JP2004063039A - Disk information processor and control method for the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To attain the compatibility of the time for resetting operation with electricity saving of a disk information processor. <P>SOLUTION: The disk information processor 1 is provided with a head 4 for reading or recording the information on a disk recording medium 2 and a servo system including a positioning mechanism and driving source of a head with respect to tracks of the medium 2. For the control of the servo system, the processor is provided with a characteristic control means for lowering the responsiveness relating to the servo control characteristics in an operation mode for temporary standby to maintain the head position after seeking so that the electric power consumption is reduced by lowering the sensitivity with torque disturbance. <P>COPYRIGHT: (C)2004,JPO

Description

【０１１５】
極方程式（あるいは特性方程式）「１＋Ｇ_ｏｐ（Ｓ）＝０」から次式が得られる。
【０１１６】
【数２】

【０１１７】
図示のように、Ｓ平面上で、半径「Ｆ１」の円周上に３個の極を配置した場合、下式のようになる。
【０１１８】
【数３】

【０１１９】
尚、ここで、Ｆ１（単位：Ｈｚ）は、系の時定数に対応する応答周波数を示す。また、「Ｃφ」は極角φの余弦関数（ｃｏｓφ）を示す。
【０１２０】
［数２］式と、［数３］式との比較から、下式を得る。
【０１２１】
【数４】

【０１２２】
各係数（Ｋｐ、Ｋｉ、Ｋｄ）は、φを一定とした場合にＦ１のみの関数である。
【０１２３】
例えば、Ｆ１＝１０００［Ｈｚ］、φ＝０［度］を指定した場合には、指数表示で、「Ｋｄ＝３×１０＾（＋３）、Ｋｐ＝３×１０＾（＋６）、Ｋｉ＝１０＾（＋９）」である（「＾」は累乗を示す。）。そのボーデ線図を示したものが図９（上方に振幅特性、下方に位相特性を示す。）であり、０ｄＢ（デシベル）カットオフ周波数が約５００Ｈｚ、位相余裕はほぼ６０度である。
【０１２４】
上記の方法を用いて、ＰＩＤパラメータを決定してサーボ系を設計する場合において、下記のシミュレーション条件を想定する。
【０１２５】
・ＲＲＯ成分（スピンドルモータの回転に伴う同期性の成分）がトラックピッチ（Ｔｐ）換算で、該ピッチの１００倍、「１００・Ｔｐ」ｐ−ｐ（Ｐｅｅｋ　ｔｏ　Ｐｅｅｋ）相当存在すること
・ＲＲＯ成分については、歪なしの正弦波（ｓｉｎ波）とされ、その周波数はスピンドルモータ回転周波数（Ｆｚ＝７５Ｈｚ）に等しいこと。また、モデルのブロック図では等価的にＶＣＭの位置指令として入力すること。
【０１２６】
・ＰＩＤ制御系の周波数特性（Ｆ特）については、Ｆ１＝１０００［Ｈｚ］、０ｄＢカットオフ周波数は５００［Ｈｚ］であり、位相余裕が６０［度］（図９参照）であること。
【０１２７】
図１０は、ＰＩＤ制御系がこの外乱に対して忠実に追従している状況下（所謂、Ｊｕｓｔトラッキングの状態）でのＶＣＭ電流値（時間的変化）を例示したものであり、この場合、約３０［ｍＡ］ｐ−ｐの電流が流れていることが分かる。
【０１２８】
前記したように、本発明のＨＤＤへの適用においては、待機モード時に、サーボ制御系のＦ特を低減して、ＲＲＯ外乱とＮＲＲＯ外乱両方に対してＶＣＭの応答感度を低減することで、該モード時に流れるＶＣＭ電流を小さくし、ＶＣＭの消費電力を低減するものである。
【０１２９】
そのための方法として、例えば、前記した方法（Ａ）に属する、下記の（Ａ１）、（Ａ２）と、前記した方法（Ｂ）が挙げられる。
【０１３０】
（Ａ１）制御Ｆ特（Ｆ１）を、所定のタイミングで一挙に「Ｆ１／ｎ」（但し、ｎ＞１）　に切り替える形態（図１１参照）
（Ａ２）ＶＣＭの過渡応答を考慮して、制御Ｆ特（Ｆ１）を、多段階に亘って切り替える形態（図１２参照）
（Ｂ）（Ａ２）の多段階切替を究極化し、制御Ｆ特（Ｆ１）を無段階、つまり連続的に変更する形態（図１３参照）。
【０１３１】
尚、図１１乃至図１３において、矢印「ｔ」が時間経過方向を示し、「ｆ１」は、ｔ＝ｔ１の時点におけるＦ１の値を示し、「ｆ２」は、ｔ＝ｔ２の時点におけるＦ１の値を示している。また、Ｆ１値変更による応答性の回復の場合には、矢印「ｔ」の向きを反転させ、ｔ１とｔ２との置換及びｆ１とｆ２とを置換を行えば良い。
【０１３２】
上記（Ａ１）、（Ａ２）は、ＣＰＵの処理負担の軽減という観点から好ましい方法であり、特に、（Ａ１）は、最も容易で簡便な方法である。また、（Ａ２）については、（Ａ１）との比較において、ＶＣＭに与える衝撃等を少なくすることができる（反面、段階数や切替間隔を決定するための、実験等が必要となり、管理すべきパラメータ数が増える。）。
【０１３３】
上記（Ｂ）は、ＣＰＵに関する演算時間への負担が許容範囲内である場合には、理想的な方法であり、過渡応答に係る変動が少なく、また、必要ならばＦ１の変更時間について調整することが可能である。
【０１３４】
尚、各形態において、変更前のｆ１からｆ２（＜ｆ１）へと値を変更する場合には、ｆ２値が小さい程、電力積の低減効果がより得られることは明らかであるが、その最小値（下限）について考慮する必要がある。つまり、ｆ２値が小さくなると、制御系はＲＲＯ外乱やＮＲＲＯ外乱を抑えられなくなりヘッド位置が大きく振れてしまう虞がある。そして、この振れ（変動）については、ＮＲＲＯ外乱の大きさにほぼ比例しその周波数成分に依存すると考えられるので、正確にこの振れを知る為には、ＮＲＲＯ成分を正確に把握する必要がある。また、現実問題として、ＨＤＤが携帯機器等に搭載された場合、人の歩行等で発生する人的外乱（ＤＴｍｂ）が最も大きな値を示すと考えられるので、この場合には、携帯機器の動作環境を良く調査してデータを収集するとともに、該データに基いてｆ２値を決める必要がある。一般的には、ディスク最外周と最内周での位置偏差（ＶＣＭの暴れ）の如何によっては、アームの衝突等、ＶＣＭの障害を起こす可能性があることを考慮して、電力積低減モードの開始時点における半径値（ディスク半径方向におけるヘッド位置）に応じてｆ２値を変更する等の対策を講じることが望ましい。
【０１３５】
また、方法（Ｂ）あるいは（Ａ２）では、変更時間「ｔ２−ｔ１」の設定にも注意する必要がある。つまり、「ｔ２−ｔ１」が短い程、過渡応答に係る変動の度合いが大きくなり、逆に、「ｔ２−ｔ１」が長い程、過渡応答に係る変動の度合いが緩やかとなる傾向を示す。従って、「ｔ２−ｔ１」を過度に長くすると、電力積低減モードから待機モードへの復帰に要する時間（復帰時間）がかかり過ぎるといった不都合が起き得る。また、復帰時間はＨＤＤ（携帯機器搭載用ＨＤＤ等）のアクセス時間にも直接的な影響を及ぼすので、ユーザーが許容できる待ち時間よりも短くする必要がある。尚、ここにいうアクセス時間の長さについては、その用途や目的からして、高速性を要求されるＨＤＤ（ＰＣ用ＨＤＤ等）のアクセス時間と比較した場合に、これよりは、かなり長めの時間で構わない。
【０１３６】
上記したいずれの形態でも制御系の安定は必要条件であるので、ここではＦ１の変更前後において周波数軸に関してスケーリング変換した特性を持たせる方法を採用する。つまり、ボーデ線図の形はゲイン及び位相共に、変更前後で全く同じであって周波数のみの大小が変化する様に変更することとする。具体的には、［数４］式において極角φの値を固定したままでＦ１値のみを連続的に変更することで実現される。尚、以下では、周波数軸に関してスケーリング変換でのみ対応する方式に限定して検討するが、必要に応じてＦ１と極角φを自由に変更する方式を採用しても構わないことは勿論である。
【０１３７】
図１４はシミュレーションに用いたブロック図を示す。
【０１３８】
ＰＩＤ制御構造に関する部分については、図６に示した通りである（よって、図６の各部と同じ部分については当該部分に付した符号と同じ符号を付すことにより説明を省略する。）ので、主に相違点や追加部分等について説明する。
【０１３９】
「ＲＥＦ」は、アーム変位に関する指令（入力）ノードを示しており、定係数の乗数器（アンプ）４９を介することで、位置変位から角度変位に変換された上で、設定部（あるいは指令値設定ノード）５０に送られる。
【０１４０】
設定部５０は、Θｒｅｆ及びその速度指令（図には、「Θｒｅｆ」における「Θ」の上にドット「・」を付して示す。）を、各部（加算器２８、３０等）に送出する。尚、速度指令は、Θｒｅｆの時間微分である（よって、前記した微分器２９は、設定部５０に含まれている。）。
【０１４１】
特性制御手段５１は、サーボ制御特性に係る応答性に関して、時定数、応答周波数、サーボパラメータ等を変更するために設けられている（実際の制御では、ソフトウェア処理としてＣＰＵの計算により実現される。）。本例では、Ｆ１値の変更によって、ＰＩＤパラメータ（Ｋｐ、Ｋｉ、Ｋｄ）を変更するとともに、該パラメータに係る時間的変化の形態について変更することにより、サーボ制御特性を制御する手段である。即ち、電力積低減モード時において、サーボ制御特性に係る応答性を低下させることにより、トルク外乱に対する感度を低下させる。
【０１４２】
特性制御手段５１の支配下で変更されるＰＩＤパラメータについては、係数「Ｋｉ」の値を示す信号が乗算器（あるいは掛算器）５２に送られる。そして、「Ｋｐ」の値を示す信号が乗算器５３に送られ、「Ｋｄ」の値を示す信号が乗算器５４に送られる。
【０１４３】
つまり、乗算器５２では、加算器２８の出力に対して、係数「Ｋｉ」の掛け算が行われ、演算結果が加算器３４に送られる（図６において、乗数器３０の係数「Ｋｉ」の値が変化すると考えれば良い。）。
【０１４４】
同様に、乗算器５３では、加算器３１の出力に対して、係数「Ｋｐ」の掛け算が行われ、演算結果が加算器３４に送られる（図６において、乗数器３２の係数「Ｋｐ」の値が変化すると考えれば良い。）。また、乗算器５４では、加算器３１の出力に対して、係数「Ｋｄ」の掛け算が行われ、演算結果が加算器３６に送られる（図６において、乗数器３３の係数「Ｋｄ」の値が変化すると考えれば良い。）。
【０１４５】
乗数器３７の出力は、切替部５５、リミッタ５６を介してＶＣＭ部３８に送られる。尚、切替部５５において、「▲１▼ｉｎ」で示す入力端子が選択された状態において外乱が供給され、乗数器３７の出力端子が選択された状態において、該乗数器の出力がリミッタ５６を介してＶＣＭ部３８に送られる。そして、「▲１▼ｏｕｔ」で示す出力端子から一巡伝達関数に係る出力が取得される。また、リミッタ５６は、モータ電流の最大値及び最小値を規定することで許容範囲（あるいは振幅）について制限するために設けられている。
【０１４６】
ＶＣＭ部３８の出力Θは、モニター部５７に送られて表示される。尚、モニター部５７には、設定部５０からのΘｒｅｆも入力されるので、ΘｒｅｆとΘとを比較しながら観測することができる。また、ＶＣＭ部３８の速度出力（「Θ」の上に「・」を付して示す。）は、モニター部５８に送られて表示され、該モニター部には、設定部５０からの速度指令も入力されるので、両者を比較しながら観測することができる。ＶＣＭ部３８の出力「ＩｍＯｕｔ」は、アンプを含むモニター部５９で表示される。
【０１４７】
この他、加算器２８の出力（「ｐｅｓ」で示すポジションエラー信号）は、アンプを含むモニター部６０で表示される。
【０１４８】
尚、「フィードフォワード制御」を考慮する場合には、例えば、Θｒｅｆに所定の係数値を掛けて、乗数器３７の出力に加算したものを切替部５５、リミッタ５６を経てＶＣＭ部３８に送出すれば良い。
【０１４９】
［数４］式に示したように、サーボ制御特性に係る応答周波数Ｆ１の値を変更すると、ＰＩＤパラメータの値が変化し、Ｆ１値を小さくすれば、トルク外乱に対するサーボ系の応答感度が低くなる。例えば、上記方法（Ａ１）の場合は、Ｆ１の値を、ｆ１からｆ２へと瞬時に変更している。
【０１５０】
図１５及び図１６は、上記方法（Ａ１）を採用した場合のシミュレーション結果を示すものである。
【０１５１】
各パラメータの設定については以下の通りである。
【０１５２】
（ｉ）待機（Ｉｄｌｉｎｇ）モード時、ｆ１＝１０００［Ｈｚ］
Ｆ１はトラッキング能力を決定する重要なパラメータであるため、変更前の値（ｆ１）は可能な限り大きいことが望ましい。
【０１５３】
（ｉｉ）電力積低減モードへの変更後のＦ１値、つまり、ｆ２＝ｆ１／１００＝１０［Ｈｚ］
ｆ２値については、ＶＣＭの動作上支障が無い範囲において極力小さい値とする（本例では、ｎ＝１００としている。）。
【０１５４】
（ｉｉｉ）待機モード時及び電力積低減モード時の極角φ＝０［度］
この値については、待機モード時（電力積低減モードでない場合）における、トラッキング性能を可能な限り引き出せる値に設定する。つまり、ＶＣＭ特性、ヘッドアームを含む機構要素の特性等を、総合的に判断してφ値を決定しなければならないが、ここでは位相余裕を十分に確保することのできる値として、０度とする。
【０１５５】
（ｉｖ）Ｆ１の変更時刻「ｔ１＝１［ｓｅｃ］　」
（ｖ）外乱ＤＴｒｒｏについては、歪なしの正弦波（ｓｉｎ関数で表される）とし、その周波数「Ｆｒｒｏ」はスピンドルモータ回転周波数（Ｆｚ）と等しく、「Ｆｒｒｏ＝Ｆｚ＝７５［Ｈｚ］」とする。
（ｖｉ）外乱ＤＴｒｒｏの大きさは、トラックピッチ（Ｔｐ）の１００倍（ｐ−ｐ値）とし、Ｔｐ＝０．５［μｍ］として、５０［μｍ］ｐ−ｐとする。
【０１５６】
（ｖｉｉ）アーム長は、４．８５［ｍｍ］とする。
【０１５７】
図１５は、切替前後におけるＶＣＭ位置偏差（ヘッドの位置偏差に相当する。）の時間的変化について一例を示すものである（モニター部６０で観測される。）。
【０１５８】
ｔ＝ｔ１の時点迄は、位置偏差が約２０［μｍ］ｐ−ｐであったものが、切替直後の過渡応答で７３０μｍにも達している。
【０１５９】
尚、本例では、見易いように、位置偏差の大きい場合を示しているが、この値が大きすぎて問題となる場合には、切替後のｆ２値を、もう少し大きい値に変更すれば良い。
【０１６０】
図１６は、切替前後のＶＣＭ電流について、時間的変化の一例を示すものである（モニター部５９で観測される。）。
【０１６１】
切替前には、６０［ｍＡ］ｐ−ｐ程度、流れていたＶＣＭ電流が、切替直後から６［ｍＡ］ｐ−ｐとなり、約１０分の１に低減していることが分かる。
【０１６２】
また、このシミュレーション結果において、電流の過渡応答は殆ど見られない。
【０１６３】
以上のシミュレーション結果より、応答周波数Ｆ１についての切替前後における各ノード（ＶＣＭ電流や、ＶＣＭ位置偏差等）の過渡応答がシステムにとって問題とならない範囲に収まっている場合には、方法（Ａ１）は簡便であって、かなり有効である。
【０１６４】
勿論、（Ａ１）の採用において、もしも何らかの問題がある場合には、切替前後のＰＩＤ制御系の周波数特性変化量（｜ｆ１−ｆ２｜）を、より小さい値に設定したり、あるいは方法（Ａ２）に示す多段階切替方式を採用することで対応が可能である。
【０１６５】
次に、方法（Ｂ）について説明する。
【０１６６】
周波数特性に係るＦ１値の変更方法については各種考えられるが、ここでは時間ｔの一次関数（これを「Ｆ１（ｔ）」と記す。）に従って変更する方式を示す。尚、Ｆ１（ｔ）の形として一次関数以外の各種関数形「Ｆ１（ｔ，ｔ１，ｔ２，ｆ１，ｆ２）」を用いることも勿論可能であるが、分かり易さを考慮して、以下では、下式に示す一次関数Ｆ１（ｔ）を用いることにする。
【０１６７】
【数５】

