NL1022038C2 - Blood pump has magnetic suspension with stator and rotor magnetic regions held by stator and rotor parts that interact so rotor part is held in radial direction to form magnetic suspension gap - Google Patents

Abstract

The arrangement has a stator part and a rotor part with adjacent stator and rotor pump regions and a magnetic suspension with stator and rotor magnetic regions held by the stator and rotor parts and that interact so that the rotor part is held in the radial direction to form a magnetic suspension gap (21) between the rotor part and the stator part. One of the magnetic regions can be moved axially with respect to the other and there are an actuator for axial positioning and a magnetic drive device.

Description

Blood pump with magnetically suspended stator and rotor element with a first and second magnet array.

The invention is a spin-off of Dutch patent application 1014329, which is based on the provisional US patent application serial number 60 / 119,356 filed on February 9, 1999, the US patent application serial number 08 / 978,670, filed November 26, 1997 and US patent applications serial numbers 09 / 273,384 and 09 / 288,413 filed on March 22, 1999 and April 8, 1999 respectively, both offshoots from aforementioned application No. 10 08 / 978,670. The descriptions of the aforementioned patent applications are hereby incorporated herein by reference.

BACKGROUND

Field of the invention

The invention relates to pumps that use magnetic suspension and rotation means to pump blood, and more particularly relates to a magnetically suspended and magnetically rotatable blood pump that has no mechanical bearings or seals and is provided with a pumping means that is magnetically is supported both radially and axially.

DESCRIPTION OF THE PRIOR ART

The use of rotary pumps (i.e. axial, centrifugal or mixed flow) for pumping liquids and in particular blood is known from the prior art. A rotary pump generally consists of an outer housing with an inlet port and an outlet port, and an impeller mounted on a shaft (with mechanical bearings and seals) in the outer housing for rotation about an axis. Mechanical bearings are subject to wear and premature failure and can generate sufficient heat and mechanical stresses to cause unacceptable damage to the blood. Shaft seals are also subject to wear and heat generation which can lead to leakage, blood clot formation, bearing failure, and bacterial growth. Examples of rotary pegs using shaft-mounted impellers with bearings and seals can be found in U.S. Pat. No. 4,135,253 to Rech et al., U.S. Pat. No. 4,403,911 to Possell, U.S. Pat. No. 4,704,121 to Moise, and U.S. Pat. from Dorman.

Numerous pumps have already been designed to cope with the above problems by using a lubricating flow for the mechanical bearings of the rotary pump. Examples of such pumps are described in US-4,944,722 to Carriker et al. And US-4S46,152 to Wampler et al. These types of pumps may have various problems such as being not fully implantable due to the need to use a percutaneous supply line and an external reservoir for realizing the bearing coil flow. The potential risk of infection and leakage also exists as a result of the flushing fluid and percutaneous pipes. Moreover, it may still be necessary to replace the mechanical bearings after a certain time because they come into direct contact with other pump structures during operation.

By using a rotary fluid pump with a magnetically suspended impeller, all of the above problems can be avoided. Examples of such pumps are described in U.S. Pat. No. 5,326,344 to Bramm et al., U.S. Pat. No. 4,688,998 to Olsen et al. And U.S. Pat. No. 4,779,614 to Moise. A problem that may be associated with all of these cited inventions is the fact that a single opening is used both as a passage for the blood flow through the pump and for the magnetic suspension and rotation of the impeller. These two functions immediately impose conflicting requirements on the dimensions of the opening. As a flow path for blood flow, the opening must be large to prevent damage to the blood. As an opening for magnetic suspension and rotation, the opening must be small to minimize the dimensions of the magnetic suspension and the rotation components and to allow efficient energy use in realizing the impeller suspension and rotation. As a result, any selected aperture size can result in an undesirable compromise between blood damage, device dimensions, and energy requirements. Another problem that may be associated with all the cited inventions is that the magnetic fields created by the permanent magnets in the described pumps may change with time, may vary with temperature, and due to the influence of external ferromagnetic objects. These changes in the magnetic field 1022038 3 may cause the equilibrium point of the pump rotor to change and may cause the rotor to make contact with the stator. Contact between pump rotor and stator can result in damage to the blood, damage to the pump and inadequate pump function.

Examples of pumps with separate openings for the primary blood flow and for the impeller rotation are described in US-5,324,177 to Golding and US-5,049,134 to Golding. However, these pumps also make use of a rotation opening for implementing hydrodynamically bearing bearings for the rotor. Such hydrodynamic bearings can subject the blood to excessive shear forces that can unacceptably damage the fragile components of the blood. Moreover, according to Golding, the pumps place the stationary magnetic components within a central bore of a rotating assembly. Such configurations generally cause the mass and rotational inertia of the rotating assembly to be greater than those in a system in which the stationary magnetic components are disposed around the outer surface of the rotating assembly. Thus, large mass rotating rotational inertia assemblies may be undesirable due to the fact that the axial and radial bearing elements must be relatively large to maintain the correct alignment of the rotating assembly due to shocks, vibrations and accelerations.

The flow rate of blood pumps capable of creating a negative inlet pressure must be dynamically adjusted as an adjustment to the blood flow rate in the ventricle of the heart, typically the left ventricle. If a too small flow is produced by the blood pump, tissues and organs of the body are inadequately flowed through and blood pressure in the left ventricle will increase, potentially causing excessive pulmonary pressure and congestion. On the other hand, if the flow rate of the blood pump becomes too high, excessive negative pressure can be generated in the left ventricle and in the pump inlet. Excessive negative blood pressure is undesirable for the following reasons: 1) unacceptable levels of blood damage can be caused by cavitation; 2) the pump can be damaged by cavitation; 3) the walls of the ventricle can collapse and be damaged; and 4) the walls of the ventricle can fall in and block the flow of blood to the pump.

1 0220 38 4

By using a control system for dynamically controlling the flow rate of the pump to avoid excessive negative blood pressure, the above problems can be avoided. An example of such a control system is described in U.S. Pat. No. 5,326,344 to Bramm et al.

Bramm describes a method for dynamically controlling the flow rate of a pump based on a signal derived from a single pressure sensor located in the pump inlet. A problem that may be associated with such a detecting system is the difficulty of achieving a long-term stability with such a sensor, in particular in view of the relatively low pressures (0 to 20 mm Hg) that are encountered and the hostile environment in which the sensor must operate. Another problem that may be associated with such a pressure sensing system is the effect of changing atmospheric pressure which may result in an inaccurate perception of the pressure needed to properly control the pump.

