 (3) combining the pre-cursor and post-cursor phase errors to produce the corresponding phase error.  
 
     
     
         24 . The method of  claim 23  wherein operations (1), (2) and (3) are performed via a lattice structure, the lattice structure comprising two delay elements, two multipliers and an adder.  
     
     
         25 . The method of  claim 24  wherein operation (3) includes the operation of combining the pre-cursor, post-cursor phase errors and an offset input from a control unit to produce the corresponding phase error.  
     
     
         26 . The method of  claim 15  wherein operation (c) comprises: 
 (1) accumulating a number of consecutive values of one of the phase errors via a first filter, resulting in a sum value;  
 (2) outputting the sum value from the first filter;  
 (3) integrating the sum value via a second filter to produce an integral value; and  
 (4) combining the sum value and the integral value to produce a filtered phase error.  
 
     
     
         27 . The method of  claim 26  wherein operation (3) includes the operation of scaling the integrated sum value by a scale factor to produce the integral value.  
     
     
         28 . The method of  claim 26  wherein operation (c) further comprises, before operation (3), the operation of multiplying the sum value by a factor different than 1 when the system is operating in a different bandwidth mode.  
     
     
         29 . The method of  claim 15  wherein operation (e) comprises the operation of filtering recursively the filtered phase errors to produce the corresponding phase control signals.  
     
     
         30 . The method of  claim 29  wherein operation (e) further comprises the operation of scaling, before filtering recursively, the filtered phase errors by a scale factor.  
     
     
         31 . The method of  claim 15  wherein operation (g) comprises the operations of: 
 (1) inputting a multi-phase input signal from a clock generator to each of the phase selectors; and  
 (2) selecting at each of the phase selectors one of the phases of the multi-phase input signal based on the phase control signal received from the corresponding oscillator.