CN1269073A - Method and apparatus for determining bit error in a sampled data system - Google Patents
Method and apparatus for determining bit error in a sampled data system Download PDFInfo
- Publication number
- CN1269073A CN1269073A CN98808666A CN98808666A CN1269073A CN 1269073 A CN1269073 A CN 1269073A CN 98808666 A CN98808666 A CN 98808666A CN 98808666 A CN98808666 A CN 98808666A CN 1269073 A CN1269073 A CN 1269073A
- Authority
- CN
- China
- Prior art keywords
- pattern
- disk
- error
- channel
- isolated
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B19/00—Driving, starting, stopping record carriers not specifically of filamentary or web form, or of supports therefor; Control thereof; Control of operating function ; Driving both disc and head
- G11B19/02—Control of operating function, e.g. switching from recording to reproducing
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B20/00—Signal processing not specific to the method of recording or reproducing; Circuits therefor
- G11B20/10—Digital recording or reproducing
- G11B20/10009—Improvement or modification of read or write signals
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B20/00—Signal processing not specific to the method of recording or reproducing; Circuits therefor
- G11B20/10—Digital recording or reproducing
- G11B20/18—Error detection or correction; Testing, e.g. of drop-outs
- G11B20/1816—Testing
- G11B20/182—Testing using test patterns
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B5/00—Recording by magnetisation or demagnetisation of a record carrier; Reproducing by magnetic means; Record carriers therefor
- G11B5/012—Recording on, or reproducing or erasing from, magnetic disks
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B5/00—Recording by magnetisation or demagnetisation of a record carrier; Reproducing by magnetic means; Record carriers therefor
- G11B5/02—Recording, reproducing, or erasing methods; Read, write or erase circuits therefor
- G11B5/09—Digital recording
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B20/00—Signal processing not specific to the method of recording or reproducing; Circuits therefor
- G11B20/10—Digital recording or reproducing
- G11B2020/10935—Digital recording or reproducing wherein a time constraint must be met
- G11B2020/10981—Recording or reproducing data when the data rate or the relative speed between record carrier and transducer is variable
Abstract
A system determines bit error rate in a sampled data system having a disc (112) formed of a magnetic medium and a data head (126) for writing data on the disc (112) and retrieving data from the disc (112). A pattern is written on the disc (112) which includes isolated instances (248) associated with a predetermined event. The pattern is retrieved from the disc (112) and the isolated instances (248) are combined to obtain a representative instance sav(n) having a reduced media noise component. The representative instance s<av>(n) is combined with each of the isolated instances (248) to obtain noise sequences. An autocorrelation component (R) is obtained based on the instances read from the disc (112) and the noise sequences. A channel filter is developed having an impulse response h(n) based on the representative instance and based on the channel requirements for a predetermined channel model. A representative instance is passed through the filter to obtain a filter output sample (smp), and an error value indicative of the bit error rate is determined based on the filter output sample and the autocorrelation component (R).
Description
Invention field
The present invention relates to sampled data system.The invention particularly relates to the system of the error rate of judging sampled data system, this sampled data system comprises disk and is used for writing data in the disk and the data head (data head) of retrieve data from this disk.
Background of invention
Typical disc driver comprises and one or morely is installed on the axle or the disk that rotates in the rotating shaft.Typical disc driver also comprises a transducer by a fluid dynamic air bearing supporting, and it flies on each disk.Transducer and fluid dynamic air bearing are referred to as data head.Usually with the commands for controlling disc driver of driving governor basis from the host computer system reception.Driver controller control disc driver, in order to from the search disk data, and with information stores on disk.
One dynamo-electric exciter is worked in a degenerative closed loop servo system.Exciter moves radially with the operation that tracks data head on magnetic disk surface, and transducer remained on the magnetic disk surface operates with tracking directly over the magnetic track.
Normally by providing write signal to encode to data head, and information is deposited in the concentric magnetic track on the magnetic disk surface so that flux directions on the magnetic disk surface of the data that will store of representative is changed.When retrieve data from disk, driving governor controller electrical driver, thus data head flies on disk, and the flux directions on the sensing disk changes, and changes the generation read signal according to these flux directions.Usually this read signal is regulated, then by driving governor decoding, the flux directions of recovering to store on the disk changes the data of representative, and represents with the read signal that data head provided.
