WO1996005596A1

WO1996005596A1 - Method and apparatus for augmented missing pulse certification of a magnetic recording medium

Info

Publication number: WO1996005596A1
Application number: PCT/US1995/010382
Authority: WO
Inventors: Gerald L. Pressman; Barry L. Shevick; Fredric A. Perez; Gerald W. O'docharty
Original assignee: Trace Mountain Products, Inc.
Priority date: 1994-08-15
Filing date: 1995-08-15
Publication date: 1996-02-22
Also published as: AU3365795A

Abstract

An apparatus and method measure the recording quality of a magnetic recording medium (21). A magnetic medium recorder (14) records signals on the recording medium (21) and detects signals recorded on the medium (21). A pattern generator (13) coupled to the magnetic medium recorder (14) produces a test signal (202) that includes at least one pulse that has an amplitude that is a minimum read amplitude relative to the amplitude of a reference condition. The sequence of pulses preferably represents the data pattern (302) that will be recorded on the medium (21) under normal operating conditions and that has such a minimun amplitude. A pulse amplitude detector (20) has a first input coupled to the magnetic medium recorder (14). The pulse amplitude detector (20) supplies a pulse detect signal (27) having a first state indicative of a detected pulse, in response to the pulse amplitude of the detected signal applied to the first input of the pulse generator (13) being greater than a threshold, and supplies the pulse detect signal (27) having a second state indicative of a missing pulse, in response to the pulse amplitude of the detected signal applied to the first input of the pulse generator (13) not being greater than a threshold.

Description

METHOD AND APPARATUS FOR AUGMENTED MISSING PULSE CERTIFICATION OF A MAGNETIC RECORDING MEDIUM

Field of the Invention

This invention relates to testing of a magnetic recording medium, and more particularly, to measuring the quality of recording signals on the magnetic recording medium.

Background of the Invention

In magnetic recording, certification is a process for ensuring the suitability of a magnetic recording medium for reliable use as a data storage or transfer medium. The certification process ensures that the magnetic and physical properties of the medium meet the requirements of the record /playback system that the medium is used in and that defects are either absent or below a predetermined level of acceptance.

One of the most important certification tests is the "missing pulse" test, which examines signals that are recorded on the magnetic recording medium and determines the presence of signal losses or "dropouts." Dropouts are caused by voids in the coating, adhered particles, or internal coating non-uniformity. The missing pulse certification test measures the loss of signal caused by these defects and the physical uniformity of the media surface and coating.

In order to provide a test that is uniform and independent of both format and data and that can be applied to all media of a given form (e.g., tape, disk, or the like), specific conditions are set under which all certifier tests, including the missing pulse test, are to be performed. For data recording in the United States, these conditions have been set by the American National Standards Institute (ANSI), and similar standards have been adopted by international industrial standards groups, such as the International Standards Organization (ISO).

The ANSI conditions specified for missing bit certification include the following:

1) A recorded test pattern of a single frequency and uniform pulse spacing.

2) A write current level determined by tests made on a standard reference medium.

3) A minimum signal level, compared to the average signal level, that is defined as a missing pulse.

These test conditions are usually adequate for use in general data recording, although, in some applications, the specification for minimum signal level has been raised to provide a higher reliability in the recording process. However, the basic form of the conditions has not been changed.

In at least one important application of magnetic recording, these standards have been found to be insufficient for providing a reliable indication of the recordability of data on magnetic media. This application is the field of flexible disk recordings for software duplication and formatting. Duplicated or pre-formatted software is intended for wide distribution. A recorded disk having such software must be readable on a variety of systems under a wide range of age, wear, and environmental conditions. As such, software duplicators and media formatters maintain a higher level of quality control and quality checks on their recorded media to ensure a high probability of readability under "field" conditions. In particular, professional, high-quality duplication /formatting systems have adopted a modified form of the missing pulse test that evaluates each recording for the amplitude of individual data pulses. In these systems, if any pulse amplitude falls below a prescribed amplitude, the recorded disk is rejected.

Users of these high-level systems have discovered that the ANSI certification process does not guarantee recordability for their application. Even when the missing pulse minimum threshold is raised above the ANSI standard, these users find a high percentage of rejects in their duplication process. The reason for this is that the amplitude reduction caused by a particular defect during ANSI certification is not the same as the amplitude reduction during the actual data duplication process.