【０１６８】
尚、本式は変更開始時刻「ｔ＝ｔ１」での周波数が「ｆ１」であって、変更終了時刻「ｔ＝ｔ２」での周波数が「ｆ２」である場合を示しており、「ｆ１＞ｆ２」が、周波数を小さくする場合（電力積低減モード時）を表すが、これとは逆に、「ｆ１＜ｆ２」は、周波数を大きくする場合（待機モードへの回復）を表す（つまり、上式は両方の場合を包含する表現である。よって、以下では、ｆ１とｆ２との大小関係に関わらず、ｆ１が変更前の値、ｆ２が変更後の値を示すものと定義する。）。
【０１６９】
各パラメータの設定について、以下に示す（ｉｖ’）以外は、上記（ｉ）乃至（ｉｉｉ）、（ｖ）乃至（ｖｉｉ）と同じである。
【０１７０】
（ｉｖ’）Ｆ１値の変更開始時刻は、「ｔ１＝０．５［ｓｅｃ］」であり、Ｆ１値の変更終了時刻は、「ｔ２＝ｔ１＋０．５＝１［ｓｅｃ］」である。
【０１７１】
Ｆ１値を連続的に変更する場合において、シミュレーションに用いるブロック図については、図１４に示す特性制御手段５１において、［数４］式及び［数５］式に従って変化するＰＩＤパラメータが、対応する乗算器５２、５３、５４にそれぞれ供給される（それ以外の事項は、既述した通りである。）。
【０１７２】
図１７乃至図１９は、上記方法（Ｂ）を採用した場合のシミュレーション結果を示すものである。
【０１７３】
図１７は、Ｆ１（ｔ）の時間的変化を示しており、ｔ＝０．５からｔ＝１にかけてＦ１の値が１００分の１に低減される様子を示す。
【０１７４】
図１８は、変更前後におけるＶＣＭ位置偏差の時間的変化について一例を示すものである。
【０１７５】
ｔ１＝０．５の時点迄の、待機モードにおける位置偏差は、約５μｍであり、変更直後（ｔ２＝１）において、それ程大きな過渡応答は見られない。また、変更後の位置偏差は最大で２５０μｍ程度とされる。
【０１７６】
尚、本例では、見易いように、位置偏差の大きい場合を示しているが、Ｆ１値変更後の偏差に関して、これがシステム上問題であるならば、ｆ２値をもっと大きい値にすれば良い（あるいは、これとは逆に、システム上許容されるのであれば、ｆ２値をもっと小さい値にすることも可能である。）。また、移行時間（あるいは変更時間）を示す、「ｔ２−ｔ１」の長さを変更することにより、詳細な調整を行うことが可能である。即ち、周波数の比率又は変更時間（移行時間）の一方、あるいは両者について調整を行えるという利点がある。
【０１７７】
図１９は、変更前後におけるＶＣＭ電流について時間的変化の一例を示すものである。
【０１７８】
Ｆ１値の変更前には、６６［ｍＡ］ｐ−ｐ流れていた電流が変更直後から２［ｍＡ］ｐ−ｐとなり、約３３分の１に低減していることが分る。
【０１７９】
また、本方式での電流の過渡応答についても、上記（Ａ１）の場合と同様に小さい。
【０１８０】
システム上許容される範囲内において、可能であればｆ２値をさらに小さくして、ＶＣＭ電流を低減しても良い。
【０１８１】
以上のシミュレーション結果より、ＰＩＤ制御での周波数特性（Ｆ特）を落とすことでＶＣＭ電流を、数十分の１程度も低減できることが分る。
【０１８２】
尚、ここでは、ＲＲＯ外乱に対する電力低減効果について検討したが、現実にはＮＲＲＯ成分（スピンドルモータの回転に対する非同期性成分）についても同様の電力低減効果が得られるので、電力積低減モードにおいて、さらに省電力化を実現することが可能である。
【０１８３】
以上の例では、待機モードにおけるサーボ系の周波数特性（Ｆ特：Ｆ１）に関して、電力積低減モードへと移行する際に、Ｆ１値を下げる方法につき説明したが、待機モードへの回復において応答性を元に戻したり、又は、Ｒ／Ｗモードへの遷移が必要になった場合には、Ｆ１値を上げる必要が生じる。
【０１８４】
図２０乃至図２２は、そのような状況について、上記方法（Ｂ）を採用した場合のシミュレーション結果を例示したものである。
【０１８５】
図２０は、Ｆ１の値について、その変更前後及び変更中の値を示す図である。
【０１８６】
図示のよう、ｆ１＝１０［Ｈｚ］からｆ２＝１０００［Ｈｚ］に向かって直線的に増加していることが分かる。尚、図示は省略するが、応答周波数の変化に伴って、ＰＩＤパラメータも増加して所定値（ｆ２に対応する値であり、［数４］式でＦ１＝ｆ２としたときの値。）に到達する。
【０１８７】
図２１は、Ｆ１値の変更前後におけるＶＣＭ位置偏差について一例を示したものである。
【０１８８】
ｔ１＝２の時点迄は位置偏差が、約１３０μｍであり、変更直後で大きな過渡応答は見られず、５μｍ程度となっている。
【０１８９】
図２２は、Ｆ１値の変更前後におけるＶＣＭ電流について一例を示したものである。
【０１９０】
変更前には、３［ｍＡ］ｐ−ｐ程度流れていた電流が、変更後からやや時間が経つとほぼ６６［ｍＡ］ｐ−ｐとなり、約２０倍に増加していることが分かる。また、電流値の過渡応答についても小さい。
【０１９１】
次に、電力積低減モードに関する計算処理とタイミングについて説明する。
【０１９２】
既述した様に、他己録メデイアを採用するＨＤＤでは、ＲＲＯ成分が大きいので高精度のトラッキングを実現する為には、ＲＲＯ成分に対応するスピンドルモータの一回転分のパターン（１周期分の検出パターン）をメモリに記憶させておき、この記憶されたパターンを用いて、ＲＲＯ成分を打ち消す制御方法（所謂Ｎｏｔｃｈ　Ｆｉｌｔｅｒ法）の採用が好ましい。
【０１９３】
ＲＲＯ成分をキャンセルする上記パターンの記録法としては、例えば、下記に示す方法がある。
【０１９４】
・パターン記録モードを別途に用意して、キャンセル用のパターンを１回だけ記録させる静的（スタティック）な方法（例えば、スタート・アップ直後にパターンの記録を実行してしまう方法等。）
・待機モードや、Ｒ／Ｗモード時には、常に上記パターンに対応したデータを更新して記録する動的（ダイナミック）な方法。
【０１９５】
いずれの方法であっても、ディスクトラック上のセクタ番号（以下、「ＳＥＣ＿Ｎｏ」と記す。）をインデックス（指標）としてＲＲＯ成分に対応するパターンがメモリに記憶される。
【０１９６】
但し、電力積低減モード時には、上記パターンに対応するデータが正確に得られなくなることに注意する必要がある。つまり、ノッチフィルタを用いたＤＴｒｒｏ圧縮制御については、本モードの開始と共に機能を停止する必要がある。従って、待機モードにおける電力積低減モードへの移行時や、該モードから別モードへの移行時において、移行の開始や終了のタイミングと、ノッチフィルタのオン動作／オフ動作（あるいは機能回復／機能停止）のタイミングとの間で同期を取ることが必要である。そのためにはセクタ番号（ここではトラック毎に、同一トラック上での起点となるセクタを、セクタ番号「０」のセクタとして定義し、該セクタをトリガセクタとする。）でトリガするのが良い。
【０１９７】
具体的には、サーボ制御処理への割り込みトリガについても、セクタ検出信号から生成されることを利用して同期取りを行うことができる。
【０１９８】
図２３乃至図２６は、電力積低減モードへの移行や、該モードの終了についてシーケンス例を示したものであり、ホスト装置、ＨＤＤ制御系、ＰＩＤ制御系（図６、図７参照）の三者間での信号や情報の流れを概略的に示したものである。尚、ＨＤＤ制御系（ＰＩＤ制御系を除く。）については、待機モード時での処理とし、また、ＰＩＤ制御系については割り込み処理とする。また、「Ｏｎ動作」とは電力積低減モードへの移行時の動作を示し、「Ｏｆｆ動作」とは電力積低減モードの終了時の動作を示す。
【０１９９】
ＨＤＤにおいて、電力積低減モードへの移行又は待機モードへの復帰を開始するためのトリガ方法については、下記の２種類が挙げられる。
【０２００】
・ユーザーがホスト装置を経由してコマンドを発行する方法
・ＨＤＤ自身が状態をチェックして自動的にモード遷移を行う方法。
【０２０１】
つまり、電力積低減モードからの復帰時に、アクセス時間を極力短くしたい場合等では、コマンドをユーザーが発行する方法が好ましい。
【０２０２】
また、ＨＤＤ制御系において、ホスト装置からＶＣＭ部へのシークモードやＲ／Ｗモードへの遷移コマンド等を常に監視していて、予め決められている時間内に、これらのコマンドがホスト装置からＨＤＤに送られて来ない場合に、ＨＤＤ自らが判断して、電力積低減モードに入るように構成すると、ユーザーの操作手間を省くことができる。そして、待機モードへの復帰時間及び別モードへの遷移に要する時間を短くすることができる（結果として、ユーザーの待ち時間を短くすることが可能となる。）。
【０２０３】
図２３は、上記の方法（Ａ１）又は方法（Ａ２）を採用する場合における、シーケンス例を示す。
【０２０４】
先ず、「Ｏｎ動作」の場合には、ホスト装置からＨＤＤ制御系に対して電力積低減モードの開始コマンドが発行され、これを受けてＨＤＤ制御系では、電力積低減モードの開始を示すフラグ（以下、これを「ＳＴＴ＿Ｆｌｇ」と記す。）をオン（あるいはセット）とする。ＰＩＤ制御系では、「ＳＥＣ＿Ｎｏ＝＝０」（「＝＝」は等値の比較演算子を意味する。）をガード条件として、該条件が成立する場合に、ＰＩＤパラメータ（Ｋｐ、Ｋｉ、Ｋｄ）を段階的（２段階以上）に変更するとともに、ノッチフィルタをオフにする（機能停止）。尚、ここで、「ガード条件」とは、「Ａ　ならば　Ｂ」において、「Ｂ」に進むための条件「Ａ」を意味し、例えば、『「Ｘ＞０」ならば「Ｙ＝１」とする』の場合に、条件「Ｘ＞０」を示す。
【０２０５】
ＰＩＤ制御系での演算結果が、ＶＣＭ部３８に出力された後、該制御系からＨＤＤ制御系に対する信号により、電力積低減モードの停止を示すフラグ（以下、これを「ＳＴＰ＿Ｆｌｇ」と記す。）がオフ（あるいはリセット）とされ、ホスト装置には、電力積低減モードが開始されたことを示す連絡がステータス情報等によって行われる。
【０２０６】
「Ｏｆｆ動作」の場合には、ホスト装置からＨＤＤ制御系に対して電力積低減モードの停止コマンドが発行され、これを受けてＨＤＤ制御系では、電力積低減モードの停止を示すフラグＳＴＰ＿Ｆｌｇをオン（あるいはセット）とする。ＰＩＤ制御系では、「ＳＥＣ＿Ｎｏ＝＝０」をガード条件として、該条件が成立する場合に、ＰＩＤパラメータ（Ｋｐ、Ｋｉ、Ｋｄ）を段階的に変更する（所定値又は電力積低減モードへの移行前の値への復帰）とともに、ノッチフィルタをオンにする（機能回復）。
【０２０７】
ＰＩＤ制御系での演算結果が、ＶＣＭ部３８に出力された後、該制御系からＨＤＤ制御系に対する信号により、ＳＴＴ＿Ｆｌｇがオフ（あるいはリセット）とされ、ホスト装置には、電力積低減モードが停止されたことを示す連絡がステータス情報等によって行われる。
【０２０８】
図２４及び図２５は、上記の方法（Ｂ）を採用する場合における、シーケンス例を示す。
【０２０９】
先ず、図２４に示す「Ｏｎ動作」の場合には、ホスト装置からＨＤＤ制御系に対して電力積低減モードの開始コマンドが発行され、これを受けてＨＤＤ制御系では、ＳＴＴ＿Ｆｌｇをオン（あるいはセット）とする。そして、時間ｔの起点（ｔ＝０）が決定される。
【０２１０】
