Self Synchronous Motor Sensor Switching Arrangement

The present invention relates to traction motors and the application thereof in electric powered vehicles. More specifically, the application relates to control systems for starting permanent magnet electric motors from an idle state.

The invention described in the present application is related in certain respects to U.S. Pat. Nos. 4,316,132 and 4,296,650, as well as U.S. Ser. No. 212,656 filed Dec. 3, 1980, U.S. Ser. No. 385,633 filed June 7, 1982, now abandoned and U.S. Ser. No. 538,859 filed Oct. 4, 1983 now abandoned.

The electric powered passenger vehicle has long been considered one of the most attractive alternatives to conventional internal combustion engine vehicles from the standpoint of overall efficiency, environmental impact and, most recently, alternative fuel capability. Many commercial enterprises and private individuals, some under the auspices of the Federal Government, have proposed various approaches to implementing an electrically powered vehicle. To date, there have been virtually no commercially successful vehicles produced on a large scale. A large number of approaches to the implementation and control of an electric vehicle are evidenced in the patent literature. Most of the approaches fall within one of three general categories of motive power source. These categories are hybrids, D.C. motor drives and induction motor drives. The first type, that is most frequently found in the patent literature, is the hybrid vehicle, comprising a small gasoline fueled internal combustion engine which mechanically drives an electrical generator which, in turn, supplies electrical energy to an AC or DC motor. With this arrangement, the gasoline engine can operate at a constant speed (at a relatively high efficiency) and achieve a substantial fuel saving compared with an engine experiencing the conventional wide range of operation. A shortcoming of many hybrids is that they are relatively heavy, requiring an electrical generator and motor as well as the gasoline engine. Additionally, the engine requires substantial amounts of volatile liquid fuel and generates exhaust emissions.

A second approach taken in the development of electric vehicles is the use of a bank of batteries which supply electrical energy to a DC motor. A variable speed motor drive circuit provides easy and versatile control of a vehicle. The principle advantage of this arrangement is that a DC motor control system requires a relatively simple power and control circuit. Unfortunately, this advantage is often more than offset by the relatively large initial cost and maintenance expenses of the motor itself. In addition, DC machinery is relatively heavy and bulky, factors which do not lend themselves well to implementation within a lightweight compact vehicle. Finally, DC motors inherently require choppers and commutators which create sparks and RF pollution which can be controlled only at additional expense.

The third, and most attractive approach from the applicant's viewpoint, is a vehicle employing a battery bank and an AC motor. AC motors are relatively lightweight, inexpensive and efficient when compared to DC motors. AC motors, with no brushes or commutators, are more rugged and reliable than their D.C. counterparts and require substantially less maintenance. Related to the power-to-weight ratio is the fact that A.C. machines can be driven at substantially greater speeds than D.C. motors. Because A.C. motors do not generate sparks, they can readily be employed in dusty, explosive and highly humid atmospheres or high altitudes. Additionally, A.C. motors can be liquid cooled if the application so requires. Although typically superior to D.C. motors in electric vehicle applications, A.C. motors often require complex control circuits which are dedicated to associated vehicle drivetrains and can be extremely bulky and expensive. To date, virtually all A.C. electric vehicles have employed multi (usually three) phase design strategies. Although three-phase machinery has many advantages as set forth hereinabove, three-phase inverter costs and complexity have proven to be extremely high. In relatively large load applications, such as that required in a passenger vehicle, appropriately sized solid state switching devices such as SCR's or transistors are often extremely expensive. In addition, three-phase inverters, by their nature, dictate a multiplicity of components, including switching devices, again increasing system cost.

The shortcomings of the prior art may be overcome by providing an electric vehicle drivetrain such as the one disclosed in co-pending application Ser. No. 385,633. This drivetrain includes a substantially fixed D.C. power source such as a battery which energizes a single-phase brushless A.C. motor which, in turn, imparts torque to at least one tractive wheel of the vehicle. The motor is characterized by a stator adapted from mechnical grounding to a relatively stationary portion of the vehicle such as its body, and a permanent magnet external rotor disposed for rotation about the stator for magnetic interaction therewith and adapted to engage the tractive wheel. An inverter provides a power input from the power source and a power output to the motor in response to switch command signals generated by a control circuit in response to an operator demand signal. This arrangement has the advantage of providing an inexpensive and simply constructed single-phase brushless permanent magnet A.C. motor traction drive for electric road vehicles. The control circuit controls the switching of the current direction through the motor winding as the permanent magnet rotor turns. A pair of sensors, in a closely-spaced angular relationship to the reluctance torque detent, detect the position of the rotor and provide signals to the control circuit which synchronizes the switching of the current direction with the rotation of the rotor.

A difficulty is encountered in the above described A.C. motor, however, when the motor is started from an at rest position. In the rest position the rotor magnets are aligned with the stator poles. This is caused by reluctance torque which is present at all times. To start the motor, the stator magnets of the motor are repulsed. The motor may begin moving in either the forward or reverse direction. One of the sensors detects the direction of movement, and if it is the wrong direction, causes the control circuit to reverse the current direction through the winding. This reverses the direction of the motor torque. Unfortunately, because the sensors must be spaced close to the reluctance torque detent for proper operation of the motor during normal operation, the motor may not develop sufficient rotational kinetic energy to overcome the inherent reluctance torque which urges the motor toward its detent position. Instead, the motor may simply oscillate in the reluctance torque detent.

The difficulties encountered in starting permanent magnet A.C. motors are overcome by providing a second pair of angularly spaced sensors in addition to the first pair. The second pair of sensors are placed in a fixed angular distance from the motor's detent position which is further than the position of the first sensor pair. The control circuit selectively receives input from either the first or second pair of sensors depending on the present motor speed. If the motor is rotating the control circuit uses the first pair of sensors to control the switching of the current direction. When the motor is starting from its at rest position, input from the second pair of sensors is used. Since the second pair of sensors are spaced farther apart, the motor, after turning past one of the sensors, can reverse direction and will obtain enough rotational kinetic energy to overcome the reluctance torque thus carrying the motor through the reluctance detent and preventing oscillation of the motor.

These and other aspects and advantages of the present invention will become apparent upon reading the following specification, which, along with the application drawings, describes and discloses a preferred embodiment of the invention, as well as modifications thereof, in detail.

A detailed description of the embodiment of the invention makes reference to the accompanying drawings.

FIG. 1, is a perspective, fragmented view of an automobile incorporating the preferred embodiment of the present invention;

FIG. 2, is a block-schematic diagram of a traction motor, inverter and control circuit, and an implementation in the drivetrain of the automobile of FIG. 1;

FIG. 3, is a cross-sectional view of the motor and frame assembly of FIG. 1;

FIG. 4, is a cross-sectional view taken on line 4--4 of FIG. 3, and further illustrating the rotor position sensing arrangement of the present invention;

FIG. 5, is a cross-sectional view taken on line 5--5 of FIG. 3;

FIG. 6a, is a schematic diagram of a forward-reverse signal conditioning circuit forming a portion of the control circuit of FIG. 2;

FIG. 6b , is a schematic diagram of a motor positioning circuit and a typical printed circuit board filter network forming a portion of the control circuit of FIG. 2;

FIG. 6c, is a schematic diagram of a current demand logic circuit and a minimum starting speed circuit forming a portion of the control circuit of FIG. 2;

FIG. 6d, is a schematic diagram of a base drive circuit forming a portion of the control circuit of FIG. 2;

FIG. 6e, is a scheamtic diagram of the energy recovery circuit of FIG. 2;

FIG. 6f, is a schematic diagram of a power supply employed within the control circuit of FIG. 2;

FIG. 7, is a drawing figure key, illustrating the proper arrangement of FIGS. 6a through 6f;

FIG. 8a, is a graph of the electrical and reluctance torque versus motor rotation characteristic of the motor of FIG. 2;

FIG. 8b, is a graph of the pre-start positioning cam profile versus motor rotation characteristic of the motor of FIG. 2;

FIG. 8c, is a graph of an applied voltage versus motor rotation characteristic of the motor of FIG. 2, illustrating the shift effected by adjustment of a leading sensor angle,

FIG. 8d, is a graph of the motor torque/EMF versus motor rotation characteristic of the motor of FIG. 2, above base speed;

FIG. 8e, is a graph of the applied voltage on motor winding A versus motor position characteristic of the motor of FIG. 2, above base speed;

FIG. 8f, is a graph of the applied voltage on motor winding B versus motor position characteristic of the motor of FIG. 2, above base speed;

FIG. 8g, is a graph of a typical motor current versus motor rotation characteristic of the motor of FIG. 2, above base speed;

FIG. 8h, is a graph of the motor torque/EMF versus motor rotation characteristic of the motor of FIG. 2, below base speed;

FIG. 8i, is a graph of the applied voltage on motor winding A versus motor position characteristic of the motor of FIG. 2, below base speed;

FIG. 8j, is a graph of the applied voltage on motor winding B versus motor position characteristic of the motor of FIG. 2, below base speed; and

FIG. 8k, is a graph of a typical motor current versus motor rotation characteristic of the motor of FIG. 2, below base speed.

