Technical Field of the Invention
This invention relates generally to the field of disk drive circuits and more particularly to a method and circuit for providing feedback in a pre-amplifier for a multi-head disk drive.
Background of the Invention
In a hard disk drive, data is stored in a medium of magnetic material on the surface of the disk. The data is generally stored in concentric tracks, each track having a sequence of regions with opposing magnetic polarities. Thus, as one travels along the track, a change in polarity represents a "one," whereas the absence of a change in polarity represents a "zero."
To read the data from the disk, a magneto-resistive head is run along the track just above the surface of the disk. A change in polarity on the disk causes a change in the resistance of the read head. This change can be used to produce an output current from the read head circuitry. This is generally accomplished by driving two initially equal currents into the read head circuit. A change in the resistance of the head, which occurs when a flux change or "one" is detected on the disk, will result in the two output currents becoming unequal. A differential current output is therefore provided on two output lines from the read head. This differential current can be converted to a voltage, which in turn can be amplified and read by signal processing electronics.
A hard disk drive (HDD) unit generally contains several disks or platters in a column, each disk having a read head on either side of the disk. Generally, the output from all of the read heads will be channeled through one feedback and preamplifier circuit. Thus, only one read head is selected for reading by a switching system at any given time.
Each read head has slightly different magneto-resistive characteristics. A feedback circuit is used to correct for the individual characteristics of the selected read head. The feedback circuit senses an offset or differential current in the read head output when no flux change is detected (a "zero") on the disk. Since there should be no output when no flux change is being detected, the feedback circuit applies a small bias voltage to the read head. This bias voltage causes the output of the read head to become equalized. Thus, no differential current is produced as output by the read head unless a flux change is detected.
Problems arise when the switching system switches from one read head to another. Because magneto-resistive characteristics vary slightly from head to head, the bias voltage produced by the feedback circuit must likewise vary depending on which read head is selected. The design of the feedback circuit is such that a certain amount of time is required to fully respond to a change in the required compensating voltage. Thus, after a read head change, the output of the read channel will be under- or over-compensated until the feedback circuit adjusts, resulting in a temporarily incorrect output signal.
It is desirable to shorten the time required to adjust the read circuit feedback after a change in read heads. It is also desirable to do so without affecting the performance of the feedback circuit during normal operation.
Summary of the Invention
A method for providing feedback in a pre-amplifier for a multi-head disk drive is disclosed. The method comprises the steps of (1) generating first and second currents in response to a read head bias voltage; (2) generating third and fourth currents in response to the read head bias voltage; (3) amplifying a selected one of the third and fourth feedback currents to produce an amplified current; (4) supplying the first current to a first charge plate of an external capacitor; (5) supplying the second current and the amplified current to a second charge plate of the external capacitor; (6) supplying a potential level associated with the first charge plate of the capacitor to a first voltage input of the read head circuit; and (7) supplying a potential level associated with the second charge plate of the capacitor to a second voltage input of the read head circuit.
A circuit for providing feedback to a selected read head in a multi-head disk drive read circuit in accordance with the invention comprises (1) a first feedback circuit for receiving an output voltage from first and second voltage output nodes of the read circuit, and for providing first and second output currents to corresponding first and second charge plates of a capacitor; (2) a second feedback circuit for receiving the output voltage from the voltage output nodes of the read circuit, and for providing third and fourth output currents to the first and second charge plates, respectively, of the capacitor; (3) first and second feedback lines for supplying voltages associated with the first and second charges plates, respectively, of the capacitor to corresponding first and second voltage inputs of the selected read head circuit; and (4) a switch for activating the second feedback circuit.
A technical advantage of the present invention is that a method and circuit for providing feedback in a pre-amplifier for a multi-head disk drive is provided. Another technical advantage is that the invention provides for quicker adjustment of feedback after a change in read heads than was provided by prior art systems. Another technical advantage is that the second feedback circuit is inactive during normal operation of the disk drive, activating only after a change in read heads. Another technical advantage is that the operation of the first feedback circuit is not detrimentally affected by the operation of the second feedback circuit.
Brief Description of the Drawings
For a more complete understanding of the present invention, the objects and advantages thereof, reference is now made to the following descriptions taken in connection with the accompanying drawings in which:
FIG. 1 is an overall block diagram of a hard disk drive read channel.
FIG. 2 is an illustration of the main feedback circuit for the read channel.
FIG. 3 is an illustration of a quick recovery feedback circuit in accordance with the invention.
FIG. 4 is a timing diagram illustrating the operation of the quick recovery feedback circuit.
Detailed Description of the Invention
The preferred embodiment of the present invention and its advantages are best understood by referring to FIGS. 1 through 4 of the drawings, like numerals being used for like and corresponding parts of the various drawings.
Referring to FIG. 1, a hard disk drive read circuit 100 which incorporates quick recovery feedback circuit 300 in accordance with the invention is shown. Read circuit 100 includes magneto-resistive read heads 111 and 121. For illustrative purposes, only two read heads are shown. However, it will be understood by those skilled in the art that any number of read heads could be incorporated into the design of read circuit 100.