ＰＩＤ制御系では、「ＳＥＣ＿Ｎｏ＝＝０　＆＆　ＳＴＴ＿Ｆｌｇ＝＝Ｏｎ」をガード条件として、該条件が成立する場合（「＆＆」は論理積演算を示す。）に、下記の処理を行う。
【０２１１】
（ａ）「ｔ＋＋」（時刻のインクリメント）
（ｂ）Ｆ１（ｔ）の計算（［数５］式参照）
（ｃ）ＰＩＤパラメータの計算（［数４］式参照）
（ｄ）ノッチフィルタのオフ（機能停止）。
【０２１２】
尚、時間「ｔ」については、サーボサンプリングタイム「Ｔｓ」を単位時間とする離散時間処理系を採用し、「ｔ２−ｔ１＝ｎ・Ｔｓ」（ｎは自然数を示す。）に同期化している。
【０２１３】
その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力される。
【０２１４】
次のガード条件は、「ｔ＞０　＆＆　ｔ＜ｔ２　＆＆　ＳＴＴ＿Ｆｌｇ＝＝Ｏｎ」であり、該条件の成立時に行われる処理は上記（ａ）、（ｂ）、（ｃ）とされ、その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力される。
【０２１５】
次に来るガード条件は、「ＳＥＣ＿Ｎｏ＝＝０　＆＆　ｔ＝＝ｔ２　＆＆　ＳＴＴ＿Ｆｌｇ＝＝Ｏｎ」であり、該条件の成立時に行われる処理は上記（ｂ）、（ｃ）とされ、その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力される。
【０２１６】
それから、ＰＩＤ制御系からＨＤＤ制御系への信号により、ＳＴＰ＿Ｆｌｇがオフ（あるいはリセット）とされ、ホスト装置には、電力積低減モードが開始されたことを示す連絡がステータス情報等によって行われる。
【０２１７】
図２５に示す「Ｏｆｆ動作」の場合には、ホスト装置からＨＤＤ制御系に対して電力積低減モードの停止コマンドが発行され、これを受けてＨＤＤ制御系では、ＳＴＰ＿Ｆｌｇをオン（あるいはセット）とする。そして、時間ｔの起点（ｔ＝０）が決定される。
【０２１８】
ＰＩＤ制御系では、「ＳＥＣ＿Ｎｏ＝＝０　＆＆　ＳＴＰ＿Ｆｌｇ＝＝Ｏｎ」をガード条件として、該条件が成立する場合に、上記（ａ）、（ｂ）、（ｃ）の処理が行われ、その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力される。
【０２１９】
次のガード条件は、「ｔ＞０　＆＆　ｔ＜ｔ２　＆＆　ＳＴＰ＿Ｆｌｇ＝＝Ｏｎ」であり、該条件の成立時に行われる処理は上記（ａ）、（ｂ）、（ｃ）とされ（但し、この場合には、Ｆ１値を経過時間とともに上げる処理を行う。）、その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力される。
【０２２０】
次に来るガード条件は、「ＳＥＣ＿Ｎｏ＝＝０　＆＆　ｔ＝＝ｔ２　＆＆　ＳＴＰ＿Ｆｌｇ＝＝Ｏｎ」であり、該条件の成立時に行われる処理は上記（ｂ）、（ｃ）の他に、下記に示す（ｅ）である。
【０２２１】
（ｅ）ノッチフィルタのオン（機能回復）。
【０２２２】
その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力され、ＰＩＤ制御系からＨＤＤ制御系に対する信号により、ＳＴＴ＿Ｆｌｇがオフ（あるいはリセット）とされ、ホスト装置には、電力積低減モードが停止（あるいは終了）されたことを示す連絡がステータス情報等によって行われる。
【０２２３】
本例のように、［数４］式及び［数５］式の計算処理については、ＰＩＤ制御系の演算処理が行われる割り込み処理の中で行うのが好ましい。こうすれば、Ｆ１値等の変更時間「ｔ１−ｔ２」の制御を精度良くリアルタイム処理に従って実現することができるとともに、ＰＩＤパラメータ（Ｋｐ，Ｋｉ，Ｋｄ）について各々の同期を確保することができる。
【０２２４】
しかしながら、ＰＩＤ制御系に係る演算処理用ＣＰＵについて充分な処理速度や処理時間が得られない場合、例えば、処理時間に課された条件が厳しい場合には、ＨＤＤ制御系における待機モード時のループ処理内で、［数４］式と［数５］式を用いてＦ１やＰＩＤパタメータの値を計算し、その結果をバッファ（ＫｄＢｕｆ，ＫｉＢｕｆ，ＫｐＢｕｆ）に先ず格納しておく。そして、３個のＰＩＤパラメータ全てについての計算が終わったら、これをＰＩＤ制御系の割り込み処理に伝え（フラグ等を用いる。）、割り込み処理の中で一挙に、バッファの格納値をＰＩＤ変数（Ｋｐ，Ｋｉ，Ｋｄ）に複写（ｃｏｐｙ）する方法を採用すれば良い。
【０２２５】
図２６は、この方式に従う処理シーケンスを例示したものである（Ｏｎ動作の場合だけを示す。）。
【０２２６】
ホスト装置からＨＤＤ制御系に対して電力積低減モードの開始コマンドが発行され、これを受けてＨＤＤ制御系では、下記に示す処理を行う。
【０２２７】
（ｆ）「ｔ＝ｔ１」及び「ＣＡＬ＿Ｆｌｇ＝０」（但し、「＝」は代入演算子を示す。）
（ｂ）Ｆ１（ｔ）の計算（［数５］式参照）
（ｃ）ＰＩＤパラメータの計算（［数４］式参照）
（ｇ）（ｃ）の計算結果についてバッファへの格納
尚、（ｆ）では、ｔ１のセットが行われ、また、（ｂ）、（ｃ）、（ｇ）の計算終了を示すフラグ「ＣＡＬ＿Ｆｌｇ」には「０」がセット（リセット）される（計算未終了を意味する。）。
【０２２８】
そして、ＳＴＴ＿Ｆｌｇをオン（あるいはセット）とし、上記計算の終了によりＣＡＬ＿Ｆｌｇが「１」にセットされて、ＰＩＤ制御系に通知される。
【０２２９】
ＰＩＤ制御系では、「ＳＥＣ＿Ｎｏ＝＝０　＆＆　ＣＡＬ＿Ｆｌｇ＝＝１」をガード条件として、該条件が成立する場合に、下記の処理を行う。
【０２３０】
（ｈ）バッファ格納値のＰＩＤパラメータへのコピー
（ｄ）ノッチフィルタのオフ（機能停止）。
【０２３１】
尚、（ｈ）では、Ｋｐ、Ｋｉ、Ｋｄに対応する各バッファをＫｐＢｕｆ、ＫｉＢｕｆ、ＫｄＢｕｆと記すとき、ＫｐＢｕｆからＫｐへ、ＫｉＢｕｆからＫｉへ、ＫｄＢｕｆからＫｄへの数値データの代入がそれぞれ行われる。
【０２３２】
その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力され、ＣＡＬ＿ＦｌｇがゼロにリセットされてＨＤＤ制御系に通知される。
【０２３３】
次に行う処理は、上記（ａ）、（ｂ）、（ｃ）、（ｇ）である。
【０２３４】
そして、上記（ｂ）、（ｃ）の計算終了によりＣＡＬ＿Ｆｌｇが「１」にセットされて、ＰＩＤ制御系に通知される。
【０２３５】
次のガード条件は、「ｔ＜ｔ２　＆＆　ＣＡＬ＿Ｆｌｇ＝＝１」であり、該条件の成立時に行われる処理は、上記（ｈ）とされ、その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力される。そして、ＣＡＬ＿ＦｌｇがゼロにリセットされてＨＤＤ制御系に通知される。
【０２３６】
次に行う処理は、上記（ａ）、（ｂ）、（ｃ）、（ｇ）である。
【０２３７】
そして、上記（ｂ）、（ｃ）の計算終了によりＣＡＬ＿Ｆｌｇが「１」にセットされて、ＰＩＤ制御系に通知される。
【０２３８】
次のガード条件は、「ＳＥＣ＿Ｎｏ＝＝０　＆＆　ｔ＝＝ｔ２　＆＆　ＣＡＬ＿Ｆｌｇ＝＝１」であり、該条件の成立時に行われる処理は、上記（ｈ）とされ、その後、ＰＩＤ演算の結果が、ＶＣＭ部３８に出力される。そして、ＣＡＬ＿ＦｌｇがゼロにリセットされてＨＤＤ制御系に通知される。
【０２３９】
ＳＴＰ＿Ｆｌｇのリセットにより、ホスト装置には、電力積低減モードが開始されたことを示す連絡がステータス情報等によって行われる。
【０２４０】
尚、Ｏｆｆ動作については、フラグ（ＳＴＴ＿Ｆｌｇ、ＳＴＰ＿ＦＬｇ）の設定等の違いを除いて基本的な手順は同様である（従って、図示及び説明を省略する。）。
【０２４１】
次に、データの読み出しや書き込みを開始するまでの待ち時間について、これを少しでも短くするための工夫として、計算処理の開始及び終了のタイミングについて説明する。
【０２４２】
先ず、上記方法（Ａ１）又は（Ａ２）を採用する場合には、ＨＤＤがホスト装置からＶＣＭに係る電力積低減モードの開始コマンドを受け取って、対象トラックにおける最初のセクタ番号０を検出した時点での割り込みの中で応答周波数Ｆ１の値を段階的に切り替える。切替後におけるＰＩＤパラメータ値は事前に知られているので、［数４］式や［数５］式のような計算を、割り込み処理の中で行う必要はない。また、ＲＲＯ用ノッチフィルタのオン動作／オフ動作についても同じタイミングをもって切り替えることで両者の同期を取ることができる（図２３には、このタイミング処理についても示してある。）。
【０２４３】
方法（Ｂ）を採用する場合の電力積低減モードの開始や終了のタイミング（割り込み処理内）については、ＨＤＤがホスト装置から電力積低減モードへの開始コマンドを受け取って、対象トラックにおける最初のセクタ番号０を検出した時点での割り込みの中で、上記［数４］式、［数５］式を用いてＰＩＤパラメータ（Ｋｐ，Ｋｉ，Ｋｄ）の値について、各々の計算を開始すると共に、ノッチフィルタをオフ動作とする（機能停止）。
【０２４４】
一方、電力積低減モードの終了時点（待機モードへと復帰する時）の後、応答周波数Ｆ１がｆ２に達した直後（この場合には、「ｆ１＜ｆ２」である。）に、最初のセクタ番号０を検出した時点での割り込み処理中でノッチフィルタをオン動作とする。
【０２４５】
ノッチフィルタのオン動作や、待機モードへの回復のタイミングをできるだけ早くするための工夫としては、電力積低減モードの終了時点（ｔ＝ｔ２）でちょうどヘッド位置がセクタ番号０を読めるように遷移時間を「ｔ２−ｔ１＝ｎ・Ｔｓ」（「Ｔｓ」はサーボサンプリングタイムを示す、「ｎ」は正整数を示す。）のように同期化しておくことが好ましい。これは、スピンドルモータの回転待ち時間をゼロにして、ＨＤＤのアクセス時間を少なくする為である。
【０２４６】
この意味では、セクタ番号０のタイミングで同期するのでは無く、ＨＤＤのアクセス時間が最小となるセクタ番号ｋで同期とった方が良い場合も勿論考えられる。
【０２４７】
尚、図２６に示す例においても、基本的には、割り込み内での処理と同様、できる限りセクタ番号０と同期を取ることが好ましい。しかし、待機モード時のループ処理内での他の処理に時間をとられる場合や、ＣＰＵの処理時間に余裕のない場合には、完全な同期取りが難しくなる。
【０２４８】
そこで、補助的な手段ではあるが、ＶＣＭに係る電力積低減モードの終了時点、つまり、「ｔ＝ｔ２」の時点が、ちょうどセクタ番号０の一つ（あるいは二つ）手前に相当する時点となるように、遷移時間「ｔ２−ｔ１＝（ｎ−１）・Ｔｓ」又は「ｔ２−ｔ１＝（ｎ−２）・Ｔｓ」（「Ｔｓ」はサーボサンプリングタイムを示し、「ｎ」は自然数を示す。）と同期化しておく方法が挙げられる。こうすれば、セクタ番号０のタイミングにて、ノッチフィルタとの間で同期化が可能となる。
【０２４９】
以上、各方法における制御シーケンスについて説明したが、例えば、バッテリ駆動式の携帯機器等では、バッテリ寿命（残量）を左右する電力積が重要な指標であり、電力積をどれくらい低減できるかが問題となる。
【０２５０】
ＶＣＭの消費電力を「Ｗｖｃｍ」とし、ＶＣＭの電流値を「Ｉｖｃｍ」、印可電圧を「Ｖｖｃｍ」と記すとき、下式が成り立つ。
【０２５１】
【数６】