The basic cause of this difference is that ANSI certification uses a uniform (constant frequency) pulse pattern, while data recording involves complex patterns involving non-uniform pulse spacing (mixed frequencies). The effect of a dropout- causing defect on a pulse amplitude is strongly dependent on the spacing of nearby pulses. Also, the reference average is not the same, because data patterns include pulse amplitudes not found in the certifier pattern. In addition, a pulse-timing adjustment known as write precompensation is used in data recording. This is an alteration of nominal pulse timing during writing to compensate for timing shifts produced when non-uniform patterns are written. The effect of write precompensation is to further increase the range of frequencies recorded on the disk. The recording frequency used for ANSI certification is usually the highest frequency (closest pulse spacing) that is expected during data recording. The use of write precompensation means that actual recorded frequencies are higher than anticipated by the ANSI certification pattern. Dropout levels are frequency-sensitive, and a given defect produces a greater amplitude reduction at increased frequencies.

The cause of excessive rejects during duplication or formatting when using certified media is dropouts that appear acceptable to the certifier, yet produce pulse amplitudes that fail the missing-pulse test in the duplication equipment. Depending on the exact parameters of recording (e.g., media resolution, write current, track location, and location of the defect with respect to recorded data), it is possible for a dropout that passes an ANSI test at one amplitude with respect to an average, e.g., 40%, to fail a duplication missing-pulse test at a lower cutoff amplitude, e.g., 25%. Raising the ANSI test missing pulse level (commonly called "clipping level") during certification can avoid these problems, but to be effective, the clipping level may be so high that the yield at the certifier is very poor, and therefore the certified disks are very expensive. Many of the disks rejected in this way actually perform satisfactorily during duplication, because their dropouts are in non-sensitive locations (e.g., outer tracks of a disk) or do not exhibit as much pattern sensitivity. Thus, this simple solution is not practical.

It would be advantageous to provide a method of certification that provides improved product quality for the software duplication and media formatting environments in particular and for demanding data recording applications in general. It would be further advantageous to provide this product quality without incurring unnecessary fallout of product during the certification process.

Summary Of The Invention

In accordance with the present invention, an apparatus and a method to measure the recording quality of a magnetic recording medium. A magnetic medium recorder records signals on the recording medium and detects signals recorded on the medium. A pattern generator coupled to the magnetic medium recorder produces a test signal having at least one pulse that has an amplitude that is a minimum read amplitude relative to the amplitude of a reference condition. The sequence of pulses preferably represents a data pattern that might be recorded on the medium under normal operating conditions and that has such a minimum amplitude. A pulse amplitude detector has a first input coupled to the magnetic medium recorder. The pulse amplitude detector supplies a pulse detect signal having a first state indicative of a detected pulse, in response to the amplitude of the detected signal applied to the first input of the pulse generator being greater than a threshold, and supplies the pulse detect signal having a second state indicative of a missing pulse, in response to the amplitude of the detected signal applied to the first input of the pulse generator not being greater than the threshold.

Brief Description of the Drawings

FIG. 1 is a graph of the conventional data pattern written on magnetic recording medium under an ANSI certification implementation.

FIG. 2 is a graph of a data pattern without write precompensation written on a magnetic recording medium according to principles of the invention.

FIG. 3 is a graph of a data pattern with write precompensation written on a magnetic recording medium according to principles of the invention.

FIG. 4 is a graph of a data pattern having the data pattern of FIG. 3 continuously repeated therein. FIG. 5 is a graph of an alternate data pattern according to the principles of the invention.

FIG. 6 is a block diagram illustrating a certification system according to principles of the invention.

FIG. 7 is a schematic diagram of the read circuit of the certification system of FIG. 6.

FIG. 8 is a diagram of the state-sequence of the precedence detector.

FIG. 9 is a block diagram illustrating a read circuit having a microprocessor for controlling the missing pulse test.

FIG. 10 is a flowchart illustrating the operation of the certification system of FIG. 9.

FIG. 11 is a graph of an alternate data pattern according to the principles of the invention.

