JPS62134806A - Demodulation system for magnetic disk device - Google Patents

Abstract

PURPOSE:To expand a slice level setting range and to attain a highly accurate demodulation by detecting the peaks of long and short period patterns through the combination of a high level slice and differentiation, and that of a low level slice and differentiation. CONSTITUTION:High and low level slice circuits 7 and 8 slice a reproduction signal at high and low levels, respectively, and the 1st and 2nd AND circuits 9 and 10 take the AND of the output signal of a pulsating circuit 3 and those of the circuits 7 and 8. A gate circuit 14 generates a gate signal corresponding to the long period pattern which a long period pattern detection circuit 11 detects from the slice signal out of the circuit 9, and the 1st and 2nd delays 12 and 13 delay the output signals of the circuits 9 and 10 by the prescribed amount. Then the 3rd and 4th AND circuits 15 and 16 and an OR circuit 17 select the pulse signals of the circuits 7 and 8 with the aid of the output of the circuit 14. Thus a slice circuit is divided into the long period pattern circuit 7 and the short pattern circuit 8, whereby the set range of a slice level can be expanded to attain a highly accurate demodulation.

Description

[Detailed Description of the Invention] [Summary] The present invention is a demodulation method that detects the peak of a reproduced signal from a magnetic disk by differentiation and level slicing, and which detects the peak of a reproduced signal from a magnetic disk at two levels, high and low. The high-level slice detects the peak of the long-period pattern of the reproduced signal, and the low-level slice detects the peak of the short-period pattern of the reproduced signal, and a gate signal corresponding to the period of the long-period pattern is created. The gate signal is configured to remove peaks in long-period patterns and extract only peaks in short-period patterns, and widen the level setting range to accurately remove extra pulses superimposed near the zero level of long-period patterns. It is possible to do so.

[Industrial application field]

The present invention relates to a demodulation method for a magnetic disk drive, and more particularly to a demodulation method for a magnetic disk drive that widens the range of slice level settings to reduce the probability of demodulation errors.

In recent years, magnetic disk drives have become increasingly denser and faster, and as a result, the resolution and output of magnetic disk reproduction signals have decreased.

In response to this reduction in the resolution and output of the reproduced signal,
There is a demand for a demodulation method that can accurately demodulate the peak position of a reproduced signal representing recorded information.

[Conventional technology]

FIG. 3 is a block diagram of a conventional demodulation method for a magnetic disk drive, and FIG. 4 is a signal waveform diagram for explaining the operation of the conventional demodulation method.

The conventional demodulation method includes an amplification system 1 that amplifies the reproduced signal, a differentiator 2 that differentiates the amplified reproduced signal and converts the peak position to a zero level, and a pulse signal at the output zero level point of the differentiator 2. a detection circuit 6 for detecting the output signal of the amplification system 1; a slicing circuit 4 for slicing the output detection signal of the detection circuit 6 at a predetermined level; It is comprised of an AND circuit 5 that ANDs the output signal and the output signal of the slice circuit 4.

The operation will be explained with reference to the signal waveform diagram in FIG.

The reproduced signal is amplified to a predetermined level by the amplification system 1, and is input to the differentiating circuit 2 and the detection circuit 6 as a signal shown by the solid line in FIG.

The differentiating circuit 2 differentiates the peak of the a signal in FIG.
It is output to the pulsing circuit 3 as a differential signal whose peak point shown in FIG. B has a zero level.

The pulsing circuit 3 corresponds to the zero level point of the differential signal.
A pulse signal shown in FIG.

In the part of the long-period pattern (1001) where the peak point does not exist for a long period, as shown in A1 in Figure 4, there is no time change, so the differential output is near the zero level, as shown in 81 in Figure 4B. Become. Therefore, as shown at 61 in FIG. 4C, the peak is detected as if it were a peak, even though it is not an actual peak.

Therefore, in order to distinguish between the original peak and the erroneous peak, the demodulated signal a is half-wave rectified by the detection circuit 6, and the negative signal is sent to the fourth
It is output to the slice circuit 4 as a g signal indicated by a dotted line in the figure.

The slicing circuit 4 slices the output signal of the detection circuit 6 at slice level b, creates a rectangular wave signal as shown in the fourth diagram corresponding to the slicing point, and outputs it to the AND circuit 5.

AND circuit 5 receives the output signal of pulse generator 3 (Fig. 4C)
and the output signal of the slice circuit 4 (FIG. 4D)
The peaks shown at 01 in FIG. 4C that are not the original peaks of the demodulated signal are removed, and only the data signal shown in FIG. 4E is outputted as a demodulated signal.

[Problem that the invention seeks to solve]

In the conventional iMflA method described above, it is necessary to detect all peak positions of the reproduced signal and exclude false peaks in the long-period pattern of AI shown in Figure 4A. The level of the signal with the minimum amplitude at the peak position (A2 in Figure 4A) and A in Figure 4A
1 and the maximum level of the false peak in the long-period pattern.

Therefore, as the resolution of the reproduced signal decreases, the amplitude shown in FIG. 4A becomes smaller and smaller, and the slice level setting range becomes narrower accordingly. Therefore, there is a problem that an extra pulse superimposed near the zero level of a long period part is detected, and the probability of a demodulation error increases, making it impossible to perform accurate demodulation.

The present invention was created in view of these points, and an object of the present invention is to provide a demodulation method for a magnetic disk device that enables highly accurate demodulation by widening the setting range of the slice level.

[Means for solving problems]

FIG. 1 shows a block diagram of a demodulation method of a magnetic disk device according to the present invention.

