Apparatus for Generating Positive and Negative Supply Rails From Operating Motor Control Circuit

This application is related to the following applications, each of which is being filed on the same day, and each of which is incorporated herein by reference: U.S. application Ser. No. 08/062,504; U.S. application Ser. No. 08/062,968; U.S. application Ser. No. 08/062,503.

This invention relates to motor control circuits and, in particular, to a technique for generating positive and negative voltage supply rails from a motor control circuit.

In motor control circuits it is often desirable to include amplifiers and comparators which operate over signal ranges from below ground up to the motor supply voltage (V.sub.CC), and in some instances even above V.sub.CC. These so-called "rail-to-rail" circuits are difficult to design and often require a number of design compromises.

One of the most difficult circuit design challenges, for example, is to build an amplifier that operates all the way from rail-to-rail. Amplifiers generally include current sources, which themselves subtract from the voltage available to the amplifier. If a rail below ground is available, however, a wide input range can be achieved without resorting to rail-to-rail construction.

A known method of solving this problem is to provide an additional voltage supply which generates supply rails below ground or above V.sub.CC. These additional voltage supplies generally add cost and complexity to the circuit because they require the integration of a charge pump or, in some cases, a switching mode power supply and an external inductor.

The present invention overcomes these problems.

This invention enables supply rails both above V.sub.CC and below ground to be provided whenever a motor is present. It relies on the voltage spikes which occur at the inputs to the motor whenever the motor control circuit switches between its high and low states.

In one embodiment, the motor control circuit includes two push-pull "halfbridges", which supply a two-phase motor. Each halfbridge includes two N-channel MOSFETs connected in totem-pole fashion. Both MOSFETs intentionally lack a source-body short. The common node between the MOSFETs, which is connected to an input terminal of the motor, is connected through a diode to a reservoir capacitor which is charged to a positive supply voltage V.sub.AUX+. The body of each MOSFET (which, as noted, is not shorted to the source nor to ground) is connected to a second reservoir capacitor which stores a negative voltage source V.sub.AUX+. A Zener diode or other voltage limiting means is connected across each of the reservoir capacitors to limit the voltage across it. A second means ties the bodies of the MOSFETs to ground during start-up so that they do not float and thereby risk turning on the intrinsic bipolar transistors in the MOSFETs.

The arrangement of this invention can be used for motors having any number of phases and can be constructed using P-channel MOSFET technology as well as N-channel MOSFET technology.

FIG. 1 illustrates an embodiment according to the invention using N-channel MOSFET technology.

FIGS. 2A-2G illustrate the voltages at various points in the embodiment of FIG. 1.

FIG. 3A illustrates an alternative circuit for the shorting unit using a CMOS inverter.

FIG. 3B illustrates another alternative circuit for the shorting unit using a depletion mode MOSFET.

FIG. 4 illustrates an embodiment of the invention using P-channel MOSFETs on the high side of the halfbridges.

FIG. 1 illustrates an embodiment according to the invention, including a two-phase motor 10 having input terminals 10A and 10B. A motor control circuit 11 includes a first halfbridge 12 and a second halfbridge 13 which provide input signals to motor 10. Halfbridge 12 contains N-channel MOSFETs M1 and M2 which are connected in a totem pole arrangement between a supply voltage V.sub.CC and ground. The common node between MOSFETs M1 and M2 delivers an output signal V.sub.OUTA to terminal 10A of motor 10. The gates of MOSFETs M1 and M2 are driven by signals on lines L1 and L2, respectively.

With reference to MOSFET M1, it will be noted that there is no source/body short. Instead the body is connected to a negative rail R2. Diodes D1A and D1B represent the drain/body and source/body junctions of MOSFET M1, respectively.

MOSFET M1 is connected between input terminal 10A and the supply voltage V.sub.CC. In a similar manner, MOSFET M2 is connected between ground and input terminal 10A. The gate of MOSFET M2 is driven by a signal on a line L2. There is no source/body short in MOSFET M2, and its body is connected to rail R2. Diodes D2A and D2B represent the drain/body and source/body junctions of MOSFET M2, respectively.