Detailed Description

Referring to FIG. 1, there is illustrated a graphical representation of a conventional data pattern written on a magnetic disk under an ANSI implementation. The ANSI standard for a 5.25-inch disk is American National Standard for Information Systems - two-sided, high-density, unformatted.5.25-inch fl30-mm . 96- tpi f3.8 tpmm.. flexible disk cartridge for 13262 ftpr use - general, physical, and magnetic requirements. ANSI X3.162-1988, by American National Standards Institute, Inc., New York, New York (1988), the subject matter of which is incorporated herein by reference. A pattern 1 is a graphical representation of a 2f test recording signal recorded on a magnetic recording medium under an ANSI certification standard. Similarly, a pattern 2 is a graphical representation of a If test recording signal recorded on the magnetic recording medium under an ANSI certification standard. The frequency of the 2f test recording signal and the frequency of the If test recording signal are each dependent on the type of magnetic recording medium. For example, for a 5.25 inch high density floppy disk, the frequency of the 2f test recording signal is 500,000 flux transitions per second and the frequency of the If test recording signal is 250,000 flux transitions per second.

The If test recording signal is produced under the same parametric conditions as the 2f test recording signal, except the If test recording signal has an amplitude larger than the amplitude of the 2f test recording signal. The "resolution" is the ratio of amplitude of the 2f test recording signal and the amplitude of the If test recording signal.

Referring to FIG. 2, there is illustrated a graphical representation of a data pattern without write precompensation written on a magnetic recording medium according to principles of the invention. A pattern 102 comprises a sequence of data signals that simulates the data that is to be recorded, or is expected to be recorded, on a magnetic recording medium 21, later described herein. The sequence of data signals preferably has at least one pulse having an amplitude that has the greatest reduction in (or mmimum) amplitude relative to a reference condition. The reference condition may be, for example, a test signal of a single frequency within the range of the frequencies of the data to be simulated. The sequence of data signals also preferably has a minimum time interval between successive pulses having the minimum read amplitude. The pattern 102 preferably comprises a sequence of pulses or waveforms of half cycles of lf-2f-2f-2f-lf test recording signals. In modified frequency modulation (MFM) recording, the pattern 102 is achieved by writing a data bit pattern of 1-0-1-1-1-0-1. A central pulse 3 of the pattern 102 has the maximum reduction in amplitude and has an amplitude that is lower than the amplitude of the continuous 2f test recording signal of FIG. 1. Therefore, the pattern 102 represents a more stringent test of media quality than does the continuous 2f test recording signal of the ANSI standard. Because the pattern 102 lacks write precompensation, later described herein, positive peaks 4, 5 of the pattern 102 are each offset from their ideal position relative to the negative peaks 6, 7 due to pulse-crowding effects in which the tails of adjacent pulses add or subtract together to shift the signals written on the magnetic recording medium 21.

Referring to FIG. 3, there is illustrated a graphical representation of a data pattern 202 with write precompensation written on a magnetic recording medium according to principles of the invention. To implement write precompensation, a certifier 10, later described herein, adjusts the pulse timing of the pulse 4 (FIG. 2) to write on the magnetic recording medium a pulse 4', which is the first pulse with a positive peak, a short time later than the corresponding pulse 4. Similarly, the certifier 10 adjusts the pulse timing of the pulse 5 (FIG. 2) to write a pulse 5', which is the second pulse with a positive peak, a short time earlier than the corresponding pulse 5. In other words, write precompensation shortens the written pulse interval between pulse 4 and the pulse 3 and the pulse interval between the pulse 3 and the pulse 5, while lengthening the interval between pulse 6 and the pulse 4 and the interval between the pulse 5 and the pulse 7.

Using write precompensation, the drive 14 writes the pulse 4' and the pulse 5' in their respective proper positions. In addition, write precompensation further lowers the amplitude of the pulse 3' as shown in FIG. 3. This now represents a "worst- case" condition, and provides the more severe test for dropouts. Referring to FIG. 4, there is illustrated a graphical representation of a data pattern 302 having the data pattern 202 of FIG. 3 continuously repeated therein. The pattern 202 of FIG. 3 is continuously repeated to form the pattern 302 of FIG. 4. The data pattern 302 is the test pattern for missing pulse certification in this invention. In an MFM recording system, this pattern is achieved by writing the repeated data-byte sequence EBAEBAEBAEBA... in hexadecimal notation (1110 1011 1010 binary).

In one embodiment, the recording medium under test 21 is certified in the certifier 10 for use in either a duplicating system or a formatting system. In such an embodiment, the derivation of the reference amplitude signal by the certifier 10 preferably is similar to the derivation method used in the duplicating system or the formatting system. For the pattern 302, the derivation of the reference amplitude signal may average, for example, the peaks 8 which are the highest amplitude peaks. Using the highest amplitude peaks provides the worst case condition for test because any other average includes peaks 9 which are at a lower amplitude. Consequently, the average reference signal is less than the average reference signal calculated using the peaks 8.