A. Electric Vehicle Drivetrain Installation

B. Inverter

C. Motor and Frame Assembly

D. Rotor Position Sensor

E. Starting

F. Control Circuit

(i) Forward-Reverse Signal Conditioning Circuit

(ii) Motor Positioning Circuit

(iii) Current Demand Logic and Minimum Starting Speed Circuit

(iv) Base Drive Circuits

(v) Snubber and Energy Recovery Circuit

(vi) Switching Power Supply Circuit

A. Electric Vehicle Drivetrain Installation

Referring to FIG. 1, packaging of the present invention within an electric vehicle 10 is conceptualized. Electric vehicle 10 is conventionally configured, having driven front tractive wheels 12 and free-running or slave rear wheels 14. An engine compartment, defined by area forward of a fire wall 16, contains a longitudinally mounted self-synchronous single-phase A.C. traction motor 18 which operates tractive wheels 12 through a multi-speed transaxle 20, output driveshafts 22 and interconnecting U-joints 24. As will be described in greater detail hereinabelow, motor 18 includes an extenal, ferrite permanent magnet rotor 26 which encloses a grounded stator (not illustrated) for rotation thereabout and magnetic interaction therewith. Rotor 26 is rotatably supported on a motor shaft 28 which is mechanically grounded at both ends thereof to a frame assembly 30. Frame assembly 30 is, in turn, mounted to a relatively stationary portion of the vehicle such as transaxle 20 or fire wall 16. Transaxle 20 includes an input shaft 32 having an axis of rotation which is maintained substantially parallel to and at a fixed distance from motor shaft 28 by frame assembly 30. Rotor 26 of motor 18 drivingly engages input shaft 32 of transaxle 20 through the provision of an appropriate pulley 34 and V-belt 36.

The details of the front suspension and vehicle steering system are deleted here for the sake of brevity. The mounting of motor 18, frame assembly 30 and transaxle 20 within electric vehicle 10 is not deemed as part of the present invention and requires no further elaboration. Input shaft 32 and output drive shafts 22 are interconnected within transaxle 20 by selectable multiple ratio gear sets engageably disposed therebetween. An example of such a transmission is disclosed in U.S. Pat. No. 4,296,650 which is hereby incorporated herein by reference. Although the transaxle disclosed in the above-mentioned patent incorporates an internal rotor transversally mounted motor, it is contemplated that it could be easily modified to accommodate the present inventive motor either in a transverse or longitudinal orientation within electric vehicle 10. Additionally, transaxle 20 includes a mechanical differential which operates to transfer torque to both front tractive wheels 12.

A ventilated compartment 38 depends forward from the fire wall 16 within the engine compartment of electric vehicle 10 and contains a D.C. power source such as a chemical battery. In the preferred embodiment of the invention, the power supply comprises a plurality of conventional lead-acid batteries 40 (see FIGS. 2 and 6e) connected to establish a 96 VDC voltage. A second compartment 42 depends forwardly from fire wall 16 of electric vehicle 10 and contains an inverter circuit, shown generally at 44, and a control circuit 46 (both shown in FIG. 2). Motor 18, inverter 44 and control circuit 46 are electrically interconnected by appropriately sized cables (not illustrated). Additionally, provision must be made to periodically connect electric vehicle 10 with a source of charging current such as the operator's residence.

The present invention can be applied to applications other than passenger vehicles. FIG. 1 is intended only as an aid to the reader in conceptualizing the packaging as it would appear in a typical commuting passenger vehicle. The details of the illustrated arrangement are therefore not to be considered limiting in any sense.

B. Inverter

Referring to FIG. 2, a simplified block diagram of the electrical circuit preferably used in conjunction with the present invention is illustrated. Although the illustrated circuit is preferred, it is understood that any appropriate circuitry may be used. A drivetrain, shown generally at 48 in electrical schematic form is adapted for use with electric vehicle 10. In the preferred embodiment of the invention, the drivetrain also includes transaxle 20 and associated apparatus for applying torque to front tractive wheels 12.

Motor 18 is provided with a single phase bifilar wound, insulated multi-strand wire armature illustrated as windings A and B, 50 and 52, respectively, which minimize winding eddy current losses, establish a high coefficient of coupling and allow for the provision of a relatively straight forward inverter 44 construction. The positive terminal of battery 40 is electrically interconnected to the point of common connection of windings 50 and 52 through a positive bus 54. The remaining end of winding 50 is electrically connected to the collector of a large power transistor 56 and interconnected to a negative bus 58 from the negative terminal of battery 40 through an inverse parallel diode 60. Likewise, the remaining end of winding 52 is electrically connected to the collector of a second large power transistor 62 and interconnected to bus 58 through a second inverse parallel diode 64. The emitters of both transistors 56 and 62 are connected to bus 58. Control circuit 46 has two control lines 66 and 68 which output switch command signals to transistors 56 and 62, respectively. Inverter 44 also includes an energy recovery circuit 70 with an output line 72 electrically connected to the point of common connection between windings 50 and 52, and an input line 74 which is interconnected to the collectors of transistors 56 and 62 through auxiliary diodes 76 and 78, respectively. Control circuit 46 has additional inputs (not illustrated).

Transistors 56 and 62 and diodes 60 and 64 effect the basic D.C. to A.C. power conversion of inverter 44. Diodes 76 and 78 operate to recover trapped inductive energy in motor windings 50 and 52. It is contemplated that thyristors or other suitable devices could be substituted for transistors 56 and 62.

The operation of inverter 44 is as follows. Control circuit 46 generates switch command signals as a function of rotor speed, rotor rotational sense, rotor position, motor current, driver (speed) demand and selected motor rotational sense (vehicle direction). In the presence of a driver demand and direction (forward/reverse) selection, control circuit 46 will output a switch command signal on line 66 whereby transistor 56 is initially turned on, allowing current to flow from winding 50. As can best be seen by referring to FIGS. 8h through 8k, inclusive, when motor 18 is stalled or turning very slowly, there will be no back EMF, and the entire battery voltage will appear across the winding impedance, causing motor current to increase rapidly according to characteristic L/R time constant. When a current limit set point (.+-.I.sub.Limit A) is reached, transistor 56 is turned off and a current is transferred by the transformer action of motor 18 to winding 52 through diode 64 back into battery 40. When the current decays or falls to a lower set point (.+-.I.sub.Limit B), transistor 56 is again turned on and current rises. Because the transformer action of motor 18 causes the peak voltage on transistor 56 to equal or exceed twice the battery voltage (V.sub.cc), trapped energy in winding 50 must be returned to a voltage source of twice the supply voltage. Lines 72 and 74 are provided to define an energy recovery bus by which energy recovery circuit 70 reconstitutes nontransformed energy and returns it to battery 40. During the second half cycle of operation, transistor 56 is held off while transistor 62 is alternately switched on and off as the current in winding 52 fluctuates between the limit set points.

When motor 18 is operating above its base speed, a back EMF exceeding V.sub.cc is present and motor current is controlled by load angle only. Accordingly, during their respective duty cycles, transistors 56 and 62 will alternately remain on as can best be seen by referring to FIGS. 8d through 8g, inclusive.

C. Motor and Frame Assembly

Referring to FIGS. 3 through 5, the structural details of motor 18 and frame assembly 30 are illustrated. Frame assembly 30 comprises a rigid base plate 80 and two vertical mounts 82 and 84 extending upwardly therefrom to cradle motor 18 therebetween. Mounts 82 and 84 are secured to base plate 80 by cap screws 86 or other suitable fastening means. The uppermost ends of mounts 82 and 84 terminate an upwardly opening cylindrical recesses 88. Clamp members 92 are disposed above mounts 82 and 84 and define downwardly opening cylindrical recesses 90 which coact with recesses 88 to receive the ends of motor shaft 28 therebetween. Suitable fasteners such as cap screws 94 extend downwardly through clamp members 92 and threadably engage mounts 82 and 84 to draw clamp members 92 downwardly against motor shaft 28 and thereby mechanically ground same to frame assembly 30.