In FIG. 1, the demodulation method of the present invention is added to the conventional demodulation method by adding a high level slicing circuit 7 and a low level slicing circuit 8 that slice the reproduced signal at high and low levels, and the output signal of the pulse generator 3 and the high level. It includes a first AND circuit 9 and a second AND circuit 1o that AND the output signals of the slice circuit 7 and the low-level slice circuit.

Further, a long-period pattern detection circuit 11 detects a long-period pattern from the slice signal of the first AND circuit 9, a gate circuit 14 generates a gate signal corresponding to the detected long-period pattern, and a first AND circuit. 9 and second A
A first delay 12 and a second delay 13 each delaying the output signal of the ND circuit 10 by a predetermined amount of delay, and a high level slice circuit 7 by the output of the gate circuit 14.
and third and fourth AND circuits 15 and 16 for selecting pulse signals from the low-level slice circuit 8, and an R circuit 16.

[Effect]

By dividing the level slice circuit into a high-level slice circuit 7 as a level slice circuit for long-period patterns and a low-level slice circuit 8 as a level slice circuit for short-period patterns, it is possible to avoid error peaks in long-period pattern parts unlike in the conventional method. Since it is not necessary to perform position removal and short-cycle original peak position detection at the same slice level, it is possible to widen the setting range of the slice level.

〔Example〕

FIG. 1 is a diagram of a demodulation system according to an embodiment of the present invention.
FIG. 2 is a signal waveform diagram of the demodulation method of the embodiment.O As shown in the block diagram of FIG. A first AND circuit 9 and a second AND circuit that AND the output signal of the high-level slice circuit 7 and the low-level slice circuit 8, and the output signal of the pulse generator 3 and the output signal of the high-level slice circuit 7 and the low-level slice circuit. 10.

Further, a long-period pattern detection circuit 11 that detects a long-period pattern from the slice signal of the first AND circuit 9, a gate circuit 14 that selects the detected long-period pattern, and A first delay 12 that delays each output signal of the AND circuit IO by a predetermined delay amount.
and a third delay 13, which selects the pulse signals from the high level slice circuit 7 and the low level slice circuit 8 based on the output of the gate circuit 14. The configuration includes fourth AND circuits 15 and 16 and an OR circuit 16.

The operation will be explained with reference to the signal waveform diagram in FIG.

As shown in FIG. 2F, the output signal F1 of the detection circuit 6 is sliced at a high level i and a low level j by a high level slice circuit 7 and a low level slice circuit 8, and each of both slice signals and a pulse generator 3 are sliced. By performing an AND operation with the outputs of , in the first AND circuit 9 and the second AND circuit 10, peak positions such as G and H in FIG. 2 are detected.

The signal G detected at a high slice level is always detected because the long period pattern has a large output.

On the other hand, in the signal H detected at a low slice level, all peaks are detected, but an erroneous peak position H1 in a long period pattern is also detected.

Therefore, using the signal G output from the first AND circuit 9, the long period detection circuit 11 uses a long period pattern (100
1) Detect Al and create signal 11 shown in Figure 2 (■).

Further, the signal G and the signal H are delayed by the first delay 12 and the second delay 13 by, for example, 3 bits as shown in the figure to create the J signal and the L signal in FIG. 2.

The J signal created by the first delay 12 and the above-mentioned ① signal are input to the gate circuit 14, and by the flip-flop operation of the gate circuit 14, the 2 signals, which are in opposite phase to 1, are input to the gate signal shown in FIG. create. The gate signals Kl and 2 are long-period patterns, and a position with no peak between them is originally found.

Then, the 1 signal and the output signal J of the first delay 12 are ANDed by the third AND circuit 15, and the 2nd signal and the output signal J of the second delay 13 are ANDed by the third AND circuit 15. Then, by ORing each AND output with the OR circuit 1117, the data pulse train J created from the high slice level is applied to the signal during 1 to 11'', and the data pulse train J created from the high slice level is applied to the signal while 2 is IT OIT. The peak position is detected by the data pulse train created from the data pulse train, and finally only the pulse position of the original peak position like the signal M is created.

〔Effect of the invention〕

As explained above, according to the present invention, the slice levels of the long-period pattern part and the short-period pattern part of the reproduced signal can be set separately, so the setting range of the slice level can be wider than in the conventional method, and the resolution can be increased. It is also strong against a drop in the value of fjtDil or an extra pulse superimposed near the zero level of a long-period pattern portion, making it possible to obtain fjtDil with high accuracy.

[Brief explanation of drawings]

FIG. 1 is a block diagram of the present invention, FIG. 2 is a signal waveform diagram of the embodiment, FIG. 3 is a block diagram of the conventional example, and FIG. 4 is a signal waveform diagram of the conventional example. In the figure, 1 is an amplification system, 2 is a differentiation circuit, 3 is a pulse circuit, 4 is a slice circuit, 5 is an AND circuit, 6 is a detection circuit, 7 is a high level slice circuit, 8 is a low level slice circuit, and 9 is a The first AND circuit 110 is a second AND circuit, 11 is a long period pattern detection circuit, 12 is a first delay, 13 is a second delay, 14 is a gate circuit, 15 is a third AND circuit, 16 indicates a fourth AND circuit, and 17 indicates an OR circuit. This opening stone's mouth 77 Figure 1 Missing Example/Episode 1 Dog @ Cause Figure 2 Conventional Autopsy Chiku 1 Other waveform people - Figure

Claims (1)

[Claims] In a configuration in which data recorded on a magnetic disk is demodulated by detecting the peak of the reproduced signal by differentiation and level slicing, the level slice has two setting levels, high and low, and the length existing in the reproduced signal is demodulated. A demodulation method for a magnetic disk device characterized in that each of the peaks of a periodic pattern and a short periodic pattern is detected by high-level slice and differentiation and low-level slice and differentiation.