Calibration of scale factor in adaptive equalizers

A computer system for use in calibrating a scalar in an adaptive equalizer during a training sequence comprising:

• a board comprising a first transmission line and a second transmission line;
• and
• a receiver coupled to the first and second transmission lines, the receiver comprising:
• a discrete-time FIR (Finite Impulse Response) filter comprising n multiplier units to implement a filter response [ h(t)]i, i=0,1, . . . , n−1, where t is a time index;
• a data generator to provide a discrete-time sequence of desired voltages d(t), t=1,2, . . . , T,
• a multiplier to provide a sequence of voltages Kd(t), t=1,2, . . . , T, where K is the scalar;
• a filter increment generator to provide, for t=1,2, . . . , T, n voltages indicative of n filter increments [δ h(t)]i, i=0,1, . . . , n−1;
• at least one summer to perform the sum [ h(t)]i, +[δ h(t)]i, i=0,1, . . . , n−1 to update the filter response;
• an overflow counter to provide an overflow count indicative of the number of numerical overflows in the at least one summer during the time period t=1,2, . . . , T;
• wherein the scalar K is increased by a first increment if after completion of the time period t=1,2, . . . , T the overflow count and a threshold satisfy a first relationship, and
• wherein the scalar K is used during the training sequence to update the adaptive equalizer.

A method to calibrate a scale factor in an adaptive equalizer, the scale factor being used to multiply a sequence of desired voltages used in updating the equalizer during a training sequence, the method comprising:

• updating the adaptive equalizer over the training sequence;
• counting the number of numerical overflows occurring while updating the adaptive equalizer over the training sequence;
• and
• increasing the scale factor by a first increment if the number of numerical overflows and a threshold satisfy a first relationship,
• wherein the scale factor is calibrated once for a communication channel.

A method to calibrate a scale factor in an adaptive equalizer, the scale factor being used to multiply a sequence of desired voltages used in updating the equalizer during a training sequence, the method comprising:

• updating the adaptive equalizer over the training sequence;
• counting the number of numerical overflows occurring while updating the adaptive equalizer over the training sequence;
• and
• increasing the scale factor by a first increment if the number of numerical overflows and a threshold satisfy a first relationship,
• wherein the scale factor is decreased by a second increment if the number of numerical overflows and the threshold satisfy a second relationship.

A method to calibrate a scale factor in an adaptive egualizer, the scale factor being used to multiply a sequence of desired voltages used in updating the equalizer during a training sequence, the method comprising:

• updating the adaptive equalizer over the training sequence;
• counting the number of numerical overflows occurring while updating the adaptive equalizer over the training sequence;
• and
• increasing the scale factor by a first increment if the number of numerical overflows and a threshold satisfy a first relationship,
• wherein
• the overflow count and the threshold satisfy the first relationship if and only if the overflow count is greater than the threshold;
• and
• the first increment is positive.