Patent Application
20050185313
This invention relates to an improved data reader and an improved method of reading data. It is particularly applicable to data storage devices, but may have wider applicability.
Early magnetic tape storage devices moved the magnetic tape past read heads at a fixed velocity. A low pass filter is provided within the decoding circuitry that removes unwanted high frequency noise. Because the tape passes the read head at a fixed velocity a signal produced by reading data read from the tape has a known maximum frequency. Therefore, the cut off frequency of the low pass filter can be set to ensure all of the signal passes the filter.
It is now known to produce magnetic tape storage devices in which the velocity of the tape past the read head is varied. Varying the velocity in this manner helps to ensure that the rate of data transfer to and from the tape can match the rate of data transfer to the storage device. This matching of data rates helps to prevent unnecessary stopping of the storage device. Stopping causes wear to the drive mechanisms and therefore, wear can be reduced if the drive can be slowed rather than stopped.
However, by altering the velocity at which the tape passes the read head, the frequency of the signal produced on reading data from the tape is altered. Therefore, it is desirable that the cut-off frequency of the low pass filter is altered accordingly. It is undesirable to have the cut-off frequency set too far above the maximum frequency of the signal since noise will not be effectively removed. Further, if the cut-off frequency is set too low then a portion of the signal will be lost. Generally, as the tape speed increases, data rate increases and it is necessary to increase the level of the cut-off frequency in order that the higher frequency data are not filtered or attenuated.
Prior solutions to this problem are known and an example is shown in
This technique relies on the matching of components in the PLL and the filter. This can be difficult to achieve over fabrication process corners and for the whole frequency range. These difficulties can result in a poor yield in the fabrication process.
It is an object of the present invention to provide a data storage device that is easier to fabricate than the prior art.
According to a first aspect of the invention there is provided a data reader arranged to produce a signal on reading a data holding medium, said data reader comprising processing circuitry arranged to process said signal, said processing circuitry including a filter having a variable cut-off frequency, a velocity signal generator arranged to produce a signal corresponding to the velocity of the data holding medium, and a processor, said processor being arranged to read the velocity signal and vary the out-off frequency of the filter in substantially linear relation to variations in the velocity signal on the basis of data generated during a calibration process of the data reader.
An advantage of such a data reader is that it is easier to fabricate than prior art data readers.
Preferably, the filter is a low pass filter.
The velocity signal generator may be provided by a read head arranged to produce the signal on reading a data holding medium. The velocity signal is preferably a clock signal produced by the read head. Conveniently the frequency of the clock signal corresponds to the velocity of the tape past the read head. It is advantageous to have a clock signal as the velocity signal since this is readily read by the processor. Alternatively, an analogue signal may be produced, but it is likely that such an analogue signal would need digitising before being able to be read by the processor.
Preferably, the read head is arranged to read markers on the data-holding medium in order to produce the velocity signal. Such an arrangement provides a simple way of allowing the velocity signal to be generated.
The processor may be arranged to determine the frequency of the velocity signal in order to vary the cut-off frequency of the filter.
In any case, the processor may be arranged to vary the appropriate cut-off frequency of the filter by referring to a look up table, the values of which look up table may have been generated during the calibration process. Use of a look up table in this manner provides a simple, yet effective system for controlling the cut-off frequency.
Alternatively, the processor may set the cut-off frequency by applying a representation of the velocity signal to a function such as a polynomial function weighted to generate an appropriate output to control the cut-off frequency of the low pass filter.
Conveniently, the processor includes an output register arranged such that the register's content controls cut-off frequency of the filter. Such an output register is convenient since is provides a simple technique to output the desired out-off frequency.
Preferably a Digital to Analogue Converter (DAC) is provided within the processing circuitry, arranged to produce an analogue signal to control the cut-off frequency of the filter.
Preferably, the DAC is arranged to have input thereto the value that is contained in the output register of the processor. Such an arrangement provides a convenient structure for controlling the filter.