【０２５２】
今、待機モード及びＲ／Ｗモードの継続時間（各モードの時間和）を「Ｔａｌｌ」と記し、これに対応する電力積を「ＷＨｖｃｍ」とすると、これは下式で定義される。
【０２５３】
【数７】

【０２５４】
尚、本式から分かるように、ＷＨｖｃｍがＴａｌｌに比例するので、この値がある程度大きくなった時点で、バッテリ切れとなる。
【０２５５】
前記した電力積低減モードを導入するとともに、待機モード及びＲ／Ｗモードの継続時間から、電力積低減モードの継続時間を除いた期間長を、Ｔａｌｌのα倍（０＜α＜１）とする。そして、電力積低減モード時におけるＶＣＭ電流が、Ｉｖｃｍのβ倍（０＜β≦１）であるとき、［数７］式を基準としたＶＣＭに係る電力積の比率を「γ」と記すと、下式が得られる。
【０２５６】
【数８】

【０２５７】
例えば、電力積低減モードとされる期間長が、Ｔａｌｌの７５％を占めるものとし（α＝０．２５）、電力積低減モードでのＶＣＭ電流の低減率が３％（β＝０．０３）である場合には、γ＝０．２７２５となる（つまり、７２．７５％もの電力積削減が可能である。）。
【０２５８】
最後に、ＨＤＤにおける、特定のモードから待機モードに移行して、電力積低減モードに入るべきであるのか、又は、特定モード又は待機モードから退避モードに入るべきであるかの判断基準について説明する。
【０２５９】
先ず、待機モードに遷移する必要性が長時間に亘って無いことが、所定の情報（ユーザー操作等により退避モードへの移行コマンドを発行する場合を含む。）や設定条件等から明らかな場合には、電力積低減モードには入らずに、退避モードに入ってしまう方がＶＣＭの電力消費の観点からは最も有利である。しかしながら、この場合、退避モードから待機モードに戻る際には、ＨＤＤの開始動作（Ｓｔａｒｔ　Ｕｐ）と殆ど同じで、データの読み出し又は書き込み動作が可能となるまでに時間がかかる。
【０２６０】
また、退避モードへの移行を許可するに足る情報が得られない場合や、当該モードに移行しても良いと判断しかねる場合には、待機モードから電力積低減モードに入る方が好ましい。即ち、電力積低減モードの終了により、元の待機モードへと比較的短時間で復帰することが可能であるため、待ち時間が短くて済む。　そして、ホスト装置側での工夫として、電力積低減モードから待機モードへの移行コマンド（ホスト装置が発行する）については、他のコマンドに優先して発行することが有効である。また、ＨＤＤ側での工夫としては前述した様に、「ｔ２−ｔ１＝ｎ・Ｔｓ」として同期をとり、スピンドルモータの回転待ち時間を極力減らすことが好ましい。
【０２６１】
尚、ＰＩＤ制御系の電流指令ノードに対してリミッタ（図１４参照）で制限する方法、つまり、ＶＣＭ部への電流指令ノードにリミッタを設けるだけでも、ＶＣＭの電流（や電力）を低減することが、ある程度は可能である。しかしながら、この方法では、ＶＣＭが制御不能の状態に陥った結果として、ＶＣＭに係る位置が大きく変動した場合（所謂非線形発振を起こした場合）、実用上の問題があるので、リミッタを用いた場合でも、本発明に係る制御方法を採用することが好ましい。
【０２６２】
しかして、以上に説明した構成によれば、例えば、下記に示す利点が得られる。
【０２６３】
・携帯機器に搭載されるＨＤＤや、高速アクセス性よりも省電力化が重視される機器等に搭載されるＨＤＤにおいて、ＶＣＭの電力積を低減し、バッテリの寿命を延ばすことが可能である。
【０２６４】
・上記の方法（Ａ２）、（Ｂ）では、ＶＣＭの電力積低減モードへの切替時間を制御することが可能であり、過渡応答量について制御することができる。
【０２６５】
・ＨＤＤの動作モードとして、最も運転時間が長い待機（Ｉｄｌｉｎｇ）モード時において、消費電力の低減が可能であるので、効果的である。
【０２６６】
・電力積低減モードの採用により、ＨＤＤの温度上昇を、ある程度抑えることができる。
【０２６７】
・ＨＤＤでは、該ＨＤＤが本質的に抱えるＲＲＯ外乱に対して、ノッチフィルタ等を用いた外乱圧縮制御系を有するが、該制御系との間で干渉を起こさず（つまり、電力積低減モード時に、外乱圧縮制御系の機能を一時的に停止させることに依る。）に電力積の低減が可能である。
【０２６８】
【発明の効果】
以上に記載したところから明らかなように、請求項１や請求項６に係る発明によれば、一時待機用の動作モードにおいて、トルク外乱に対する感度を低下させて消費電力を低減することができる。そして、該動作モードでは、シーク後のヘッド位置を維持したままの状態であり、動作復帰を短時間に行うことができるので、退避モードからの復帰に比してアクセス時間の増加を防止することができる。
【０２６９】
請求項２に係る発明によれば、一時待機用の動作モードにおいてサーボ制御特性に係る応答周波数を低減することによって、制御構造の複雑化等を伴わずに制御を行うことができる。
【０２７０】
請求項３や請求項７に係る発明によれば、応答性についての段階的な変更により、計算処理上の負担を軽減することができる。
【０２７１】
請求項４や請求項８に係る発明によれば、応答性についての連続的な変更によって、円滑な特性制御を行えるので、過渡応答性の悪化等を防ぐことができる。
【０２７２】
請求項５に係る発明によれば、バッテリ駆動の携帯型機器への適用において、バッテリの消耗を抑えることができる。
【図面の簡単な説明】
【図１】ディスク情報処理装置の基本構成例を示す説明図である。
【図２】装置構成の一例を示す図である。
【図３】ＦＡＴファイルシステムの説明図である。
【図４】モード間の遷移図である。
【図５】サーボ応答制御についての概念的な説明図である。
【図６】図７とともに、サーボ制御系の構成例を示すブロック図であり、本図はＰＩＤ制御構造をもった例を示す。
【図７】ＶＣＭ部の構成例を示す図である。
【図８】Ｓ平面上での極配置例を示す図である。
【図９】ボーデ線図の一例を示す。
【図１０】ＶＣＭ電流値の時間的変化を例示した図である。
【図１１】図１２、図１３とともに、Ｆ１値の変更方法について説明するための図であり、本図は２段階の変更方法を示す。
【図１２】多段階のＦ１値変更方法を示す図である。
【図１３】連続的なＦ１値変更方法を示す図である。
【図１４】シミュレーションに用いた制御構成例を示すブロック図である。
【図１５】図１６とともに、段階的なＦ１値の変更に関するシミュレーション結果を示すものであり、本図は位置偏差の時間的変化を示す。
【図１６】ＶＣＭ電流の時間的変化を示す図である。
【図１７】図１８、図１９とともに、連続的なＦ１値の変更に関するシミュレーション結果を示すものであり、本図はＦ１値の時間的変化を示す。
【図１８】位置偏差の時間的変化を示す。
【図１９】ＶＣＭ電流の時間的変化を示す図である。
【図２０】図２１、図２２とともに、連続的なＦ１値変更（応答性の回復）に関するシミュレーション結果を示すものであり、本図はＦ１値の時間的変化を示す。
【図２１】位置偏差の時間的変化を示す。
【図２２】ＶＣＭ電流の時間的変化を示す図である。
【図２３】段階的なＦ１値の変更方法を採用した場合において、電力積低減モードへの移行又は該モードの終了について処理の流れを例示した説明図である。
【図２４】連続的なＦ１値の変更方法を採用した場合において、電力積低減モードへの移行について処理の流れを例示した説明図である。
【図２５】連続的なＦ１値の変更方法を採用した場合において、電力積低減モードの終了について処理の流れを例示した説明図である。
【図２６】図２４とは別の処理例を示す説明図である。
【符号の説明】
１…ディスク情報処理装置、２…ディスク状記録媒体、４…ヘッド、６…駆動源、５１…特性制御手段[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a technology for realizing low power consumption in an information processing device using a disk-shaped recording medium.
[0002]
[Prior art]
A hard disk drive (hereinafter, referred to as “HDD”) is used as a device used in a computer device and constituting a part of the architecture. The device is applied to various devices such as a device for processing data such as video information and audio information, a mobile communication terminal device, a game machine, and particularly to portable electronic devices (so-called mobile devices and the like). , Its applications are expanding.
[0003]
Under these circumstances, low power consumption of HDDs has become one of the important requirements.
[0004]
In the positioning control of the magnetic head with respect to the disk-shaped recording medium, the magnetic head is controlled by a seek (SEEK) control for moving the magnetic head to a desired track on the disk, and by moving the magnetic head to a recording track after the seek. (Tracking) control for causing the disk to follow, but if such control is always performed without a break, the power consumption (power time product) increases. Means are taken to reduce power consumption by retracting the head.
[0005]
For example, there is a method of preparing a retreat mode (or a SLEEP mode) for retreating the head to a loader (a member necessary for loading the head arm) so that power is not consumed, and coping with an unused state. .
[0006]
[Problems to be solved by the invention]
By the way, when the HDD is used, for example, mounted on a portable device or the like, an effective method or the like for reducing power consumption and returning to a required operation (read / write operation or the like) in a short time. However, it has not been proposed so far, and for example, there is a problem in the following points.
[0007]
In general, in HDDs, it is essential to shorten the access time for the purpose of speeding up. Therefore, it cannot be mounted on a portable device or the like without considering any restrictions on the power supply capacity.
[0008]
In addition, while the rotation of the spindle motor is stopped, the control of the drive source (VCM) is also stopped, and the arm is retracted to the loader (Loader). Although power consumption can be reduced, according to this method, the time required to return to the operation is almost the same as the time required to start the operation (Start Up) of the HDD, so that the access time (access time) is significantly increased. (Such as a longer waiting time of the user).
[0009]
Therefore, when the apparatus is driven by a battery as in a portable device or the like, there is a possibility that it is difficult to effectively use the limited power supply capacity or the convenience may be impaired.
[0010]
Accordingly, it is an object of the present invention to achieve both power saving and shortened operation return time in a disk information processing apparatus.
[0011]
[Means for Solving the Problems]
In order to solve the above-described problem, the present invention provides a temporary standby operation mode for maintaining a head position after a seek without retracting the head, and in this mode, a response related to a servo control characteristic is improved. It is provided with a characteristic control means for lowering it.
[0012]
Therefore, according to the present invention, the response sensitivity to the torque disturbance is reduced to reduce the power consumption, and the operation is returned from the state where the head position after the seek is maintained, so that the operation is shifted to the next operation in a short time. Can be done.
[0013]
BEST MODE FOR CARRYING OUT THE INVENTION
The present invention relates to an information processing apparatus for reading and writing information by using a disk-shaped recording medium (such as a magnetic disk or an optical disk), and a head positioning mechanism for a head unit or a track of the disk-shaped recording medium. An object of the present invention is to provide a servo system including a drive source and realize power saving in control of the servo system.
[0014]
In the case of the HDD, mechanical elements having relatively large power consumption, such as a spindle motor provided for rotationally driving the magnetic disk and a VCM (voice coil motor) constituting an actuator of the head, are incorporated in the apparatus as main components. However, in the present invention, high-speed access is prioritized as in the case of an HDD used in a device that has a relatively long access time. When it is not necessary to reduce the power consumption, the responsiveness related to the servo characteristics of the VCM is controlled in a specific operation mode.
[0015]
Therefore, the present invention is preferably applied to a case where the apparatus is driven by a battery, such as a portable apparatus. However, the present invention is not limited to this, and can be applied to various apparatuses that require power saving. That is, as long as the device does not always require high-speed access performance to the disk-shaped recording medium, it can be applied not only to portable devices but also to a wide range of applications (for example, operating modes using the power reduction method according to the present invention are A device that allows the user to freely specify or select, etc.).
[0016]
FIG. 1 conceptually shows a basic configuration example of a disk information processing apparatus 1 according to the present invention, and can be applied to, for example, a drive device for a magnetic disk (hard disk or the like).
[0017]
The disk information processing apparatus 1 includes a head unit 3 for reading (or reading) information from the disk-shaped recording medium 2 or recording information on the disk-shaped recording medium 2. For example, examples of the disk-shaped recording medium include a magnetic disk, an optical disk, a magneto-optical disk, and a phase-change disk.
[0018]
The head unit 3 is provided with a mechanism for moving the head by a seek operation (SEEK) with respect to the disk-shaped recording medium 2 and positioning the head at a desired position. For example, in the HDD, a rotating mechanism including a support arm 5 of a head (magnetic head) 4 and a VCM (voice coil motor) as a driving source 6 are provided. That is, a servo system including a drive mechanism 6 and a positioning mechanism of the head 4 with respect to tracks on the disk-shaped recording medium 2 is provided.
[0019]
The control system of the head unit 3 includes a command processing and setting unit 7 and a control unit 8 of a head seek system.
[0020]
The command processing and setting unit 7 receives a command from a higher-level command unit (such as a host device) (not shown), and sends a signal to the control unit 8 according to the content of the command.
[0021]
The control means 8 generates a speed command profile including an acceleration section and a deceleration section, or an acceleration section, a constant speed section, and a deceleration section, and controls the head seek system using the profile as a control command. It has a unit 9 and a servo control unit 10.
[0022]
The command value generator 9 generates a command profile such as acceleration, speed, and position, and the output is sent to the servo controller 10.
[0023]
The servo controller 10 includes a drive circuit (driver) for the drive source 6, a low-pass filter, and the like, and has a configuration depending on a control target. In the case of an HDD, an output signal is supplied to the VCM to control the driving of the rotating mechanism including the head arm. That is, the head 4 is positioned with respect to the disk-shaped recording medium 2 using the driving force generated by the motor system including the driving source 6. Although not shown, a driving unit for rotating the disk-shaped recording medium 2 is, of course, provided. Further, the present invention is not limited to a magnetic disk, and can be applied to, for example, a drive device using an optical head (or an optical pickup). In this case, the servo control unit 10 mounts the optical head. Position control of the moving base (slide base) is performed.
[0024]
FIG. 2 is a diagram schematically illustrating a configuration example 11 of an information processing system including an HDD and its host device (computer or the like).
[0025]
The HDD 12 is provided with rotation means 14 for a magnetic recording medium (magnetic disk) 13, and a motor is used as a drive source 15. Further, a head unit 16 for writing and reading data to and from the magnetic recording medium 13 and a driving mechanism 17 thereof are provided, and a motor (VCM) is used as a driving source 18.
[0026]
The drive control of the drive sources 15 and 18 is performed according to a control signal from the servo control unit 19.
[0027]
A data channel section 21 and a disk control section 22 are provided as signal processing means 20 for transferring data to and from the head section 16.
[0028]
The data channel unit 21 performs a recording and encoding process for writing an information signal to the magnetic recording medium 13, and converts a processing result into a digital bit sequence suitable for characteristics of a recording and reproducing system. . At the time of reproduction, the reproduction signal read from the magnetic recording medium 13 is subjected to conversion reverse to the conversion at the time of recording, and is also subjected to error detection and error correction processing.
[0029]
The disk control unit 22 is provided for managing data relating to the magnetic recording medium 13, and exchanges data and necessary information with the data channel unit 21 or exchanges data with the servo control unit 19. Thus, information necessary for rotation control of the magnetic disk, seek operation of the head, and the like are transmitted to each other. The disk control section 22 and the servo control section 19 constitute the control means 8 described above.
[0030]
The storage unit 23 includes a buffer memory provided for temporarily storing data.
[0031]
The interface unit (hereinafter, referred to as an “I / F unit”) 24 with the host device, for example, when connected to a computer device or the like, is a small computer system interface (SCSI) controller or an IDE (intelligent drive electronics). It is configured using a controller and the like.
[0032]
A command (command) from the host device 25 is sent to the HDD 12 via the I / F unit 24, and data and parameters are exchanged between the host device 25 and the I / F unit 24. For example, when the I / F unit 24 receives a command issued from the host device 25, the instruction content of the command is transmitted to the disk control unit 22.
[0033]
FIG. 2 shows only a CPU (central processing unit) 26 as a control center and a storage unit 27 (a system memory for storing system data including management information and the like) for the host device. (That is, it does not matter whether the configuration is other than these.)
[0034]
A plurality of tracks are formed concentrically on the recording surface of the magnetic recording medium 13, and each track is divided into a plurality of fixed-length sectors. Each track is provided with a track identification number (track number starting from 0) along the direction from the outermost circumference to the innermost circumference of the magnetic disk.
[0035]
In a hard disk, logical addresses are continuously allocated and arranged from the outer circumference of the disk to the inner circumference, and time-series data is sequentially recorded from tracks on the outer circumference of the disk to tracks on the inner circumference.
[0036]
Also, in the HDD, a CAV (Constant Angular Velocity) method is adopted with priority given to the access speed to the disk. Recently, in order to further increase the recording capacity, the recording area is divided into a plurality of zones from the outer circumference to the inner circumference of the disk, and the number of sectors per track is changed for each zone. A method in which the number of sectors per hit is increased toward the outer periphery of the disk, the recording frequency is variably controlled, and data is efficiently packed (so-called zone bit recording) is employed.
[0037]
Next, the operation of reading and writing data in the HDD will be briefly described. In the following example, an explanation will be given using a file allocation table (FAT) compatible with MS-DOS (a trademark of Microsoft Corporation) as a format for managing files on a disk-shaped recording medium (in the application of the present invention). Of course, since it does not depend on the file system, it is possible to support various formats.)
[0038]
FIG. 3 is a simplified illustration of the configuration (disk layout) of the FAT file system.
[0039]
The system area is composed of three types of areas, and the boot data area stores data defining the structure of the disc.
[0040]
Information indicating the state of the cluster is recorded in the FAT area. For example, the meaning of the set value (hexadecimal notation) in a 16-bit FAT is as follows.
[0041]
“0000h” = indicating that the corresponding cluster is in the “free” state
“0002h to FFF6h” = indicates that the corresponding cluster is in the “assigned” state (the value means the next cluster number).
・ Indicates that “FFF7h” = “defective cluster”
“FFF8h to FFFFh” = indicates that the corresponding cluster is in the “assigned” state (the value means the end of file (EOF)).
[0042]
Further, file management information is recorded in the directory area. Further, a directory area can be configured in the data area. The structure of the directory entry includes a file name (base name), extension, attribute, reserved area, recording time, recording date and time, cluster number, and file length.
[0043]
“Sector” is the minimum unit for recording data (usually set to 512 bytes). In the disk operating system (DOS) shown in this example, two recording units of “logical sector” and “cluster” are used. I use it to manage the HDD. The logical sector number is represented by a continuous number with the start of the drive being 0 sector, and the relationship between the logical sector number and the physical address (surface number, track number, sector number) is expressed by the following equation. Become like
[0044]
"Logical sector number = number of sectors per track x (surface number + number of surfaces x track number) + sector number-1"
On the other hand, a cluster is composed of a plurality of logical sectors, and is a unit for specifying a file position in a FAT (File Allocation Table). The size of the cluster must be a power of 2 (= 2 ＾ n = 1, 2, 4, 8, 16,...) Of logical sectors. Note that the cluster number is represented by a serial number with the head of the file area being “2”.
[0045]
In the data area, the data body of the file and the information of the sub-directory are divided and recorded in cluster units, so the minimum size of the file is one cluster.
[0046]
The file access procedure is briefly described as follows.
[0047]
First, the basic flow of file reading is shown below.
[0048]
(1) Specify the execution file name
(2) Search for a file name in the specified directory (directory search)
(3) When a file is found, the FAT is accessed according to the specified value of the head FAT.
(4) Check the value of FAT, and if it is not EOF (end of file), read a file for one cluster size and move the head to the next designated FAT position
(5) The process of (4) is repeated until the EAT is reached, and when the FAT position becomes the EOF, the file reading ends.
[0049]
More specifically, with reference to FIG. 3, the file information in the directory area of the target “File 1” is found to have the FAT of the first cluster as “1234h”. I see. It contains "1235h", which indicates that the data is continuing. The FAT address position to be connected next can be determined by looking at the stored value of the FAT address “1235h” (in this example, “1236h”). Hereinafter, when the same processing is performed, the FAT address reaches “1240h”. Since the location indicated by the address contains EOF as a data value, this is the final position of the file (“File 1”).
[0050]
Regarding the file writing procedure, first, in order to detect the file writing position, the FAT area is searched in order from the start address (search for a free area). In other words, the value “0000h” indicating an empty cluster is searched from the first cluster, and the address detected first becomes the write position information. Then, by continuing this search, the write operation is sequentially performed (the difference from the above-described file read is that a chain of clusters already allocated to the FAT area is linked so that the written file can be searched later). Need to be recorded and managed.)
[0051]
If the above FAT file system is used, for example, even if stream data related to video information and audio information is recorded on a disc, data of an arbitrary length can be filed and registered in the directory area by file registration. It is possible to search the head position of the file. Further, data can be reproduced based on the linked address information (the above-mentioned “0002h to FFF6h”) recorded in the FAT area. As described above, even when the FAT system is used, recording and reproduction of stream data can be performed.
[0052]
After the system part such as the operating system is loaded, data of the system area (FAT, directory) on the disk is read and written into the system memory area constituting the storage unit 27 in the host device 25. This is to obtain in advance management information and the like necessary at the time of issuing a command. For example, when issuing a file read command, the host device 25 sets the FAT entry number of the directory (the head of the file). The cluster address) and the transfer length are transmitted to the HDD 12 as parameters. As described above, the read / write control is performed by the host device 25 based on the information of the FAT and the directory.
[0053]
With respect to HDDs, the density has been further increased in the future, and accordingly, the track pitch of magnetic disks tends to be further narrowed. Therefore, it is necessary to increase the bandwidth of the servo control for both the seek operation and the tracking operation.
[0054]
This means that a high suppression capability against various disturbances entering the VCM is required, which leads to an increase in the power consumption of the VCM.