Read head 111 is associated with read head circuit 115, which includes a select input 116, voltage inputs 117 and 118 and current outputs 112 and 113. Likewise, read head 121 has an associated read head circuit 125 which includes select input 126, voltage inputs 127 and 128 and current outputs 122 and 123.
Only one read head circuit at a time may be activated by its select input 116 or 126. Therefore, only the operation of read head 111 and associated read head circuit 115 will be described. However, it should be noted that read head circuit 125, and any other read heads incorporated into the design of read circuit 100, would operate according to the same principles.
When read head circuit 115 is activated by a signal on select input 116, a current is sunk on select input 116, and the voltages on voltage inputs 117 and 118, as well as the resistance of read head 111, cause output currents on output lines 112 and 113. As read head 111 passes over the surface of the hard disk (not shown), the resistance of read head 111 changes, resulting in a differential current output on output lines 112 and 113. Thus, when read head 111 senses a flux change in one direction, the current output on line 112 will be greater than the current output on line 113. Conversely, when read head 111 senses a flux change in the other direction, the current on output line 113 will be greater than the current on output line 112.
Current-to-voltage converter 140 converts the differential current on lines 112 and 113 to a potential difference between nodes 103 and 104. Post-amplifier 150 amplifies this voltage, and the amplified voltage is read out on lines 101 and 102.
Referring to FIG. 2, a diagram of main feedback circuit 200 is shown. The voltages of nodes 103 and 104 are supplied to the bases of matched transistors 220 and 230, respectively. A constant current sink 210 is connected to and draws a constant current from the emitters of transistors 220 and 230. The collectors of transistors 220 and 230 are connected to current mirrors 240 and 250, respectively.
The operation of the main feedback circuit is as follows. When no flux change is detected on the hard disk by read head 111, nodes 103 and 104 should be at the same potential. Any potential difference between nodes 103 and 104 would be a read head bias which must be corrected. The primary object of the main feedback circuit is to provide feedback which will correct for this bias.
If there is a potential difference between nodes 103 and 104, then one of the transistors 220 or 230 will be caused to conduct slightly less than half the current sunk by current sink 210, while the other transistor 220 or 230 will be caused to conduct slightly more than half the current sunk by current sink 210. Current mirror 240 will cause the amount of current conducted by transistor 220 to be driven to node 105. Likewise, current mirror 250 will cause a current equal to that conducted by transistor 230 to be driven to node 106.
Referring back to FIG. 1, it can be seen that, as a result of the action of main feedback circuit 200, nodes 105 and 106 will receive slightly different currents. Capacitor 190, which is connected between nodes 105 and 106, will act as a low pass filter, filtering out any current spike which occurs at node 105 when a complementary current spike occurs at node 106. Capacitor 190 also effectively converts a current difference between nodes 105 and 106 into an increasing voltage difference between lines 196 and 197. When the current driven to node 105 is slightly less than the current driven to node 106, the potential of line 196 will slowly drop relative to the potential of line 197 as charge accumulates on the plates of capacitor 190. Capacitor 190 is an external capacitor, not residing on the same substrate as the other components of read circuit 100.
Current mirror 195, which is connected between the outputs of the main feedback circuit 200, operates to subtract an equal current from both output lines. As a result, capacitor 190 sees only a small differential current, which it converts into a voltage difference on lines 196 and 197 as previously described.
This increasing voltage difference on lines 196 and 197 is fed back via lines 117 and 118 to read head circuit 115. When the voltage difference between lines 196 and 197 has reached a level sufficient to correct for the output bias of read head circuit 115, nodes 103 and 104 will reach the same potential, and the current outputs of main feedback circuit 200 will equalize. At this point, the charge held by capacitor 190 will stabilize, holding the read head circuit input voltages on lines 117 and 118 at the desired values.
One limitation of main feedback circuit 200 is that it cannot respond quickly to a change in the bias voltage between nodes 103 and 104. External capacitor 190, which has a large capacitance, carries a stable amount of charge during normal operation. When the output currents at nodes 105 and 106 change, as when a new read head is selected for output, capacitor 190 must be charged or discharged to reach a new charge level associated with the new feedback level.
The charging or discharging of capacitor 190 causes a delay before the feedback voltages carried by lines 196 and 197 attain their desired values. Thus, after a read head change, the output of the read channel will be under-or over-compensated until the charge carried by external capacitor 190 adjusts to a new level. This results in a temporarily incorrect output signal.
To shorten the time required to adjust after a read head change, it is desirable to temporarily amplify the current difference between nodes 105 and 106, so as to more quickly charge or discharge capacitor 190. This current amplification is accomplished by quick recovery feedback circuit 300, shown in FIG. 1. Quick recovery feedback circuit 300 receives analog voltage inputs from nodes 103 and 104 and a logic input from AND gate 180. Circuit 300 drives output currents to nodes 105 and 106. Quick recovery feedback circuit 300 is therefore connected into circuit 100 in parallel with main feedback circuit 200.
Referring to FIG. 3, quick recovery feedback circuit 300 is shown in detail. The operation of this circuit is as follows. AND gate 180 receives input signals indicating when operation of quick recovery feedback circuit 300 is necessary. One input signal comes from the read head switching system (not shown), indicating that a read head change is occurring or has occurred. This signal remains high for a period of time after the read head change. The other input signal comes from offset detect block 160 and delay block 170. This signal indicates that there is a voltage difference between lines 101 and 102. Thus, after a read head change occurs, the output of AND gate 180 will remain high until some time after the output voltages on lines 101 and 102 have been equalized.
The output of AND gate 180 is inverted and supplied to the base of transistor switch 306. During normal operation (without read head changes), the output of AND gate 180 is low and switch 306 acts to shunt the current from current source 360 to ground, thus rendering quick recovery feedback circuit 300 inactive. During or after a transition between read heads, the output of AND gate 180 is high, and switch 306 is open. This forces the current from current source 360 through programming transistor 351 of current mirror device 350, triggering the operation of quick recovery feedback circuit 300.
The components of current mirror device 350 are configured so that a constant current (I) through transistor 351 will result in an equal current (I) being conducted by transistors 352 and 354, and twice the current (2I) being conducted by transistor 353. Transistor 353 is therefore a constant current sink which functions in a matter analogous to constant current sink 210 in main feedback circuit 200. The voltages from nodes 103 and 104 are supplied to the bases of transistors 301 and 302, respectively.
As in the main feedback circuit, a difference between the voltages on nodes 103 and 104 will result in currents slightly more or less than I being passed by each of the transistors 301 and 302. The currents through transistors 301 and 302 will be duplicated by the currents through current mirrors 310 and 320 respectively. Since one of the transistors 301 or 302 will be drawing slightly more than I, that same current will be driven by one of the current mirrors 310 or 320.
Transistors 352 and 354 act as constant current sinks, each sinking a current approximately equal to I. Thus, one of the current mirrors 310 or 320 will be driving more current than is sunk by its respective current sink 352 or 354. Therefore, an excess current will be driven through one of the current boost devices 330 or 340.
It should be noted that one of the current mirrors 310 or 320 will drive less current than its respective current sink transistor 352 or 354 seeks to sink. The current sink transistor 352 or 354 will therefore be saturated, and its base current will be increased. As a result, there is no current flow to or from the respective current boost device 330 or 340.
Each current boost device 330 or 340 is configured in a manner similar to a current mirror, but with a lower resistance on the collector of the mirroring transistor 331 or 341 than is on the collector of the programming transistor 332 or 342. As a result, a much larger current is drawn through mirroring transistor 331 or 341 than is forced through the programming transistor 332 or 342. The output currents from current boost devices 330 and 340 are driven through nodes 105 and 106 respectively.
It can be seen by comparison of FIGS. 3 and 4 that quick recovery feedback circuit 300, when activated, operates in a manner similar to main feedback circuit 200, the difference being that the output current driven to node 105 or 106 by quick recovery feedback circuit 300 is much greater than the current generated by main feedback circuit 200. As a result, when quick recovery feedback circuit 300 is activated and there is a potential difference between nodes 103 and 104, circuit 300 will charge or discharge capacitor 190 much more quickly than main feedback circuit 200 would alone. Thus, the feedback voltage will be corrected and nodes 103 and 104 will be driven toward an equal potential much more quickly when quick recovery feedback circuit 300 is in operation.
Referring to FIG. 4, a timing diagram illustrating the operation of quick recovery feedback circuit 300 is shown. The graph shows the total feedback current as a function of time. Before time (a), main feedback circuit 200 conducts a small current to correct the bias of a first read head. At time (a), a read head change occurs, and quick recovery feedback circuit 300 is activated. Circuit 300 conducts a maximum current until a time (b), when the potential difference between nodes 103 and 104 approaches zero. After time (b), the voltages of nodes 103 and 104 rapidly equalize, reaching an equal potential at time (c).
At this point, two things occur: First, current sink transistors 352 and 354 sink the entire current being produced by current mirrors 310 and 320, respectively, so that current boost devices 330 and 340 receive no input current. Thus, no current is output from current boost devices 330 and 340 to nodes 105 and 106, and quick recovery feedback circuit 300 has no further effect on circuit 100 so long as nodes 103 and 104 remain at the same potential. Second, offset detect block 160 triggers delay block 170. At time (d), delay block 170 ceases to send a high signal to AND gate 180, thereby deactivating quick recovery feedback circuit 300.
While the invention has been particularly shown and described by the foregoing detailed description, it will be understood by those skilled in the art that various other changes in form and detail may be made without departing from the spirit and scope of the invention.