The processor may be arranged to perform a self-test routine in which the values that are contained in the look up table are adjusted. Such an arrangement is convenient because it allows minor discrepancies in the values that are contained in the look up table to be corrected. Therefore, the control of the cut-off frequency of the filter should be more accurate.
The filter may be arranged to have an increased gain in the region of the cut-off. This is advantageous because it provides the necessary equalisation to achieve the desired signal characteristics.
Preferably, the reader is arranged to cause the velocity of the data-holding medium to be varied over any velocity within a predetermined range. In one embodiment the maximum velocity is limited to roughly three times the minimum velocity. Other ratios of maximum to minimum are equally possible: For instance roughly any of the following may be suitable: 2 to 1, 4 to 1, 5 to 1, 6 to 1, 8 to 1, or 10 to 1, or indeed any value in between these ranges.
In the preferred embodiment the maximum tape velocity is roughly 4.1 m/s. However, the maximum tape velocity may be roughly any of the following values: 1 m/s, 2 m/s, 3 m/s, 5 m/s, 6 m/s, 7 m/s, 8 m/s, 9 m/s.
Alternatively, or additionally, the reader is arranged to cause the velocity of the data-holding medium to be varied over a number of predetermined velocities within a predetermined range.
According to a second aspect of the invention there is provided a data storage device including a data reader according to the first aspect of the invention.
Advantageously, the storage device is provided with a buffer arranged to receive data sent to the device (and/or buffer data sent from the device). The data reader may be arranged to cause the velocity of the data-holding medium to be varied according to the amount of data present in the buffer.
In one embodiment the data storage device is arranged to receive magnetic tapes wherein the magnetic tape provides the data-holding medium. However, the device may be arranged to read data from a hard disk wherein the disk platter is the data-holding medium. The storage device may be arranged to read data from other forms of data-holding medium.
The storage device may be any of the following types of tape drive and for example may be any of the following: DAT (Digital Audio Tape), DLT (Digital Linear Tape), DDS (digital Data Storage), or LTO (Linear Tape Open), or any other type.
The storage device may be arranged to communicate with other devices via any form of bus. The bus may be SCSI, Firewire, USB, Fibrechannel, etc.
According to a third aspect of the invention there is provided a method of reading data from a data-holding medium to produce an output signal, the method comprising determining the velocity of the data holding medium and varying the cut-off frequency of a filter in substantially linear relation to variations in the velocity of the data holding medium on the basis of data generated during a calibration process of the data reader An advantage of such a method is that it is easier to perform than prior art methods of reading data from a data-holding medium.
Conveniently, the processor consults a look up table to vary the cut-off frequency. Such a method provides a convenient way of determining the cut-off frequency.
Alternatively, other methods of determining the cut-off frequency may be utilised by the processor. For instance, the processor may apply a predetermined function to a representation of the velocity of the data-holding medium to determine the cut-off frequency.
The method may comprise performing a self calibration routine in which the processor adjusts values contained in the look up table/adjusts the predetermined function to help ensure that the cut-off frequency of the filter is correctly controlled relative to the velocity of the data holding medium. This helps to ensure that data can be accurately read from the data-holding medium.
Conveniently, the filter is arranged to provide gain to a signal fed thereto in a region of the cut-off frequency. This helps to equalise the signal to the desired signal characteristics.
Preferably, the processor utilises a clock signal generated by a read head to determine the velocity of the data-holding medium. Read heads of data storage devices generally produce such a clock signal. Therefore, utilising this signal is a convenient way of determining the velocity.
There now follows by way of example only a detailed description of the invention with reference to the accompanying drawings of which:
This invention will be described in relation to a magnetic tape data storage device, although it may have wider applicability. The basic components of a magnetic tape storage device 2 are shown in
However, in a tape storage device 2 in which the velocity of the tape 4 is varied to vary the data rate the frequencies contained in the envelope 16 also vary. Therefore, to ensure that the cut-off frequency fc is not too high or too low (ie let through too much noise, or remove some of the wanted frequencies respectively) the processing circuitry 8 is arranged to vary it.