Motor shaft 28 is constructed of ferrous metal such as magnetic steel and supports a stator assembly 96 at a point intermediate mounts 82 and 84. As is best seen in FIG. 5, stator assembly 96 includes an armature member 98 defining a hub portion 100 and six circumferentially spaced radially outwardly extending protuberant armature poles or teeth 102. Motor shaft 28 extends through an axial bore 104 within hub portion 100 of armature member 98. Stator assembly 96 is secured from rotation about motor shaft 28 by an elongated square key 106 nestingly disposed within cooperating notches within shaft 28 and armature member 98. As viewed in FIG. 3, stator assembly 96 is assembled upon motor shaft 28 from the right and has its leftward displacement limited by a step 108 formed in shaft 28. Induction coils 110 are wound around each armature tooth 102 within spaces 112 provided between each pair of adjoining teeth 102. The six induction coils 110a through 110f are series connected to collectively comprise single phase windings 50 and 52 (see FIG. 2) and are bifilar-wound insulated multi-strand wire. Induction coils 110 are suitably insulated and mechanically retained by armature member 98 employing methods well known in the art. Accordingly, the entire stator assembly 96 is rigidly secured to motor shaft 28 in the position illustrated. Electrical connection to the motor coils 110 is provided via terminals 122, 124 and 126 (FIG. 3). The terminals are insulated from each other by an insulating material 128 (FIG. 3).

Rotor 26 of motor 18 includes a soft steel cylindrical shell 130 which carries six arcuate ferrite permanent magnet segments 132 circumferentially spaced about the radially innermost surface thereof. Magnet segments 132a through 132f are affixed to the inner surface of shell 130 by adhesives or other suitable material. Shell 130 serves as the machine flux return path as well as to protect magnets 132 and to absorb hoop stresses during operation. The radially innermost surface of magnet segments 132 are closely spaced radially outwardly from the outwardmost face defined by armature teeth 102 by a gap 133 and are substantially circumferentially coextensive therewith. Magnet segments 132 are radially polarized in a sense opposite to that of adjoining magnet segments 132. Thus, shell 130 and magnet segments 132 rotate about stator assembly 96 in close proximity thereto. Adjoining magnet segments 132 have a circumferential space 134 disposed therebetween.

Referring again to FIG. 3, the radial spacing of rotor 26 is maintained by aluminum end bells 136 and 138. End bell 136 closes the lefthand end of shell 130 and is secured thereto about its outer circumference by suitable fastening means such as cap screws 140. Likewise, the outer circumferential portion of end bell 138 is affixed to the righthand opening of shell 130 by cap screws 141.

Although V-belt 36 and conventional pulley designs 34 and 150 were employed by the applicant and disclosed in the preferred embodiment of the invention, it is contemplated that larger applications may dictate the use of chains, cogged belts, direct gearing or the like. Accordingly, the V-belt 36 arrangement is only intended to be illustrative in nature and not to be considered limiting.

A ball bearing assembly 142 has the inner race thereof press fit on the righthand end of motor shaft 28. The radially innermost surface of end bell 138 defines an axially aligned stepped bore having a first portion 144 press fit over the outer race of ball bearing 142 and second portion 146 circumscribing the outer surface of a black delrin sleeve bearing or bushing 148 press fit upon motor shaft 28 intermediate ball bearing 142 and mount 84. End ball 138 has a pulley 150 integrally formed therein which coacts with pulley 34 of transaxle 20 to carry V-belt 36. A suitable passageway 152 is provided through base plate 80 for routing V-belt 36.

End bell 136 also has an axially aligned stepped bore having a first portion 154 press fit over the outer race of a second ball bearing 156. The inner race of ball bearing 156 is press fit over motor shaft 28 near the lefthand end thereof.

Thus, rotor 26 is carried on motor shaft 28 in the orientation shown and is free to rotate thereabout, supported by ball bearings 142 and 156. Bushing 148 is provided to counter radial loading imposed by V-belt 36. In addition, bushing 148 provides axial spacing between ball bearing 142 and mount 84. Likewise, a tubular sleeve 158, carried by motor shaft 28, interspaces ball bearing 156 and mount 82.

An annular cam ring 160 is rigidly mounted to the outer circumferential portion of end bell 136 by screws 140 and defines a cam surface 162 on the outer circumference thereof. Cam ring 160 is carried by rotor 26 for rotation therewith with respect to stator assembly 96 and frame assembly 30.

D. Rotor Position Sensor

A rotor position sensing assembly is illustrated in FIGS. 3 and 4 generally at 164. Assembly 164 includes an encoded washer shaped mild steel vane 166 mounted to the lefthandmost surface of end bell 136 by screws 168 which pass through an annular aluminum vane spacer 170 and threadably engage end bell 136. Vane 166 has a central opening 172 for providing radial spacing from sleeve 158. A flangeless sleeve 158 is provided to interspace the inner race of ball bearings 156 and the rightmost surface of mount 82. Vane 166 and vane spacer 170 rotate with rotor 26. The circumferential profile of vane 166 can best be seen in FIG. 4.

A set of vane actuated sensors or switches 212, 214, 216, and 218 are rigidly carried by mount 82. The sensors include sensing elements which straddle vane 166 near the outer circumference thereof. These sensing elements may be, for example, four type AV12a Hall sensors. The sensors function to detect the presence of an object within the sensing region 184 thereof. The sensors are provided in a fixed angular relationship to each other and are radially spaced from motor shaft 28.

Vane 166 has three circumferentially spaced surface portions 166a, 166b and 166c which project outwardly far enough to pass through sensing region 184 of sensors 212-218 during rotation of rotor 26. As the leading or trailing edge portions 166a, 166b and 166c pass through sensing region 184, the electrical output of sensors 212-218 will go through transistion between low and high states. The vane portions 166a, 166b and 166c each extend through a 60.degree. arc and are spaced from adjoining portions by a relieved portion of vane 166 which does not pass through sensing region 184. Vane portions 166a, 166b and 166c are in precise rotational alignment with magnet segments 132c, 132a and 132e.

Another vane actuated sensor or switch 186 is provided with its sensing region (not illustrated) in register with vane 166. Sensor 186 is nominally offset angularly from sensors 212-218 by 90.degree.. Sensor 186 is carried by a mount (not shown) and is fixed with respect to stator assembly 96. Sensor 186 is used to sense when the rotor is in its highest torque position as will be described in detail below. Electrical leads, shown generally at 211 in FIG. 3, are routed from sensors 212-218 and sensor 186 to control circuit 46. Control circuit 46 will be described in detail hereinbelow.

E. Starting

As a single-phase machine, permanent magnet traction motor 18 shares a characteristic of all single-phase A.C. motors in that no starting torque may be available, depending upon at rest rotor position vis-a-vis stator pole position. However, the motor 18 employed in the present application is better than induction or reluctance type single-phase motors in two important respects. First, it is not possible for motor 18 to run other than in the desired direction. This is insured by the angular position of sensors 212, 214, 216 and 218 and 186 which determine current polarity as a function of position. Additionally, starting torque is available over most of the possible rotor positions. Starting torque will be unavailable in an ideal motor only when the salient armature poles or teeth 102 line up with rotor magnet segments 132. This, however, is the position to which the unexcited reluctance torque of the machine will return the rotor 26 if no restraining forces is applied. In a small practically constructed machine, small asymmetries will normally allow a small starting torque, even in these reluctance torque detent positions. This is best appreciated by referring to FIG. 8a. An actual motor 18, which is inherently imperfectly constructed, will slightly shift the zero reluctance torque point from the zero electrical torque point. This results in an asymmetry with respect to the zero torque line. The applicant has surmised that this asymmetry is probably the predominent reason that motor 18 will start most of the time on its own when it is in the nominal zero torque position. However, because of the intended application of the present invention in a passenger vehicle, it is imperative that 100% starting reliability be achieved.

