Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Servo Scheme for Integrating Positive Error Signal Detector

IP.com Disclosure Number: IPCOM000035189D
Original Publication Date: 1989-Jun-01
Included in the Prior Art Database: 2005-Jan-28
Document File: 3 page(s) / 33K

Publishing Venue

IBM

Related People

Best, JS: AUTHOR [+2]

Abstract

The quality of the position error signal (PES) for a disk file servo system has a significant effect on both servo performance and capacity. One method for encoding and detecting the PES is by amplitude detection, where the PES is derived from the peak amplitudes of transitions. Theoretically, it is possible to improve the signal-to-noise (S/N) for such a system by using an integrating detector, and measuring the area under the peaks. The main problem with integration PES detection is eliminating errors due to timing. It is possible to eliminate timing errors from integrating PES detection without requiring a phase synchronization between the servo data and the clock.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 3

Servo Scheme for Integrating Positive Error Signal Detector

The quality of the position error signal (PES) for a disk file servo system has a significant effect on both servo performance and capacity. One method for encoding and detecting the PES is by amplitude detection, where the PES is derived from the peak amplitudes of transitions. Theoretically, it is possible to improve the signal-to-noise (S/N) for such a system by using an integrating detector, and measuring the area under the peaks. The main problem with integration PES detection is eliminating errors due to timing. It is possible to eliminate timing errors from integrating PES detection without requiring a phase synchronization between the servo data and the clock.

Integrating a waveform will have a greater S/N ratio than averaged peak detection when the channel bandwidth is greater than the servo data bandwidth. Most servo systems operate well below this limit since the data are written at frequencies well below the cut-off frequency of the channel filter, which typically has a gradual roll-off. The following integrating servo detection scheme will accurately measure the area of a servo waveform without requiring a phase synchronous clock. The technique is to write a constant frequency pattern of n bits, full-wave rectify the signal and to integrate for m cycles, m < n. m is chosen such that the waveform is symmetric for m + 1 pulses, and the integration is started near (n-m)/2 cycles from the beginning of the first pulse. The principle is that any timing error in the starting of the integration causes a gain (or loss) in the area of the first peak being measured, but this is exactly offset by the loss (or gain) in the area of the last peak being measured. There is no loss in S/N due to the timing error.

The figure shows an implementation using a simple A/B servo pattern, although the concept can be extended to other patterns (quadrature, etc.). The full-wave rectified pattern is shown for two adjacent tracks. The first peak is present on all tracks, and is the start, or synch, bit. This pulse will be used to synchronize the integration with a system clock. A 2F clock has been shown as an example. Here the first pulse occurs within the middle of a cl...