It should be noted that, in the data pattern 302 of FIG. 4, all of the maximum- amplitude pulses 8 have the same polarity (negative in this example, but absolute polarity is not a limitation of the invention). As described above, the high-amplitude pulses 8 determine the reference average used to compute the signal loss. In embodiments in which rectifiers 15, later described herein, used to derive the average amplitude are half-wave rectifiers 15, these rectifiers 15 respond to either the positive or negative peaks only. A pattern generator 13, later described herein, preferably controls and adjusts the polarity of the written signal so that the polarity of the high- amplitude peaks 8 corresponds to the conducting cycle of the rectifier 15 used for the averaging circuit. If such control of signal polarity is not feasible, than an alternative test pattern, shown in FIG. 5 may be used.

Referring to FIG. 5, there is illustrated a graph of an alternate data pattern according to principles of the invention. As an alternative to adjusting the polarity of the written signal in response to the conducting cycle of the rectifier 15, the pattern generator 13 provides a test data pattern having dual polarity. In particular, a data pattern 502 has two large amplitude peaks, one of each polarity so that it does not matter which half of the waveform (positive or negative) that the rectifier 15 uses. The data pattern 502 includes the repeated hexadecimal byte sequence EAEAEA... (or in binary the data sequence 0101110101011101....).

Referring to FIG. 11, there is illustrated a graph of another alternate data pattern according to principles of the invention. As an alternate to the pattern 502 of -FIG. 5, the pattern generator 13 provides a data pattern 602 similar to the data pattern 302 of FIG. 4. The pattern 602 has blocks of pulses. A first group of such blocks includes pulses having a positive polarity of the peaks of the pulses having the largest amplitude. A second group of such blocks includes pulses having a negative polarity of the peaks of the pulses having the maximum amplitude. The blocks of the first and second groups of blocks are arranged so that the polarity of the peaks in the blocks alternates. The pattern 602 has an additional pulse 11 after a pulse 10, which is the half cycle waveform of the If test recording signal. The pulse 11 is a half cycle waveform of a If test recording signal. The pulse 11* causes subsequent high amplitude pulses 12 to have positive peaks. Such additional pulses 11 may be included in a plurality of locations in the pattern to be recorded so that the number of high amplitude pulses having positive peaks equals the number of high amplitude pulses having negative peaks. In the averaging process later described herein, the differences between the response of the system to positive and negative peaks may be averaged out.

Referring to FIG. 6, there is illustrated a certification system or certifier 10. A pattern generator 13 supplies a write data pattern 22 to a magnetic medium recorder or read /write drive 14 which writes a write data signal 40 on a magnetic recording medium under test 21. The write data pattern 22 includes a data pattern that has a repeating pattern of bits having a hexadecimal value of EBA (1110 1011 1010 binary). The read /write drive 14 reads a read data signal 41 from the recording medium under test 21 and applies an amplified analog read data signal 19 to an input of a read circuit 35, later described herein.

The write data pattern 22 includes the data pattern 302 of FIG. 4 or alternatively the data pattern 502 of FIG. 5. The sequence of bits of the data pattern,

101110101110101110 ( or 01011101010111010101110...), has preferably a duration to completely write to the area of the magnetic recording medium under test 21 that is used operationally. If the medium under test 21 is a flexible disk, the write data pattern 22 has preferably the duration of one track.

The pattern generator 13 has a memory 44 for storing the bit sequence of the write data pattern 22. In response to a pattern request command 48 from the drive 14, the memory 44 provides the bit sequence to a parallel-to-serial converter 46 for providing the write data pattern 22. Alternatively, discrete logic or state-sequence devices may be used to create the bit sequence. The pattern generator 13 preferably applies the same write precompensation as used in the duplicators or formatters.

The read /write drive 14 may be, for example, a certifier drive, such as a 001-43410-02 model certification/formatting drive module manufactured by Trace Mountain Products, Inc. of San Jose, California. The drive 14 is preferably the same type of drive used by software duplicators or media formatters, if the medium under test 21 is to be used in such a duplicator or formatter. The drive 14 has write current control necessary for full ANSI certification. Alternatively, the read /write drive 14 may be a conventional drive for reading and recording that is modified to provide test signals ordinarily used for alignment and testing as the analog read signal 19.

