JPS6138548B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a head crash prevention method for floating heads such as magnetic disk packs.

The head of a magnetic disk or magnetic disk pack device floats above the disk at a height of about 3 microns by utilizing the air flow generated as the disk rotates. In this air, dust particles ranging from 1 micron to several tens of microns cannot be removed and adhere to the disk head through the air flow around the disk, especially on the disk surface, and as a result, they are caught between the floating head and the disk. This causes head crash and damages the head and disk surfaces.

The air sent to prevent the drop in air pressure on the disk surface caused by the rotation of the disk has an important effect on the disk, so air filters are installed at key points. is about 1 micron and will pass through the air filter, and items that are replaced frequently, such as disk packs, must be removed and dirty air and dust may enter the shroud where the disk is attached when installing. It is something. Conventionally, various types of dust countermeasures have been adopted. For example, cleaning the surface of the disk pack after it is installed and before it can be used is not effective as a countermeasure once the disk is ready for use, for the reasons mentioned above. It was hot. The present invention has been made in view of this point,
The purpose of this is to take preventive measures against magnetic disk head crashes, avoid damage caused by head and disk damage, and improve reliability. In a disk device with this function, when there is no seek operation within a certain period of time, the dust attached to the head is removed by seeking from the head stop position to the disk center cylinder and outer peripheral cylinder at least once. This is achieved by a magnetic disk head crash prevention method.

The details of the present invention will be explained below with reference to the drawings.

FIG. 1a is a system configuration diagram showing a control system of a conventional disk pack device. The disk storage device DP is generally connected to the computer's central processing unit CPU via an IO channel controller IOC.

The disk storage device DP connected to the service processor SVP is also connected as shown in the figure.

FIG. 1b is a block diagram showing one embodiment of the present invention, and this example shows a method of processing by hardware within a disk pack device.
The disk pack device DP is a disk control device DPC that is located between the disk pack unit DPU, which includes the disk mechanical part and the electrical control system around it, and the IO channel control device IOC, and controls and checks the writing and reading of information. It becomes more. This control device DPC writes IO channel control device IOC,
They are connected by a plurality of signal lines 1 indicating read control signals, information bit signals for transferring data in parallel, alarms, and the like. Furthermore, various types of modulated and demodulated serial bit information for writing to or reading from the disk are sent and received to and from the disk pack unit DPC, and cylinder addresses that indicate the movement position of the disk head and control signals that control these are sent and received from each other. The signal lines 2 are connected to each other by a plurality of signal lines 2. By the way, in order for the hardware to detect the state in which the disk head is stopped on the same cylinder for a certain period of time and seek between the center cylinder and the outermost cylinder of the disk, the conventional control unit 1 is further provided with crash prevention control. Part 2 has been added. This additional portion will be explained with reference to FIGS. 2a and 2b. The same parts as in FIG. 1 are given the same reference numerals.

In the figure, a crash control unit 2 has a cylinder address register 3 that stores a cylinder address signal ADR instructed when writing or reading a disk, and a discrepancy between the stored address and the address to be written or read from now. The timer counter 5 is cleared by detecting the mismatch. The timer counter 5 receives the clock pulse CLK that is generated when the disk pack controller DPC is powered on and becomes in use, and when the first read and write commands are received and control is started, the disk heads are on the same cylinder. The elapsed time during which the clock is stopped is counted by the clock, and a time overflow signal 5 is generated by counting a certain period of time, for example, about one minute.
When this time overflow signal is generated, the control circuit 6 instructs the conventional controller 1 so that the disk drive unit performs at least one seek operation to the centermost cylinder and outermost cylinder of the disk. A seek request signal and a seek address signal SA are sent in response to a busy signal BGY indicating that an operation such as a seek, read or write operation is in progress, and a completion signal CMP indicating the end of the command for the operation. The time limit for timer over and the specified number of times for seeking are not shown, but they can be made variable by sending a signal from the selection switch to the timer counter 5 and the control section 6.

FIG. 2b explains the above operation by means of a flow chart.

When the disk pack controller DPC receives a disk read or write command, the control shown in the flow chart starts at A.

When a cylinder address is entered, it is compared with the cylinder address specified by the previous instruction, and if a mismatch is detected, the timer counter is reset; if the addresses match, the timer counter continues. count the time.

When it is detected that the timer counter has overflowed and a certain period of time has elapsed, the disk head seeks to move across the entire surface of the disk at least once. When this seek operation is completed, if the disk pack device is in a state where a command is received, it returns to the initial state, point A, and when it is in the state of waiting for a command, the timer counter starts counting the elapsed time and repeats the seek operation. This prevents the disk head from being stopped on the same cylinder for a long time. Dust adhering to the head is blown away by the airflow by moving transversely to the airflow that lifts the head. Although this example explains the method using hardware, it is also possible to use software that monitors the elapsed time of the cylinder address of each disk device attached to the IO channel control device by using the timer function attached to the central processing unit CPU. You can also create clothing.

As is clear from the above, according to the present invention, head crash can be efficiently prevented by monitoring the stoppage time of disk heads on the same cylinder and moving the heads.

[Brief explanation of the drawing]

FIG. 1a is a system configuration diagram showing a control system of a conventional disk pack device, FIG. 1b is a block configuration diagram showing an embodiment of the present invention, and FIGS. FIG. 2 is a detailed diagram and flow chart. CPU...Central processing unit, IOC...IO channel control device, SVP...Service processor, DP...Disk pack device, DPC...Disk pack control device, DPU...Disk pack unit, 1...Conventional control unit, 2...Crash prevention control unit , 3... cylinder address register, 4... mismatch detection section, 5... timer counter, 6... control section, 7... NOR gate, 1, 2, 3... signal line, 4... timer reset signal line, 5... overflow signal line.

Claims (1)

[Claims] 1. In a magnetic disk device that has a control function for writing and reading data on a magnetic disk, if there is no seek operation within a certain period of time, the disk center cylinder and outer peripheral cylinder are moved at least once from the head stop position. A magnetic disk head crash prevention method characterized by removing dust attached to the head by seeking.