From the torque-angle curves of FIG. 8a, it is self-evident that motor 18 will tend to stop in one of six "detent" positions. These positions correspond to a position where armature teeth 102 are in line with rotor magnet segments 132. The plus and minus 4.degree. and 15.degree. angular offsets of sensors 212, 214, 216 and 218 were selected so that in case there where rotor 26 came to rest at 2.degree.-3.degree. to the left of the zero electrical torque point (refer to FIG. 8a) starting can be effected by electrically driving rotor 26 backwards to a higher electrical torque point and then reversing direction to gain momentum of rotor 26 to shoot past this zero torque point. In the angular offset selected in the present invention, rotor 26 would rotate through negative 15.degree. revolution and then experience a current reversal that would cause motor 18 to slow and start rotating forward past minus 4.degree., where the motor would begin normal current switching. Typical driveline backlash will assist this type of operation.

F. Control Circuit

Referring to FIG. 7, a key to the arrangement of the pages of drawings containing FIGS. 6a through 6f, inclusive, is given. FIGS. 6a through 6f define the complete electrical circuit of drivetrain 48 including inverter circuit 44, control circuit 46, battery 40 and motor 18.

Motor 18 has a cross-section (see FIG. 5) having structure similar to that of a conventional D.C. permanent magnet motor. In addition to the external permanent magnet rotor 26, the major differences are electrical in that the motor is electronically driven and the armature current is electronically commutated within the stator assembly 96 while the field rotates. Furthermore, the electronic commutation is effected by events controlled by sensors.

As outlined in the discussion of operation of inverter 44 hereinabove, control circuit 46 functions to sense certain operating parameters and to generate switch command signals on lines 66 and 68 as a function thereof to reciprocally energize windings 50 and 52, and thus motor 18, by alternately switching transistors 56 and 62 between conductive and nonconductive states. The switch command signals are characterized as being width modulated when motor 18 is below its base speed, as best viewed in FIGS. 8i and 8j. The measured input parameters to control circuit 46 are rotor speed, rotor rotational sense, rotor position, driver speed demand and driver directional demand. Rotor speed, rotor rotational sense and rotor position are deduced by control circuit 46 from the cumulative inputs of sensors 186, 212, 214, 216 and 218. Motor current is measured directly, as are the driver inputs. Because motor 18 is single-phase bifilar wound, system operation is self-synchronous and relatively straight forward as will be described in detail hereinbelow.

(i) Forward-Reverse Signal Conditioning Circuit

Referring to FIG. 6a, a forward-reverse signal conditioning circuit, shown generally at 332, is illustrated. One input of each of sensors 212, 214, 216 and 218 are connected to a +15 VDC power supply which will be described in detail herein below, through a local printed circuit board filter 334 (see FIG. 6b). The remaining output of sensors 214, 216 and 218 are electrically interconnected to terminals XIII, XII and X, respectively, of a type 4539 1 of 4 CMOS selector gate 336 through separate current limiting resistors 338, 340 and 342, respectively. The terminals designated by Roman numerals herein represent those specified by the manufacturer of a particular solid state device employed by the applicant. It is understood that any number of suitable alternative devices could be substituted and thus the specific devices called out herein and their characteristic terminal configurations are not to be considered limiting in any sense.

The output of sensor 212 is electrically interconnected to one input of an exclusive OR (XOR) gate 344 through a current limiting resistor 346. The remaining input of gate 344 is connected to the +15 VDC power supply. Terminals XIII, XII and X of selector 336 are interconnected to the +15 VDC power supply through 4.7k pull-up resistors 348, 350 and 352, respectively, and to ground through 0.01 microfarad capacitors 354, 356 and 358, respectively. The input of gate 344 connected to resistor 346 is also interconnected to the +15 VDC power supply through a 4.7k resistor 360 and to ground through 0.01 microfarad capacitor 362. The output of gate 344 is connected to input XI of selector 336.

The input of gate 344 connected to resistor 346 also electrically connected to one input of an OR gate 364. Terminal X of selector 336 is electrically connected to the remaining input of OR gate 364. The output of OR gate 364 is electrically connected to one input of an XOR gate 366. The remaining input of XOR gate 366 is interconnected to ground through a 0.01 microfarad capacitor 368 and interconnected to the output of OR gate 364 through a 5.1k resistor 370. Resistor 370 and capacitor 368 operate to establish a phase lag.

The output of XOR gate 366 is connected to S terminal VI of a type 4043 flip-flop 372 and interconnected to the base of an NPN transistor 374 through a 10k resistor 376. The emitter of transistor 374 is electrically connected to ground and the base thereof is interconnected to ground through a 1k resistor 378. The collector of transistor 374 is interconnected to the +15 VDC power supply through a 100k resistor 380, to ground through a 50 microfarad capacitor 382 and directly connected to R terminal VII of flip-flop 372. Flip-flop 372 as well as its associated componentry function as a zero rotor speed detector, shown generally at 384, which outputs the zero speed select signal from Q terminal IX of flip-flop 372 to terminal II of selector 336.

Forward and reverse directional mode select switches, 386 and 388, respectively, are provided within the passenger compartment area of electric vehicle 10. Both switches 386 and 388 are normally open, momentary contact single pole types, having one terminal thereof connected to a +24 VDC tap 389 (see FIG. 6e) in battery 40. The remaining terminal of switch 386 is interconnected to one input of an AND gate 390 through a series combination of a 6.2k resistor 392 and a 20k resistor 394. The point of common connection between the resistors 392 and 394 is interconnected to ground through a 10k resistor 396. The terminal of AND gate 390 connected to resistor 394 is interconnected to ground through a 20 microfarad capacitor 398. The output of AND gate 390 is interconnected to S terminal IV of a type 4043 flip-flop 400 through a forward biased signal diode 402. Terminal IV of flip-flop 400 is interconnected to the +15 VDC power supply through a 2.2 microfarad capacitor 404 and to ground through a 100k resistor 406.

The remaining terminal of reverse switch 388 is interconnected to one input terminal of an AND gate 408 through a series combination of a 6.2k resistor 410 and 20k resistor 412. The point of common connection between resistors 410 and 412 is connected to ground through a 10k resistor 414. The input terminal of AND gate 408 connected to resistor 412 is also interconnected to ground through a 20 microfarad capacitor 416. The output of AND gate 408 is connected to R input terminal III of flip-flop 400. An XOR gate 418 has one input connected to a terminal A. Terminal A is connected to a complimentary terminal A in FIG. 6c. The remaining input terminal to XOR gate 418 is connected to the +15 VDC power supply. The output of XOR gate 418 is connected directly to the remaining input of AND gate 390 and the remaining input of AND gate 408. Terminals V and XVI of flip-flop 400 are connected to the +15 VDC power supply and terminals VIII, XII, XI, XIV and XV are connected to ground. Flip-flop 400 and its associated componentry operates as a latch, indicated generally at 420, which outputs a forward/reverse select signal from Q terminal II of flip-flop 400 to terminal XIV of selector 336.

Terminal II of flip-flop 400 is also connected to one input of an XOR gate 424 of a display driver circuit, shown generally at 426. The remaining input of XOR gate 424 is connected to the +15 VDC power supply. The output of XOR gate 426 is connected to terminal I of a type ULN2004A lamp and relay driver 428. Terminal II of flip-flop 400 is directly connected to terminal II of driver 428. Terminals X and XI of driver 428 are interconnected to the +24 VDC battery tap 389 through reverse and forward directional mode display lights 430 and 432, respectively, located within the passenger compartment of electric vehicle 10. Terminal VIII of driver 428 is connected to ground. Terminal IX of driver 428 is interconnected to ground through a parallel combination of a 1.4 micorfarad capacitor 434 and a 4.7k resistor 436.

Terminal XIII of selector 336 is connected to a terminal D and terminal IX of selector 336 is connected to a terminal E. Finally, terminals F and G are directly interconnected with one another.

The one of four selector 336 operates to connect or enable one of the sensors 212, 214, 216 or 218 as a function of vehicle (motor) speed and directional mode selection via the forward/reverse select signal. If electric vehicle 10 is below a predetermined threshold speed and forward switch 386 is momentarily closed, the output of AND gate 390 goes high. Likewise, if reverse switch 388 is momentarily closed, the output of AND gate 408 goes high.

