Browse Prior Art Database
Quick Links

# Feedforward control for magnet latch release

IP.com Disclosure Number: IPCOM000013768D
Original Publication Date: 2001-Nov-17
Included in the Prior Art Database: 2003-Jun-18
Document File: 2 page(s) / 40K

IBM

## Abstract

Disclosed is 'Feedforward control for magnet latch release'. 1) Introduction We invent feedforwad control for ramp loading of HDD and show you what it is. 2) Conventional method Ramp loading needs velocity control. Velocity can be monitored by BEMF(Back Electro -Magnetic Force). We should control velocity by VCM current to load actuator properly

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 74% of the total text.

Page 1 of 2

Feedforward control for magnet latch release

Disclosed is 'Feedforward control for magnet latch release'.

1) Introduction

We invent feedforwad control for ramp loading of HDD and show you what it is.

2) Conventional method Ramp loading needs velocity control. Velocity can be monitored by BEMF(Back Electro -Magnetic Force). We should control velocity by VCM current to load actuator properly

to disk.

Conventioanl method for ramp loading is PI control whose formula is Output = Kp * Verr + Ki *sum(Verr) (formula-A) Here Verr is velocity error to a target, Kp is proportional coefficient and Ki is integral coefficient.

3) Feedforward control We offer a feedforward term such as Km/(x^2) to compensate magnet force, because magnet force is inversely proportional to the squre of distance. When the term is add to

formula-A, we get a following formula.

Output = Kp * Verr + Ki *sum(Verr) + Km/(x^2) (formula-A) (fromula-B) Here x is distance that is actually accumulated of velocity(BEMF) and Km is magnet force coefficient.

4) Advantage of feedforward control Then we introduce the advantage of feedforward control. Using conventional method, we needs sevral sets of Kp and Ki according to strength of magnet force. For example,

if we have three sets of Ki and Kp for formula-A, they are as follows.

(Kp1, Ki1), 0<x<x1 (Kp2, Ki2), x1<=x<x2 (Kp3, Ki3), x2<=x Of course , coefficients becomes larger at smaller distance. Kp1>Kp2>Kp3
Ki1>Ki2>Ki3 But this coventional method is weak at switc...