Halfbridge 13 is connected in an identical manner to terminal 10B of motor 10. MOSFETs M3 and M4 are connected between Vcc and ground, with the common node between them coupled to terminal 10B. The body of each of MOSFETs M3 and M4 is connected to rail R2. Diodes D3A and D3B in MOSFET M3 and diodes D4A and D4B in MOSFET M4 represent the drain/body and source/body junctions of MOSFETs M3 and M4. The gates of MOSFETs M3 and M4 are driven by signals on lines L3 and L4, respectively. Halfbridge 13 delivers an output signal V.sub.OUTB to terminal 10B of motor 10.

Terminals 10A and 10B are connected via diodes D5A and D5B, respectively, to a positive rail R1. A capacitor C1 connects rail R1 to ground, in parallel with a Zener diode D6.

Rail R2 is connected to ground via a capacitor C2, which is in parallel with a Zener diode D7. Rail R2 is also connected to a grounding unit 14, which includes a depletion mode N-channel MOSFET M5, which has the customary source/body short. The source and gate terminals of MOSFET M5 are both connected to rail R2. A diode D8 represents the intrinsic "antiparallel" diode in MOSFET M5.

Lines L1-L4 are driven in a normal manner to provide the V.sub.OUTA and V.sub.OUTB signals to inputs 10A and 10B of motor 10. Signals on lines L1 and L4 turn MOSFETs M1 and M4 on simultaneously, while the signals on lines L2 and L3 turn MOSFETs M2 and M3 off. Terminal 10A is thus connected to V.sub.CC while terminal 10B is grounded. In the next half cycle, signals on lines L2 and L3 turn MOSFETs M2 and M3 on, while the signals on lines L1 and L4 turn MOSFETs M1 and M4 off. In this half cycle, terminal 10A is grounded while terminal 10B is connected to V.sub.CC.

The waveforms of V.sub.OUTA and V.sub.OUTB are illustrated in FIGS. 2A and 2B, respectively, which show the positive and negative voltage spikes that are generated by the inductances in motor 10 whenever halfbridges 12 and 13 switch between their high and low output states. Each time one of terminals 10A and 10B goes high, the corresponding diode D5A or D5B is forward-biased, and a current flows through it to charge capacitor C1. The voltage spike which occurs whenever halfbridge 12 or 13 goes high is clamped by the forward biasing of diodes D5A and D5B to a voltage equal to V.sub.AUX+ plus the voltage drop across the conducting diodes D5A and D5B, represented as a in FIGS. 2A and 2B. FIGS. 2C and 2D illustrate the current in diodes D5A and D5B, respectively, which charges capacitor C1 repetitively and incrementally. This process continues until capacitor C1 is charged to the breakdown voltage of Zener diode D6. This is represented in FIGS. 1 and 2 as V.sub.AUX+, which is the steady-state voltage of the positive rail R1.

When terminal 10A goes low, diodes D1B and D2A are forward biased and current flows through them to produce a negative voltage on capacitor C2. When terminal 10B goes low, diodes D3B and D4A are forward biased, and capacitor C2 is driven further downward. The current in diodes D1B/D2A and diodes D3B/D4A which charge capacitor C2 repetitively and incrementally are illustrated in FIGS. 2E and 2F, respectively. This process continues until capacitor C2 is driven to the breakdown voltage of Zener diode D7. Thus V.sub.AUX-, the steady-state voltage on the negative rail R2, is equal to the negative of the breakdown voltage of Zener diode D7.

The voltages on the positive rail R1 and negative rail R2 as the system reaches steady state are illustrated in FIG. 2G. As shown, the voltage on rail R1 increases at the beginning of each half cycle until it reaches V.sub.AUX+, which is equal to the breakdown voltage of Zener diode D6 (V.sub.BD6). Conversely, the voltage on rail R2 decreases at the beginning of each half cycle until it reaches V.sub.AUX-, which is the negative of the breakdown voltage of Zener diode D7 (V.sub.BD7).

During startup, prior to the time when V.sub.AUX- is driven below ground, the voltage on rail R2 must be held near 0 volts. Otherwise, if the bodies of MOSFETs M2 and M4 are allowed to float, it may be difficult to get these MOSFETs to turn on and it is even possible that the intrinsic bipolar transistors in these MOSFETs would be turned on. The latter event would destroy the MOSFET.