This clock signal 21 is fed to a processor 22, which determines the frequency of the clock signal 21 and consequently determines the velocity of the tape 4 passing the read head 6. In the present embodiment, the processor 22 has associated therewith a look up table 24, which contains a list of register values for various velocities of tape 4. The look up table 24 could of course be provided with memory external to the processor 22 such as E2PROM, or other non-volatile memory or possibly within dedicated memory provided within the processor 22. The values contained in the look-up table 24 are determined at the time of device manufacture and are specific to each data reader.
The processor 22 includes an output register 26, which is arranged to receive the value which, when applied to the low pass filter, causes the cut-off frequency fc of the filter 14 to be set to the appropriate value. A Digital to Analogue Converter (DAC) 28 is provided and arranged to convert the digital value, placed by the processor, into the output register 26 into an analogue signal. The analogue signal 30 produced by the DAC 28 is fed to the filter 14 such that the cut-off frequency fc is varied appropriately. The analogue signal is presented as a voltage or current respectively depending on whether the filter cut-off frequency is voltage or current controlled.
In some embodiments the filter 14 is arranged such that there is an amount of gain in the region of the cut-off frequency fc. An example of this is shown in
In use, the tape 4 is inserted into the storage device 2 such that the read head 6 can read it. As the tape 4 passes the read head 6 the markings 18 on the tape are read 32 and used to produce the clock signal 21. The velocity of tape 4 is varied to alter the rate at which data is moved to/from the tape 4 to/from a device connected to the storage device 2 via the port 10.
As the velocity of the tape 4 varies the frequency of the clock signal 21 varies. Therefore, as the clock signal 21 is input to the processor 22 (block 34) the processor 22 can determine the velocity of the tape 4 by determining the frequency of the clock signal 21. Once the frequency has been determined the processor looks up in the look up table 24 (block 36) the output value to generate the required filter cut off frequency.
The output value determined by looking in the look up table 24 is placed into the output register 26, and converted to an analogue signal 30 by the DAC 28. This analogue signal 30 is input to the filter 14 to control the cut-off frequency fc (block 38).
According to the present embodiment, the relationship between the tape velocity and the cut-off frequency is substantially linear throughout the required range. The values in the look up table are set during a drive calibration process at the time of manufacture, whereby non-linearities are factored-out by setting appropriate values in the look up table. Because the calibration process maps the tape velocity to the required value of fc, compensation can be made for any of these non-linearities so that velocity versus fc is linear. One example of a calibration process would be to increment the tape velocity in a linear fashion and, for each increment, vary the respective value in the look up table in order to obtain the required cut-off frequency. Other appropriate calibration procedures would be apparent to the skilled person.
If the relationship is not linear then the drop out level in data read by the reader may increase significantly (due to increased noise because the cut-off frequency is set too high, or to loss of signal because the cut-off frequency is set too low). Compensating for non-linearity allows the cut-off frequency to be set at the correct level.
The skilled person will appreciate that the term processor is envisaged to cover a range of different types of circuit: micro-controllers, microprocessors, ASIC's, Programmable Logic Arrays (PLA), hardwired circuitry of discrete components, etc.
In alternative embodiments, the look up table may be replaced by a function for generating a processor output value, which causes the tape velocity to have a linear relationship with the cut-off frequency. In effect, the function would compensate for any intrinsic non-linearity. One function would be a polynomial of the form a+bx+cx2+dx3+ . . . +nXm, where x represents the tape velocity signal value and the coefficients a, b, c, d, . . . , n are set at appropriate values to compensate for any non-linear relationships between the tape velocity and the resultant fc. A similar calibration procedure as the one described above to calibrate the look up table could be used to calibrate the polynomial. Instead of varying the look up table values, however, calibration of the polynomial would require varying the coefficients a, b, c, d in order to arrive at the correct output.
| Number | Date | Country | Kind |
|---|---|---|---|
| 0026622.1 | Oct 2000 | GB | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 09984619 | Oct 2001 | US |
| Child | 11111867 | Apr 2005 | US |