The write current of the drive 14 effects the certification effectiveness because the write current affects resolution. In particular, both the amount of amplitude loss and the amount of precompensation required to correct pulse position errors increase as resolution decreases. The effect is compounded because both of these factors affect the overall precompensated pattern amplitude loss. Thus, write current differences indirectly effect the correlation between the certifier test and data-recording performance. The write current of the drive 14 during certification preferably equals the write current used for actual data recording, rather than ANSI specification. For disks recorded on by duplication systems, the write current of the certifier preferably equals the write current of the duplication system. For certified disks where the intended user is unknown (and consequently the write current is unknown), the write current of the certifier 10 preferably equals the setting of the write current specified by the drive manufacturer as shipped. Referring to FIG. 7, there is illustrated a schematic diagram of the read circuit

35. The analog read signal 19 from the drive 14 is applied to an input of a filter 12 which provides a filtered read signal 37 to an input of a read channel processor 16 and to an input of an amplifier 34. The filter 12 may be, for example, a IMP42C451 model filter manufactured by IMP, Inc. of San Jose, California. The read channel processor 16 includes a peak detector for recovering data pulses from the filtered read signal 37. The read channel processor 16 supplies a peak detect output signal 32 in response to a recovered data pulse. The read channel processor 16 may be, for example, a 32P541 model read data processor manufactured by Silicon Systems of Tustin, California. The amplifier 34 amplifies the filtered read signal 37 and provides a filtered and amplified signal 26 to a first input of a window comparator 18, later described herein, and to an input terminal of a rectifier 15. The rectifier 15 is preferably an active precision rectifier. The rectifier 15 provides a rectified signal 36 to the input terminal of a filter 28.

The filter 28 comprises a resistor 110 and a capacitor 112, each coupling both the input terminal and an output terminal of the filter 28 to ground. The filter 28 preferably has a short time constant. The time constant is selected so that the rectifier 15 tracks the fluctuations in the overall amplitude of the filtered and amplified signal 26 caused by physical variations in the recording medium under test 21. For magnetic recording disks, the time constant is preferably approximately equal to several hundred cycle periods of the If test recording signal. The filter 28 provides the filtered rectified signal 33 to an input terminal of an integrator 23.

The integrator 23 integrates the filtered rectified signal 33 to provide an average amplitude reference signal 39 against which a signal reduction during a dropout is measured. The average amplitude reference signal 39 may be, for example, an average amplitude signal which is the average of the amplitude of the read signal. The integrator 23 comprises, for example, a resistor 114 coupling the input terminal of the integrator 23 to the negative input of an operational amplifier 116. A positive input of the operational amplifier 116 is coupled to ground. An integrating capacitor 118 and a reset switch 24 each separately couple the negative input of the operational amplifier 116 to an output terminal of the operational amplifier 116. The reset switch 24 is closed, in response to a reset signal 54, later described herein, to reset the integrator 23 before taking another average. The integrator 23 provides the average amplitude reference signal 39 to a first input of an attenuator 17. The attenuator 17 attenuates the average amplitude reference signal 39 to a predetermined level (or "clipping level") that defines the mini-mum peak amplitude of a read pulse that constitutes a passing pulse test.

When the precompensated test pattern 302 of this invention is used in place of the ANSI test frequency, the same clipping level creates a higher rejection rate of media under test 21. Accordingly, a lower clipping level may be appropriate. When the pattern 302 is used, the reference average against which pulse amplitude is measured includes many pulses of the If test recording signal, resulting in an overall higher reference average than the ANSI average under the same conditions. Therefore, even in the absence of defects, some pulses may fall to 60% or less of the reference average. In applications where the certification is being performed on media to be used in duplicating /formatting systems that incorporate missing pulse tests, the certifier 10 preferably has a clipping level that is nominally at the same level used in the duplication/formatting system. For example, if the duplication system is set at a 30% pulse amplitude threshold, the certifier 10 uses the same level for certification. However, because there is always some variability in drives 14, the certifier 10 may use a small guard-band during certification. For this example, the certifier 10 is set to a clipping level (for pattern 302 testing only) of 33%. The attenuator 17 is preferably a potentiometer coupling the first input of the attenuator 17 to ground. A tap on the potentiometer provides an attenuated reference signal 43 to the output of the attenuator 17. The attenuator 17 provides the attenuated reference signal 43 having an amplitude that is indicative of a threshold for defining the minimum acceptable peak amplitude of a pulse read from the recording medium under test 21. The attenuator 17 provides the attenuated reference signal 43 to an input of an inverter 25 and to a second input of the window comparator 18.