Typical readback system comprises data head, preconditioning logical circuit (as pre-amplification circuit and filter circuit), data detector and restore circuit, and error-detecting and correcting circuit.Readback system can be a discrete circuit, or incorporates in the relevant driving governor of disc driver.
The error rate of the recorded bit of each determined number is remained on the quite low level.In a function disc driver, can come estimating bit error rate with several approach.For example, can continuously various data patterns be write on the disk in the disc driver and, and the margin of error that runs in can reading data is easily counted from this disk reading of data.Yet, in the magnetic head of the exploitation of disc driver of future generation is selected, concerning drive manufacturers, before selecting to be used for the data head of disc driver of future generation, several different data heads or different data head types are tested or assessed not is rare.The exploitation of data head is normally before the exploitation of the read channel circuit that finally uses with these data heads.So, be difficult in when not having the read channel circuit, estimate the error rate of the system of these data heads of employing.
Once the someone attempted estimating to adopt by the measurement that is considered to have at least some helps when estimating the error rate of system's (not possessing the read channel circuit in this system) of so a kind of magnetic head in the past, but not success.Generally speaking, this measurement comprises the signal that writes down on the disk and the measurement of noise in measured degree.Yet, the measurement of this prior art also not well with adopt disk drive system in this data head the time error rate that run into relevant well.
The purpose of this invention is to provide a kind of system, and this system has than the better advantage of prior art at these or other problems.
Summary of the invention
The invention provides a kind of system that is used for determining the error rate in the sampled data system, this sampler has the disk that is formed by magnetic medium and is used for data are write on the disk and the data head of retrieve data from this disk.(pattern) writes on the disk with pattern, and it comprises M the isolated example relevant with scheduled event.Retrieve this pattern from disk, and M isolated example combined, obtain the representative example that an electronic noise component reduces.Each that this representative is isolated in example and the isolated example combines, and obtains M noise sequence.According to read from disk M example and M noise sequence, obtain an autocorrelative component.The channel model of exploitation has according to representing example and according to the impulse response of the desired channel of predetermined channel model.One represents example by this filter, obtains a filter output sampling, and according to the output sampling of filter and the error amount of the definite expression of the auto-correlation component error rate.
The present invention relates to method and apparatus simultaneously.
The accompanying drawing summary
Fig. 1 be removed top cover disc driver vertical view.
Fig. 2 is a high position (high level) block diagram of disc driver shown in Figure 1.
Fig. 3 determines the error rate relevant with data head and the block diagram that do not adopt the system of read/write channel electronic circuit.
What Fig. 4 described is main error event.
Fig. 5 is the flow chart of the operation of system shown in Fig. 3.
Fig. 6 is the more detailed block diagram of system shown in Fig. 3.
Fig. 7-1 describes is pattern according to the isolated pulse of one embodiment of the present of invention.
What Fig. 7-2 described is the computing of isolated example shown in Fig. 7-1, so that obtain the noiseless sampling.
Fig. 8 is the flow chart according to system operation shown in Fig. 6 of another aspect of the present invention.
Fig. 9 describes is autocorrelation matrix according to the filter input of one aspect of the present invention.
Figure 10 describes is autocorrelation matrix according to output place of one aspect of the present invention filter.
The detailed description of preferred embodiment
The invention provides a kind of like this system, adopt this system, can determine and adopt the relevant error rate of sampled-data system of data head, and need not the relevant read/write channel circuit of disc driver common and the employing data head.Yet,, once disc driver and relevant read/write channel circuit thereof are described here for the ease of understanding the present invention clearly and completely.
Referring to Fig. 1, drawn the spinning disk drive system among the figure in schematic form, and unification is represented with label 110.A plurality of magnetic information memory discs 112 are coaxial with the rotating shaft electric motor assembly 114 in the shell 116.A plurality of concentric circles recording tracks are arranged on each disk 112, and signal is represented with label 118 in the drawings, in order to recorded information.Each magnetic track 118 is divided into a plurality of sectors, and signal is represented with label 120 among the figure.Can be to disk 112 with storage, perhaps retrieve data from the disk 112, shown in particular track among the figure 118 and sector 120 like that.Actuator arm assembly 122 preferably is installed in rotation on the Yi Jiaochu of shell 116.Actuator arm assembly 122 has a plurality of head gimbal assemblies 124, and each head gimbal assembly has a slide block 125 with a W head, or transducer 126, be used for reading information from disk 112, and with on the information channel disk 112.The voice coil motor of installing 128 can accurately make actuator arm assembly 122 rotate back and forth, thereby transducer 126 is crossed over disks 112 and moved along camber line 130.