The function of shorting unit 14 is to keep the voltage at rail R2 near ground during startup. Depletion mode MOSFET M5 acts as a resistor for small values of V.sub.AUX-, and thereby performs a shorting function. As the magnitude of V.sub.AUX- falls, MOSFET M5 saturates and acts as a current limiter to minimize the current drain on the negative rail R2. Motor 10 must, however, continue to supply this current during the operation of the device.

Alternative circuitry for shorting unit 14 is illustrated in FIGS. 3A and 3B. In FIG. 3A, a P-channel MOSFET M6 is connected between rail R2 and ground. The gate of MOSFET M6 is connected to a CMOS inverter 30, which includes a P-channel MOSFET M7 and an N-channel MOSFET M8. At startup, MOSFET M6 is turned on and grounds rail R2. As the voltage on rail R2 begins to fall, a shunt disable signal is applied to the gates of MOSFETs M7 and M8, turning MOSFET M6 off and eliminating the undesirable current draw on negative rail R2. The shunt disable signal can be provided from any circuit signifying startup is complete such as a microprocessor, a timer or other comparator which switches startup after a predetermined voltage level is reached by V.sub.CC, V.sub.AUX+ or V.sub.AUX-. The disadvantage of using enhancement mode MOSFET M6 with a gate to drain short (via MOSFET M8) is its high resistance and low current until its gate voltage exceeds 1 volt (i.e., its threshold voltage V.sub.t) supplied from rail R2 via MOSFET M8.

In FIG. 3B, a depletion mode P-channel MOSFET M9 is connected between rail R2 and ground. Again, the gate of MOSFET M9 is controlled by inverter 30. The depletion mode MOSFET M9 is fully on during startup because its gate is tied to its drain via conducting MOSFET M8. As is known, depletion mode MOSFETs are turned on whenever the gate-to-source voltage V.sub.GS =0. After startup, gate 30 is instructed to switch states and supplies a gate voltage (V.sub.CC) more positive than the source voltage (OV) of MOSFET M9, resulting in V.sub.GS <V.sub.t. In this situation, the depletion node MOSFET M9 becomes pinched off.

In summary, during startup V.sub.GS =the drain-to-source voltage V.sub.DS =V.sub.AUX- and MOSFET M9 is on (even when V.sub.AUX- is at zero volts), and after startup, V.sub.GS =+V.sub.CC >V.sub.t and MOSFET M9 is pinched off.

FIG. 4 illustrates an alternative embodiment in which P-channel MOSFETs are used as the high side MOSFETs in the halfbridges. Motor control circuit 21 is similar to motor control circuit 11 (FIG. 1) except that halfbridges 22 and 23 contain P-channel MOSFETs M11 and M13 in place of N-channel MOSFETs M1 and M3, respectively. The bodies of MOSFETs M11 and M13 are not shorted to the sources or to ground. Instead, they are coupled together and connected directly to the positive rail R1. Diodes D5A and D5B (FIG. 1) are omitted. Otherwise, the circuit elements are identical to those shown in FIG. 1. When V.sub.OUTA goes high, the intrinsic drain-body diode D11B in MOSFET M11 is forward-biased, and a current flows toward positive rail R1, charging capacitor C1. Similarly, when V.sub.OUTB makes a positive transition, diode D13B in MOSFET M13 is forward-biased, and capacitor C1 is further charged. The charge across capacitor C1 continues to increase in this manner until the breakdown voltage of Zener diode D6 is reached. Otherwise, the circuit shown in FIG. 4 operates similarly to the circuit shown in FIG. 1.

While the above embodiments describe a two-phase motor with two halfbridges, it will be apparent that the principles of this invention are applicable to three-phase motors or motors having any number of phases. In each instance, the output terminals of the motor are connected to the high side rail through a diode, and the bodies of the MOSFETs in each phase-bridge are connected to the low side rail.

The foregoing examples are intended to be illustrative and not limiting. Many alternative embodiments in accordance with this invention will be apparent to those skilled in the art. The broad principles of this invention are defined in the following claims.