A method for reducing the effects of thermal asperities

Thermal asperity suppression based on least-squares fitting in perpendicular magnetic recording systems

“…Typically, a TA signal has a short rise time (60-150 ns) with a long decay time (1-5µs), and its peak TA amplitude could be 2 to 3 times the peak of the readback signal [1], [2]. The TA effect can cause a burst of errors in data detection, which could easily exceed the correction capability of the error correction code (ECC), and thus results in a sector read failure.…”

“…Thus, Klaassen and van Peppen [3] proposed the TA detection by looking at the baseline of the averaged readback signal, while the TA correction was performed by use of a high-pass filter. Dorfman and Wolf [2] proposed a method to combat with the TA effect by passing the TA-affected readback signal through a filter (1 − D), where D is a delay operator. This method has been tested with an EPR4 target in longitudinal recording channels, where the number of bits corrupted by the TA effect is dramatically reduced.…”

The effect of bandpass filters for thermal asperity suppression in perpendicular recording systems

“…Practically, the TA signal described by Stupp et al [2] has a short rise time (50 -160 ns) with a long decay time (1 -5 μs), and its peak TA amplitude could be 2 -3 times the peak of the readback signal [2,3]. If precautions are not taken, the TA effect can cause a burst of errors in data detection, which could easily exceed the correction capability of the errorcorrection code (ECC), and thus results in a sector read failure.…”

“…Most of TA detection and correction algorithms proposed in the literature attempt to filter out the TA, lessen its duration, or employ a suitable equalization target to reduce the TA at the detector input [3]. Because the TA causes a shift in the baseline of the readback signal, the average value of the normal readback signal is zero, whereas that of the TA-affected readback signal is not.…”

“…Therefore, Klaassen and van Peppen [4] proposed the TA detection by looking at the baseline of the averaged readback signal, while the TA correction was performed by use of a high-pass filter. Dorfman and Wolf [3] proposed a method to combat with the TA effect by passing the TA-affected readback signal through a filter (1 -D), where D is a delay operator. This method has been tested with an EPR4 target in longitudinal recording channels, where the number of bits corrupted by the TA effect is dramatically reduced.…”