What Fig. 2 provided is the high-level block diagram of the control circuit 132 of disk drive system 110.Disk drive system 110 comprises position and the control circuit 126 to writing disk 112 or handling from the information that disk 112 receives of controlling transducer 126.Microcontroller 134 is directly carried out all basic functions of disk drive system 110.Being labeled as 136 read/write support links to each other with microcontroller 134 by conventional data, address and control bus 140 with exciter controller 138 with motor with interface control circuit.Circuit 136 provides usually by the hardware interface of communication bus 142 between disk drive system 110 and host computer system (not shown).Simultaneously, circuit 136 provides the interface between motor and exciter controller 138 and the read/write channel 144 usually.The interface that read/write channel 144 is used as between microcontroller 134 and the transducer 126 by circuit 145.Read/write channel 144 also is provided to motor and exciter controller 138 by circuit 146 with signal.Controller 138 provides interface between microcontroller 134 and the electric machine assembly 114 by circuit 148, and provides interface between microcontroller 134 and the actuator arm assembly 122 by circuit 150.
The data that are written to disk 112 are provided to read/write supporting interface circuit 136, and this circuit 136 then is provided to read/write channel 144 with data.Read/write channel 144 is sent to data head 126 with data, and data head 126 is used for disk 112 lip-deep data are encoded.
For reading of data from disk 112, the disk 112 lip-deep magnetic tracks that magnetic head 126 is written into by data, and produce the signal that flux directions changes on the expression magnetic track.Read signal is provided to read/write channel 144 from data head 126, and this channel generally includes channel model and detector.A kind of channel model of universal class is finite impulse response (FIR) (FIR) filter with a plurality of taps (tap).A kind of detector of universal class is a Viterbi type detector, and it detects data in a kind of known mode according to a kind of network.Filter is sent to data the Vitebi detector that detects data.Detected data are provided to usually carries out decoders for decoding to data, and continues to transmit, and further processes.
When manufacturer carried out analysis and assessment to new or different data heads 126, drive manufacturers may also not have contacted read/write channel circuit 144, and this circuit 144 will use with new data head.Read/write channel circuit 144 is not normally also developed ripe when testing with analysis and evaluation to the data magnetic head.
Fig. 3 is the block diagram of error rate test system, wherein contains the error rate of the sampled data system that comprises magnetic head 126 and disk 112.System shown in Figure 3 comprises controller 200, servo controller 202, data head 126 and disk 112.Controller 200 comprises mode generator (patterngenerator) 204 and the error rate (BER) element 206.The error rate directly calculates, and need not the read/write channel electronic circuit with magnetic head 126 and medium (or disk) 112 from measurement.
Make it possible to calculate the error probability of some certain errors incident according to measurement of the present invention.
What Fig. 4 described is first waveform 208 and second waveform 210.It has been observed that a kind of main error event in such system is to cause owing to the dipulse of describing of waveform among Fig. 4 208 being write the surperficial of disk 112 and being read waveform 210 represented constant magnetization.In other words, main error event owing to the expectation sampling+/-(10-1) and read sampling (000) and cause.So although can measure the error rate of any error event, the description of this specification is carried out at writing dipulse and reading constant magnetized error event.
For brevity, derive the probability expression formula that adopts the error that the grid detector run into now.The probability of some specific error event when consideration is judged by grid is for example in the detector that adopts the Viterbi type.Such error event is to take place in the path sizes in the incorrect path path sizes less than the correct path by grid.Suppose that correct path is path " a ", its path sizes is M under merging phase (merging state)
aFurther the incorrect path of hypothesis is path " b ", and its path sizes is M under merging phase
bSo,
Prob (Error)=Prob[M
b＜M
a]=Prob[(s
k-b
k)
2＜(s
k-a
a)
2] (equation 1)
In the formula, s
kRepresent noise samples, and a
kAnd b
kThe noise-free value of difference delegated path a and b.
Prob (Error)=Prob[-2 ∑ n
ke
k＞∑ e
k 2] (equation 2)
In the formula, e
k=a
k-b
k, and n
kRepresent the noise samples at k place constantly.Note sequence [e
k] be the error event that causes error.Under the situation of PR4 channel, [e
k]=[10-1] or [101].For an EPR4 channel, [e
k]=[11-1-1] or [1-111].