The one of four selector 336 will thus receive both the forward/reverse select and zero speed select signals and interconnect one of sensors 212, 214, 216 or 218 to output terminal E to effect a particular applied voltage leading angle, according to the following truth table:

(ii) Motor Positioning Circuit

Referring to FIG. 6b, a motor positioning circuit indicated generally at 442, as well as typical P.C. board filter 334 is illustrated. Filter 334 is reproduced on each logic P.C. board within control circuit 44. Each filter 334 has input terminals 441 connected to .+-.15 VDC output terminals of the main logic power supply and .+-.15 VDC output terminals 443 connected to each separate +15 VDC and -15 VDC terminal on its associated board. Each filter 334 includes a positive bus 444 and a negative bus 446. Buses 444 and 446 are interconnected by series combination of two 0.022 microfarad capacitors 448 and 450. The point of common connection between capacitors 448 and 450 are connected to ground. Likewise, buses 444 and 446 are interconnected by series combination of two 15 microfarad tantalum capacitors 452 and 454. The point of common connection between capacitors 452 and 454 are also connected to ground.

Hall effect sensor 186 has one terminal connected directly to the +15 VDC power supply and a second terminal interconnected to the TR+ terminal IV of a type 4098 CMOS timer circuit 462 through a current limiting resistor 460. Terminal IV of timer 462 is interconnected to the +15 VDC power supply through a 4.7k resistor 456, and to ground through a 0.01 microfarad capacitor 458. As in the case of sensors 212, 214, 216 and 218, resistors 456 and 460 and capacitor 458 provide signal conditioning. The point of common connection of resistor 456 and capacitor 458 is connected to an input terminal of an exclusive OR gate 464. The Q terminal VII of timer 462 is connected to the clock terminal III of a type 4013 flip-flop 466. Terminals VIII and XIII of timer 462 are connected to ground and terminals V, XVI and III of timer 462 are connected to the +15 VDC power supply. Terminal I of timer 462 is interconnected with the +15 VDC power supply by a series combination of a 0.22 microfarad capacitor 468 and a 390k resistor 470. The point of common connection between resistor 470 and capacitor 468 is connected to terminal II of timer 462. Data terminal V of flip-flop 466 is connected to the +15 VDC power supply, and terminals VI, VII and VIII of flip-flop 466 are connected to ground. Terminals X and XIV of flip-flop 466 are connected to the +15 VDC power supply. R terminal IV of flip-flop 466 is connected to R terminal X of a type 4013 flip-flop 472 and to terminal F. Q terminal I of flip-flop 466 is connected to terminal I of a type 2905 timer 474. Terminal VII of timer 474 is connected to clock terminal XI of flip-flop 472. Q terminal XII of flip-flop 472 is interconnected to the gate of a power MOSFET 476 through a 100 ohm resistor 478. The source of FET 476 is connected to the +24 VDC battery tap 389 through a parallel combination of solenoid coils 268 and 308. Additionally, the drain of FET 476 is connected to the anode of a free wheeling diode 480, the cathode of which is connected to the +24 VDC battery tap 389. Flip-flop 472, FET 476, resistor 478, diode 380 and coils 308 and 268 constitute a solenoid actuation circuit 481. Diode 480 is provided for protection of FET 476. Additionally, it is contemplated that any production embodiment of the present invention would include a fuse disposed intermediate coils 308 and 268 and +24 VDC tap 389.

Terminals V and VI of timer 474 are connected to the +15 VDC power supply. Terminal II of timer 474 is interconnected to ground through a series combination of an 826k resistor 482 and a 2.2 microfarad capacitor 484. The point of common connection between resistor 482 and capacitor 484 is connected to terminal III of timer 474. Terminals IV and VIII of timer 474 are directly connected to ground. Terminal VII of timer 474 is connected to one input of an AND gate 486, to one input of AND gate 499, to ground through a 10k resistor 488 and to both inputs of a NOR gate 490. The output of NOR gate 490 is connected to one of the inputs of an AND gate 492. The other input of AND gate 492 is interconnected to a terminal H through a 10k resistor 494. The output of AND gate 492 is connected to one of the inputs of a NOR gate 496. The output of AND gate 486 is connected to the remaining input of NOR gate 496. The output of NOR gate 496 is connected to a terminal I. The remaining input terminal of AND gate 486 is connected to the input terminal of exclusive OR gate 464 which is connected to terminal IV of timer 462. The output of exclusive OR gate 464 is connected to the remaining input of AND gate 499. The output of AND gate 499 is connected to an input of XOR gate 501. The other input to XOR gate 501 is coupled to the junction of an RC network formed by a 15 microfarad capacitor 438 and a 51K resistor 440. The opposite end of capacitor 438 is tied to +15 VDC. The other end of resistor 440 is tied to ground. Capacitor 438 and resistor 440 operate to output an intialization or start-up pulse on terminal B and C (FIG. 6a). The output of XOR gate 501 is interconnected to terminal C.

Sensor 186 is designated the electrical stop sensor inasmuch as it is used only for electronically stopping motor 18. Specifically, sensor 186 is employed for centering rotor 26 on a high torque parking position and to control the rotor current toggling as described hereinabove. In addition to being tied into minimum RPM timer circuit 462, sensor 186 is input to exclusive OR gate 464 along with the -4.degree. sensor 214. Exclusive OR gate 464 selects direction reversal and is connected to one of the inputs of exclusive OR gate 500 in FIG. 6c.

Q output VII of timer circuit 462 goes low when rotor speed falls below 20 RPM. Q output I of flip-flop 466 will then go high when speed is below 20 RPM and the accelerator pedal has been released. This will output a two second pulse from timer 474 and a two second low output from OR gate 490. During this two second interval, terminal XI of flip-flop 472 will go high and Q terminal XII will be latched low, deenergizing solenoid coils 268 and 308. Additionally, the output of OR gate 490 is input to AND gate 492 along with the output of one of four selector 336. The pulse from timer 474 is also fed to AND gate 486 with the output of sensor 186. AND gate 486 will output a two second pulse to the input of OR gate 496 along with the output of AND gate 492 to feed a control signal to the remaining input of exclusive OR gate 500 of FIG. 6c.

(iii) Current Demand Logic and Minimum Starting Speed Circuit

Referring to FIG. 6c, terminal E is directly connected to terminal H. Terminal B is connected to one of the inputs of an exclusive OR gate 500. The remaining input of exclusive OR gate 500 is connected to terminal I. The output of exclusive OR gate 500 is interconnected with the input of an inverter 506 through a parallel combination of a signal diode 508 and an 80.6k resistor 510. The input of an inverter 506 is tied to ground through a 470 microfarad capacitor 512. The output of inverter 506 is connected to the input of inverter 514 and to one input of a AND gate 516. The output of inverter 514 is interconnected with the negative of an operational amplifier 518 through a 15k resistor 520. The output of amplifier 518 is interconnected with its negative input through a 10k feedback resistor 522.

The output of exclusive OR gate 500 is also connected to the input of an inverter 524. The output of inverter 524 is interconnected to the input of another inverter 526 through a parallel combination of an 80.6k resistor 528 and a signal diode 530. The input of inverter 526 is connected to ground through a 470 picofarad capacitor 532. The output of inverter 526 is connected to the input of another inverter 534 and to one of the inputs of a AND gate 536. The output of inverter 534 is inteconnected to the positive input of operatonal amplifier 518 through a 15k resistor 538. The positive input of amplifier 518 is also connected to ground through a 10k resistor 540.

The output of amplifier 518 is interconnected with the positive input of a comparator 542 through a 2k resistor 544. The output of comparator 542 is interconnected with the positive input thereof through a series combination of two variable output voltage inverters 546 and 548 and a 100k resistor 550. The output of comparator 542 is also interconnected to a +15 VDC power supply through a 11k resistor 552 and to the remaining input of AND gate 516. The output of AND gate 516 is connected to the base of a transistor 554. The collector of transistor 554 is connected to the +15 VDC power supply and the emitter of transistor 554 is interconnected to a terminal J through a 470 ohm resistor 556. The output of comparator 542 is interconnected to the remaining input of AND gate 536 through an inverter 558. The output of NAND gate 536 is connected to the base of a transistor 560. The collector of transistor 560 is connected to the +15 VDC power supply and the emitter of transistor 560 is interconnected to a terminal K through a 470 ohm resistor 562.