[0055]
On the other hand, as the use environment of HDDs for portable devices and the like increases more and more, lower power consumption becomes an important requirement for the operation of VCM.
[0056]
The operation modes of the VCM include, for example, the following modes.
[0057]
a) Evacuation (or SLEEP) mode
This is a mode for reducing waste of power consumption by retracting the head (moving the head to a predetermined retracted position). In the HDD, as the flying height of the head becomes smaller and smaller as the density increases, the conventional CSS (Contact Start Stop) method is replaced with a ramp load (RampLoad) method (“Ramp” is derived from the mechanical shape of the loader). And has a uniform gradient in the structure of the loader). Therefore, in this mode, no power is consumed by retracting the head unit to the loader.
[0058]
b) Standby (or idling) mode
This is a temporary standby operation mode for maintaining the head position after the seek in the control of the servo system without retreating the head unlike the retreat mode. This mode is a mode in which tracking is simply performed so that the head does not deviate from the target track on the recording medium, and various disturbances that affect the control of the VCM are suppressed and controlled. Therefore, a current flows through the VCM according to the magnitude of the disturbance. Note that the VCM current in this mode is not particularly large, but the duration of this mode is relatively long, so that the load on the power product [W · H] that affects the life of the battery is large.
[0059]
Here, the “power product” (or power time product) is an amount ([joule]) obtained by multiplying power consumption W (having a dimension of [joule / hour]), which indicates energy consumption per unit time, by time. , And is represented by “WH” hereinafter. For example, the size of the battery life is expressed by the total energy consumption [unit: joules], which is equivalent to “power consumption W × time”.
[0060]
c) Transition between save mode and standby mode (mode)
This is an operation mode when a transition (or transition) is made between the two modes. For example, in the case of the “RampLoad” method, it is necessary to supply a current to the VCM, and a certain amount of power is consumed when the head is unloaded.
[0061]
d) SEEK mode
This is a mode in which the head moves from one track to another track on the recording medium, and the largest current flows when the VCM accelerates or decelerates. However, since the seek duration is generally several tens of milliseconds or less, the power product consumed is not as large as the current value. As a method of reducing the power product in the seek mode, for example, a method of saving power by generating a speed profile according to a variable setting of the seek speed (intentionally extending the seek time) and the like can be mentioned. Can be
[0062]
e) R / W mode
In this mode, data is read (Read) or written (Write). The power consumption of the VCM is almost the same as the power consumption in the standby mode.
[0063]
FIG. 4 shows a transition diagram between the modes.
[0064]
For example, if the HDD is provided with a standby mode (motor not started) and the following two idling modes, the second mode corresponds to the standby mode.
[0065]
・ First idling mode = mode in which the head is retracted with the motor started
Second idling mode = mode in which the head is loaded and tracked with the motor started.
[0066]
What is problematic in the standby mode, the R / W mode, and the like is various disturbances to the VCM. When these disturbances are viewed in terms of torque, the following disturbances mainly occur. No.
[0067]
-Electric disturbance (hereinafter referred to as "DTe")
A disturbance that converts noise from power supplies and various electric elements into VCM torque.
[0068]
・ Spindle motor rotation synchronous disturbance (hereinafter referred to as “DTRo”)
The deviation between the center of rotation of the magnetic disk and the geometric center of the concentric track (Track) and a disturbance (geometric disturbance) based on the geometric deviation of the track are converted into VCM torque, and the frequency is determined by the spindle. It is the same as the motor rotation speed (this is described as “Fz”).
[0069]
・ Human disturbance (hereinafter referred to as “DTmb”)
For example, when carrying around a portable device or the like on which an HDD is mounted, a disturbance that occurs as a result of a person's motion (eg, walking). Generally, the frequency component is low, but the torque disturbance may be considerably large. Since the gain in the low frequency range is large due to the servo characteristics, the compensating ability is high, and the component is a burden on the power product.
[0070]
-Spindle motor rotation asynchronous disturbance (hereinafter referred to as "DTnrro")
Disturbances other than the above-mentioned rotation synchronization disturbance "DTRo". Note that the above DTe or DTmb may be included here depending on the classification method.
[0071]
As described above, the torque disturbance to the VCM can be roughly classified into two types, “DTRo” and “DTnro”. Then, in the servo system including the VCM, in the standby mode, the seek mode, and the R / W mode, the current is compensated by flowing a current to compress these disturbances.
[0072]
Unlike devices in which high-speed accessibility is one of the important requirements, for example, HDDs for computer devices (HDDs for PCs), HDDs for portable use, etc., are required to start reading and writing data. In many cases, there is some time margin.
[0073]
Therefore, with respect to the servo system including the VCM, the servo frequency characteristic (F characteristic) in the standby (Idling) mode with respect to the servo control characteristic is greatly reduced within a range where there is no hindrance, thereby making the servo system insensitive to the VCM torque disturbance. In this case, VCM power consumption can be reduced. That is, the present invention proposes a method of reducing the power product in the standby mode, and intentionally lowers (reduces) the responsiveness of the servo characteristics during the mode, thereby significantly increasing the power consumption of the VCM. Reduction is possible.
[0074]
FIG. 5 is a conceptual explanatory diagram of the servo response control, in which an arrow “t” shown in a horizontal direction of the figure indicates a direction of elapse of time, and a vertical arrow indicates a direction related to responsiveness ( The direction of the arrow indicates the direction of high responsiveness.)
[0075]
The meaning of each symbol shown in the figure is as follows.
[0076]
"Tidl" = period during standby mode
"Twh" = a period indicating a control mode in which the responsiveness related to the servo control characteristics is reduced (hereinafter, referred to as "power product reduction mode").
・ “Ta” = Transition period from “Tidl” to “Twh”
"Tb" = transition period from "Twh" to "Tidl".
[0077]
In the standby mode, the responsiveness (time constant, response frequency, servo parameter, etc.) related to the servo control characteristics is reduced, so that the sensitivity to the torque disturbance can be reduced and the power consumption can be reduced. In the case where the responsiveness is changed in "Tb" and "Tb", the method is not limited to the method of performing the response continuously as shown in FIG.
[0078]
That is, for the servo system including the VCM, the following methods can be used as a method for changing the servo control characteristics (including both a decrease in response and an increase (recovery) in response).
[0079]
(A) Stepwise change method
(B) Continuous change method
(C) A method using both (A) and (B).
[0080]
First, the method (A) is a method of changing the response over a plurality of stages. For example, in the transition period “Ta”, the response frequency or the servo parameter related to the servo control characteristic is reduced stepwise. In addition, in the transition period “Tb”, the response is restored by increasing the response frequency or the servo parameter related to the servo control characteristic in a stepwise manner.
[0081]
This method involves the simplest two-step modification. That is, this is a method in which the response frequency and the servo parameters are switched at once with a certain timing. Regarding the “certain timing”, the trigger factor does not matter. For example, in a method of issuing a command based on a user operation, characteristics are changed from the time when the HDD receives the command. Alternatively, in a method in which the HDD automatically determines the characteristic change timing based on a result of collation with a predetermined condition (for example, a case where a seek or reading / writing is not performed within a certain time): The change timing is determined by the judgment function of the HDD itself.
[0082]
Regarding the frequency characteristics (F characteristic) before and after the switching, it is preferable that there is a scaling relationship between the frequencies. This is because, instead of having separate parameters with no correlation before and after switching, it is better to easily calculate and derive the other from one side, in terms of characteristics, parameter management, and system configuration. This is convenient and suitable for simplification.
[0083]
Here, "scaling (frequency scaling)" means that the servo frequency characteristics (gain and phase) before and after the change are the same regardless of the frequency, regardless of the control law. I do. Various VCM control rules for HDDs are employed, for example, PID (P: proportional, I: integral, D: differential) control, optimal control, observer (observer) control, other (H infinity control, etc.) ) Or a combination of these. Frequency scaling before and after changing the servo frequency characteristic is very convenient and practical in servo parameter management and servo parameter calculation, but in some cases, different characteristics (scaling relation Of course, it is possible to adopt a characteristic which does not exist in the above.
[0084]
In the transition state due to the switching of the servo parameters and the like, there is a possibility that the behavior of the VCM may fluctuate due to the transient response (so-called servo rampage). As a result, various inconveniences may occur. When it is necessary to consider a coping method, it is effective to switch servo parameters and the like in three or more stages. In this case, for example, the number of stages, the width of each stage, the switching time, and the like are determined based on whether or not the position of the head falls within an allowable range, and changes in the drive current (VCM current and the like) before and after switching. Is preferred. That is, fluctuations in the head position cause collisions and interference between the head arm and peripheral components, and an increase in drive current leads to a result contrary to the intended purpose (reduced power product). It is necessary to set the switching condition with care.
[0085]
The method (B) is a method of continuously changing the response. For example, in the transition period “Ta”, the response frequency or the servo parameter related to the servo control characteristic is continuously reduced. In the transition period “Tb”, the response frequency or the value of the servo parameter related to the servo control characteristic is changed. The responsiveness is restored by continuously increasing the responsiveness.
[0086]
According to this method, when the value of a servo parameter or the like is gradually changed to another value at a certain timing, a method of designating this as a linear change (linear change according to a linear function) and a method of specifying a curve There is a method of designating as a change. Further, when designating the power product reduction mode or when issuing a command to shift to the mode, for example, a method of changing the responsiveness according to a predetermined continuous function, and some obtained information, for example, information relating to the state of the servo system There is a method of changing the responsiveness according to the situation while monitoring (servo error signal and the like) (in this case, control that does not change the responsiveness depending on the state of the servo system is also included). .
[0087]
Also in this method, it is convenient to set the servo parameters before the start of the change in a scaling relationship on the frequency axis for the servo parameters at each moment when the characteristic is changed. In addition, it is preferable to change the response frequency and the like by using the fluctuation of the head position after the change related to the response and the change of the drive current before and after the change as a guideline.
[0088]
The method (C) includes, for example, a method of selectively using the methods (A) and (B) depending on the situation, and a method of selectively using both methods according to the temporal division. For example, as an example of the latter, a method using the method (A) when decreasing the responsiveness and using the method (B) when increasing the responsiveness, or a certain section when decreasing (or increasing) the responsiveness Therefore, various modes can be adopted according to the actual situation, such as using the method (A) and using the method (B) in another section.
[0089]
The head control system of the HDD includes, for example, a seek speed control system and a tracking position control system, and a configuration form employing various control rules is known.
[0090]
In recent years, the track pitch (Tp) has been reduced to less than 0.1 μm (micron) as the density of recording media has increased. Also, after a disk is fixed to the spindle motor shaft, a stamper in which the servo format is stamped using a mastering technique or a semiconductor manufacturing technique is made from the conventional method of writing a servo format using a STW (Servo Track Writer), and injection is performed. Research and development of a method for manufacturing media by molding or magnetic transfer (pre-embossing method) is under way.
[0091]
When viewed from the servo format pattern on the recording medium, the conventional STW method is referred to as “self-recording” and is referred to as a pre-emboss method (or PERM method) as a term expressing a measure relating to the magnitude of disturbance (DTro) to the VCM. ) Is referred to as “other self-recorded”. Incidentally, the disturbance DTro of another self-recording may be about 50 times as large as that of the self-recording.
[0092]
In consideration of the above situation, in the control of the VCM, for example, the following control rules are examined and adopted.
[0093]
(A) Combination of PID control law, DTro compression control, and feed forward control
(B) Combination of observer control law, DTro compression control and feed forward control
(C) Combination of H-infinity control law, DTRo compression control and feedforward control.
[0094]
Here, the “DTro compression control” is control using a kind of a notch filter, and when its function is properly performed, it is necessary to cancel the disturbance DTro. The VCM is driven. As a result, accurate tracking by the head becomes possible. However, since this is realized by supplying a current corresponding to DTRo to the VCM, the burden on power consumption is heavy.
[0095]
The present invention does not relate to the control law itself, and does not care about the control law as far as the present invention is applied. However, in the following description, for convenience, the configuration of (a) will be described, and the VCM will be described. The effect of reducing the electric power product will be examined (even if a control law other than this is adopted, it is needless to say that the following concept can be applied as it is).
[0096]
Hereinafter, the power consumption of the VCM when the PID control rule is adopted will be described using simulation results. Note that, for the simulation results shown in the figure, since a prototype VCM is used as a model constant of the VCM, the absolute value of the result itself has little meaning, and only the difference (or ratio) from the result when this example is applied. It makes sense. In addition, a DTro compression control model (a model using the Notch Filter method) is not included in the simulation, but this is a simulation result for more rigorous evaluation (that is, the improvement effect of the DTro compression control is not significant). No.).
[0097]
First, a simulation model will be described with reference to FIGS. FIG. 6 is a block diagram of a model of a PID control system (included in the servo control unit 10) related to the VCM unit, and FIG. 7 is a block diagram of a model of the VCM unit itself. In these models, the current control drive method is used for the VCM, and for the main part of the VCM model, "1 / S ² (S: Laplace variable).
[0098]
In the configuration example shown in FIG. 6, “$ ref” indicates a command value, which is supplied as a positive input to the adder 28 and is differentiator (a first-order differential element based on time t. d / dt ”).
[0099]
An output (Θ) of a VCM unit (38), which will be described later, is supplied to the adder 28 as a negative input (negative-phase input), and the output of the adder is added via a multiplier 30 of a coefficient “Ki”. (Positive-phase input). Note that the operation by the "adder" includes not only addition but also subtraction (addition of a negative value). In the case of addition, "+" is added to each of the two inputs, and in the case of subtraction, Are distinguished by adding “+” as one input and “−” as the other input (this is the same in an adder described below). For example, the output of the adder 28 is “Θref−Θ”.
[0100]
The output of the differentiator 29 is supplied to the adder 31 as a positive input of the adder 31. The speed output of the VCM unit (38) (indicated by adding a dot “•” over “$”) is supplied to this adder as a negative input. The output of the adder 31 is a positive input to the adder 34 via the multiplier 32 of the coefficient “Kp”, and is also a positive input to the adder 36 via the multiplier 33 of the coefficient “Kd”. Input.
[0101]
The output of the adder 34 is obtained as the sum (arithmetic sum) of the outputs of the multipliers 30 and 32, and is provided as a subsequent integrator (a first-order integral element with respect to time. Shown). Then, the output of the integrator 35 is supplied to the adder 36 as a positive input to the adder 36.
[0102]
The output of the adder 36 is sent as an input (Vin) to a VCM unit 38 via a multiplier 37 of a predetermined coefficient (normalized gain).
[0103]
Note that “Kp”, “Ki”, and “Kd” indicate PID parameters, respectively (see equations [1] to [4] described later), “Kp” is proportional control, “Ki” is integral control, “Kd” is each parameter related to the differential control.
[0104]
In the configuration example of the VCM unit 38 shown in FIG. 7, the input indicated by "Vin1" is supplied as a positive input to an adder 39, and the output of the multiplier 40 for the coefficient "Kt" is input to the negative input. Supplied as Note that “Kt” corresponds to a motor torque coefficient related to VCM.
[0105]
The output (subtraction result) of the adder 39 is sent to the switching unit “SW” via the multiplier 41 of the coefficient “K”. This “K” indicates a gain (reciprocal of inertia) corresponding to the inertia (moment of inertia) of the VCM.
[0106]
The input indicated by “Vin2” is sent to the switching unit “SW” via the multiplier 42 of the coefficient “1”.
[0107]
The switching unit “SW” selects one of the outputs sent from the multipliers 41 and 42 and sends it to the multiplier 43 of the coefficient “Kt”. Then, the output of the multiplier 43 is supplied as a positive input to the adder 44. The output of the multiplier 48 of the coefficient “Dm” is supplied to the adder 44 as a negative input, and the output (subtraction result) of the adder 44 is sent to the multiplier 45 of the coefficient “K”. Can be Note that “Dm” corresponds to the viscosity coefficient of the shaft related to the VCM (because a viscous torque is generated in proportion to the speed).
[0108]
The output of the multiplier 45 becomes the output “Θ” through the integration elements 46 and 47 indicated by “1 / S” in this order. That is, the integration elements 46 and 47 connected in series are connected to the ideal VCM “1 / S ² It is equivalent to.
[0109]
The speed output obtained between the integration elements 46 and 47 (indicated by adding “•” to Θ) is input negatively to the adder 39 via the multiplier 40 of the coefficient “Kt”. And is sent as a negative input to adder 44 via multiplier 48 of coefficient "Dm".
[0110]
Since the signal selected in the switching unit “SW” can be extracted from “ImOut”, for example, when “Vin2” is connected as “Vin” of the VCM unit shown in FIG. If the output of the multiplier 42 is selected by the switching unit “SW”, this can be obtained from “ImOut” as it is.
[0111]
As a method of designing the PID parameters (Kp, Ki, Kd) shown in FIG. 6, as an example, a pole arrangement method of arranging three poles on the circumference on the S plane as shown in FIG. 8 is adopted. I will.
[0112]
In this example, one of the poles is on the real axis (negative axis), and the other two are symmetrically arranged with respect to the real axis with a polar angle “φ”.
[0113]
Regarding VCM (motor section), "1 / S ² ”, The transfer function relating to the open characteristic of the PID control system (this is referred to as“ G _op (S) ". ) Is as follows.
[0114]
(Equation 1)

[0115]
Polar equation (or characteristic equation) "1 + G _op (S) = 0, the following equation is obtained.
[0116]
(Equation 2)

[0117]
As shown in the figure, when three poles are arranged on the circumference of the radius “F1” on the S plane, the following equation is obtained.
[0118]
[Equation 3]

[0119]
Here, F1 (unit: Hz) indicates a response frequency corresponding to the time constant of the system. “Cφ” indicates the cosine function (cosφ) of the polar angle φ.
[0120]
The following equation is obtained from the comparison between the equation [2] and the equation [3].
[0121]
(Equation 4)

[0122]
Each coefficient (Kp, Ki, Kd) is a function of only F1 when φ is constant.
[0123]
For example, when F1 = 1000 [Hz] and φ = 0 [degree] are specified, in exponential notation, “Kd = 3 × 10 ＾ (+ 3), Kp = 3 × 10 ＾ (+ 6), Ki = 10 ＾ (+ 9) ”(“ ＾ ”indicates a power). FIG. 9 shows the Bode diagram (the amplitude characteristic is shown in the upper part and the phase characteristic is shown in the lower part). The 0 dB (decibel) cutoff frequency is about 500 Hz, and the phase margin is about 60 degrees.
[0124]
The following simulation conditions are assumed when designing a servo system by determining PID parameters using the above method.
[0125]
The RRO component (synchronous component accompanying the rotation of the spindle motor) is 100 times as large as the track pitch (Tp) and is equivalent to “100 · Tp” pp (Peek to Peak).
The RRO component is a sine wave without distortion (sin wave), and its frequency is equal to the spindle motor rotation frequency (Fz = 75 Hz). In the model block diagram, it should be equivalently input as a VCM position command.
[0126]
Regarding the frequency characteristic (F characteristic) of the PID control system, F1 = 1000 [Hz], the 0 dB cutoff frequency is 500 [Hz], and the phase margin is 60 [degrees] (see FIG. 9).
[0127]
FIG. 10 exemplifies a VCM current value (temporal change) under a situation where the PID control system faithfully follows this disturbance (a so-called Just tracking state). It can be seen that a current of 30 [mA] pp flows.
[0128]
As described above, in the application of the present invention to the HDD, in the standby mode, the F characteristic of the servo control system is reduced, and the response sensitivity of the VCM to both the RRO disturbance and the NRRO disturbance is reduced. This is to reduce the VCM current flowing in the mode and reduce the power consumption of the VCM.
[0129]
As a method therefor, for example, the following (A1) and (A2) belonging to the above method (A) and the above method (B) can be mentioned.
[0130]
(A1) A mode in which the control F characteristic (F1) is switched to "F1 / n" (where n> 1) at a predetermined timing at a time (see FIG. 11).
(A2) A mode in which the control F characteristic (F1) is switched in multiple stages in consideration of the VCM transient response (see FIG. 12).
(B) A form in which the multi-step switching of (A2) is ultimately performed, and the control F characteristic (F1) is continuously changed, that is, continuously changed (see FIG. 13).
[0131]
11 to 13, the arrow “t” indicates the time lapse direction, “f1” indicates the value of F1 at the time of t = t1, and “f2” indicates the value of F1 at the time of t = t2. Indicates the value. In the case of recovering the responsiveness by changing the F1 value, the direction of the arrow "t" may be reversed, and the replacement of t1 and t2 and the replacement of f1 and f2 may be performed.
[0132]
The above methods (A1) and (A2) are preferable methods from the viewpoint of reducing the processing load on the CPU, and in particular, the method (A1) is the easiest and simplest method. Further, regarding (A2), in comparison with (A1), it is possible to reduce the impact on the VCM, etc. (On the other hand, an experiment or the like for determining the number of steps and the switching interval is required, and should be managed. The number of parameters increases.)
[0133]
The above (B) is an ideal method when the load on the calculation time for the CPU is within an allowable range, and the fluctuation relating to the transient response is small, and if necessary, the change time of F1 is adjusted. It is possible.
[0134]
In each embodiment, when the value is changed from f1 before the change to f2 (<f1), it is clear that the smaller the f2 value, the more the effect of reducing the power product can be obtained. It is necessary to consider the value (lower limit). That is, when the value of f2 is reduced, the control system cannot suppress the RRO disturbance or the NRRO disturbance, and the head position may be largely shaken. Since the shake (fluctuation) is considered to be substantially proportional to the magnitude of the NRRO disturbance and depends on its frequency component, it is necessary to accurately grasp the NRRO component in order to know the shake accurately. In addition, as a practical problem, when the HDD is mounted on a portable device or the like, it is considered that a human disturbance (DTmb) generated when a person walks or the like has the largest value. It is necessary to collect data by investigating the environment well and determine the f2 value based on the data. In general, the power product reduction mode is set in consideration of a possibility that a VCM failure such as an arm collision may occur depending on a positional deviation between the outermost circumference and the innermost circumference of the disk (running of the VCM). It is desirable to take measures such as changing the f2 value according to the radius value (head position in the disk radial direction) at the start of the process.
[0135]
In the method (B) or (A2), it is necessary to pay attention to the setting of the change time “t2−t1”. In other words, as “t2−t1” is shorter, the degree of fluctuation related to the transient response increases, and conversely, as “t2−t1” is longer, the degree of fluctuation related to the transient response tends to be gentler. Therefore, if “t2−t1” is excessively long, there may be a problem that a time required for returning from the power product reduction mode to the standby mode (return time) becomes too long. In addition, the recovery time directly affects the access time of an HDD (such as an HDD mounted on a portable device), so that it is necessary to make the waiting time shorter than the waiting time that the user can tolerate. Incidentally, the length of the access time mentioned here is considerably longer than the access time of an HDD (such as a HDD for a PC) which is required to have a high speed due to its use and purpose. It does not matter in time.
[0136]
Since the stability of the control system is a necessary condition in any of the above-described embodiments, a method of giving a scaling-converted characteristic with respect to the frequency axis before and after the change of F1 is adopted here. That is, the shape of the Bode diagram is exactly the same before and after the change in both the gain and the phase, and is changed so that the magnitude of only the frequency changes. Specifically, this is realized by continuously changing only the F1 value while fixing the value of the polar angle φ in the equation (4). In the following, a study will be limited to a method corresponding only to scaling conversion with respect to the frequency axis, but a method of freely changing F1 and the polar angle φ may be adopted as needed. .
[0137]
FIG. 14 shows a block diagram used for the simulation.
[0138]
The parts related to the PID control structure are as shown in FIG. 6 (thus, the same parts as those in FIG. 6 are denoted by the same reference numerals as those corresponding parts, and the description thereof will be omitted). The differences and additional parts will be described.
[0139]
“REF” indicates a command (input) node related to arm displacement, which is converted from a position displacement into an angular displacement by way of a constant coefficient multiplier (amplifier) 49 and then set by a setting unit (or a command value). Setting node) 50.
[0140]
The setting unit 50 sends the Θref and its speed command (in the figure, a dot “•” is added above “Θ” in “Θref”) to each unit (adders 28, 30 and the like). . The speed command is the time derivative of Θref (therefore, the differentiator 29 is included in the setting unit 50).
[0141]
The characteristic control means 51 is provided to change the time constant, the response frequency, the servo parameters, and the like with respect to the response related to the servo control characteristics (in actual control, the processing is realized by calculation of the CPU as software processing. ). In this example, the PID parameters (Kp, Ki, Kd) are changed by changing the F1 value, and the servo control characteristics are controlled by changing the form of the temporal change of the parameters. That is, in the power product reduction mode, the sensitivity to the torque disturbance is reduced by reducing the responsiveness related to the servo control characteristics.
[0142]
For the PID parameter changed under the control of the characteristic control means 51, a signal indicating the value of the coefficient “Ki” is sent to the multiplier (or the multiplier) 52. Then, a signal indicating the value of “Kp” is sent to the multiplier 53, and a signal indicating the value of “Kd” is sent to the multiplier 54.
[0143]
That is, in the multiplier 52, the output of the adder 28 is multiplied by the coefficient “Ki”, and the calculation result is sent to the adder 34 (in FIG. 6, the value of the coefficient “Ki” of the multiplier 30 is Should change.).
[0144]
Similarly, in the multiplier 53, the output of the adder 31 is multiplied by a coefficient “Kp”, and the calculation result is sent to the adder 34 (in FIG. 6, the coefficient “Kp” of the multiplier 32 is calculated). Just think that the value changes.). Further, the multiplier 54 multiplies the output of the adder 31 by a coefficient “Kd” and sends the result to the adder 36 (in FIG. 6, the value of the coefficient “Kd” of the multiplier 33 is shown). Should change.).
[0145]
The output of the multiplier 37 is sent to the VCM unit 38 via the switching unit 55 and the limiter 56. In the switching unit 55, a disturbance is supplied when the input terminal indicated by “1” in is selected, and when the output terminal of the multiplier 37 is selected, the output of the multiplier turns on the limiter 56. The data is sent to the VCM section 38 via the VCM section 38. Then, an output related to the loop transfer function is obtained from the output terminal indicated by “1” out. The limiter 56 is provided to limit the allowable range (or amplitude) by defining the maximum value and the minimum value of the motor current.
[0146]
The output の of the VCM unit 38 is sent to the monitor unit 57 and displayed. Since 尚 ref from the setting unit 50 is also input to the monitor unit 57, it is possible to observe while comparing Θref and Θ. The speed output of the VCM unit 38 (indicated by adding “•” to “Θ”) is sent to the monitor unit 58 and displayed, and the monitor unit 58 displays a speed command from the setting unit 50. Is also input, so that the two can be observed while comparing. The output “ImOut” of the VCM unit 38 is displayed on a monitor unit 59 including an amplifier.
[0147]
In addition, the output of the adder 28 (position error signal indicated by “pes”) is displayed on the monitor unit 60 including the amplifier.
[0148]
When “feedforward control” is considered, for example, a value obtained by multiplying Θref by a predetermined coefficient value and adding the result to the output of the multiplier 37 is transmitted to the VCM unit 38 via the switching unit 55 and the limiter 56. Good.
[0149]
As shown in Expression 4, when the value of the response frequency F1 related to the servo control characteristic is changed, the value of the PID parameter changes. When the value of F1 is reduced, the response sensitivity of the servo system to torque disturbance is reduced. Become. For example, in the case of the above method (A1), the value of F1 is instantaneously changed from f1 to f2.
[0150]
FIGS. 15 and 16 show simulation results when the above method (A1) is adopted.
[0151]
The setting of each parameter is as follows.
[0152]
(I) In standby (Idling) mode, f1 = 1000 [Hz]
Since F1 is an important parameter that determines the tracking ability, it is desirable that the value (f1) before the change is as large as possible.
[0153]
(Ii) F1 value after changing to the power product reduction mode, that is, f2 = f1 / 100 = 10 [Hz]
The f2 value is set to a value as small as possible in a range where there is no problem in the operation of the VCM (n = 100 in this example).
[0154]
(Iii) Polar angle φ = 0 [degree] in standby mode and power product reduction mode
This value is set to a value that can extract the tracking performance as much as possible in the standby mode (when not in the power product reduction mode). In other words, the φ value must be determined by comprehensively judging the VCM characteristics, the characteristics of the mechanical elements including the head arm, and the like. Here, 0 ° is set as a value that can sufficiently secure the phase margin. I do.
[0155]
(Iv) F1 change time “t1 = 1 [sec]”
(V) The disturbance DTRo is a sine wave (represented by a sin function) without distortion, and its frequency “Fro” is equal to the spindle motor rotation frequency (Fz), and “Fro = Fz = 75 [Hz]”. I do.
(Vi) The magnitude of the disturbance DTro is 100 times the track pitch (Tp) (pp value), Tp = 0.5 [μm], and 50 [μm] pp.
[0156]
(Vii) The arm length is 4.85 [mm].
[0157]
FIG. 15 shows an example of a temporal change of the VCM position deviation (corresponding to the head position deviation) before and after the switching (observed by the monitor unit 60).
[0158]
Until the time t = t1, the position deviation was about 20 [μm] pp, but has reached 730 μm in the transient response immediately after switching.
[0159]
Note that, in this example, the case where the position deviation is large is shown for easy viewing, but if this value is too large to cause a problem, the f2 value after switching may be changed to a slightly larger value.
[0160]
FIG. 16 shows an example of a temporal change of the VCM current before and after the switching (observed by the monitor unit 59).
[0161]
It can be seen that the VCM current flowing about 60 [mA] pp before the switching becomes 6 [mA] pp immediately after the switching, which is reduced to about 1/10.
[0162]
Also, in this simulation result, almost no transient response of the current is observed.
[0163]
From the above simulation results, when the transient response of each node (VCM current, VCM position deviation, etc.) before and after the switching of the response frequency F1 is within a range that does not cause a problem for the system, the method (A1) is simple. It is quite effective.
[0164]
Of course, if there is any problem in adopting (A1), the frequency characteristic change amount (| f1-f2 |) of the PID control system before and after switching is set to a smaller value, or the method (A2) is used. It is possible to cope with this by adopting the multi-stage switching method shown in FIG.
[0165]
Next, the method (B) will be described.
[0166]
There are various methods for changing the F1 value related to the frequency characteristic. Here, a method of changing the F1 value according to a linear function of time t (hereinafter, referred to as “F1 (t)”) will be described. It is of course possible to use various function forms “F1 (t, t1, t2, f1, f2)” other than the linear function as the form of F1 (t). , A linear function F1 (t) shown below is used.
[0167]
(Equation 5)

[0168]
Note that this equation shows a case where the frequency at the change start time “t = t1” is “f1” and the frequency at the change end time “t = t2” is “f2”, and “f1> “f2” represents a case where the frequency is reduced (in the power product reduction mode), and conversely, “f1 <f2” represents a case where the frequency is increased (recovery to the standby mode) (ie, The above expression is an expression encompassing both cases, and hence, hereinafter, it is defined that f1 indicates a value before change and f2 indicates a value after change, regardless of the magnitude relationship between f1 and f2.) .
[0169]
The setting of each parameter is the same as the above (i) to (iii) and (v) to (vii) except for (iv ′) shown below.
[0170]
(Iv ′) The change start time of the F1 value is “t1 = 0.5 [sec]”, and the change end time of the F1 value is “t2 = t1 + 0.5 = 1 [sec]”.
[0171]
In the case where the F1 value is continuously changed, in the block diagram used for the simulation, in the characteristic control means 51 shown in FIG. The components are supplied to the containers 52, 53, and 54, respectively (other items are as described above).
[0172]
17 to 19 show simulation results when the above method (B) is adopted.
[0173]
FIG. 17 shows a temporal change of F1 (t), and shows a state where the value of F1 is reduced to 1/100 from t = 0.5 to t = 1.
[0174]
FIG. 18 shows an example of a temporal change of the VCM position deviation before and after the change.
[0175]
The position deviation in the standby mode up to the point of time t1 = 0.5 is about 5 μm, and immediately after the change (t2 = 1), no significant transient response is observed. The position deviation after the change is set to about 250 μm at the maximum.
[0176]
In this example, the position deviation is large for easy viewing, but if the deviation after changing the F1 value is a system problem, the f2 value may be set to a larger value (or Conversely, if the system permits, the f2 value can be made smaller.) Further, it is possible to perform detailed adjustment by changing the length of “t2−t1” indicating the transition time (or the change time). That is, there is an advantage that adjustment can be performed for one or both of the frequency ratio and the change time (transition time).
[0177]
FIG. 19 shows an example of a temporal change in the VCM current before and after the change.
[0178]
It can be seen that the current flowing 66 [mA] pp before the change of the F1 value becomes 2 [mA] pp immediately after the change, and is reduced to about 1/33.
[0179]
Also, the transient response of the current in this method is small as in the case of (A1).
[0180]
If possible, the fCM value may be further reduced within the allowable range of the system to reduce the VCM current.
[0181]
From the above simulation results, it can be seen that the VCM current can be reduced by about one-tenth of a degree by lowering the frequency characteristic (F characteristic) in the PID control.
[0182]
Although the power reduction effect on the RRO disturbance has been examined here, the same power reduction effect can be actually obtained on the NRRO component (a component asynchronous with respect to the rotation of the spindle motor). It is possible to realize power saving.
[0183]
In the above example, with respect to the frequency characteristic (F characteristic: F1) of the servo system in the standby mode, the method of decreasing the F1 value when shifting to the power product reduction mode has been described. When it is necessary to return to the original state or to make a transition to the R / W mode, it is necessary to increase the F1 value.
[0184]
FIGS. 20 to 22 illustrate simulation results in the case where the above method (B) is employed in such a situation.
[0185]
FIG. 20 is a diagram showing the value of F1 before and after and during the change.
[0186]
As shown in the figure, it can be seen that it increases linearly from f1 = 10 [Hz] to f2 = 1000 [Hz]. Although not shown, the PID parameter also increases with a change in the response frequency, and becomes a predetermined value (a value corresponding to f2, a value when F1 = f2 in Expression 4). To reach.
[0187]
FIG. 21 shows an example of the VCM position deviation before and after the change of the F1 value.
[0188]
Until the point of time t1 = 2, the position deviation is about 130 μm. Immediately after the change, a large transient response is not seen and is about 5 μm.
[0189]
FIG. 22 shows an example of the VCM current before and after the change of the F1 value.
[0190]
It can be seen that the current that had flowed about 3 [mA] pp before the change became approximately 66 [mA] pp after a little time after the change, and increased about 20 times. Also, the transient response of the current value is small.
[0191]
Next, calculation processing and timing regarding the power product reduction mode will be described.
[0192]
As described above, in the HDD employing the self-recording media, since the RRO component is large, in order to realize high-accuracy tracking, a pattern for one rotation of the spindle motor corresponding to the RRO component (for one cycle). It is preferable to store a detection pattern) in a memory and use a control method (so-called Notch Filter method) for canceling the RRO component using the stored pattern.
[0193]
As a recording method of the above-mentioned pattern for canceling the RRO component, for example, there is a method shown below.
[0194]
A static method in which a pattern recording mode is separately prepared and a canceling pattern is recorded only once (for example, a method of executing pattern recording immediately after start-up).
A dynamic method of constantly updating and recording data corresponding to the pattern in the standby mode or the R / W mode.
[0195]
In either method, a pattern corresponding to the RRO component is stored in the memory using a sector number (hereinafter, referred to as “SEC_No”) on the disk track as an index (index).
[0196]
However, it should be noted that in the power product reduction mode, data corresponding to the pattern cannot be obtained accurately. In other words, the function of the DTro compression control using the notch filter needs to be stopped when the mode is started. Therefore, at the time of transition to the power product reduction mode in the standby mode or at the time of transition from this mode to another mode, the timing of the start and end of the transition and the ON / OFF operation of the notch filter (or the function recovery / function stop) ) Must be synchronized. For this purpose, it is preferable to trigger with a sector number (here, for each track, the sector which is the starting point on the same track is defined as a sector with a sector number "0" and the sector is set as a trigger sector).
[0197]
Specifically, synchronization can be performed also for an interrupt trigger to the servo control process by utilizing the fact that it is generated from the sector detection signal.
[0198]
FIGS. 23 to 26 show sequence examples of transition to the power product reduction mode and termination of the mode. The three examples of the host device, the HDD control system, and the PID control system (see FIGS. 6 and 7) are shown. 1 schematically shows the flow of signals and information between persons. The HDD control system (excluding the PID control system) is a process in the standby mode, and the PID control system is an interrupt process. “On operation” indicates an operation at the time of transition to the power product reduction mode, and “Off operation” indicates an operation at the end of the power product reduction mode.
[0199]
In the HDD, the following two types of trigger methods are provided for starting the transition to the power product reduction mode or the return to the standby mode.
[0200]
-How the user issues commands via the host device
A method in which the HDD itself checks the state and automatically performs mode transition.
[0201]
That is, when the access time is to be reduced as much as possible when returning from the power product reduction mode, a method in which the user issues a command is preferable.
[0202]
Also, in the HDD control system, commands for transition from the host device to the VCM unit to the seek mode or the R / W mode are constantly monitored, and these commands are transmitted from the host device to the HDD within a predetermined time. If it is determined that the HDD is not sent to the HDD and the HDD itself determines and enters the power product reduction mode, it is possible to save the user's operation labor. Then, the time required to return to the standby mode and the time required to transition to another mode can be shortened (as a result, the waiting time of the user can be shortened).
[0203]
FIG. 23 shows a sequence example when the above method (A1) or method (A2) is adopted.
[0204]
First, in the case of "On operation", a start command of the power product reduction mode is issued from the host device to the HDD control system, and in response to this, the HDD control system issues a flag (start) indicating the start of the power product reduction mode. Hereinafter, this is referred to as “STT_Flg”.) Is turned on (or set). In the PID control system, if “SEC_No == 0” (“==” means an equality comparison operator) is used as a guard condition and the condition is satisfied, the PID parameters (Kp, Ki, Kd) Is changed stepwise (two or more steps), and the notch filter is turned off (function stops). Here, the "guard condition" means a condition "A" for proceeding to "B" in "B if A", for example, "Y = 1 if"X> 0 ". In this case, the condition “X> 0” is indicated.
[0205]
After the calculation result in the PID control system is output to the VCM unit 38, a flag indicating the stop of the power product reduction mode is given by a signal from the control system to the HDD control system (hereinafter, this is referred to as "STP_Flg"). Is turned off (or reset), and a notification indicating that the power product reduction mode has been started is sent to the host device based on status information or the like.
[0206]
In the case of “Off operation”, a stop command of the power product reduction mode is issued from the host device to the HDD control system, and in response thereto, the HDD control system turns on the flag STP_Flg indicating the stop of the power product reduction mode. (Or set). The PID control system uses “SEC_No == 0” as a guard condition and changes the PID parameters (Kp, Ki, Kd) stepwise when the condition is satisfied (transition to a predetermined value or power product reduction mode). With the return to the previous value), the notch filter is turned on (functional recovery).
[0207]
After the calculation result in the PID control system is output to the VCM unit 38, STT_Flg is turned off (or reset) by a signal from the control system to the HDD control system, and the host device stops the power product reduction mode. A notification indicating that the notification has been made is made based on status information or the like.
[0208]
FIG. 24 and FIG. 25 show sequence examples when the above method (B) is adopted.
[0209]
First, in the case of “On operation” shown in FIG. 24, a start command of the power product reduction mode is issued from the host device to the HDD control system, and in response to this, the HDD control system turns on (or sets) STT_Flg. ). Then, the starting point of time t (t = 0) is determined.
[0210]
The PID control system performs the following processing when “SEC_No == 0 && STT_Flg == On” is a guard condition and the condition is satisfied (“&&” indicates a logical product operation).
[0211]
(A) "t ++" (time increment)
(B) Calculation of F1 (t) (refer to [Equation 5])
(C) Calculation of PID parameters (refer to equation 4)
(D) The notch filter is turned off (function stopped).
[0212]
For the time “t”, a discrete time processing system using the servo sampling time “Ts” as a unit time is adopted, and is synchronized with “t2−t1 = n · Ts” (n indicates a natural number). .
[0213]
After that, the result of the PID calculation is output to the VCM unit 38.
[0214]
The next guard condition is “t> 0 && t <t2 && STT_Flg == On”, and the processes performed when the condition is satisfied are (a), (b) and (c) described above. The result of the operation is output to the VCM unit 38.
[0215]
The next guard condition is “SEC_No == 0 && t == t2 && STT_Flg == On”, and the processes performed when the condition is satisfied are (b) and (c) described above. Is output to the VCM unit 38.
[0216]
Then, STP_Flg is turned off (or reset) by a signal from the PID control system to the HDD control system, and a notification indicating that the power product reduction mode has been started is sent to the host device based on status information and the like.
[0219]
In the case of the “Off operation” shown in FIG. 25, a stop command of the power product reduction mode is issued from the host device to the HDD control system, and in response to this, the HDD control system turns on (or sets) STP_Flg. I do. Then, the starting point of time t (t = 0) is determined.
[0218]
In the PID control system, when “SEC_No == 0 && STP_Flg == On” is set as a guard condition and the condition is satisfied, the processing of the above (a), (b), and (c) is performed. The result of the operation is output to the VCM unit 38.
[0219]
The next guard condition is “t> 0 && t <t2 && STP_Flg == On”, and the processes performed when the condition is satisfied are the above (a), (b), and (c). In this case, a process of increasing the F1 value together with the elapsed time is performed.) Then, the result of the PID calculation is output to the VCM unit 38.
[0220]
The next guard condition is “SEC_No == 0 && t == t2 && STP_Flg == On”, and the processing performed when the condition is satisfied is shown below in addition to the above (b) and (c). (E).
[0221]
(E) Turn on the notch filter (function recovery).
[0222]
Thereafter, the result of the PID operation is output to the VCM unit 38, the STT_Flg is turned off (or reset) by a signal from the PID control system to the HDD control system, and the host device stops (or terminates) the power product reduction mode. ) Is communicated by status information or the like.
[0223]
As in this example, it is preferable that the calculation processing of the equations [4] and [5] be performed in the interrupt processing in which the arithmetic processing of the PID control system is performed. This makes it possible to accurately control the change time “t1−t2” of the F1 value or the like according to the real-time processing, and to ensure synchronization of the PID parameters (Kp, Ki, Kd).
[0224]
However, when a sufficient processing speed and processing time cannot be obtained for the arithmetic processing CPU related to the PID control system, for example, when the conditions imposed on the processing time are severe, the loop processing in the standby mode in the HDD control system is performed. The values of the F1 and PID parameters are calculated using the equations [4] and [5], and the results are first stored in buffers (KdBuf, KiBuf, KpBuf). When the calculation for all three PID parameters is completed, this is transmitted to the interrupt processing of the PID control system (using a flag or the like), and the stored value of the buffer is changed at once in the interrupt processing by the PID variable (Kp , Ki, Kd) may be adopted.
[0225]
FIG. 26 illustrates a processing sequence according to this method (only the case of the On operation is shown).
[0226]
A start command of the power product reduction mode is issued from the host device to the HDD control system, and in response to this, the HDD control system performs the following processing.
[0227]
(F) “t = t1” and “CAL_Flg = 0” (where “=” indicates an assignment operator)
(B) Calculation of F1 (t) (refer to [Equation 5])
(C) Calculation of PID parameters (refer to equation 4)
(G) Store the calculation result of (c) in the buffer
In (f), t1 is set, and "0" is set (reset) in the flag "CAL_Flg" indicating the end of the calculation in (b), (c), and (g) (calculation). Means unfinished).
[0228]
Then, STT_Flg is turned on (or set), and CAL_Flg is set to “1” by the end of the above calculation, and the PID control system is notified.
[0229]
The PID control system performs the following processing when “SEC_No == 0 && CAL_Flg == 1” is a guard condition and the condition is satisfied.
[0230]
(H) Copying buffer storage value to PID parameter
(D) The notch filter is turned off (function stopped).
[0231]
In (h), when the buffers corresponding to Kp, Ki, and Kd are denoted as KpBuf, KiBuf, and KdBuf, respectively, numerical data are substituted from KpBuf to Kp, from KiBuf to Ki, and from KdBuf to Kd. .
[0232]
Thereafter, the result of the PID operation is output to the VCM unit 38, CAL_Flg is reset to zero, and the CAL_Flg is notified to the HDD control system.
[0233]
The processing to be performed next is the above (a), (b), (c), and (g).
[0234]
Then, upon completion of the calculations in (b) and (c), CAL_Flg is set to “1” and notified to the PID control system.
[0235]
The next guard condition is “t <t2 && CAL_Flg == 1”, and the process performed when the condition is satisfied is (h). After that, the result of the PID operation is output to the VCM unit 38. You. Then, CAL_Flg is reset to zero and notified to the HDD control system.
[0236]
The processing to be performed next is the above (a), (b), (c), and (g).
[0237]
Then, upon completion of the calculations in (b) and (c), CAL_Flg is set to “1” and notified to the PID control system.
[0238]
The next guard condition is “SEC_No == 0 && t == t2 && CAL_Flg == 1”, and the processing performed when the condition is satisfied is the above (h). After that, the result of the PID calculation is It is output to the VCM unit 38. Then, CAL_Flg is reset to zero and notified to the HDD control system.
[0239]
By resetting STP_Flg, a notification indicating that the power product reduction mode has been started is sent to the host device based on status information and the like.
[0240]
The basic procedure of the Off operation is the same except for the difference in setting of the flags (STT_Flg, STP_FLg) and the like (thus, illustration and description are omitted).
[0241]
Next, the start and end timings of the calculation process will be described as a measure for shortening the waiting time until the start of reading or writing of data as much as possible.
[0242]
First, when the above method (A1) or (A2) is adopted, when the HDD receives a start command of the power product reduction mode related to the VCM from the host device and detects the first sector number 0 in the target track, , The value of the response frequency F1 is changed stepwise. Since the PID parameter value after the switching is known in advance, it is not necessary to perform a calculation such as Expression 4 or Expression 5 in the interrupt processing. Also, by switching the ON / OFF operation of the RRO notch filter at the same timing, both can be synchronized (FIG. 23 also shows this timing processing).
[0243]
Regarding the start and end timings (within the interrupt processing) of the power product reduction mode when the method (B) is adopted, the HDD receives a start command to the power product reduction mode from the host device, and the first sector in the target track is used. In the interrupt at the time point when the number 0 is detected, the calculation of the value of the PID parameter (Kp, Ki, Kd) is started using the above [Equation 4] and [Equation 5], and the notch is started. Turn off the filter (stop function).
[0244]
On the other hand, after the end point of the power product reduction mode (when returning to the standby mode), immediately after the response frequency F1 has reached f2 (in this case, "f1 <f2"), the first sector. The notch filter is turned on during the interrupt processing when the number 0 is detected.
[0245]
In order to make the notch filter ON operation and the recovery timing to the standby mode as early as possible, the transition time is set so that the head position can read the sector number 0 just at the end time (t = t2) of the power product reduction mode. Is preferably synchronized as “t2−t1 = n · Ts” (“Ts” indicates a servo sampling time, “n” indicates a positive integer). This is to reduce the access time of the HDD by setting the rotation wait time of the spindle motor to zero.
[0246]
In this sense, there may of course be cases where it is better to synchronize with the sector number k that minimizes the HDD access time, instead of synchronizing with the timing of the sector number 0.
[0247]
In the example shown in FIG. 26, it is basically preferable to synchronize with the sector number 0 as much as possible, similarly to the processing in the interrupt. However, when time is required for other processing in the loop processing in the standby mode or when the processing time of the CPU is not sufficient, complete synchronization is difficult.
[0248]
Therefore, as an auxiliary means, the end time of the power product reduction mode related to the VCM, that is, the time point of “t = t2” is exactly the time point corresponding to one (or two) just before the sector number 0. The transition time “t2−t1 = (n−1) · Ts” or “t2−t1 = (n−2) · Ts” (“Ts” indicates a servo sampling time, and “n” indicates a natural number. To be synchronized). This enables synchronization with the notch filter at the timing of the sector number 0.
[0249]
As described above, the control sequence in each method has been described. For example, in a battery-operated portable device or the like, the power product that determines the battery life (remaining amount) is an important index, and the problem is how much the power product can be reduced. It becomes.
[0250]
When the power consumption of the VCM is “Wvcm”, the current value of the VCM is “Ivcm”, and the applied voltage is “Vvcm”, the following equation is established.
[0251]
(Equation 6)

[0252]
Assuming that the durations of the standby mode and the R / W mode (the sum of the times in each mode) are denoted by “Tall” and the corresponding power product is denoted by “WHvcm”, this is defined by the following equation.
[0253]
(Equation 7)

[0254]
As can be seen from this equation, WHvcm is proportional to Tall, so that when this value increases to some extent, the battery runs out.
[0255]
In addition to the introduction of the power product reduction mode, the period length excluding the duration time of the power product reduction mode from the duration time of the standby mode and the R / W mode is set to α times Tall (0 <α <1). . When the VCM current in the power product reduction mode is β times Ivcm (0 <β ≦ 1), the ratio of the power product related to the VCM based on Expression 7 is denoted as “γ”. The following equation is obtained.
[0256]
(Equation 8)

[0257]
For example, assume that the period length in the power product reduction mode occupies 75% of Tall (α = 0.25), and the reduction rate of the VCM current in the power product reduction mode is 3% (β = 0.03). In this case, γ = 0.2725 (that is, a power product reduction of 72.75% is possible).
[0258]
Finally, a criterion for determining whether to shift from the specific mode to the standby mode and enter the power product reduction mode or to enter the evacuation mode from the specific mode or the standby mode in the HDD will be described. .
[0259]
First, when it is clear from predetermined information (including a case where a command to shift to the evacuation mode is issued by a user operation or the like) or setting conditions that the necessity of shifting to the standby mode is not maintained for a long time. It is most advantageous to enter the evacuation mode without entering the power product reduction mode from the viewpoint of VCM power consumption. However, in this case, returning to the standby mode from the evacuation mode is almost the same as the HDD start operation (Start Up), and it takes time until the data read or write operation becomes possible.
[0260]
If it is not possible to obtain sufficient information to permit the transition to the evacuation mode, or if it is difficult to determine that the transition to the evacuation mode is allowed, it is preferable to enter the power product reduction mode from the standby mode. That is, by ending the power product reduction mode, it is possible to return to the original standby mode in a relatively short time, so that the waiting time can be reduced. As a contrivance on the host device side, it is effective to issue a command for shifting from the power product reduction mode to the standby mode (issued by the host device) prior to other commands. As described above, it is preferable that the HDD side synchronizes with “t2−t1 = n · Ts” to minimize the rotation waiting time of the spindle motor as described above.
[0261]
A method of limiting the current command node of the PID control system with a limiter (see FIG. 14), that is, reducing the current (or power) of the VCM simply by providing a limiter at the current command node to the VCM unit However, it is possible to some extent. However, this method has a practical problem when the position related to the VCM fluctuates greatly (so-called nonlinear oscillation occurs) as a result of the VCM falling into an uncontrollable state. However, it is preferable to employ the control method according to the present invention.
[0262]
According to the configuration described above, for example, the following advantages can be obtained.
[0263]
In an HDD mounted on a portable device or an HDD mounted on a device where power saving is more important than high-speed accessibility, it is possible to reduce the power product of the VCM and extend the life of the battery.
[0264]
In the above methods (A2) and (B), the switching time of the VCM to the power product reduction mode can be controlled, and the transient response amount can be controlled.
[0265]
-The operation mode of the HDD is effective because the power consumption can be reduced in the standby (Idling) mode where the operation time is longest.
[0266]
-By adopting the power product reduction mode, the temperature rise of the HDD can be suppressed to some extent.
[0267]
The HDD has a disturbance compression control system using a notch filter or the like for the RRO disturbance inherent in the HDD, but does not cause interference with the control system (that is, in the power product reduction mode). The power product can be reduced by temporarily stopping the function of the disturbance compression control system.)
[0268]
【The invention's effect】
As is apparent from the above description, according to the first and sixth aspects of the present invention, in the temporary standby operation mode, it is possible to reduce the sensitivity to torque disturbance and reduce power consumption. In the operation mode, the head position after the seek is maintained, and the operation can be returned in a short time. Therefore, it is possible to prevent an increase in access time as compared with the return from the evacuation mode. Can be.
[0269]
According to the invention according to claim 2, control can be performed without complicating the control structure by reducing the response frequency related to the servo control characteristic in the temporary standby operation mode.
[0270]
According to the third and seventh aspects of the present invention, the load on the calculation processing can be reduced by changing the responsiveness stepwise.
[0271]
According to the fourth and eighth aspects of the present invention, since smooth characteristic control can be performed by continuously changing the response, deterioration of the transient response can be prevented.
[0272]
According to the invention of claim 5, in application to a battery-driven portable device, consumption of a battery can be suppressed.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram illustrating a basic configuration example of a disk information processing apparatus.
FIG. 2 is a diagram illustrating an example of a device configuration.
FIG. 3 is an explanatory diagram of a FAT file system.
FIG. 4 is a transition diagram between modes.
FIG. 5 is a conceptual explanatory diagram of servo response control.
FIG. 6 is a block diagram showing a configuration example of a servo control system together with FIG. 7, and this figure shows an example having a PID control structure.
FIG. 7 is a diagram illustrating a configuration example of a VCM unit.
FIG. 8 is a diagram showing an example of pole arrangement on an S plane.
FIG. 9 shows an example of a Bode diagram.
FIG. 10 is a diagram illustrating a temporal change of a VCM current value.
FIG. 11 is a diagram for explaining a method of changing the F1 value together with FIGS. 12 and 13, and FIG. 11 shows a two-stage changing method.
FIG. 12 is a diagram showing a multi-step F1 value changing method.
FIG. 13 is a diagram showing a continuous F1 value changing method.
FIG. 14 is a block diagram illustrating an example of a control configuration used for a simulation.
FIG. 15 shows a simulation result regarding a stepwise change of the F1 value together with FIG. 16, and this figure shows a temporal change of the position deviation.
FIG. 16 is a diagram showing a temporal change of a VCM current.
FIG. 17 shows a simulation result regarding a continuous change of the F1 value together with FIGS. 18 and 19, and this figure shows a temporal change of the F1 value.
FIG. 18 shows a temporal change of a position deviation.
FIG. 19 is a diagram showing a temporal change of a VCM current.
FIG. 20 shows a simulation result regarding continuous F1 value change (recovery of responsiveness) together with FIG. 21 and FIG. 22, and this figure shows a temporal change of the F1 value.
FIG. 21 shows a temporal change of a position deviation.
FIG. 22 is a diagram showing a temporal change of a VCM current.
FIG. 23 is an explanatory diagram exemplifying a flow of processing for transition to the power product reduction mode or termination of the mode when a method of changing the F1 value stepwise is adopted.
FIG. 24 is an explanatory diagram exemplifying a flow of processing for shifting to a power product reduction mode when a method of continuously changing the F1 value is adopted;
FIG. 25 is an explanatory diagram exemplifying a processing flow for terminating a power product reduction mode when a method of continuously changing the F1 value is adopted;
FIG. 26 is an explanatory diagram showing another processing example different from FIG. 24;
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Disc information processing apparatus, 2 ... Disc-shaped recording medium, 4 ... Head, 6 ... Drive source, 51 ... Characteristic control means

Claims (8)

A head for reading or recording information on the disk-shaped recording medium, and a servo system including a head positioning mechanism and a drive source for a track of the disk-shaped recording medium, and in controlling the servo system, without retracting the head, A disk information processing apparatus having a temporary standby operation mode for maintaining a head position after a seek,
A disk information processing apparatus comprising a characteristic control means for reducing responsiveness of a servo control characteristic in the temporary standby operation mode. The disk information processing apparatus according to claim 1,
In the temporary standby operation mode, the response frequency or the servo parameter relating to the servo control characteristic is reduced by the characteristic control means, and the response sensitivity of the servo system to torque disturbance to the drive source is reduced. Disk information processing device. 3. The disk information processing apparatus according to claim 2,
A disk information processing apparatus, wherein a response frequency or a value of a servo parameter relating to a servo control characteristic is reduced stepwise. 3. The disk information processing apparatus according to claim 2,
A disk information processing apparatus wherein a response frequency or a value of a servo parameter relating to a servo control characteristic is continuously reduced. The disk information processing apparatus according to claim 1,
A disk information processing apparatus applied to a portable device driven using a battery. A head for reading or recording information on the disk-shaped recording medium, a servo system including a head positioning mechanism and a drive source for the track of the disk-shaped recording medium, and in the control of the servo system, without retracting the head, A method for controlling a disk information processing apparatus having an operation mode for temporary standby for maintaining a head position after a seek,
A control method for a disk information processing apparatus, characterized in that, in the temporary standby operation mode, sensitivity to torque disturbance to the drive source is reduced by reducing responsiveness related to servo control characteristics. 7. The control method for a disk information processing apparatus according to claim 6,
A control method for a disk information processing apparatus, wherein a response frequency or a value of a servo parameter relating to a servo control characteristic is reduced stepwise. 7. The control method for a disk information processing apparatus according to claim 6,
A control method for a disk information processing apparatus, wherein a response frequency or a value of a servo parameter related to a servo control characteristic is continuously reduced.

Images