Spacing loss and nonlinear distortion in digital magnetic recording

The exponentially decaying spacing loss function in digital magnetic recording has traditionally been treated as a linear rather than a nonlinear loss. With the combination of realistic iterative recording-process simulation and skillful laboratory measurements, it becomes possible to segregate nonlinear and linear distortions from the composite readback signal, and investigate their respective causes. For a typical 3350 equivalent head/disk interface, using a Triplet test pattern at 3730 flux changes per cm, it was predicted that the spread of nonlinear timing error (after linear equalization) would be reduced from 10 to 6 ns while the corresponding nonlinear amplitude loss would be reduced from 30% to 20% by decreasing the flying height from 0.5 μm to 0.35 μm. This shows that non-linearities are notably spacing loss dependent. Furthermore, this separation-dependent nonlinear distortion is primarily contributed by the recording and demagnetizing process. This finding partially explains the discrepancy in the observed signal non-linearities between disk and tape technologies where their flying height differences may have hitherto discouraged a scientific unification.