The inverter 25 has a first resistor 120 coupling the input of the inverter 25 to a common node of a negative input of an operational amplifier 122 and to a second resistor 124, which couples the common node to another common node of an output of the operational amplifier 122 and an output of the inverter 25. A positive input of the operational amplifier 122 is coupled to ground. The inverter 25 inverts the voltage level of the attenuated reference signal 43 to generate an inverted attenuated reference signal 45 at the output of the inverter 25, which is coupled to a third input of the window comparator 18.

An output signal 31 of the window comparator 18 switches state if the filtered and amplified read signal 26 applied to the input of the window comparator 18 exceeds the attenuated reference signal 43 in the positive direction, or is more negative than the inverted attenuated reference signal 45. Thus both positive and negative peaks switch the window comparator 18 if the peak amplitude exceeds the level of the attenuated threshold. The window comparator 18 has a positive amplitude comparator 126 having a first input coupled to the first input of the window comparator 18 and a second input coupled to the second input of the window comparator 18 for providing an output signal in response to the amplified read signal 26 applied to the first input of the comparator 126 having an amplitude greater than the amplitude of the attenuated reference signal 43 applied to the second input. In other words, the positive amplitude comparator 126 provides an output signal if the amplitude of the positive peaks of the filtered and amplified read signal 26 are above the threshold or clipping level set by the attenuator 17. Similarly, a negative amplitude comparator 128 has a first input coupled to the first input of the window comparator 18 and a second input coupled to the third input of the window comparator 18 for providing an output signal in response to the amplifier read signal 26 having an amplitude greater than the amplitude of the inverted attenuated reference signal 45 applied to the second input of the comparator 128. In other words, the negative amplitude comparator 128 provides an output signal when the amplitude of the negative peaks of the filtered and amplified read signal 26 are below the threshold set by the attenuator 17 and the inverter 25. The output signals from the first and second comparators 126, 128 are applied to corresponding first and second inputs of an OR 130 gate for providing the output signal 31 to an output terminal of the window comparator 18 in response to either of the output signals from the comparators 126, 128. In other words, the OR gate 130 provides an output signal if either a positive or negative peak of the filtered and amplified signal 26 exceeds the threshold set by the attenuator 17.

The peak detect output signal 32 from the read channel 16 processor is applied to a first input of a precedence circuit 20. The precedence circuit 20 detects the presence of pulses having an amplitude lower than the threshold. The output signal 31 from the window comparator 18 is applied to a second input of the precedence circuit 20.

The peak detect output signal 32 of the read channel processor 16 is a data pulse that occurs in response to the peak value of the filtered read signal 37 applied to the input of the read channel processor 16. If a peak amplitude of the filtered read signal 37 exceeds the clipping level (attenuated reference amplitude), the window comparator 18 provides an output signal 31 before the peak detect output signal 32 occurs. Conversely, if the peak falls below the clipping level, a peak detect output signal 32 occurs without an output signal 31 from the window comparator 18.

In response to the application of a peak detect output signal 32 that is not preceded by the application of an output signal 31, the precedence circuit 20 outputs a missing pulse (MP) detect signal 27 that indicates the detection of a pulse below the clipping threshold (for example, Missing Pulse or dropout). A reset signal 53 is applied to a reset input of the precedence circuit 20 to cause the precedence circuit 20 to clear the dropout indication once recognized. The precedence detector 20 may be implemented, for example, with programmed array logic or discrete logic devices in accordance with the state-sequence diagram of FIG. 8.

Referring to FIG. 8, there is illustrated the state-sequence of the precedence detector 20. In a zero state 350, the value of the MP Detect Signal 27 is zero. The value of the MP Detect Signal 27 remains at zero if the reset signal 53 is applied to the reset input of the precedence circuit 20. If the output signal 31 of the window comparator 18 changes state, the MP Detect Signal 27 remains at zero in an intermediate state 352. If a peak detect output signal 32 or a reset signal 53 is received, the precedence detector 20 returns to state 350. If a peak detect output signal 32 occurs in state 350, the peak detector enters a one state 351 in which the value of the MP Detect Signal 27 changes to a one, in response to a peak detect output signal 32 from the read channel processor 16. The MP Detect Signal 27 remains in a one state 351 until a reset signal 53 is applied to the reset input of the precedence circuit 20.