Certainly, it is also conceivable that other error event.We will suppose n
ke
kSummation be very approaching Gaussian function.So, can adopt Q (error) function of standard.Like this, equation 2 just becomes:
Prob (Error)=BER=Q[ESNR] (equation 4)
Particularly, for the PR4 channel:
For an EPR4 channel:
Here,
DEN＝
R(t
1，t
1)+R(t
2，t
2)+R(t
3，t
3)+R(t
4，t
4)
+2R(t
1，t
2)-2R(t
1，t
3)-2R(t
1，t
4)
-2R(t
2，t
3)-2R(t
2，t
4)+2R(t
3，t
4)
After having obtained generally being used for the expression formula of expression formula especially for PR4 channel and EPR4 channel of the error rate, can be described as obtaining the system of the measurement that ber value carries out now in more detail.
Fig. 5 is the flow chart (square frame 230-272) of the operation of the BER element 206 shown in the depiction 3.In addition, Fig. 6 is the more detailed functional block diagram of BER element 206.
At first, at the isolated pulse that is used for signal adjustment alignment at the beginning, just write suitable pattern.In order to discuss, wherein error event is to write a dipulse and read constant magnetized incident, and suitable pattern is an isolated dipulse.The isolated example of reflection error event that the pattern that writes preferably contains a plurality of (for example M).The lip-deep one routine pattern of writing disk 112 is shown in Fig. 7-1.Among Fig. 7-1, the longitudinal axis 242 expression voltages, transverse axis 244 express times.Adjusting alignment pulse 246 is to provide in first frame of pattern.Subsequently, write down a plurality of isolated dipulses 248.
Certainly, if error event under discussion is three pulses of a displacement, then M three isolated pulses are written in the disk 112.
In a descriptive embodiment, pattern contains the isolated example of 220 reflection error event.Pattern is write situation on the disk 112 see square frame 230 among Fig. 5.
Then, pattern retrieve element 212 is from disk 112 search modes.In a descriptive embodiment, pattern retrieve element 212 comprises traditional amplification and the circuit for signal conditioning that is used for from data head reception read signal.Subsequently, pattern retrieve element 212 will represent that the signal of the signal that reads from disc driver 112 is provided to data storage elements 214, and among the figure, this memory element comprises one or more solid-state memories.Read with memory module shown in the square frame among Fig. 5 232.
In a kind of descriptive embodiment, can be independent of the error rate of electronic noise basically.For this reason, more than 212 (as 20 times) read mode of pattern search circuit, and with each pattern storage in the pattern that reads in data storage 214.See shown in the square frame 234 among Fig. 5.Repeatedly the reason of read mode is because can the pattern that read for more than 112 time from disk be averaged, so that suppress the electronic noise with the signal correction of reading.Therefore, average mode has the relative electronic noise component that significantly reduces.Certainly, many more to the number of times that reads and be averaged of pattern, relevant electronic noise component is just more little.If electronic noise is included in the calculating of the error rate, then just only read the one-off pattern formula from disk 112.
In either case, as long as the number of times of read mode is greater than once, then each pattern in these patterns is all by the mode adjustment alignment be averaged element 216 and obtain from data storage 214.Mode adjustment is alignd and is averaged element 216 and adopts adjustment aligned signal 246 (in a descriptive embodiment is isolated pulse) to adjust all patterns of alignment rightly.Element 216 then is averaged all patterns, obtains an average mode.If pattern has been read N time and N pattern is averaged, relevant with this pattern so electronic noise component has just reduced
Doubly.In addition, support the speed of the rotating shaft rotation of disk 112 also can change.So before being averaged, element 216 can be adjusted the length of alignment pattern selectively to adapt to the variation of spindle speed.The adjustment of pattern alignment and be averaged and the adjustment alignment of modal length is seen shown in the square frame 236,238 and 240 among Fig. 5.
Refer now to the pattern of writing on the disk, it contains M isolated example.After having obtained average mode, with the M in the average mode isolated example separately, and temporarily it is adjusted alignment, thereby can be averaged, to suppress and this M the relevant All Media noise component(s) of isolated example to them.Among Fig. 7-2, transverse axis 250 is the time, and the longitudinal axis 252 is a voltage.Drawn a plurality of among Fig. 7-2 for being averaged through adjusting the isolated example of alignment.For these isolated examples 254 are averaged, represent example generator 218 from time memory 214, to retrieve average mode, pattern is divided into M isolated pulse, adjust these pulses of alignment, and they are averaged.This represents with square frame 256 in Fig. 5.Average signal s
Av(n) expression, the media noise component of this signal is suppressed.Therefore, just obtained signal s in so a kind of mode
Av(n), electronic noise and media noise promptly from this signal, have been removed.