An amplifier shown generally at 564 receives an input from a fast response current sensor 566 (see FIG. 6e) through terminals L, M and N and outputs a sensed motor current signal from the output of an amplifier 568. A line 569 interconnects the output of amplifier 568 with the negative input of comparator 542 through a 2k resistor 570. The negative input of comparator 542 is also connected to ground through a 0.068 microfarad capacitor 572. Terminal N is directly connected to a constant current power supply 574 and interconnected with a second terminal of power supply 574 through a 390 ohm current control resistor 576. Power supply 574 is connected to the +15 VDC power supply and operates in a manner well known in the art. Terminal M is interconnected to the negative input of an operational amplifier 578 through an 11k resistor 580. The output of amplifier 578 is interconnected with the negative input thereof through a parallel combination of a 110k resistor 582 and a 0.001 microfarad filter capacitor 584. Terminal L is interconnected to the positive input of amplifier 578 through a 11k resistor 585. The positive input of amplifier 578 is also interconnected to ground through a parallel combination of a 0.001 microfarad filter capacitor 586 and a 110k resistor 588. The output of amplifier 578 is interconnected with the negative input of amplifier 568 through a 11k resistor 590. The output of amplifier 568 is interconnected with its negative input through a 110k feedback resistor 592. The negative input of amplifier 568 is interconnected to the wiper of a 100k ohm potentiometer 594 having one end tap connected to the +15 VDC power supply and the other end tap connected to the -15 VDC power supply, through a 110k resistor 596. The positive input of amplifier 568 is interconnected with ground through a 4.99k resistor 598.

A type 4043 flip-flop 600 has its S terminal VII connected to sensor 212 and its R terminal XI connected to sensor 218. Q terminal X is connected directly to one input of an exclusive OR gate 602. Terminal X of flip-flop 600 is also interconnected to the remaining input of exclusive OR gate 602 through a 50k resistor 604. The remaining input of exclusive OR gate 602 is also interconnected to ground through a 0.1 microfarad capacitor 606. Terminals XVI and V of flip-flop 600 are connected to the +15 VDC power supply and terminal VIII thereof is connected to ground. The output of exclusive OR gate 602 is connected directly to S terminal XIV of a type 4043 flip-flop 608 and interconnected to ground through a series combination of a 20k resistor 610 and 1k resistor 612. The point of interconnection between resistors 610 and 612 is connected to the base of a transistor 614. The collector of transistor 614 is interconnected to the +15 VDC power supply through a 100k resistor 616 and to ground through a 10 picofarad capacitor 618. The emitter of transistor 614 is connected directly to ground. The collector is transistor 14 is connected directly to R terminal XV of flip-flop 608. Terminal VIII of flip-flop 608 is connected directly to ground and terminal VI thereof is connected to the +15 VDC power supply. Terminal I of flip-flop 608 is connected to terminal A as well as one input of an exclusive OR gate 620. The remaining input of exclusive OR gate 620 is connected to the +15 VDC power supply. The output of exclusive OR gate 620 is connected directly to one of the inputs of each of separate AND gates 622 and 624. The remaining inputs of AND gates 622 and 624 are commonly interconnected to +15 VDC power supply through a 10k resistor 626. The outputs of AND gates 622 and 624 are commonly connected to one end tap of a 20k potentiometer 628. The remaining end tap of potentiometer 628 is connected to ground and its wiper is connected to the anode of a signal diode 630. The cathode of diode 630 is connected to a node, designated as 632. Node 632 is interconnected to ground through a 100k resistor 634 and directly to the supply inputs of variable output voltage inverters 514 and 534 as well as inverters 546 and 548. Terminal G is connected to the output of an operational amplifier 636 and to the input of AND gate 622 connected to resistor 626.

A 20k, voltage present potentiometer 638 has one end tap thereof connected to the +15 VDC power supply and other end tap connected to ground. The wiper of potentiometer 638 is interconnected to the negative input of amplifier 636 through a 100k resistor 640. An accelerator pedal actuated 100k potentiometer 642 has one end tap connected to the +15 VDC power supply and the other tap connected to the -15 VDC power supply. The wiper of potentiometer 642 is connected to the anode of a signal diode 644. The cathode of diode 644 is interconnected to the plus input of amplifier 636 through a 100k resistor 646 and to the positive input of an operational amplifier 648 through a 110k resistor 650. The output of amplifier 636 is interconnected with its positive input through a 10M feedback resistor 652. The output of amplifier 648 is interconnected with its negative input through a 1k feedback resistor 654. The negative input of amplifier 648 is also interconnected to ground through a 1k resistor 656. The output of amplifier 648 is also connected to the anode of a signal diode 658 whose cathode is connected to node 632.

During normal operation, the output of one of four selector 336 is effectively connected to the upper input of exclusive OR gate 500 and the lower input thereof is high. During stopping of motor 18, the upper input of gate 500 is tied to sensor 186 and the lower input is toggled so motor 18 oscillates around the 30.degree. or stopping sensor 186.

The operator demand is input from potentiometer 642 into noninverting amplifier 648 which, in turn, outputs a current command signal through node 632 to buffering inverters 514 and 534. The driver demand signal is also input into comparator 636 whose output is low when no pressure on the accelerator pedal is present. Potentiometer 638 establishes a predetermined reference level.

During normal operation, the output of exclusive OR gate 500 outputs a signal to a fast off, delayed-on delay circuit for the prevention of cross-firing. From there the signal is fed through successive buffering inverters 506 and 514, and into one of the inputs of amplifier 518. Likewise, the output of exclusive OR gate 500 is inverted and passed through a delay circuit and successive buffering inverters 526 and 534. Inverters 506 and 526 are not accelerator pedal sensitive, and inverters 514 and 534 are accelerator pedal sensitive by virtue of their receiving the current command signal as their supply voltage. The outputs of buffers 514 and 534, which are proportional in amplitude to current demand and at a frequency determined by the motor position sensors, are input into a gain circuit composed of amplifier 518 and resistors 522 and 540. The output of the gain circuit is fed to the input of a hysteresis circuit including comparator 542 which also inputs the current sense signal from amplifier 564. The positive input feedback loop of comparator 542 includes inverters 546 and 548, which are accelerator pedal sensitive. The interconnection of inverters 548 and 546 with the accelerator circuit is not illustrated. The hysteresis circuit is output to complimentary AND gates 516 and via inverter 558, and AND gate 536, which also include inputs from the output of buffering inverters 506 and 526, respectively, as a verification of the receipt of a control signal to eliminate false triggering by verifying that the signal was sent from the delay circuit. The output of AND gates 516 and 536 are then fed to complimentary base drive circuits as will be described in detail hereinbelow.

Power supply 574 outputs a constant current signal to terminal N which subsequently energizes a motor current sensor. Two lines are fed back from the sensor via terminals L and M to the inputs of amplifier 578. The difference of these signals is output to amplifier 568 which generates the motor current signal.

The minimum starting speed circuit 599 has inputs from the +1.5.degree. and -15.degree. sensors 212 and 218 to flip-flop 600, which outputs a signal from Q terminal X to one of the inputs of exclusive OR gate 602 and to the remaining input of exclusive OR gate 602 through a phase delay circuit composed of resistor 604 and capacitor 606. Minimum starting speed circuit 599 eliminates cogging and prevents entrapment of rotor 26 in the zero electrical torque position. Exclusive OR gate 602 outputs a signal to flip-flop 608 which functions as a zero speed detector circuit to output a high signal on Q terminal I when speed exceeds 20 RPM. That signal is also fed to latch 420 through an exclusive OR gate 418 in FIG. 6a. Exclusive OR gate 620 will then output an override demand signal to node 632 through gates 622, 624 and potentiometer 628 which, in some circumstances, will override the torque demand signal. The remaining inputs for gates 622 and 624 are from the output of amplifier 636 and receive a high signal when pressure exits on the throttle pedal.

If there is pressure on the accelerator pedal and rotor speed is below 20 RPM, the control signal will be established by potentiometer 628 which forces a minimum torque demand current until the operator releases the accelerator pedal or rotor speed exceeds 20 RPM.

Amplifier 648 outputs torque command signals through node 632. However, the larger of the torque signal or the minimum start torque signal from minimum starting speed circuit 599 will be applied to the gain circuit. When rotor speed is below 20 RPM, Q terminal I of flip-flop 608 goes low. Thereafter, flip-flop 600 and sensors 212 and 218 comprise an anti-oscillation circuit. If motor 18 does not have enough kinetic energy to move the vehicle load, rebound will occur and potentially could cause an oscillating condition. The +15.degree. and -15.degree. sensors 218 and 212, respectively, mandate the oscillation through an unreal 30.degree. range, which effectively prohibit oscillation from occurring. It can be seen that the +15.degree. and -15.degree. sensors 218 and 212, respectively, are only required if the motor stops at zero torque point.