Referring to FIG. 9, there is illustrated a read circuit having a digital processor 29. The read circuit of FIG. 9 is similar to that of FIG. 7 except the functions of the attenuator 17 of FIG. 7 are done digitally by a processor 29. For simplicity and clarity, only a portion of the identical components are shown in FIG. 9. Like elements have like reference numbers. The amplified read signal 26 from the amplifier 34 is applied to the input of the rectifier 15 which provides the rectified signal 36 to a filter 28. The filter 28 provides a filtered rectified signal 33 to the integrator 23.

The integrator 23 provides the average amplitude reference signal 39 to an input te-rminal of a conventional analog-to-digital converter 30. The analog-to-digital converter 30 samples the average amplitude reference signal 39 and provides a digital signal representative of the average amplitude reference signal 39 to a conventional processor 29. The processor 29 digitally applies an attenuation to the average amplitude reference, in response to a predetermined clipping level, and provides a digital attenuated reference signal 42 to a conventional digital-to-analog converter 38. As described below in conjunction with FIG. 10, for a disk under test 21, the processor 29 uses the average amplitude reference signal 39 of the previously tested track for testing the current track under test. The processor 29 provides a reset signal 54 to the reset switch 24 of the integrator 23 to reset the integrator 23 for calculating another average amplitude and a reset signal 53 to the precedence circuit 20 to clear the dropout indication of the MP Detect Signal 27. The digital-to-analog converter 38 converts the digital attenuated reference signal 42 into the attenuated reference signal 43 and provides the signal 43 to the second input of the window comparator 18 and the input of the inverter 25. The window comparator 18 and the inverter 25 function as described earlier herein. The processor 29 has a threshold register for storing the calculated clipping level or threshold. For certifiers for testing disks, the processor 29 has a track 0 first revolution for storing a state value of whether the drive 14 is reading track 0 of the disk under test 21 for the first time.

In application to flexible disk certification, the processor 29 determines the average amplitude reference signal 39 for one revolution on each track. Because of the lower head-media relative velocity at tracks near the center of the disk, the average amplitude of these tracks is lower than those at the outer edge. To determine the track average amplitude for each track takes one complete revolution. Similarly, the certification operation for each track takes another complete revolution. It is common practice in the certification art to avoid the extra time required by using the average amplitude from the previous track as a reference for the track being certified, because there is typically very little amplitude difference between adjacent tracks. Consequently, it is possible to certify and compute average amplitude in the same revolution. The first track, of course, requires an extra revolution to acquire its own average, which is then also used for the next track.

Referring to FIG. 10, there is illustrated the operation of the certifier 10 of FIG. 9. For FIG. 10, the operation of the certifier 10 is described for certifying a magnetic disk 21 for exemplary purposes only. The invention is not limited to certifying magnetic disks. The processor 29 commands 130 the disk drive 14 to move its write and read heads to the first track (Track 0). The processor 29 sets 130 the threshold (THRESH) register to zero and sets the track 0 first revolution (TKFirst Rev) register to a TRUE state. The processor 29 then commands 132 the pattern generator 13 to repetitively provide the test pattern to the disk drive 14 which writes the test pattern throughout the track of the disk under test 21. The processor 29 outputs 134 the contents of the THRESH register. The processor 29 resets 134 the integrator 23 by applying a reset signal to the switch 24. The processor 29 commands 134 the disk drive 14 to read the medium under test 21 for testing. As the drive 14 reads the data recorded on the track under test. As described earlier herein, the read signal from the drive 14 is processed by the rectifier 15 and the filter 28 and the resultant signal is applied to the integrator 23. The analog to digital converter 30 samples 134 the integrator and provides the digital signal to the processor 29. The processor 29 calculates 134 a threshold and stores it into the THRESH register. If the register TKOFirst Rev is set 136 true for the first read of the first track, the processor 29 outputs 138 the value stored in the THRESH register and commands 138 the drive 14 to reread the track of the magnetic recording medium under test 21.