Subsequently, subtraction signal s from M isolated (and being noise) sampling
Av(n), so that obtain M noise sequence.Noise sequence generator 220 retrieves M isolated noise sequence and noiseless sequence and (is included in sequence s from data storage 214
AvAnd from another, deduct one (n)).This produces M noise sequence, and these sequences are provided to autocorrelation matrix generator 222.This represents with square frame 258 in Fig. 5.
In the formula, on M noise sequence, k is sued for peace from 0 to M-1.Index i and j refer to the point in the specific noise sequence.The length of noise sequence is L.The generation of autocorrelation matrix is shown in the square frame among Fig. 5 260.
Then, channel model generator 224 produces a channel model.For this reason, channel filtering generator 224 signal s
Av(n) import as channel model, and output claims to channel model.Channel requires to require (being the channel equalization target) corresponding with the channel model of selecting.For example, in a descriptive embodiment, channel requires corresponding to PR4 channel, EPR4 channel and E
2In the pR4 channel one.According to the channel requirement of channel model input and channel model output place, channel model generator 224 produces the channel model with impulse response h (n).The length of impulse response h (n) also is L.See shown in the square frame 262 among Fig. 5.
Then, R
0Generator 226 is at channel model output (R
0) locate to produce an autocorrelation matrix.For this reason, R
0Generator 226 at first produces a matrix H, and it is the matrix (noticing that R is the matrix that a L takes advantage of L) that a 2L-1 takes advantage of L.In the following manner, produce the H matrix from h (n):
When 0≤(j-k)≤(i-1), and H (j, k)=h (j-k)
Otherwise, H (j, k)=0
Here, 0≤j≤(2L-2), and 0≤k≤(L-1) (equation 8)
And L is the length of noise sequence.
Can from the output of above-mentioned R and H calculating channel filter the auto-correlation of Viterbi grid input noise.The autocorrelation matrix of channel model output is defined as follows: R
0(i, j)=(i) n ' is (j) for E[n '] (equation 9) here, E is the expectation operator, n ' is a noise sequence through filtering (i).Channel model output place is provided by following formula through the noise of filtering:
Here, n (k) is a channel model input noise sequence, and h (n) is the unit-sample response of channel model.The length of n (k) and h (n) is 1.Equation 10 is substituted in the equation 9, then obtains following formula:
Here, 0≤i≤2L-2,0≤j≤2L-2,0≤k≤L-1, and 0≤1≤L-1
The form that equation 11 can be write as matrix notation is as follows:
R
0=HRH
T(equation 12)
Here,
For 0≤(j-i)≤L-1, and H (i, j)=h (j-i)
Otherwise, H (i, j)=0
At channel model output R
0The situation that the place forms the H matrix and obtains autocorrelation matrix is seen shown in the square frame 264 and 266 among Fig. 5.
Subsequently, signal s
Av(n) by having the filter of impulse response h (n).This channel model output place provide noiseless sampling+/-(smp, 0 ,-smp), here, smp represents a sampling value.Therefore, now just there has been enough information to come the calculating chart 5 and the error rate shown in Figure 6.Particularly, for the PR4 channel, can calculate the error rate as getting off:
In the formula, Q is a standard Gauss q function.Therefore, the error rate determines that element is at R
0The channel model output R of generator 226
0The place retrieves the desired value of auto-correlation element, and retrieves the noiseless sampling from channel model 224, and calculates the error rate 229.This is illustrated by square frame 268,270 and 272 in Fig. 5.Should be noted that this can realize under the significant approximate situation not having the read/write channel circuit and do not have.
The present invention also can consider a total BER who includes the BER corresponding with more than one error event.Fig. 8 is a flow chart of describing the calculating of such total bit error rate.At first, a plurality of patterns are written in the disk 112, each pattern has the isolated sampling relevant with different error event.This is by shown in square frame 274 and 276.Select in a plurality of patterns first, and determine with by the relevant error rate of the error event of this pattern representative.See square frame 278 and 280.Determining basically of the error rate to finish with Fig. 5 described same mode.If also leave pattern, then select these patterns, and also determine the error rate relevant with these patterns.See square frame 282 and 284.
