CA2265984A1 - Enhanced transcutaneous recharging system for battery powered implantable medical device - Google Patents

Abstract

An improved transcutaneous energy transmission device is disclosed for charging rechargeable batteries in an implantable medical device and to minimize peak temperature rises in the implantable device. A current with a sinusoidal waveform is applied to a resonant circuit comprising a primary coil and a capacitor. Current is induced in a secondary coil attached to the implantable medical device. Two solid state switches are used to generate the sinusoidal waveform by alternately switching on and off input voltage to the resonant circuit. The present invention charges the batteries using a charging protocol that either reduces instantaneous charging current or duty cycle of a fixed charging current as the charge level in the battery increases. Peak temperature rises are less while delivering comparable electrical charge to the battery than for prior charging systems. A controller preferably is constructed as a pulse width modulation device with enable and reference voltage features to effectuate variable duty cycle control of the current level applied to the primary coil. An alignment indicator also is provided to insure proper alignment between the energy transmission device and the implantable medical device. The implantable device maximizes transcutaneous energy transmission at different current levels by placing different capacitors into the circuit depending on the magnitude of charging current.
Peak temperature rises are reduced while delivering comparable electrical energy charge to the battery than for prior charging systems.

Description

CA 0226~984 1999-03-17 Description Enhanced Transcutaneous Recharging Svstem for Batterv Powered Implantable Medical Device Technical Field The present invention relates generally to a power source for an implantable medical device.
More particularly, the present invention relates to an external energy transmission device for recharging batteries inside an implantable medical device. Still more particularly, the present invention relates to a charging device for remotely recharging a battery in an implanted medical device. The battery may be of the type disclosed in commonly assigned U.S. P:~tent No. 5,411,537, issued May 2, 1995. entitled "Rechargeable Biomedical Battery Powered Devices With Recharging and Control System Theret'or".
Background Art Currently. battery operable implantable medical devices principally comprise cardiac pacemakers, but also may include heart assist systems. drug infusion and dispensing systems,~
defibrillators, nerve and bone growth stimulators, organ stim~ tors. pain suppressers and implanted sensors, to name a few. The basic cardiac pacemaker generally comprises an electrode, in contact with the heart. that connects by a fle,Yible lead to a pulse generator. The pulse generator includes a microelectronics package. which implements the pacemaker functions, and a power source for supplying operating power to the microelectronics package and other peripheral devices and 'O components. A 'fixed rate" pacem:lker continuously provides timed pulses to the heart. irrespective of proper beating, while a demand inhibited pacPm~ker provides pulses only when the heart fails to deliver a natural pulse. Depending upon the various sensed events, the pacemaker stimlll~tes the right atrium. the right ventricle, or both chambers of the heart in succession. The pacemakers in current use incorporate circuits and antennae to communic~re noninvasively with external programming transceivers. Most pacemakers currently used are of the demand inhibited type, and are programmable.
Early pacem~kers and defibrillators typically were powered by disposable primary zinc-mercuric oxide cells. Although the popularity of this power system persisted for many years, the system suffered from high self-discharge and hydrogen gas evolution. Several mech:lnicm~
contributed to battery failure, most of which were related to cell ch~omi~try. In addition, the open-circuit voltage of each cell was only 1.5V, with several cells connected in series to obtain the voltage required for pacing. Furthermore, because of gas evolution, these prior art pa~em~ker could not be hermetically sealed, and had to be encapsulated in heavy epoxy. In 1970, the average life of the REP LACEMENT

,, ... . , .. . ... , . ~ .......... . . . .....