After 20 RPM is exceeded, the RC constant of resistor 616 and capacitor 618 is such that it cannot ramp up, preventing flip-flop 608 from being reset. Q output I of flip-flop 608 will thus always be low, as will the outputs of exclusive OR gate 620 and AND gate 622, which commands minimum gain.

If the output of exclusive OR gate 500 is high, the output of buffering inverter 506 goes low, and the upper input of AND gate 516 goes low. Also, buffering inverter 526 output will go high and the lower input of AND gate 536 will go high. The outputs of buffering inverters 514 and 534 are opposites by virtue of inverter 524 and are multiplied by the gain circuit. If the output of buffering inverter 514 is high, the output of buffering inverter 534 is low and the output of amplifier 518 goes negative. This output is compared with motor current. In the starting mode, the output of comparator 542 is low, the lower input of AND gate 516 is low and the upper input of AND gate 536 is high. Thus, the control signal imparted to the base drive circuits is a function of gain or pedal position, sensed motor current and hysteresis. When the sensed motor current exceeds commanded torque current, comparator 542 is shut off to disable the transmission of further control signals and the generation of switch command signals.

(iv) Base Drive Circuits

Referring to FIG. 6d, identical first and second base drive circuits, 660 and 662, respectively, are illustrated. During normal operation, base drive circuit 660 and base drive circuit 662 will alternatingly receive control signals from current demand logic circuit 499 via terminals J and K. Each circuit 660 and 662 processes a received control signal and inputs a switch command signal to the base of transistor 56 or 62, respectively. Because circuits 660 and 662 are identical, the specific configuration and operation of only one will be given for the sake of brevity.

Terminal J is connected to the anode of an LED of a type 6N135 optical coupler 664. It is contemplated that other fast response type commercially available equivalents can be substituted. The cathode of the LED is interconnected to ground through a 470 ohm resistor 666. The anode of the photodiode portion of optical coupler 664 is connected to the base of a transistor 668. In the components selected by the applicant for the illustrated embodiment of the invention, transistor 668 is integrally formed with optical coupler 664. However, it is shown as a discrete element for purposes of clarity. The cathode of the photodiode is connected to the +4 VDC power supply. The collector of transistor 668 is interconnected to the +4 VDC power supply through a 10k resistor 670 and interconnected to both inputs of a NOR gate 672 through a 47k resistor 674. The emitter of transistor 668 is interconnected to ground through a parallel combination of a 1,000 microfarad capacitor 676 and a 0.1 microfarad capacitor 678. Capacitors 676 and 678 comprise a negative base drive power supply filter circuit shown generally at 680. The output of NOR gate 672 is connected to both inputs of another NOR gate 682. The output of NOR gate 682 is directly connected to one of the inputs of a NOR gate 684 and interconnected to the inputs of NOR gate 672 through a 1M hysteresis resistor 686. The output of NOR gate 682 is interconnected to the +4 VDC power supply through a series combination of a 200 picofarad capacitor 688 and a 100k resistor 690. The +4 VDC power supply is also interconnected to ground through a parallel combination of 1000 microfarad capacitor 692 and a 0.1 microfarad capacitor 694. The point of common connection between capacitor 688 and resistor 690 is connected to one of the inputs of a NAND gate 696. The +4 VDC power supply is connected to the emitter of a transistor 698 whose collector is interconnected to the -4 VDC power supply through a 10k resistor 700. The collector of transistor 698 is also connected to both inputs of a NAND gate 702 whose output is connected to the remaining input of NAND gate 696. The output of NAND gate 696 is connected to both inputs of a NAND gate 704 whose output is connected to the remaining input of NOR gate 684. The output of NOR gate 684 is connected to both inputs of a NOR gate 706 whose output is interconnected to the base of a power transistor 708 through a 2.7k resistor 710.

The emitter of transistor 708 is connected to the emitter of transistor 668 and to the -4 VDC power supply. The collector of transistor 708 is interconnected to the +4 VDC power supply through a 27 ohm resistor 712. The collector of transistor 708 is connected to the anode of a small powder diode 714 and to the base of a power transistor 716. The cathode of diode 714 is interconnected to the base of transistor 698 through a 100k resistor 718 and to the collector of the driver transistor 56. The base of transistor 698 is also connected to the anode of a signal diode 720 whose cathode is connected to the +4 VDC power supply. The collector of transistor 716 is connected to the +4 VDC power supply. The emitter of transistor 716 is interconnected to the emitter of a power transistor 722 through a 3.9 ohm resistor 742. The collector of transistor 708 is connected to the base of transistor 722 and the collector of transistor 722 is connected to the -4 VDC power supply. The emitter of transistor 716 is connected to the base of power transistor 56 through line 66. The collector and emitter of power transistor 56 are connected to terminals O and P, respectively.

Terminal K of base driver circuit 662 is likewise connected to an optical coupler 726. The collector and emitter of power transistor 62 are connected to terminals Q and R, respectively. Finally, terminals L, M and N are directly connected to terminals S, T and U, respectively.

During normal operation, base drive circuits 660 and 662 will alternately be on when a control signal is received at terminals J or K. The power transistor, 56 or 62, associated with the base driver circuit which is "on", remains in saturation (after initial turn on). During such times, the input to that base drive circuit 660 or 662 is high. When optical coupler 664 is on, power transistor 56 will be on and when optical coupler 726 is on, power transistor 62 will be on.

By way of example, when power transistor 56 is on or conducting, and is to be turned off, the control signal received at optical coupler 664 is terminated. The collector voltage of transistor 668 will rise to four volts. The output of NOR gate 682 will then go high. The output of NOR gate 706 will also go high turning on transistor 708. Transistor 708 going on will turn off transistor 716 and turn on transistor 722, which, in turn, will pull current from power transistor 56 to the -4 VDC power supply, turning off transistor 56.

Base drive circuits 660 and 662 provide protection when power transistors 56 and 62 are out of saturation. Continuing the example, when power transistor 56 is off, its collector voltage will be at 96 VDC (battery voltage). Transistor 698 will be off, connecting the -4 VDC supply to both inputs of NAND gate 702. The output of NAND gate 702 will be high. During steady state conditions, the input of NAND gate 696 connected to the point of common connection between resistors 690 and capacitor 688 will have risen to +4 VDC due to the charging of capacitor 688. If power transistor 56 comes out of saturation after steady state condition has been achieved, the output of NAND gate 702 will also go high. If the input of NOR gate 684 connected to the output of NAND gate 704 is high, then the output of NOR gate 706 will also be high, latching power transistor 56 off.

In the case where saturation of power transistor 56 is desired, the failsafing provided by base drive circuits 660 and 662 described immediately hereinabove is bypassed. When a control signal is received at optical coupler 664, it goes high, turning on transistor 668. The collector voltage of transistor 668 will go low as will the output of NOR gate 682. The voltage across capacitor 688 will instantaneously be low, causing the input of NAND gate 696 associated therewith to go low for one R.C. time constant (20 ms). When the input of NAND gate 696 goes low, the output of NAND gate 704 will also go low. The input of NOR gate 684 connected to the output of NAND gate 704 will thus go low for one time constant. Simultaneously, the other input of NOR gate 684 is also low whereby power transistor 56 will, by default turn on into saturation. When this happens, the collector voltage of power transistor 56 goes low, turning transistor 698 on. The inputs of NAND gate 702 will go high and its output low. The input of NAND gate 696 connected to the output of NAND gate 702 will go high and its output low shutting off NAND gate 704 and driving the input of NOR gate 684 associated therewith low. As long as the other input (that connected to resistor 686) remains low, power transistor 56 cannot turn off. Power transistor 56 can only be turned off by (1) receiving a logic signal through terminal J bringing the input of NOR gate 684 high, or (2) the collector current of power transistor 56 increases to the point that the transistor comes out of saturation. In such a case, transistor 698 will lose its bias and turn off. The exemplified operation of base drive circuit 660 hereinabove can be applied equally to that of base drive circuit 662.