After having calculated all error rates,, they are weighted according to they dominance relations to the whole error rate.Weighting function is preferably determined by experience.See shown in the square frame 286.All error rates through weighting are added together, so that obtain being used for total error rate of each error event of being considered.See shown in square frame 288 and 290.
Fig. 9 is the autocorrelation matrix of channel model input.With the nanosecond express time, represent voltage with transverse axis 300 and 302 with the longitudinal axis 304.The transition of isolated sampling interval (transitionspacing) is 10 microinch (254nm) in the pattern, and the unit of axle 304 is volts
2* 10
-5
Figure 10 is drawn in similar axle shown in Fig. 9 to go up filter output (R
0) autocorrelation matrix located.
Therefore can see, can obtain the error rate of sampled-data system with the present invention, and need not to develop the read/write channel circuit, and need not to carry out approximate significantly.Can see that whole measuring process needs about three minutes, can be used for improving different types of data head of data head structure and more follow-on disc driver.
The present invention can determine the method for the error rate as having the disk 112 that is formed by magnetic medium and writing data and retrieve the sampled data system of data head 126 of data from disk 112 on disk 112.This method is included in the step 230 of the pattern that writes on the disk 112, and this pattern comprises M isolated example 248 that is associated with the error event of being scheduled to.This method also comprises from disk 112 and retrieves that pattern and the example 248 that M is isolated combine (in step 256) and the representative example s that obtains having the media noise component that reduces
Av(n) step 232.This method also comprises step 258, to represent each combination in the individual isolated example 248 of example and M and obtain M noise sequence, obtain (in step 260) auto-correlation components R according to the M that reads from disk 112 example and M noise sequence, according to representing example and requiring to obtain (at step 262 place) channel model with impulse response h (n) according to channel to the predetermined channel model, represent example to obtain filter output sampling (smp) by filter transmission (at step 268 place) with impulse response h (n), and the error amount of exporting the sampling and auto-correlation element R definite (at step 270 place) the expression error rate (BER) according to filter.
In a kind of exemplary embodiment, this method also comprises the step 232 of the pattern that storage obtains from disk 112 retrieval.Searching step also comprises (at step 234 place) and repeatedly retrieves pattern obtaining the pattern of a plurality of retrievals from disk 112, and get up the mode combinations of retrieval and obtain having the representative search modes of the electronic noise component that reduces at (at step 238 place).
In one exemplary embodiment, the step 238 that the mode combinations of retrieval is got up also comprises the pattern of temporary transient adjustment alignment alignment retrieval and the pattern of retrieving is averaged.
In a kind of exemplary embodiment, the step 256 that example is combined also comprises M isolated example 248 of temporary transient adjustment alignment and M isolated example 248 is averaged to obtain representing the step 256 of example.In a kind of preferred embodiment, obtaining M temporary transient sequence is to realize by deduct the step 258 of representing example from M example.In addition, the step that obtains described auto-correlation element also comprises at M 248 couples of M of isolated example temporary transient sequences and is averaged to obtain the step 260 of autocorrelation matrix R.
In addition, a plurality of patterns can be write on the disk 112, each pattern comprises a plurality of isolated example 248 that is associated with different error event.Represent that the error event of the total bit error rate corresponding with error event determines according to weighted error value.
The present invention can also realize with the form of device.Pattern write element 204 is write pattern on the disk 112.Pattern retrieve element 212 search modes.First composition element 218 obtains to represent example, and second composition element 220 obtains noise sequence, and auto-correlation component (component) generator 222 produces the auto-correlation components R.Channel model generator 224 produces channel model, and error is determined the error amount 229 of element 228 according to the filter output sampling and the definite expression of the auto-correlation component error rate.In one embodiment, pattern retrieve element 212 also comprises data storage 214, and repeatedly from disk 112 search modes.
Be to be understood that, although in preamble, described various feature and advantage of various embodiment of the present invention, and the detailed structure of each embodiment of the present invention and function, but these descriptions only are descriptive, in detail particularly in concept of the present invention to the structure of parts with arrange various changes on the represented degree of the wide meaning of term doing in the claim to be stated.For example, can be according to desired specific equalization target of BER and error event, and keep identical functions substantially, under the situation that does not depart from scope and spirit of the present invention, change specific element.
