摘要 |
In a synchronous read channel for magnetic recording, a timing recovery phase-locked loop (PLL) comprises a technique for smoothing a timing gradient DELTA t computed from estimated sample values and actual sample values. If an estimated sample value for computing the timing gradient is zero, then the timing gradient is increased, and if all of the estimated sample values for computing the timing gradient are zero, then the timing gradient is copied from its prior value. Smoothing the timing gradient reduces gain variations in the PLL and results in more effective timing recovery.
|