If a missing pulse is detected 140, the processor 29 rejects 142 the medium under test 21 and applies a reset signal 53 to the precedence detector 26. The processor 29 terminates 142 the certification test. On the other hand, if a missing pulse is not detected 140, the processor 29 determines 144 whether the track that was tested is the last track of the disk under test 21. If it is the last track, the processor 29 accepts 146 the disk under test 21 and terminates the test. Otherwise, the processor 29 commands 148 the drive 14 to step to the next track. The processor 29 also sends a reset signal 54 to reset the integrator 23 for the average amplitude of the next track and sends a reset signal 53 to reset the precedence circuit 20.

In an alternative embodiment, certification of media under test 21 may be performed at several different clipping levels. The tested media 21 may be separated by the highest clipping level that the media passes to establish different quality levels. In yet other embodiments, the pattern test of this invention may also be employed with different quality levels directed to different duplication systems or to specific customer requirements. In summary, this invention applies a specially-determined "worst-case" (for amplitude) pattern during missing pulse testing. In combination with duplication write current and write precompensation, this pattern simulates actual data-recording conditions more closely than standard ANSI missing pulse certification. As a result, a product certified with this technique meets the performance requirements of very demanding data recording environments, in particular flexible-disk formatting and duplication systems in which individual pulse amplitude is measured as a quality factor.

Claims

We claim:

1. A method for measuring the recording quality of a magnetic recording medium comprising the steps of:

generating a test signal including a pulse sequence that produces at least one pulse having an amplitude that is a minimum read amplitude relative to the amplitude of a reference condition;

writing the test signal on the magnetic recording medium;

reading the test signal written on the magnetic recording medium to produce a read signal therefrom;

rejecting the magnetic recording medium, if the amplitude of any pulse peak of the read signal is less than a threshold value; and

accepting the magnetic recording medium, if the amplitude of all pulse peaks of the read signal is not less than the threshold value.

2. The method of claim 1 wherein the test signal includes a data sequence having a hexadecimal value of EBA.

3. The method of claim 1 wherein the test signal includes a data sequence having a hexadecimal value of EA.

4. The method of claim 1 further comprising the steps of

reading a portion of the test signal written on the magnetic recording medium;

measuring the amplitude of the portion of the test signal read from the magnetic recording medium; and

averaging the measured amplitude of the portion of the test signal to generate the threshold value.

5. The method of claim 1 wherein the test signal comprises a sequence of pulses of half cycles of lf-2f-2f-2f-lf test recording signals.

6. The method of claim 1 wherein the at least one pulse of the test signal has a minimum time interval between successive pulses having the minimum read amplitude.

7. The method of claim 1 wherein the test signal includes a sequence of pulses wherein the number of high amplitude pulses having a positive polarity equals the number of high amplitude pulses having a negative polarity.

8. Apparatus for measuring the recording quality of a magnetic recording medium comprising:

a magnetic medium recorder for recording signals on the recording medium and for detecting signals recorded on the recording medium;

a pattern generator coupled to the magnetic medium recorder for producing a test signal including a pulse sequence that produces at least one pulse having an amplitude that is a minimum read amplitude relative to the amplitude of a reference condition; and

a pulse amplitude detector having a first input coupled to the magnetic medium recorder and having an output for supplying a pulse detect signal having a first state indicative of a detected pulse in response to the pulse amplitude of the detected signal applied to the first input of the pulse generator being greater than a threshold, and for supplying the pulse detect signal having a second state indicative of a missing pulse in response to the pulse amplitude of the detected signal applied to the first input of the pulse generator not being greater than the threshold.

9. The apparatus of claim 8 wherein the sequence of pulses include a data sequence having a hexadecimal value of EBA.

10. The apparatus of claim 8 wherein the sequence of pulses include a data sequence having a hexadecimal value of EA.

11. The apparatus of claim 8 further comprising a reference generator having an input coupled to the magnetic medium recorder and having an output for supplying a reference signal representative of the average peak amplitude of a predetermined portion of the detected signals recorded on the recording medium; and

said pulse amplitude detector having a second input coupled to the output of the reference generator and the pulse detector for determining the threshold in response to the reference signal.

12. The apparatus of claim 8 wherein the test signal includes a sequence of pulses of half cycles of lf-2f-2f-2f-lf test recording signals.

13. The apparatus of claim 8 wherein the at least one pulse of the test signal has a minimum time interval between successive pulses.

14. The apparatus of claim 8 wherein the test signal includes a sequence of pulses wherein the number of high amplitude pulses having a positive polarity equals the number of high amplitude pulses having a negative polarity.