Claims (10)
- Have the disk that forms by magnetic recording medium and be used on disk writing events and from disk the method for a kind of definite error rate the sampled data system of the data head of retrieve data, it is characterized in that described method comprises following step:(a) write a pattern on described disk, described pattern comprises and the predetermined relevant isolated example of error event;(b) the described pattern of retrieval from described disk;(c) described isolated example is combined, to obtain having the representative example of the media noise component that reduces;(d) in described representative example and the described isolated example each is combined, obtain corresponding noise sequence;(e) according to the isolated example and the described noise sequence that read from described disk, obtain an auto-correlation component;(f), and, obtain having the channel model of impulse response according to the channel requirement of predetermined channel model according to described representative example;(g) transmit described representative example by filter, to obtain filter output sampling with impulse response; And(h), determine the error amount of the described error rate of expression according to described filter output sampling and auto-correlation component.
- 2. the method for claim 1 is characterized in that, described searching step comprises following step:(b) (i) repeatedly from the described pattern of described search disk, to obtain the pattern of a plurality of retrievals; And(b) (ii) the mode combinations of described retrieval is got up, to obtain having the representative search modes of the electronic noise component that reduces.
- 3. method as claimed in claim 2 is characterized in that, described combination step (b) (ii) also comprises following step:(d) (i) the temporary transient pattern of adjusting the described retrieval of alignment; And(d) (ii) the pattern of described retrieval is averaged.
- 4. method as claimed in claim 3 is characterized in that described disk rotates under a speed, and described temporary transient adjustment alignment step comprises the step of the length of the pattern of regulating described retrieval with the speed-adaptive variation.
- 5. the method for claim 1 is characterized in that, described combination step (c) comprises following step:(c) (i) the described isolated example that temporarily aligns; And(c) (ii) described isolated example is averaged, to obtain described representative example.
- 6. the method for claim 1 is characterized in that, the step of described acquisition (e) comprises at described isolated example described noise sequence is averaged to obtain an autocorrelative step.
- 7. the method for claim 1 is characterized in that, the described step of writing is included in the step that writes a plurality of patterns on the described disk, and each pattern comprises a plurality of isolated example relevant with different error event.
- 8. method as claimed in claim 7 is characterized in that, it also comprises each the pattern repeating step (b) in a plurality of patterns to (h).
- 9. method as claimed in claim 8 is characterized in that, it also comprises following step:(i) to each weighting in the described error amount of determining, to obtain through weighted error value; And(j), determine the aggregated error value of the total bit error rate that expression is corresponding with described error event according to described weighted error value.
- 10. the device of the error rate in the sampled data system of determining to have the disk that is formed by magnetic recording medium is characterized in that described device comprises:The pattern write element, described element is arranged with respect to described disk, and in order to write pattern on described disk, described pattern comprises and a predetermined relevant isolated example of error event;The pattern retrieve element, it is arranged with respect to described disk, in order to from the described pattern of described search disk;First composition element, it and described pattern retrieve element are coupled, and with described isolated example combination, obtain having the representative example of the media noise component that reduces;Second composition element, it and described first composition element are coupled, with each combination in described representative example and the described isolated example, to obtain corresponding noise sequence;The auto-correlation component generator, it and second composition element are coupled, and produce an auto-correlation component according to described isolated example that reads from described disk and described noise sequence;The channel filtering generator, it and described first composition element are coupled, require to produce the channel model with impulse response according to described representative example with according to the channel to the predetermined channel model, wherein, described channel is used for receiving described representative example and filter output sampling is provided; AndError is determined element, and the error amount of the described error rate is determined to represent in it and described auto-correlation component generator and the coupling of described channel model according to described filter output sampling and described auto-correlation component.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US5800997P | 1997-08-28 | 1997-08-28 | |