CA 0226~984 l999-03-l7 W O 98/11942 PCT~US97/16319 pulse generator was only about two years, and about 80 percent of explants were nPcessit,qted by battery failure.
Because of such limitations, many other power generation and power storage devices have been considered as possible alternates. Research and development efforts have focused primarily on rhPmi~.ql batteries, nuclear batteries, and l~ch~lg~able batteries. Additional developmental efforts were directed at dividing the pacing system into two packages, with a power pack located outside the patient's body for tr,qn.~mhting electrical signals through wires to a passive receiver implanted in the body. Cardiac p~emqkPrs based on rechargeable nickel-cq~millm and zinc-mercuric systems also were developed. Examples of such devices are disclosed in U.S. Patent Nos. 3,454,012;
0 3,824,12g; 3,867,950; and 4,014,346. These rechargeable p,q~ernqkPrs incorporated a charging circuit which typically was ene~ d by electromiqgnPtic induction from a device external to the body. The electrom,qgnPti~ induction produced a current in the pacemaker's charging circuit which was converted to a direct current (DC) voltage for charging the battery. Although this system was incorporated in many cardiac pq~PmqkPrs, it was unpopular among patients and physicians because frequent recharging was usually nPcecsqry (somPtimPs on a weekly basis), and because the nickel-cfl~lminm system suffered from memory effects which reduced the battery capacity exponentially after each recharge. In ,q,~l~lition, the specific energy density of both types of rechargeable batteries was poor, cell voltage was low, the state-of-charge condition was difficult to ascertain, and hydrogen gas liberated during overcharge was not properly scavenged either through a recombination reaction, or hydrogen getters.
Charging nickel-cq 1millm cells and zinc-mercuric oxide cells is problematic. Cells of these types have a relatively flat voltage versus time curve during the charging process. The flat slope of the voltage time curve during charging provides little resolution to ascertain accurately the in.~t,qnt,qnPous percentage of full charge. Accordingly, these cells (especially nickel-ca~mi-lm), provide a poor inf1icqtion of the charge level. Additionally, overcharged nickel-c~-lmillm cells liberate oxygen exothermically at the nickel which migrates to the cq~lminm electrode and recombines to form ca-lmil~m hydroxide. In some situations, particularly during an overcharge condition, the rate of oxygen evolution is higher than the rate of oxygen r~coll~billdlion. This may lead to an excessive internal pressure, forcing the cell to vent the excess gas. The overcharge reaction also heats the cell which in turn lowers cell voltage. Therefore, charging is t~ r.d when the battery voltage begins to decrease, thus in-lir,qting the beginning of an overcharge condition.
Other techniques for controlling the charging operation have been employed. For example, U.S. Patent No. 3,775,661 teaches that the pres~ule build-up internally can be sensed by a diaphragm that is external to the battery. As the pressure within the cell casing increases, the CA 0226~984 1999-03-17 diaphragm is flexed to actuate an associated mPrh:lni~l switch located in the battery charging circuit.
The closure of the switch dee.lelgi~es the charger when the battery internal pressure in~iir~Ps a fully charged state.
In somewhat similar fashion, U.S. Patent No. 4,275,739 employs a diaphragm internal to the cell. Deflection of the diaphragm during an episode of incre~sing internal pressure indicates the cell has reached a full charge. Other examples of systems which control charge operation are U.S.
Patent Nos. 3,824,129; 3,888,260; 3,942,535; and 4,082,097.
Today, systems employing nickel-c~millm cells control battery charging use a variety of different charging techniques. Common parameters for ascertaining the end-of-charge condition include maximum voltage, maximum time, maximum temperature, a reduc~ion in cell voltage with respect to time, (dV/dt,) change in temperature ~T, and increase in temperature with respect to time (dT/dt). Details of these end-of-charge indicators can be found in EDN, Mav 13, 1993.
Both zinc-mercuric o,Yide and nickel-~a~millm cells suffer from additional problems such as memory effect and high self-discharge. Fast recharge often is implemented by charging the battery to some preselected voltage with a relatively high current followed by a smaller trickle charge. It is known that nickel-cadmium batteries that are fast charged cannot be charged to 100 percent of rated cell capacity. This loss of capacity is called the memory effect. Each time the battery is discharged at some low current rate, and then recharged at a higher current rate, a loss in capacity results. The capacity loss of each recharge cycle accllmlll~es. Cells affected by the memory etfect then have to be fully discharged and "reconditioned" before full capacity can be recovered. Because of this loss of capacity and high self-discharge, p~rPm~k~rs with cells that suffered a memory effect had to be frequently recharged, sometimes on a weekly basis. In theory, rechargeable battery powered p~-~P~ kers were designed for a 10 year usable life. Battery chemistry problems. however, reduced the device's usable life to two or three years, the same lifetime as that of the earlier devices that employed disposable primary cells. As a result of the inherent limitations in zinc-mercuric oYide and nickel-c:lllmillm battery cells, the assignee of the present invention has suggested the use of lithium batteries. See U.S. Patent No. 5,411,53~, issued May 2, 1995.

An additional problem arises with respect to recharging an implanted device's rechargeable battery. Due to the health risks and costs associated with surgical intervention, it is highly undesirable to perform surgery to access the implanted device and recharge the batteries. Some noninvasive methods for l~cha~ g implanted batteries are disclosed in the prior art. Cert~in patents disclose a technique for delivering electrical energy through the skin between a transcutaneous energy transfer device and an implanted medical device. For example, U.S. Patent No. 5,350,413 REPLACEMENT

CA 0226~984 1999-03-17 discloses a tr:lngrut~nPous energy transfer device Co~ isillg an external primary coil located on or near the skin, and a secondary coil for illl~,ld..lalion under the skin. The primary and secon-l~ry coils form a transformer so that electrical current in the primary induces current in the secondary coil.
An approximation to a half-wave sinusoidal voltage is developed across the primary winding by the action of a field effect lla~ lOl (FET) ~wil~;llillg a direct current (DC) voltage source across a tuning capacitor. Because of the construction of the energy llA~ c~ion device, high frequency harmonic c~Jlllpolle-lL~ are present in the waveform. These high frequency cu~ one~ induce eddy currents in the housing of the implanted device. The t~ eral~lre of the housing hlCI eases in response to the eddy currents, and can also increase in response to the elevated l~ alui~ of the battery during charging. A rise in l~lllpel~lule of the outer surface of the housing may be detrimental to operation of the medical device and harmful to ~Illluu~ lg body tissue. Industry standards suggest maximum allowable temperature rises. l imi~ing a temperature rise (i.e., peak temperature) is desirable to minimi7P the harmful effects on surrounding body tissues. Prior art systems, however, for the most part have not examined methods for reducing temperatures rises.
The prior art rechalgi.. g devices suffered seveMl drawbacks. First, many prior art systems either did not have a feature for properly aligning the external device on the patient's body over the irnplanted device or had such a feature but required additional circuitry (and therefore increased cost, volume, and power drain) and lacked sufficient accuracy. In addition, some prior art devices have an :~lignml~n~ mPrh~nicm which requires the recharging device to be turned off while ~lignmPnt is measured. Also, the charging distance between the external and internal devices was limited requiring the implanted medical device to be located relatively close to the skin. Further, some systems inrlnded on ill.l)lalllable device COl~i~Li~g of two separately housed parts, an electronics unit and a receiving coil, which ill~;reased the difficulty and risk associated with surgical implantation as well as reducing the quality of the herrnetic seal.
It would be desirable to provide a battery charging system that overcomes these and other problems associated with rechargeable implantable devices. In particular, it would be desirable to construct a battery charging device which can efficiently charge a battery in an implanted medical device at a relatively high power transfer rate while reducing the peak Lt~ Jel~lUle generated by the device. Similarly, it would be desirable to develop an ~lignmPnt mech~ni.~m that is located in the recharging device, which requires no extra components in the irnplanted device besides those components needed for charging. It would further be advantageous to develop an energy tr:lncmi~gion system that "~ ;"~ s the size of the receiving coil and perrnits the coil to be located inside the housing of the implantable device. Despite the apparent advantages of such a system, to date no such system has been developed.

CA 0226~984 1999-03-17 WO 98/11942 PCT/US97tl6319 Disclosure of the Inventi- n The present invention overcomes the shortcomings and deficiencies of the prior art by providing a trAnccutAnPous energy llA~ sion system with two external solid state switches that produce a sllhs~ntiqlly si,l.lsoilal power waveform. The solid state switches connect a regulated DC
voltage across an inductor and capacitor resonant circuit. The inductor forms a primary coil of a l,d~ro"~ in which current is induced in a secondary coil attached to an implanted medical device.
The implanted medical device receives the induced current for charging leclla.~able batteries. Two different charging protocols are implPmPntPd to minimi7e the peak t~lllpelaLule produced on the outer surface of the housing of the implanted device housing or can.
Using a first charging protocol, an initial relatively high charging current is generated by an external charging device, followed by a lower charging current. The present invention includes a primary current control circuit that provides control signals to an inverter. Based upon the status of the control signals, the invention produces charging current at either a high or low level to provide efficient charging without an excessive t~ lalule rise in the implanted device.
Charging efficiency is maximized through the use of two dirrere"L resonant circuit configurations in the implanted medical device. An antom~fir ~lviLcl~ing mrrhAnicm is used to switch between the two resonant circuit configurations based on the m~gnitllde of charging current. In one configuration, a capacitor connected in parallel across the secondary coil is used for relatively low levels of charging current, whereas a series connected capacitor is used for higher charging current levels. A current sensor provides an input to a switch drive that activates switches to connect one or the other of the capacitors to the charging circuit based upon the m:lEnitn(l~P. of the charging current.
In a second charging protocol, the trAnccutAnPous energy Ll~ inn device produces a relatively high charging current to the battery, but is periodically illl~llu,uled by periods without any 25 - ~,hdl~ g current. The resulting duty cycle of the charging current is adjustable to allow for different levels of average charging current to the battery. An effective current step is thus generated by reducing the duty cycle of the charging current from an initial high level to a lower level. The initial charging duty cycle preferably is 100% (constant DC current). Because only one relatively high charging current is produced, albeit illlell~lpted by periods of no current, only the series connected capacitor is used and no automatic switching mPc.h:.ni.cm and parallel capacitor are nPcesc~ry ~
The coils of the external energy Lldl~ ion device and the implanted medical device must be properly aligned for efficient energy Llall~lllission. Accordingly, an Alignm~nt circuit and indicator are provided to indicate whether the coils are properly aligned The AlignmPnt circuit . .

