Writing Process Modeling and Identification for Heat-Assisted Magnetic Recording

Written-in signal quality for heat-assisted magnetic recording (HAMR) is dominated by the hot spot thermal profile on the media. The thermal profile is determined by the head and media design, the laser power, and the near-field transducer (NFT) tip to media spacing (NMS). NMS keeps changing during the writing process due to protrusions generated by the laser-related heat sources, on top of the usual head-to-media spacing change due to thermal fly-height control and writing current. In this paper, data pattern writing process is modeled as a dynamic system consisting of NFT protrusions of different time constants, energy transfer efficiency versus NMS, media heat transient process, and power-to-signal conversion function. The parameters of this model are identified by minimizing the differences between the model output and the measured read-back signal magnitude of an experimental data pattern written on an HAMR drive. The Gaussian-Newton method is used to find the parameters of the non-linear system. Using the identified parameters, our model is able to predict the written-in signal magnitude, given the writing conditions such as laser power, laser current overshoot, and various thermal actuator operating conditions.