| US60/058,009 | 1997-08-28 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1269073A true CN1269073A (en) | 2000-10-04 |
| CN1132331C CN1132331C (en) | 2003-12-24 |
Family
ID=22014099
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN988086662A Expired - Fee Related CN1132331C (en) | 1997-08-28 | 1998-08-27 | Method and apparatus for determining bit error in a sampled data system |
Country Status (6)
| Country | Link |
|---|---|
| JP (1) | JP2002519801A (en) |
| KR (1) | KR100424131B1 (en) |
| CN (1) | CN1132331C (en) |
| DE (1) | DE19882647T1 (en) |
| GB (1) | GB2343289B (en) |
| WO (1) | WO1999010990A1 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100442864B1 (en) | 2001-09-04 | 2004-08-02 | 삼성전자주식회사 | Method and apparatus for performing a recovery routine by instability head detection in a data storage system |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5121263A (en) * | 1989-10-31 | 1992-06-09 | International Business Machines Corporation | Method and apparatus for determining the error rate of magnetic recording disk drives having a amplitude sampling data detection |
| EP0593021B1 (en) * | 1992-10-14 | 1998-09-16 | Sony Corporation | Magnetic disk apparatus |
| JPH0877726A (en) * | 1994-09-06 | 1996-03-22 | Hitachi Ltd | Digital signal processing device, error detecting method and recording medium reproducing device |
| US5600500A (en) * | 1994-11-14 | 1997-02-04 | Seagate Technology, Inc. | Performance based write current optimization process |
| US5771127A (en) * | 1996-07-29 | 1998-06-23 | Cirrus Logic, Inc. | Sampled amplitude read channel employing interpolated timing recovery and a remod/demod sequence detector |
-
1998
- 1998-08-27 GB GB0004048A patent/GB2343289B/en not_active Expired - Fee Related
- 1998-08-27 DE DE19882647T patent/DE19882647T1/en not_active Withdrawn
- 1998-08-27 CN CN988086662A patent/CN1132331C/en not_active Expired - Fee Related
- 1998-08-27 KR KR10-2000-7002064A patent/KR100424131B1/en not_active IP Right Cessation
- 1998-08-27 WO PCT/US1998/017768 patent/WO1999010990A1/en active IP Right Grant
- 1998-08-27 JP JP2000556558A patent/JP2002519801A/en active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| GB2343289A (en) | 2000-05-03 |
| GB2343289B (en) | 2001-09-12 |
| DE19882647T1 (en) | 2000-08-03 |
| WO1999010990A1 (en) | 1999-03-04 |
| JP2002519801A (en) | 2002-07-02 |
| KR20010023422A (en) | 2001-03-26 |
| CN1132331C (en) | 2003-12-24 |
| GB0004048D0 (en) | 2000-04-12 |
| KR100424131B1 (en) | 2004-03-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5233482A (en) | Thermal asperity compensation for PRML data detection | |
| US6657803B1 (en) | Method and apparatus for data error recovery using defect threshold detector and viterbi gain | |
| CN1294564C (en) | Method and device of caculating retry times in optical drive | |
| US6606211B1 (en) | Method and apparatus for detecting media defects in a disc drive | |
| US6252731B1 (en) | Parametric optimization using disc drive read channel quality measurements | |
| US5392295A (en) | Error measurement circuit | |
| US7307807B1 (en) | Disk servo pattern writing | |
| US9019642B1 (en) | Synchronization mark detection for multi-dimensional magnetic recording | |
| US6873488B2 (en) | Enhanced MR offset with dynamic tuning range | |
| US8164846B1 (en) | Hybrid defect detection for recording channels | |
| CN1945718A (en) | Dynamic method for asymmetry compensation in a storage read channel | |
| WO2003030168A1 (en) | Method and apparatus for detecting media defects | |
| JP2003022623A (en) | Method for deciding position of conversion head on disk surface, and disk drive | |
| US6031672A (en) | Method and apparatus for servo data pattern detection | |
| CN1664931A (en) | Data storage device, storage device control method, and magnetic disk drive with parameter setting | |
| US6947232B2 (en) | User data wedge media certification apparatus and method | |
| US6583942B2 (en) | Method and apparatus for determining equalized signal-to-noise ratio in a storage device | |
| CN1355921A (en) | Accumulating read channel performance data | |
| CN1577501A (en) | Method and apparatus for servo writing using incorporated servo writer in a disk drive | |
| US7245444B2 (en) | Method and apparatus for providing a read channel having imbedded channel signal analysis | |
| US7760454B1 (en) | Sequence-permutation control information detection and signal polarity determination | |
| CN1132331C (en) | Method and apparatus for determining bit error in a sampled data system | |
| US6735029B2 (en) | Equalized response extraction from a read channel | |
| US6877116B1 (en) | Method and apparatus for determining bit error rate in a sampled data system without requiring read channel circuitry | |
| CN1920993A (en) | Data storage device and a method of controlling write of user data thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent for invention or patent application | ||
| CB02 | Change of applicant information |
Applicant after: Seagate Technology, Inc. Applicant before: Sichater Tehc. Co., Ltd. |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: APPLICANT; FROM: SEAGATE TECHNOLOGY, INC. TO: SEAGATE TECHNOLOGY LLC |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |