Enhanced voltage drive circuit for HDD write driver

Cross reference is made to commonly assigned U.S. patent application Ser. No. 09/772,776 entitled “Ultra Fast Voltage Drive,” filed herewith and the teachings incorporated herein by reference.

The present invention is generally related to mass media drive controllers, including HDD drives, and more particularly to a write drive control circuit controlling the head of the drive.

The coil current waveform characteristics that are important in a HDD voltage write driver circuit is depicted at 10 in FIG. 1 . In particular, the current rise time, overshoot amount and duration, undershoot, and settling time are of particular interest. Desired characteristics for the coil current waveform is a fast rise time and settling time, a controllable amounts of overshoot, and very little undershoot. By achieving these characteristics, a hard disk drive storage capacity can be improved.

A conventional voltage source based write coil driver circuit is shown at 20 in FIG. 2. A coil is modeled as LS, and is connected to four transistors via coil node HX and node HY. The connection of the four transistors to the coil is known conventionally as an H-bridge. The H-bridge controls the direction of the coil current I w , by raising the voltage on one node (HY or HX), this voltage level is referred to as the high side voltage, and lowering the voltage on the other node, this voltage level is referred to as the low side voltage. The current direction can be changed by raising the low side voltage to the level required to become the high side voltage and lowering the high side voltage to level required for it to become the low side voltage. Changing the direction of the coil current I w changes the magnetic flux through the coil.

One limitation of this conventional voltage source based write coil driver 20 is the maximum differential transient voltage generated at nodes HX and HY, during a current reversal. This limitation can be appreciated in conventional circuit 20 shown in FIG. 2, whereby the voltage of the low side voltage node can be momentarily pulled down, such as node HX, to near the second voltage rail V ee using transistor Q 1. However, during this same time there is no circuit to pull up the high side voltage node HY. Thus, the voltage at each node can go no higher than that expressed by the following equation:

V node =V cc −( V be +RS×I w ).

Assuming the coil current I w has a nominal current of about 40 milliamps. It is controlled by signals QL, {overscore (QL)}, QH, QH, {overscore (QH)}, QB and {overscore (QB)}. FIG. 8 shows the voltage levels for all four signals for a current transition with current initially flowing through the coil from node HX to node HY and then reversing and flowing from node HY into node HX. These signals are not complements in the strict sense but they are in the sense that the voltage levels for a current transition in the opposite direction can be accurately represented by the same wave forms but swapping the names, QL for {overscore (QL)}, QH for {overscore (QH)}, and QB for {overscore (QB)}. As is shown in FIG. 8 during current reversal of the coil, transistors Q 1 or Q 2 are temporarily turned on using signals QB and {overscore (QB)}to “boost” normal current and increase the differential voltage, decreasing the rise time and fall time commonly referred to as (TRTF). It can be appreciated that when voltage level of HX is transitioned from the high side voltage level to the low side voltage level, commonly referred to as pulled down, transistor Q 1 will momentarily turn on and bypass half of the matched impedance resistors RS. During this “boost,” however, HY is transitioning from the low side voltage level to the high side voltage level or pulled up. The series resistor RS connected to node HY is still in the current path. Thus, the above equation represents the maximum pull up voltage on node HY when the node HX is pulled down close to V ee . Conversely if HX was being pulled up it could only achieve the maximum voltage level of the above equation where Q 2 would momentarily turn on and bypass half of the matched impedance resistors RS allowing HY to be pulled down close to V ee . Conventionally, the maximum pull up voltage at either node is approximately 2.7 volts below the positive voltage rail V cc .

There is desired an improved write coil driver circuit that improves the maximum differential transient voltage across nodes HX and HY during a current reversal of the coil, decreasing TRTF, and reducing the width of the overshoot to achieve faster write data rates. This can be achieved by providing a high side voltage boost that bypasses the matched impedance resistors RS during the pull up. This boost is realized by including a “high side” drive using large PMOS FETs one connected between HX and VCC and the other connected between HY and VCC, 32 in FIG. 3 .

The present invention achieves technical advantages as an enhanced voltage drive circuit for a HDD write driver by providing circuitry that pulls up the voltage at one end of the coil to approximately V cc effectively bypassing the associated series resistor RS. During a current reversal, the high side boost devices preferably PMOS FETs, significantly improve the maximum differential transient voltage at nodes HX and HY. This increased voltage decreases the TRTF and has the potential to greatly increase the overshoot with out increasing the overshoot pulse width or the settling time thereby obtaining a faster data rate.

FIG. 1 illustrates the important current characteristics of a coil current waveform for a write coil driver;

FIG. 2 is a schematic diagram of a conventional voltage source based write control driver circuit;

FIG. 3 is a schematic diagram of the present invention comprising a voltage based write coil driver including two FETs, one connected between the positive rail V cc and each of the nodes HX and HY;

FIG. 4 is a waveform diagram depicting the conventional pull up and pull down voltage waveforms at nodes HX or HY during a current reversal, illustrating a moderate peak voltage and width of the transient;

FIG. 5 is a waveform diagram depicting improved pull up and pull down voltage waveforms at node HX and HY for the circuit of the present invention shown in FIG. 3, illustrating an improved peak voltage and narrowed transient;

FIG. 6 illustrates a current waveform through the coil for the conventional circuit of FIG. 2;

FIG. 7 is a waveform diagram depicting the current waveforms through the coil for the present invention shown in FIG. 3, illustrating increased peak current and a reduced transient width;

FIG. 8 illustrates signal waveforms for the circuit of FIG. 2; and

FIG. 9 illustrates signal waveforms for the circuit of FIG. 3 .

Referring now to FIG. 3, there is depicted generally 30 a HDD write driver circuit. Circuit 30 is seen to include the basic elements of the conventional H-bridge circuit 20 illustrated in FIG. 2, and advantageously also includes a pair of large PMOS FETs that generally shown at 32 each being selectively switched in conjunction with a respective transistor Q 1 and Q 2 . The control signals are shown on FIG. 9 . When on, FETs 32 advantageously pull-up the voltage at the respective side of the coil LS up to substantially V cc , and thus effectively bypass the associated and proximate resistor RS. PMOS FETs are preferably used instead of NMOS FETs since PMOS FETs have a smaller drain-to-source “on-resistance,” and hence, include a smaller drain-to-source voltage drop. However, the use of NMOS FETs is within the scope of the present invention.

During a current reversal cycle, two boost devices will now be on temporarily, either transistor Q 1 and M 2 , or, transistor Q 2 and M 1 , depending on which way the inductor current is to be steered. The maximum differential voltage that can be achieved is increased by more then 2 volts over the conventional write driver due to the addition of these PMOS FETs 32 . The increased maximum differential voltage during the current reversal cycle will decrease the TRTF of the coil current and allows for a much higher overshoot to be generated. The minimum width of the overshoot is directly related to the TRTF and inversely related to the height of the overshoot. Therefor holding the overshoot constant would result in a decrease in the minimum width of the overshoot. This ability to have more overshoot for a given overshoot width then the conventional design is essential to achieving faster data rates.

FIG. 4 and FIG. 5 depict voltage waveforms at the nodes HX and HY shown as waveforms 40 and 50 , illustrating several advantages of implementing the additional PMOS FET devices 32 , including an increased pull up voltage over that shown in FIG. 2, thus producing an increased differential voltage V D across the coil, and a narrowed width W v for the overall transient as depicted.

In a first instance when transistor Q 1 is turned on by corresponding signal QB to responsively pull-down the voltage at node HX to approximately V ee , the associated PMOS FET M 2 is also turned on to pull-up the voltage at node HY to substantially V cc . During this cycle, transistors Q 2 and M 1 are off.

At the next current reversal, conversely, transistors M 1 and Q 2 are in the on-state such that node HX is pulled substantially to V cc , with a small drop of only about 0.2 volts being that across the drain-to-source terminals of PMOS FET M 1 . At the same time, node HY is pulled low to substantially V ee by the transistor Q 2 . The voltage across PMOS FET is inversely related to its size. A larger PMOS device will have a smaller voltage drop. The maximum size of the device is limited by the circuit used to switch it. Without improvements to the driving circuit such as the “Ultra Fast Voltage Drive” a large PMOS FET could not be switched fast enough.

A previously mentioned regarding the conventional circuit shown in FIG. 2 at 20 , conventionally, the highest voltage that nodes HX or HY could obtain was defined by the equation V cc −(V be +RS×I w ), whereby I w is nominally set around 40 milliamps which equates to about V cc −2.7 volts. The present invention achieves technical advantages in that the voltage drop across the associated resistor RS is eliminated and the voltage at nodes HX and HY when pulled up can be very near V cc depending on the size of the PMOS FET used. The improvement increases the differential voltage V D across the coil. A voltage drop across the properly sized PMOS FET is typically no more than about 0.2 volts. To achieve this the FETs M 1 and M 2 are large PMOS FETs that are driven hard by ultra fast rail-to-rail voltage drive circuit described in the cross-referenced commonly assigned U.S. Pat. No. 09/772,776 “Ultra Fast Voltage Drive Circuit,” the teachings of which are incorporated herein by reference. This circuit shown in the cross-referenced application quickly drives the large PMOS FET devices M 1 and M 2 with a maximum voltage swing on the Gate to minimize the voltage drop across the FET.

Referring now to FIG. 6 and FIG. 7, there is depicted the coil current waveform I w for the conventional circuit showing FIG. 6, depicted as waveform 60 and at 70 according to the present invention comprising circuit 30 as shown in FIG. 3 . The differential voltage waveforms 40 at node HX and HY of conventional circuit 20 are shown in FIG. 4, compared with the differential voltage waveforms shown at 50 in FIG. 5, depicting the advantage of the present invention. As shown in FIG. 7, the peak current of the coil current waveform is substantially improved, due to the improved voltage range of HX and HY FIG. 5 . The width of the transient is substantially narrowed as shown at Wc.

In summary, the present invention shown as circuit 30 meets the objectives of a desired HDD write driver circuit in that it has a fast current reversal, a maximum pull up voltage and an increased differential voltage across the coil. It also shows increased overshoot and controlled undershoot, with a narrower overshoot width. The properly sized additional PMOS FETs M 1 and M 2 occupy nominal silicon space on a die, and the gate can be driven with the “ultra-fast rail-to-rail voltage drive circuit.” to achieve quick switching time.

Though the invention has been described with respect to a specific preferred embodiment, many variations and modifications will become apparent to those skilled in the art upon reading the present application. It is therefore the intention that the appended claims be interpreted as broadly as possible in view of the prior art to include all such variations and modifications.