The natural heart normally creates a pulsating blood flow. Artificial blood pumps that do not generate a pulsating flow can cause thrombosis and will not adequately flow through the organs and tissues of the body. The speed of the rotating blood pumps can be varied to create a pulsating flow. Examples of pumps that can vary their speed to create a pulsed flow are described in US-5,211,546 by Isaacson et al., US-5,174,726 by Findlay and US-4,135,253 by Reich et al. One problem that may be related to varying the speed of pumps as quoted in all inventions is: if the speed variation is excessive, the blood flow in the pump may be too low and / or misdirected. If the blood flow in the pump becomes too low, the blood can be damaged by constant shearing of the blood pool in the pump. If the blood flow in the pump changes direction, the ventricle cannot be adequately discharged.

There is, therefore, a need for a blood pump to overcome the aforementioned problems, which may be associated with conventional blood pumps, and for a system for dynamically controlling such a blood pump in order to overcome the above-described problems that may occur with control systems utilizing blood pumps.

022033 5

BRIEF OVERVIEW OF THE INVENTION

The invention therefore relates to a blood pumping device implantable in a patient, comprising: a. A stator member; b. a rotor member positioned adjacent to the stator member for the purpose of rotation relative to the stator member; c. a magnetic suspension with a first magnetic array supported by the stator member and a second permanent magnetic array supported by the rotor member, said first and second arrays of permanent magnets cooperating to radially magnetically support said rotor member such that an annular magnetic suspension gap is created between the rotor element and the stator element; d. each of the first and second permanent magnetic arrays consisting of at least three adjacent permanent magnetic rings, numbered of the at least three adjacent permanent magnetic rings being polarized in a direction perpendicular to a polarization direction of odd-numbered of the at least three adjacent permanent magnetic rings, and wherein the odd-numbered are polarized in a direction parallel to each other such that magnetic flux between the first and the second permanent magnetic arrays is radially directed, and wherein at least one of the first and second permanent magnetic arrays are wholly by permanent magnets is formed, and e. a magnetic drive with a stator motor part carried by the stator member and a rotor motor part supported by the rotor member, the stator motor part and the rotor motor part cooperating to magnetically rotate the rotor element relative to the stator element to pump blood through the stator member stator and rotor pump parts.

In such a device, the blood pump comprises a magnetic suspension of a stator and a rotor with a first magnet array and second magnet array, each comprising a number of adjacent permanent magnet arrays with a predetermined polarity direction.

1022038 6

The blood pump device is provided with a stator element with which a magnetically suspended and rotating rotor element is maintained. The rotor can preferably be magnetically suspended within the stator, both in radial and in axial direction. The blood pump may also have an associated magnetic suspension control system and a flow rate control system through the blood pump. The blood pump may preferably be a centrifugal pump in which an impeller withdraws blood from the left ventricle of the heart and supplies it to the aorta thereby reducing the pressure to be generated in the left ventricle. The blood pump can also have relatively small dimensions such that it can be completely implanted in the human body. If bi-ventricular cardiac assistance is needed, a second blood pump can be implanted to assist the right ventricle. The impeller of the centrifugal pump can be an integral part of a rotor assembly. The rotor assembly can preferably be suspended via permanent magnetic radial bearings and an axial bearing acting on Lorentz force. The Lorentz force axial bearing can generate bidirectional axial forces in response to an applied current. The blood pump can include an axial position sensor and an axial position controller. The axial position sensor can monitor the axial position of the rotor and provide a feedback signal to the controller to maintain the axial position of the rotor. The axial position controller can also adjust the axial position of the rotor such that axial loads in the normal operating condition due to gravitation, acceleration or the impeller of the centrifugal pump are counteracted by inherent axial forces generated by the permanently magnetic radial bearings. By counteracting these axial forces in the operating condition using the axial position controller, the power required by the axial bearing acting with Lorentz force is minimized. The rotor assembly can be rotated by an electric motor. The blood pump can also be provided with an actuator to adjust the axial position of the magnetic low ring and rotation components in the stator, for example with respect to the stator itself. Such mechanisms can compensate for changes in the magnetic fields 30 of the permanent magnets used in the blood pump.

A primary blood flow entry path preferably runs through the relatively large central bore in the rotor. A first blood flow path in the back direction can run through an annular gap formed between the rotor and the stator of the pump as a result of the radial magnetic suspension. In order to minimize the dimensions of the device as a whole, all magnetic suspension and rotation forces can be supplied over the relatively narrow annular gap. A second blood flow path in the reverse direction can extend through an axial gap provided between the impeller wall and the cochlea chamber. All surfaces of the pump that come into contact with blood are continuously flushed to prevent blood clots and protein from precipitating.

The speed of the centrifugal pump can be controlled dynamically in order to avoid excessive negative pressure in the left ventricle. The blood pump flow rate control system may include an electronic heart pacer. The cardiac pacer may be operatively coupled to the outer surface of the heart and provide a feedback signal to the control system for controlling the flow rate of the blood pump. The cardiac pacer can be used to monitor the external dimension of the left ventricle. The blood pump flow rate control system may preferably operate in two modes, continuous and pulsed. In continuous operation mode, the pump speed can be controlled so that the sensed left ventricle size is kept at a first preset set point. In the pulsed operating mode, the pump can be dynamically adjusted such that the sensed left ventricle size alternates between a first predetermined set point and a second predetermined set point. A flow sensor can be used to avoid excessively low or reverse flow through the blood pump when operated in a pulsed mode.

Other details, objectives and advantages of the invention will become apparent from the following detailed description and accompanying figures of certain preferred embodiments thereof shown.

BRIEF DESCRIPTION OF THE FIGURES

A more complete understanding of the invention can be obtained with reference to the accompanying detailed description in conjunction with the accompanying figures in which: 1022038

FIG. 1a is a cross-sectional view of an embodiment of a blood pump with magnetically suspended and rotating rotor assembly and a mechanism for adjusting the axial position of the magnetic suspension components and rotation components mounted in the stator; FIG. 1b is a cross-sectional view of another embodiment of the blood pump shown in FIG. 1a;

FIG. 2 is a view of the blood pump in FIG. 1A on the line II - II;

FIG. 3a is a view of an embodiment of the permanent magnetic arrays used for the radial suspension and the axial biasing force; FIG. 3b is a view of a second preferred embodiment of the permanent magnetic arrays used for radial suspension and axial biasing force;

FIG. 4 is a perspective view of the blood pump of FIGS. 1a and 1b coupled to a circulation system; FIG. 5 is a schematic diagram of a circuit for detecting the axial position of the magnetically suspended rotor assembly;

FIG. 6 is a simplified schematic diagram of an axial position controller;

FIG. 7 is a graphic illustration of an axial position control method operating with minimum power; FIG. 8a is a partial view of a cardiac fitting attached to a swollen ventricle;

FIG. 8b is a partial view of a cardiac pacemaker attached to a contracted ventricle;

FIG. 9 is an enlarged partial view of a device for electronically measuring the angle between the two caliper arms shown in FIGS. 8a and 8b;

FIG. 10a is a partial view of a sonomicrometry-based cardiac pacemaker attached to a swollen ventricle;

FIG. 10b is a partial view of a sonomicrometry-based cardiac pacemaker attached to a contracted ventricle; FIG. 11 is a graphical illustration of a method for controlling the flow rate of the blood pump in the operating state; and

FIG. 12 is a graphical illustration of a method for controlling the flow rate of the blood pump in a pulsed mode of operation.

* 022038 9

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Reference is now made to the figures in the drawing in which like reference numbers indicate like parts in all figures, the presently preferred embodiments of a blood pumping device being shown in figures 1a and 1b and being provided with a stator assembly 1 and a rotor assembly 2.

The stator assembly 1 can be provided with an outer stator casing 3, a cochlea housing 4, an end cap 5 and a thin-walled stator casing 6. The stator casing 3, the cochlea casing 4, the end cap 5 and the stator casing 6 are each made of titanium. The stator coating 6 can have a thickness of about 0.005 to 0.015 inch (0.0127 cm to 0.0381 cm) and preferably about 0.010 inch (0.0254 cm). The outer stator cover 3, the cochlea housing 4 and the stator cover 6 may preferably be soldered together to form a hermetically sealed annular stator chamber 54. The stationary magnetic suspension and the motor components can preferably be accommodated in the stator chamber 54.

The rotor assembly 2 may have a relatively large bore 20 which may be the primary blood flow path 20 'through the pump. Preferably, the central bore 20 is 0.50 inch. The rotor assembly 2 may comprise an internal rotor support sleeve 7, a rotor end cap 8, and a thin-walled rotor liner 9. The inner rotor support sleeve 7, the rotor end cap 8 and the rotor lining 9 can also be made of titanium. The rotor coating 9 can have a thickness between about 0.005 and 0.015 inch (0.0127 cm to 0.0381 cm) and can preferably be about 0.010 inch (0.0254 cm). The rotor support sleeve 7, the rotor end cap 8 and the rotor lining 9 can preferably be soldered together to form a hermetically sealed annular rotor chamber 55. The magnetic rotor suspension and the motor components may preferably be housed in the rotor chamber 55. The internal rotor support sleeve 7 may be fabricated with an integral impeller 10, or alternatively, the impeller 10 may be independently fabricated and soldered or bonded to the rotor support sleeve. 7.

The blood-contacting surfaces of the blood pump 10 can be chemically polished or mechanically polished by pumping a polishing slurry through the pump. The blood-contacting surfaces of the 1022038 blood pump can also be coated with a diamond-like carbon film or a ceramic film. Such films prolong the long-term biocompatibility of the surfaces by improving the surface finish and the wear resistance. Companies capable of providing such films include Diamonex Performance Products, Allentown, PA, and Implant Sciences Corporation, Wakefield, MA.

The impeller 10 can be a closed impeller with an upper wall 56 and a lower wall 57 that enclose the impeller blades 36.

The primary blood flow path 20 'can pass through the central bore 20 of the inner rotor support sleeve 7, through the closed impeller 10 to the output flow port 38.

In order to minimize the formation of blood clots, blood flow can be maintained along all blood contacting surfaces. A first return blood flow path 2Γ may extend through the radial magnetic support gap 21, which is the radial magnetic support gap between the stator cover 6 and the rotor cover 9. The radial magnetic support gap 21 may preferably be approximately 0.010 inches. A second blood flow path 59 in the reverse direction can extend through an axial gap 60 disposed between the lower impeller wall 57 and the end cap 5, and through an opening 67 in the lower wall 57 of the impeller 10. The axial gap 60 can preferably be about 0.005 inch (0.0127 cm).

If polarized as shown in Fig. 3a, the radial magnetic repulsive forces are generated between permanent magnetic array 11 mounted in the stator chamber 54, and the permanent magnetic array 12 mounted in the rotor chamber 55. The permanent magnetic array 11 consists of the ring magnets 64, 65 and 66. The ring magnet 64 is radially polarized from its outer to its inner diameter, the ring magnet 66 is radially polarized in the opposite direction from its inner to its outer diameter and the ring magnet 65 is axially polarized from right to left. With the ring magnets 64, 65 and 66 polarized as described, the magnetic flux is directed essentially in the radial direction to the permanent magnetic array 12 mounted in the rotor chamber 55. Similarly, the permanent magnetic array 12 exists. from ring magnets 61, 62 and 63. The ring magnet 63 is radially polarized from its outer to its inner diameter, the ring magnet 61 is radially polarized in the opposite direction from its inner to its outer diameter and the ring magnet 62 is axially polarized from the right to the left. If the ring magnets 61, 62 and 63 are polarized as described, the magnetic flux is directed substantially in the radial direction to the permanent magnetic array 11 mounted in the stator chamber 54.

In the FIG. 1a and 1b, when the rotor assembly 2 is moved radially with respect to the stator assembly 1, the repulsion force between the lower part of the permanent magnetic arrays 11 and 12 will increase while the repulsion force between the upper part of the permanent magnetic arrays 11 and 12 decreases . A net upward force is thus created which tends to return the rotor assembly 2 to a radially aligned position. Similarly, if the rotor assembly 2 is moved radially upwardly relative to the stator assembly 1, the repulsion force between the upper parts of the permanent magnetic arrays 11 and 12 will increase while the repulsion force between the lower parts of the permanent magnetic arrays 11 and 12 decreases. Thus, a net downward force is created which tends to return the rotor assembly 2 to its radially aligned position. The radial repulsion forces described tend to maintain the rotor assembly 2 radially suspended with respect to the stator assembly 1. The ring magnets 61, 62, 63, 64, 65 and 66 are preferably made of a magnetically hard material with a relatively high energy product. such as neodymium iron boron. The ring magnets 61, 62 and 63 can be bonded together to form a permanent magnetic array 12 that can be glued to the rotor support sleeve 7. The ring magnets 64, 65 and 66 can be glued together to form the permanent magnetic array 12 which can be glued to the stator support sleeve 68.

In addition, more than three adjacent ring magnets can be used. For example, as shown in FIG. 3b, the two permanent magnetic arrays 1Γ and 12 'can each be made from five ring magnets glued together. The magnetic array 1Γ comprises the rings 664-668 each with polarities indicated by arrows. Similarly, the magnetic array 12 'includes the rings 659-663 each with a polarity indicated by the arrows. Two separate magnetic flux paths are realized in the composition shown. The rings 664-666 and 659-661 direct the magnetic flux in the same way as previously described for 1022038 12 the rings 64-66 and 61-63 shown in Fig. 3. In addition, the rings 66-668 and 661-663 create a second magnetic flux path that directs the magnetic flux in the same manner as described above with the exception of the opposite direction thereof.

It will be clear that other permanent magnet compositions can be used to implement a radial bearing. Such an alternative embodiment is described by Wasson in US-4,072,370 which is hereby incorporated by reference.

A composition of permanent magnets 13, 14, 15, coils 16, 17 and an associated iron rear element 18 together form a Lorentz force 10 actuator that can be used as an axial bearing for axially supporting the rotor assembly 22. The permanent magnets 13, 14 and 15 cause the magnetic flux 19 to flow radially from the outer surface of the magnet 13, radially through the secondary blood flow path 21, radially through the coil 16, axially through the stationary iron back element 18 of the actuator, radially through the coil 17, radially through the secondary blood flow path 21, radially through the magnet 15, axially through the magnet 14 and radially through the magnet 13. The permanent magnets 13, 14 and 15 are preferably made from a magnetically hard material with a relatively high maximum energy product such as neodymium-iron-boron and may, for example, be attached to the rotor support sleeve 17. The stationary iron rear element 18 of The actuator can preferably be made of a magnetically soft material with a high saturation flux density. Such a material is 48% iron - 48% cobalt -2% vanadium, available under the name HIPERCO® 50A from Carpenter Technology Corporation, Reading PA. The coils 16 and 17 may be made of copper or silver wire and may, for example, be attached to the iron rear element 18 of the stationary actuator, which in turn may be attached to the stator support sleeve 68. It will be appreciated that the coils 16 and 17 can be made from round wire, square wire, rectangular wire or flat tape. If the coils 16 and 17 are energized in such a way that the currents start running in the clock direction in the coil 16 and in the counter-clockwise direction in the coil 17, viewed from the pump inlet 20, a net axial Lorentz -power generated which tends to move the rotor assembly to the right. When the direction of the currents in the coils 16 and 17 is reversed, so that the current flows counter-clockwise through the coil 16 and in the clockwise direction in the coil 17, seen from the pump inlet 20, then a net axial Lorentz force is generated which tends to move the rotor assembly 2 to the left. A Lorentz force actuator as described is preferably used with attractive ferromagnetic actuators because a single Lorentz force actuator is capable of producing bidirectional forces; the output force is a linear function of the input current; the bandwidth is wider; the radial attractive force between the moving and stationary parts of the actuator is relatively low; and the generated force is parallel to the air gap formed between the moving and stationary parts of the actuator.

A permanent magnet 31, anchor windings 32 and an iron rear element 33 work together to form a trenchless and brushless DC motor with a coreless anchor. Such trenchless, coreless motors are known to those skilled in the art and are described in U.S. Pat. No. 4,130,769 which is incorporated herein by reference. A two-pole permanent magnetic ring 31 provides a magnetic flux radially from its north pole 34, through the radial magnetic suspension gap 21, radially through the anchor winding 32, circulating through the iron rear element 33 of the stator, radially through the anchor winding 32 , radially through the radial magnetic suspension gap 21 to the south pole 35 of the permanent magnetic ring 31. Interaction between the axial current passing through the armature winding 32 and the radial magnetic flux produces a moment between the rotor assembly 2 and the stator assembly 1. The permanent magnetic ring 31 can preferably be made of a magnetically hard material with a relatively high magnetic energy product such as neodymium iron boron. Although the permanent magnetic ring 31 can be displaced by a permanent magnetic ring assembly with more than two poles to reduce the dimensions and / or increase the efficiency of the motor. The iron rear assembly 33 of the stator can be fabricated from a stack of magnetically soft lamination rings that preferably have high resistivity and high saturation flux density. Such a material is 48% iron -48% cobalt - 2% vanadium and is supplied under the name HIPERCO® 50A 30 by Carpenter Technology Corporation, Reading PA.

Electrically insulated laminates are used in the iron back element assembly 33 to minimize power losses caused by eddy currents induced by the rotating magnetic field of the permanent magnetic 1022038 14 ring 31. It will be clear that a conventional brushless DC motor with protruding poles can be used instead of the motor described, however, a trenchless and coreless motor may be preferred because in trenchless motors the jerky torque can be eliminated whereby a more evenly quieter operation is possible in comparison with brushless DC motors with protruding poles, and trenchless, coreless motors generally have a greater radial distance between the permanent magnets in the rotor and the rear iron element in the stator resulting in lower radial attractive forces. Radial attraction forces generated by the motor may be undesirable because they tend to counteract the repulsive radial suspension forces generated by the permanently magnetic radial bearing magnets resulting in reduced radial suspension stiffness. Such slotless, brushless and coreless DC motors are supplied by companies such as Electric Indicator Company, Inc., Norwalk CT; Portescap U.S., Inc., Hauppauge NY; Maxon Precision 15 Motors, Inc., Fall River MA and MicroMo Electronics, Inc., Clearwater FL.

A composition of coils 23, 24 and ferromagnetic rings 25, 26 cooperates to form an axial position sensor used to monitor the axial position of the rotor assembly 2 with respect to the stator assembly 1. The two coils 23, 24 can be made from copper wire. A first ferromagnetic ring 25 ensures that the inductance of a first coil 23 is increased and the inductance of the second coil 24 is reduced as it moves to the left. Similarly, the inductance of the first coil 23 is reduced and the inductance of the second coil 24 is increased as the first ferromagnetic ring 25 moves to the right. A second ferromagnetic ring 26 can serve both for magnetic shielding and for increasing the Q of the coils 23, 24. The two ferromagnetic rings 25, 26 can preferably be made from a ferrite material with a high permeability at the excitation frequency of the coils 23, 24. Such a material is MATERIAL-W® available from Magnetics, Division of Spang & Co., Butler PA. A pair of spacer rings 27, 28 can be used to radially locate the two ferromagnetic rings 25, 26.

Tf 04203a 15

As shown in Fig. 2, the impeller 10 rotates within a cochlea 4. A series of impeller blades 36 centrifugally drives the blood around the cochlea passage 37 and out through the outflow port 38.

It is clear that other impeller-cochlear gait compositions can also be developed by those skilled in the art and the invention is therefore not limited to the specific configurations illustrated and described herein.

Fig. 4 schematically illustrates a method for connecting the blood pump 51 to the circulation system. Various cannulas 44, 46, 49, 90 may be required to connect the pump 51 between the left ventricle and the aorta. An opening is made in the top 10 of the left ventricle at position 43 and a cannula 44 guides the blood from the left ventricular cavity to the flow sensor 91. The flow sensor 91 is preferably provided to monitor blood flow through the pump during the low flow phase of the pump if the pump is used to provide a pulsating blood flow. A pump flow rate controller 92 can use the output signal from the flow sensor 91 to adjust the rotation speed of the blood pump so that the forward blood flow through the pump 51 is maintained. Excessive damage to the blood can occur if the blood flow through the pump becomes too low. The blood flow sensor 91 can be an electromagnetic sensor or an ultrasonic sensor. The construction and operation of electromagnetic flow sensors is assumed to be known to those skilled in the art and is described, for example, in U.S. Pat. No. 2,149,847 to Kolin, which is incorporated herein by reference. The construction and operation of ultrasonic flow sensors is assumed to be known to those skilled in the art and described, for example, by Drost in US 4,227,407, which is hereby incorporated by reference. The cannula 90 guides the blood from the flow sensor 91 to the inlet 45 of the pump. Another cannula 46 guides the blood from the outlet 47 of the pump to an in-line valve assembly 48. Alternatively, a solenoid-actuated valve may be used in place of the valve assembly 48. The valve assembly 48 may preferably be arranged to prevent a backflow of blood from the aorta through the pump to the left ventricle in case the blood pump or an associated control system fails. From the exit of the valve assembly 48, a cannula 49 guides the blood to the ascending aorta 50. The pump flow rate controller 92 may also provide an output signal to the cardiac pacer 400/500 for 1022038 16 to adjust the rotation speed of the blood pump such that an excessive expansion or contraction of the ventricle is prevented. Excessive expansion of the ventricle can occur if the pump speed and the resulting flow speed are too low. An excessive contraction of the ventricle can occur if the pump speed and the resulting flow speed are too high. For biventricular heart assistance, a second pump can be connected in a similar manner between the right ventricle and the pulmonary artery.

In some patient applications, it is highly desirable to provide a pulsating flow rate. Although the pumping mechanism shown may function as a continuous pump at randomly varying rates of speed, suitable for use in a patient, it may be desirable to achieve a cyclic or pulsed flow rate. In this way the pump can simulate the varying current that may be desirable for some patients. The flow rate controller 92 as shown in Fig. 4 can be of various construction. It may be located separately from the pump 15 as shown in FIG. 4 or may in fact be integrated within the pump 51. Another placement or other embodiments of the flow rate controller can be used in practicing the invention. The flow rate controller is preferably based on a microprocessor that uses feedback control systems via the flow sensors. In other systems, feedback of the flow is not required and other modulation techniques can be used. Any modulation of the speed control of an electric motor can be applied in the flow speed controller 92. These systems can use an open loop or can use combinations of flow sensors, pressure sensors or both. As shown, the sensor 91 is located externally of the pump housing. In other embodiments, it may be desirable to place the position sensors in or adjacent to the elements that transport the blood or the heart itself. The flow rate controller 92 may include a software programming that controls a different number of sensors such that regulation and modulation occurs at one group of sensors while another group of sensors 30 such as a flow rate sensor or another sensor indicative of the performance in the circular system is used as a limit controller. These sensors can also be stored within the blood pump 51.

1022038 17

An axial position sensing subsystem 141 may have a circuit as shown in Fig. 5. The subsystem 141 may use the ratio between the inductances of the coils 23 and 24 to measure the axial position of the rotor assembly 2. The subsystem 141 can also be provided with an amplitude-stabilized sine wave oscillator 100, which is used to excite the coils 23 and 24 in a half-bridge circuit 101, and a synchronous demodulator 102. Synchronous demodulation is used to detect the relatively low amplitude signals at the output of the half-bridge circuit 101 because synchronous demodulation technically effectively filters out the electrical noise at all frequencies except those centered around the excitation frequency. An oscillator 103 generates a square wave 104 which is used to control the analog switch 105. The output signal 106 from the analog switch 105 is a square wave that alternates between the output voltage 107 of the operational amplifier 108 and ground. A capacitor 109 and a resistor 110 form a high-pass filter that removes the DC offset from the signal 106. A low-pass filter 111 attenuates the higher harmonics in the input signal 112 resulting in a sinusoidal output signal 113. The low-pass filter 111 is of sufficient order and type to attenuate the third harmonic of the input block wave 112 by 40 dB or more. A possible embodiment of the low-pass filter 111 is a fifth-order Butterworth type. A capacitor 114 removes any direct voltage offset from the output 113 of the low-pass filter 111. A sinusoidal alternating voltage 115 is used to excite the half-bridge network 101. A pair of resistors 116, 117 and an operational amplifier 118 form an inverting circuit with a gain of -1. A comparator 119 detects the sign of the sinusoidal excitation signal 115. The output signal 120 from the comparator 119 is used to control an analog switch 121. If the sign of the sine wave 115 is negative, then the output 122 of the analog switch is 121 coupled to the non-inverting sine signal 115. If the sign of the sine 115 is positive then the output 122 of the analog switch 121 is coupled to the inverted sine 123. The output signal 122 is thus the inverted and double-sided rectified representation of the sine shaped excitation signal 115. An operational amplifier 108, a pair of resistors 124, 144 and a capacitor 125 form an integrating differential amplifier. The output signal 107 of the operational amplifier 108 increases if the average 1022038 18 double-sided rectified representation of the sinusoidal excitation signal 115 is smaller than the accurate reference voltage 126 presented. Similarly, the output signal 107 of the operational amplifier 108 decreases if the average double-sided rectified representation of the sinusoidal excitation signal 115 is greater than the accurate reference voltage 126 presented. By the described integrating action, the amplitude of the alternating voltage signal 106 is controlled if necessary for maintaining the average double-sided rectified representation of the sinusoidal excitation signal 115 equal to the accurate reference voltage 126 presented. As previously described, the ratio of the inductances of the coils 23 and 24 is a function of the axial position of the rotor assembly 2 shown in Figs. 1a and 1b. The amplitude of the output signal 127 of the half-bridge circuit 101 formed by the coils 23 and 24 thus varies with the axial position of the rotor assembly 2. A pair of resistors 128, 129 and an operational amplifier 130 form an inverting circuit with a gain -15 factor of -1. The output signal 120 from the comparator 119 is used to control an analog switch 131. If the sign of the sine wave 115 is negative, then the output signal 132 from the analog switch 131 is coupled to the non-inverting output signal 127 from the half-bridge rectifier. 101. If the sign of the sine wave 115 is positive, then the output signal 132 of the analog switch 131 is coupled to the inverted output signal 133 of the half-bridge circuit 101. The output signal 132 is thus the inverted double-sided rectified representation of the output signal 127 of half-bridge circuit 101. A low-pass filter 134 attenuates the a.c. components of the output signal 132. A possible configuration for the low-pass filter 134 is an eighth-order Butterworth type. Various resistors 135, 136, 137 together with an operational amplifier 138 and an accurate reference voltage 126 provide for shifting and resizing the output signal 139 of the low-pass filter 134 as required for the subsequent circuits. Thus, the output signal 140 from the operational amplifier 138 is a representation of the axial position of the rotor assembly 2. Thus, changes in the output signal 140 constitute a measure of the axial movement of the rotor assembly 2. The circuit illustrated in FIG. 5 is only one example of a circuit that can be used to detect changes in the ratio of the inductances of the coils 23 and 24. It will be understood that other acceptable circuits can also be designed by those skilled in the art.

Using the output signal 140 of the axial position observer sub-system 141, an axial position controller 200 shown in FIG. 6 can be used to maintain both the axial position of the rotor at a defined set point for this axial position and the set point for the to adjust the axial position to minimize the power dissipation in the Lorentz power actuator. The axial position controller 200 may have basic circuitry as shown in Fig. 6 including the circuit 201, which maintains the rotor at a set point for a defined axial position and circuit 202 which post-adjusts the axial position set point to a minimum displacement power dissipation in the Lorentz power actuator coils 16, 17. The axial position set point maintenance circuit 201 is composed of the aforementioned axial position sensor subsystem 141, a gain factor and servo compensation circuit 203, a switching power amplifier 204, the Lorentz force actuator 205 and the rotor assembly 2. The axial position sensing subsystem 141 outputs a signal 140 proportional to the axial position 215 of the rotor assembly 2. Different resistors 207, 208, 209 together with a capacitor 210 and an operational amplifier 211 form a gain and advance compensating network 203, with which the gain and the phase of the signal 140 are modified as necessary to prevent unstable oscillation of the rotor assembly 2. Designing such a gain and advance compensation network is considered within the scope of a person skilled in the art. The output voltage 212 of the gain and advance compensating network 203 is input to the switching power amplifier 204. The switching power amplifier 204 outputs a current signal 213 that is proportional to the input voltage 212. The designs of such transconductance switching Amplifiers are known to those skilled in the art. The current signal 213 is applied to the coils 16, 17 of the Lorentz force actuator 205. The Lorentz force actuator 205 produces an axial force 214 proportional to the current current signal 213. The axial force 214 is applied to the rotor assembly 2. The axial position 215 of the rotor assembly 2 changes in response to the presented axial force 214. The overall polarity of the described servo loop 201 is such that the force produced by the Lorentz force actuator counteracts the displacement of the rotor assembly from the set point defined. The skilled artisan will recognize herein that the function of the analog, gain, and servo compensating circuit 203 can be implemented by software running on a microprocessor or by a digital signal processor.

In Fig. 7 the described axial control method for realizing a minimum power is illustrated. The x-axis 300 of the graph represents the axial position of the rotor assembly 2 relative to the stator assembly 1. The y-axis 301 of the graph represents the axial force exerted on the rotor assembly 2. The line 302 represents the inherent axial forces generated by the permanent magnetic arrays 11, 12 for small axial movements of the rotor assembly 2. At point 303 of the graph, the permanent magnets 11, 12 are magnetically aligned and do not generate axial force. The slope of the curve 302 depends on the design of the permanent magnetic arrays 11, 12 and can be between 0.2 lb / 0.001 inch to 1.0 lb / 0.001 inch (35.716 grams / cm to 178.580 grams / cm) . Line 304 of Fig. 7 represents the axial load that is presented to the rotor assembly 2 in the stable state. The axial load 304 in the stable state can be caused by gravity, acceleration, the centrifugal pump impeller, etc. The line 305 of Fig. 7 forms the addition of lines 302 and 304 and represents the net force with respect to the axial position of the rotor assembly 2 when the load 304 is presented in a stable state. The point 306 defines the axial position of the rotor assembly 2 when the stable state load force is eliminated by the axial force produced by the permanent magnets 11, 12. By adjusting the set point from the axial position of the rotor assembly 2 to the axial position defined by point 306, the stable state actuator output force required to maintain the axial set point becomes zero. Because the power dissipated by the Lorentz force actuator is proportional to the square of this output force, the net power dissipated by the actuator is also minimized if the rotor assembly is operated at the axial position defined by point 306. When there is if no load forces are offered for the stable state, then the net power dissipated by the actuator is minimized if the rotor assembly is operated at the axial position defined by point 303.

? O · .. 7 Λ 3 3 1 W ^ V Q.

21

The circuit shown in Fig. 6 can be used to effectively set the axial position set point of the rotor assembly 2 for minimal power dissipation in the Lorentz power actuator 205 using the method described above. The set point of the stable state of the axial position can be controlled by means of the output voltage 216 of the operational amplifier 217 and the resistor 218. The circuit formed by the resistor 219, the capacitor 220 and the operational amplifier 217 inverts and integrates the amplification factor output voltage 212 and advance compensating network 203. The signal 212 is directly proportional to the current flowing through the Lorentz force actuator 10 coils 16.17. If the average voltage of the signal 212 is positive, indicating that a net positive current flows through the actuator coils 16, 17, the output signal 216 of the operational amplifier 217 decreases and the axial position set point of the rotor assembly 2 shifts until the average current through the actuator coils 16, 17 is zero. Similarly, if the average voltage of the signal 212 is negative, indicating that a net negative current flows in the actuator coils 16, 17, the output signal 216 of the operational amplifier 217 will increase and the axial set point of the rotor assembly 2 will shift until the axial current flowing through the coils 16, 17 is zero. The axial position set point for the stable state of the rotor assembly 2 is thus controlled as necessary for a minimum power dissipation in the Lorentz power actuator 205. The skilled person will recognize herein that the function of the analog automatic set point control circuit 202 can be implemented with software that runs on a microprocessor or with a digital signal processor.

The magnetic fields created by the permanent magnets located in the stator assembly 1 and the rotor assembly 2 can change with time, with variations in temperature, and due to the influence of external ferromagnetic objects. If the change in the magnetic fields causes the preferred axial operating point 306 (Fig. 7) to fall outside the axial positional limitation of the rotor assembly 2, contact may occur between the rotor assembly 30 and the cochlea housing 4 or the end cap 5 at points 71 and 72. To prevent such contact between rotor and stator, an axial position control actuator 70 as shown in Fig. 1a can be used. The permanent magnetic array 11, the Lorentz force actuator components 16, 17 and 18 and the motor components 32 and 33 1022038 22 may be attached to the stator support sleeve 68. The stator support sleeve 68 may be axially disposed with respect to the stator sleeve 3, the cochlea housing 4, the end cap 5 and stator cover 6 move. The post-adjustment of the axial position of the axially movable stator support sleeve 68 can therefore be performed using the axial position actuator 70 for post-adjusting the position of the sleeve 68 and thus of the magnet array 11 which is carried within the sleeve 68. An embodiment of the currently preferred axial position control actuator 70 is shown in more detail in Fig. 1b in which the actuator 70 preferably has an output shaft 74 of an engine-gearbox combination 73 attached to the screw 75. The screw 75 matches the thread 76 in the stator support sleeve 68. The motor-gearbox combination 73 can preferably be provided with a stepper motor, although other types of motors such as a DC brush motor or a brushless DC motor can be used. Combinations of a miniature stepper motor and a gearbox are available from companies such as Donovan Micro-Tek, Ine., Simi Valley, CA. A c-15 clamp 77 can be used to axially retain the screw 77. A tooth 78 on the stator support sleeve 68 fits at a slot 79 in the stator cover 3 to limit the rotation of the stator support sleeve 68 within the stator cover 3. The axial position of the stator support sleeve 68 and the magnetic suspension and rotation components engaging on it can be adjusted relative to the stator cover 3, the cochlea housing 4, the end cap 5 and the stator cover 6 by electrically controlling the motor gear 73 By controlling the axial position of the stator support sleeve 68 and the magnetic suspension and rotation components, the preferred operating point 306 (FIG. 7) can be maintained within the axial position limits of the rotor assembly 2 to allow contact between the rotor assembly 2 and the rotor assembly 2. 25 snail shell 4 or the end cap 5 in points 71 and 72. It will be appreciated that additional linear, rotary and impact bearings can be used to improve the accuracy, effectiveness and reliability of the described axial adjustment mechanism. A magnet 80 and hall effect sensors 81 and 82 can be used by a controller to monitor and limit the movement of the stator support sleeve 68. It will be appreciated that other mechanisms can be used to adjust the axial position of the magnetic supporting components. These mechanisms may use electromagnetic linear actuators, piezoelectric actuators, magnetostrictive actuators or actuators based on formed memory alloys.

The blood flow from any blood pump capable of creating a negative inlet pressure must be dynamically adjusted as an adjustment to the blood flow rate in the left ventricle. If too low a flow is produced by the blood pump, the tissues and organs of the body cannot be adequately flowed through and the blood pressure in the left ventricle will increase which could potentially cause excessive pulmonary pressure and congestion. On the other hand, if the flow rate of the blood pump is too high, an excessive negative pressure can be created in the left ventricle and in the pump inlet. Excessive negative blood pressure is undesirable for the following reasons: 1) unacceptable levels of blood damage caused by cavitation, 2) the pump can be damaged by cavitation, 3) the walls of the ventricle can collapse and be damaged and 4) the walls of the ventricle may collapse and block blood flow to the pump. Preferably, the flow rate of the pump is dynamically controlled such that these problems are prevented.

As shown in Figure 4, the blood pump blood flow rate controller 92 may operate the pump such that the flow rate does not cause the ventricle to contract excessively. Preferably, a heart measurement device can provide the flow rate controller 92 with information about the dimensions of the ventricle during normal contraction and swelling. Such a heart measuring device can be an electronic heart fiter, two types of which are illustrated in Figs. 8a-10b.

In Fig. 8a a cross-section through a heart is illustrated comprising a right ventricle 405 and a left ventricle 404 that is maximally swollen by the blood pressure prevailing therein. In Fig. 8b, the left ventricle 404 is partially relieved of pressure. When blood is withdrawn from the left ventricle 404, the radial dimension of the outer surface 418 of the heart is reduced. By dynamically controlling the flow rate of the blood pump, excessive swelling or contraction of the left ventricle is prevented as indicated by the radial dimension of the outer surface of the left ventricle, which allows the average blood pump flow rate to be controlled as an adjustment to the blood velocity in the left ventricle. An embodiment of an electronic heart fitting 400 is shown that can be used to measure the radial dimension of the outer surface 418 of the heart. The heart pacer 400 includes two arms 401 and 402 suitably attached to the outer surface 418 of the heart and hinges around a point that is preferably positioned within a hermetically sealed sheath 403. A measure of the radial dimension of the left ventricle 404 can be achieved by electronically measuring the angle between the compass arms 401, 402. An angle measuring device that can be used to measure the book between the arms 401, 402 is illustrated in Fig. 9. The angle measuring device can preferably be included within a hermetically sealed envelope 403 to protect the internal components from the tissues and fluids of the body. A bellows 407 and end caps 408, 409 may preferably be welded together to form a hermetically sealed chamber. The bellows 407 and the end caps 408, 409 are preferably made of titanium. A hermetic electrical lead-through 410 for which glass or a brazed ceramic insulator 411 can be used can be installed in the titanium end cap 409. The caliper arms 401, 402 can be effectively coupled to a hinge element 415 via the respective end caps 408, 409 and the respective control arms 416, 417. The caliper arm 401, the end cap 408 and the control arm 416 may be made of a single piece of titanium or may be individually constructed and welded or glued together. Similarly, the caliper arm 402, the end cap 409, and the control arm 414 may be made of a single piece of titanium or may be individually constructed and welded or glued together. A hinge 415 limits the movement of the compass arms 401, 402 into an arc within a single plane. Because the caliper arms 401, 402 move due to the contraction or swelling of the left ventricle, the extension 419 moves from the control arm 416 to respectively an eddy current position sensor 417 that may be glued to the control arm 414. The alternating current-based position sensor 417 can be fabricated from a miniature ferrite pot core 412 and a copper coil 413. Such a miniature pot core is available from Siemens Components, Inc., Iselin, NJ. The eddy current sensor coil can be connected between two electrical passages (one of the two passages 410 being illustrated in FIG. 9). As the metal control arm extension 419 moves closer to the eddy current position sensor 417, the effective resistive load of the coil increases causing a reduction of the coil Q (1022038) (Q is defined in the art as the ratio between the reactance and the effective series resistance of a coil). An electronic circuit can be used to measure the changes in the Q of the coil and supplies a signal corresponding to the relative position of the compass arms 401, 402. Such electronic circuits, as described for measuring changes in the Q1 are known from the prior art.

An alternative embodiment of an electronic cardiac adapter 500 is illustrated in FIGS. 10a and 10b. In a similar manner as in Figs. 8a and 8b, Fig. 10 shows a left ventricle that is maximally expanded due to the blood pressure present therein and Fig. 10b shows a left ventricle that is partially relieved of pressure. The cardiac pacer 500 may have a pair of arms 501, 502 suitably attached to the outer surface of the heart and pivot about a point that may preferably be located within a hermetically sealed envelope 503. A measure of the radial dimension of the left ventricle can be obtained by measuring the time it takes for an ultrasonic pulse to propagate from a sonomicrometer transducer 504 on one caliper arm 501 to an opposite sonomicrometer transducer 505 on the other caliper arm 502. Sonomicrometer transducers suitable for such a heart fitting are available from companies such as Triton Technology, Ine., San Diego CA and Etalon Inc., Lebanon IN. The details of sonomicrometry are known to those skilled in the art. It will be appreciated that other suitable methods for measuring the relative swelling and contraction of the ventricles, such as impedance and conductance measurements of the ventricle, may be used by those skilled in the art and that the invention is not limited to the methods particularly described herein.

A way to implement the flow rate controller 92 to control the flow rate of the blood pump to prevent excessive expansion or contraction of the left ventricle is graphically illustrated in FIG. 11. The x-axis 600 represents time. The line 601 represents a radial dimension setpoint of the left ventricle that is defined upon initial implantation of the system and can be periodically updated in a non-invasive manner using ultrasonic transmission. The curve 602 represents the radial dimension of the left ventricle as observed by one of the previously described electronic cardiac passers 400, 500. The curve 603 represents the angular velocity of the described centrifugal pump impeller. The flow rate of the centrifugal pump described varies with the angular velocity of its impeller. If the observed radial dimension is greater than the radial dimension set point as illustrated by point 604, the angular speed of the centrifugal pump may increase as illustrated by point 5605. The increased angular speed of the centrifugal pump causes its flow speed to increase and that blood is removed more quickly from the left ventricle, which in turn causes the radial dimension of the left ventricle to be reduced toward the radial dimension set point 601. Similarly, the sensed radial dimension is smaller than the radial dimension radial dimension set point 10 as illustrated by point 606 then the angular velocity of the centrifugal pump may be lower than illustrated by point 607. The decreased angular velocity of the centrifugal pump causes the flow velocity to decrease and reduces the speed at which the blood from the left ventricle is removed, which in turn causes the radial dim measurements of the left ventricle increase in the direction of the radial dimension set point 601.

Another way to implement the flow rate controller 92 such that the flow rate of the blood pump is controlled to prevent excessive contraction or expansion of the left ventricle and also to create a pulsating blood flow is graphically illustrated in Fig. 12. A pulsing flow -20 velocity is much closer to the characteristics of the blood flow characteristics of a natural heart. In FIG. 12, the x-axis 608 represents time. The lines 609 and 610 respectively represent the upper and lower radial dimension set points of the ventricle that were defined when the system was initially implanted and that can be periodically operated in a non-invasive manner using ultrasonic transmission. The curve 611 represents the radial dimensions of the left ventricle as observed by one of the previously described electronic cardiac passers 400, 500. The curve 612 represents the angular velocity of the described centrifugal pump impeller. The angular velocity of the centrifugal pump can be increased periodically as indicated during the time period 613. The increased angular velocity of the pump during the period 613 causes the blood to be removed from the heart more rapidly, which in turn causes the radial dimensions of the left ventricle are reduced in the direction of the lower radial dimension set point 610. The reduced angular velocity of the 1022038 27 pump during the time period 615 ensures that the speed at which the blood is removed from the left ventricle is reduced, which in turn causes the radial dimensions of the left ventricle to increase toward the upper radial dimension set point 609. The angular velocity of the centrifugal pump 5 may be increased again as indicated during the time period 616 once the observed radial dimension of the left ventricle becomes practically the same as the upper radial dime nsie setpoint 609. The described pulsating pump flow rate serves to mimic the blood flow characteristics of the natural heart. It may be desirable to maintain a forward blood flow through the pump to prevent blood damage caused by constant shearing of blood pools in the vicinity of the impeller blades. A flow sensor 91 in FIG. 4 can be used to control the pump speed during the time period 615 in FIG. 12 such that adequate forward blood flow in the pump is maintained.

Although certain exemplary embodiments of the invention have been described in detail, it will be apparent to those skilled in the art that various modifications of the details thereof may be developed within the scope of this specification. Certain embodiments described herein are intended for illustrative purposes only and are not limitative of the scope of the invention which is apparent from the full width of the following claims and all possible embodiments thereof.

"022038

Claims (4)

A blood pumping device implantable in a patient comprising: a. A stator member; 5 b. a rotor member positioned adjacent to the stator member for the purpose of rotation relative to the stator member; c. a magnetic suspension with a first magnetic array supported by the stator member and a second permanent magnetic array supported by the rotor member, said first and second arrays of permanent magnets cooperating to radially magnetically support said rotor member such that a annular magnetic suspension gap is created between the rotor element and the stator element; d. wherein each of the first and second permanent magnetic arrays consists of at least three adjacent permanent magnetic rings, wherein even-numbered of the at least three adjacent permanent magnetic rings are polarized in a direction perpendicular to a polarization direction of odd-numbered of the at least three adjacent permanent magnetic rings magnetic rings, and wherein the odd-numbered are polarized in a direction parallel to each other such that magnetic flux between the first and second permanent magnetic arrays is radially directed, and wherein at least one of the first and second permanent magnetic arrays is wholly by permanent magnets formed, and e. a magnetic drive with a stator motor part supported by the stator member and a rotor motor part supported by the rotor member, the stator motor part and the rotor motor part cooperating to magnetically rotate the rotor element relative to the stator element to pump blood by the stator and rotor pump parts. 2. Blood pumping device according to claim 1, further characterized in that: a. A first ring of the three adjacent rings is polarized in a first direction; b. wherein a second ring adjacent to the first ring is polarized in a second direction perpendicular to the first direction; and 1022038 c. wherein a third ring adjacent to the second ring on a side opposite to that of the first ring is polarized in a third direction parallel to but opposite to the first direction such that the magnetic flux is directed radially through the at least three adjacent rings between the first and 5 second permanent magnetic arrays. A blood pump device according to claim 1, further characterized in that: a. The at least three adjacent permanent magnetic rings consist of five adjacent permanent magnetic rings; 10 b. a first ring of said five adjacent rings polarized in a first direction; c. a second ring positioned adjacent to the first ring is polarized in a second direction perpendicular to said first direction; d. a third ring is centrally positioned adjacent to the second ring 15 and is polarized in a third direction parallel to but opposite to the first direction; e. a fourth ring is adjacent to the third ring on a side opposite to that of the second ring wherein the fourth ring is polarized in a fourth direction parallel to but opposite to the second direction; 20 f. a fifth ring is polarized in the first direction; g. wherein a first magnetic flux path is created by said first, second and third adjacent rings and the first magnetic flux path is radially directed between the first and second arrays of permanent magnets; and H. a second magnetic flux path is created by the third, fourth and fifth adjacent rings and the second magnetic flux path is radially directed between the first and second arrays of permanent magnets. The blood pumping device of claim 3, wherein the first and second magnetic flux paths are radially polarized in opposite radial directions. 30 * 022038