A novel sensitivity function for MR heads and an application

A two-dimensional sensitivity function, describing the down-track and cross-track responses of a MR head, is derived starting from a microtrack formed with a maximal-length pseudo random binary sequence (PRBS). This sensitivity function is built up from a set of `extracted dipulse´ responses each taken with the readback head in a different off-track position with respect to the microtrack. The resulting two-dimensional function represents the sensitivity to a `unit magnet´ with a length of one channel bit, rather than to the vertical field component alone. There are two advantages to this approach: first, the sensitivity to high frequency signal components is much enhanced and the signals used are generally much more representative of written data. Second, this method has the ability to quantify both linear and nonlinear readback distortions as a function of off-track position. This paper initially describes the practical procedure used to obtain the sensitivity function. An advantage of this procedure is that you do not need to have a clock channel for the timing alignment of the dipulses as they are already tied to the bit-cell positions. In the latter half of the paper, this procedure is used to analyze the effects of “read-write non-parallelism” a condition wherein, the MR stripe reads the transitions at a skew due to a manufacturing defect or otherwise. In other words, it is probed whether a cross-track variation in head sensitivity would contribute towards the so-called “deterministic off-track peak jitter” and also whether a relative skew between read and write operations can influence the head sensitivity function and aggravate the deterministic peak-jitter. The answers we got for both the questions were in the negative. There was no deterministic trend in peak-position variation (as against the trends that we got with full-width tracks) with or without a skew and the head sensitivity function was also not affected by a skew