(v) Snubber and Energy Recovery Circuit

Referring to FIG. 6e, terminals O and Q are directly connected to windings 50 and 52, respectively, of motor 18, passing through current sensor 566 in opposite directions as illustrated at loop 728. Terminals S, T and U are connected to current sensor 566 which can be, for example, a linear hall effect sensor. Terminal P is interconnected to terminal O by a snubber circuit indicated generally at 730, and terminal R is interconnected to terminal Q through an identical snubber circuit 732. Snubber circuit 730 comprises a series connection of a 4 microfarad capacitor 734 and a 10 ohm resistor 736 interconnecting terminals P and O. The point of common connection between resistor 736 and capacitor 734 is connected to the cathode of a power diode 738. The anode of diode 738 is connected to terminal O. Likewise, snubber circuit 732 comprises a series combination of a 4 microfarad capacitor 740 and 10 ohm resistor 742 interconnectomg terminals R and Q. The point of common connectin between resistor 742 and capacitor 740 is connected to the cathode of a diode 744. The anode of diode 744 is connected to terminal Q.

Terminal O is interconnected to input line 74 through diode 76, and terminal Q is interconnected to input line 74 through diode 78. Input line 74 and output line 72 are connected to energy recovery circuit 70. Output line 72 is also connected to ground through a 7500 microfarad filter capacitor 746.

Energy recovery circuit 70 is constructed as follows. A 2500 microfarad electrolytic capacitor 748 interconnects output and input lines 72 and 74, respectively. Input line 74 is interconnected to the center tap of a primary winding 750 of an energy recovery transformer, shown generally at 752, through a line 754. Primary 750 comprises thirty-two turns of 14AWG wire. Transformer 752 has a secondary winding 756 constructed of fifteen turns of 18AWG wire. Secondary winding 756 is connected to diagonally opposed corners of a bridge 758 constructed of four power diodes 760. One remaining corner of bridge 758 is connected to output line 72 and interconnected to ground through a 7.47 mircofarad capacitor 762. Output line 72 is interconnected to a terminal V through a line 764. Lines 754 and 764 are interconnected by a 0.82 microfarad capacitor 766. Line 754 is interconnected to the cathode of a zener diode 768 through a 3.5k resistor 770. The anode of zener diode 768 is connected to line 764. The point of common connection between diode 768 and resistor 770 is interconnected to line 764 through a parallel combination of a 0.1 microfarad capacitor 772 and a 15 microfarad capacitor 774. The point of common connection between the resistor 770 and diode 768 is also connected to terminals IV, VI and XIV of a type 4047 low power CMOS multivibrator 776. Terminals VII, VIII, IX and XII of multivibrator 776 are connected to line 764. Terminals II and I of multivibrator 776 are interconnected by a series connected 45.5k resistor 778 and a 220 picofarad capacitor 780. The point of common connection between capacitor 780 and resistor 778 is connected to terminal III of multivibrator 776. Terminal X of multivibrator 776 is interconnected to the base of a transistor 782 through a 2.2k resistor 784. Terminal XI of multivibrator 776 is also interconnected to the base of a transisitor 786 through a 2.2k resistor 788. The point of common connection between resistor 770 and diode 768 is interconnected to the gate of a power MOSFET 790 through a 470 ohm resistor 792. The collector of transistor 782 is also connected to the gate of FET 790. The point of common connection between resistor 770 and diode 768 is interconnected the gate of a second MOSFET 794 through a 470 ohm resisitor 796. The collector of transisitor 786 is connected to the gate of FET 794. The emitters of transistors 782 and 786 are connected to line 764. The sources of FETs 790 and 794 are commonly interconnected to the center tap of primary winding 750 of transformer 752 through a 60 microfarad capacitor 798.

Energy recovery circuit 70 was implemented primarily due to motor 18 being bifiler wound, and because the transformer action between windings 50 and 52 is not 100% efficient. As energy is transformed between windings 50 and 52, the presence of inefficiencies due to imperfect coupling will cause some energy to remain in the winding presently transferring energy to the other. Because the point of common connection of of windings 50 and 52 is connected to the positive terminal of battery 40, energy not transformed will be reflected by a potential in excess of two times the battery voltage (V.sub.cc), no energy is recovered and transferred back to line 72. During normal operation, however, stray energy will be reflected in a primary voltage exceeding V.sub.cc which will cause recovered energy to flow back to battery 40 during operation.

(vi) Switching Power Supply Circuit

Referring to FIG. 6f, a switching power supply, shown generally at 800 is illustrated which powers the inverter circuit 44 and the control circuit 46. Power supply 800 has terminal V (direct from battery 40) interconnected to the center tap of a primary winding 802 of a power transformer 804 through a fuse 806. Primary winding 802 comprises 72 turns of 26AWG wire. Transformer 804 has three secondary windings 808, 810 and 812. Secondary winding 808 comprises 16 turns of 26AWG wire and has a center tap connected to ground. Both secondary windings 810 and 812 comprise 8 turns of 16AWG wire with grounded center taps. Secondary winding 808 is diagonally connected to a diode bridge 814 composed four type A114 diodes 816. One opposite corner of bridge 814 is interconnected to a type 7815 regulator 818 through a 1 ohm resistor 820. The remaining corner of bridge 814 is interconnected to a type 7915 regulator 822 through a 1 ohm resistor 824. The corner of bridge 814 connected to resistor 820 is interconnected with the corner connected to resistor 824 through a series combination of two 0.0056 microfarad capacitors 826 and 828. The point of common connection between capacitors 826 and 828 is also interconnected to the point of common connectin between resistor 820 and regulator 818 through a 0.1 microfarad capacitor 830. Likewise, point of common connection between capacitors 826 and 828 is interconnected to the point of common connection between resistor 824 and regulator 822 by a 0.1 microfarad capacitor 832. The output of regulators 818 and 822 are interconnected by a series connection of two 0.1 microfarad capacitors 834 and 836 as well as a pair of series connected 15 mircrofarad capacitors 838 and 840. The point of common connection between capacitors 826 through 840 is connected to ground. The output of regulator 818 represents the +15 VDC power supply output terminal 841 and the output of regulator 822 represents the -15 VDC output terminal 843.

Secondary winding 810 is diagonally connected across a diode bridge 842 composed of two type A115 diodes 844 and two type A114 diodes 846. The remaining corners of bridge 842 are interconnected by series connected 15 microfarad capacitors 848 and 850, and comprise the +4 VDC and -4 VDC output terminals 849 and 851, respectively, of power supply 800. The point of common connection between capacitors 858 and 860 is connected to ground.

Each end of primary winding 802 of transformer 804 is connected to the drain of a power MOSFET 862 and 864. LThe sources of FETs 862 and 864 are commonly tied to ground. The gate of FET 862 is interconnected to terminal X of a type 4047 CMOS multivibrator 866 through a 4.7k resistor 868. The gate of FET 862 is also connected to the anode of a signal diode 870. The cathode of diode 870 is connected to terminal X of multivibrator 866. Terminals I and II of multivibrator 866 are interconnected by a series connected 45.5k resistor 876 and a 220 picofarad capacitor 878. The point of common connection between resistor 876 and capacitor 878 is connected to terminal III of multivibrator 866. Terminals VII VIII, IX, and XII of multivibrator 866 are connected to ground. Terminals IV, V, VI and XIV of multivibrator 866 are interconnected to ground through a 0.01 microfarad capacitor 880. Terminals IV, V, VI and XIV of multivibrator 866 are interconnected to the center tap of primary winding 802 of transformer 804 through a 3.5k resistor 882, and are connected to the cathode of a 12 volt zener diode 884. The anode of zener diode 884 is connected to ground. Finally, the center tap of primary winding 802 is connected to ground through a 0.982 microfarad capacitor 886.

Power supply 800 operates by including a fused line 888 connected through resistor 882 and zener diode 884 to energize multivibrator 866 as a free running oscillator shown generally at 890. Oscillator 890 toggles FETs 862 and 864 to energize the transformer 804 and the secondary circuit shown generally at 892 thereof.

It is to be understood that the invention has been described with reference to a specific embodiment which provides the features and advantages previously described, and that such specific embodiment is susceptible of modificatin, as will be apparent to those skilled in the art. For example, instead of providing 2 pairs of position sensors, motor position may be determined by generating a pulse for each increment of movement and counting pulses or by generating a voltage proportional to the distance the motor has rotated. It is also to be understood that although described in conjunction with a permanent magnet motor in a passenger vehicle, in its broadest sense, the present invention can be adapted for use in other devices which use permanent magnet motors. Accordingly, the foregoing description is not to be construed in a limiting sense.