CA 0226~984 1999-03-17 continuously senses current in the primary coil to determine whether the angular and lateral ~lignmPnt is optimal by sensing a peak DC current. A visual and/or audible signal is provided only when the charging coil is sl-hst~nti~lly in ~lignment with the receiving coil in the implanted device thereby in~lic~fing proper ~lignment Brief Descr~tion of the Dl~wi.~.~
Other objects and advantages of the invention will become apparent upon reading the following detailed description and upon reference to the accompanying drawings in which:
Figure 1 is a drawing showing the external l~ ;u(~ ous energy tr~n~mi~ion device and implantable medical device of the present invention positioned to charge the batteries in the implantable device;
Figure 2 is a sçhem~tir block diagram of the plerelled circuit i.,.~l; "P~ ion of the present invention;
Figure 3 is a ~l~phical illustration depicting the telll~Je-dtule rise which occurs during battery charging for constant current prior art chargers relative to the charging protocols of the present invention;
Figure 4A shows a charging protocol for a typical prior art charging system;
Figure 4B shows a ~lefelled charging protocol for use with the invention of Figure 2;
Figure 4C shows an alternative charging protocol for use with the invention of Figure 2;
Figure 5 is a s~hem~tic block diagram providing additional details regarding the inverter, PWM controller, and ~lignmPnt indicator of the ~ n~ nPous energy transmission device shown in Figure 2;
Figure 6 is a timing diagram depicting the voltage at switches SW1 and SW2 in Figure 5;
Figure 7 is timing diagram depicting the use of a voltage reference to the PWM controller of Figure 5 to change the m~gnitllde of the primary coil current;
Figure 8 is an electrical srh~om~tic diagram showing additional details of the inverter circuit of Figure 5;
Figure 9 is an electrical srhPm~tie diagram of an amplifier constructed in accordance with a preferred embodiment of Figure 5;
Figure 10 is an electrical sch~m~tic diagram showing the preferred peak detector circuitry of Figure 5;
Figure 11 is an electrical s. h~ ir diagram depicting the ~;lirrelellce amplifier and associated circuitry of Figure 5;

CA 0226~984 1999-03-17 Figure 12 is an electrical schPm~tir diagram of the comparator and LEn circuit of Figure 5;
Figure 13 shows a cu-l-pa~isoll of charging efficiency versus current m~gnitl--le for different tuning capacitor configurations; and Figure 14 is an electrical sCllpm~tic of for the implantable device in charging circuitry accordance with the prerel.~d embodiment.
While the present invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the dl~willgs and detailed description thereof are not intended to limit the invention to the particular form disclosed, but on the contrary, the intention is to cover all modifications, equivalents and alternatives falling within the spirit and scope of the present invention as defined by the appended claims Best Mode for Carryin~ Out the Invention Referring now to Figure 1, the l~ c~ nPous recharging system of the present invention includes a L~ Pous energy tr~ ;on (TET) device 50 and an impl~ntPd medical device 14.
In Figure 1, the medical device 14 is shown implanted beneath the skin in the chest or pectoral region of the patient, as might be the case with a defibrillator device, with external TET device 50 positioned on, or near, the surface of the skin and placed proximally to the implanted device 14.
One skilled in the art will understand, however, that the energy tran~mi.c.~inn device 50 may be used to charge any implanted medical device, wherever located. In accordance with the preferred embodiment, the implanted medical device 14 is housed in a housing or "can" made of titanium or stainless steel. Although the energy tr~n~mi~.~ion device 50 is shown with a generally rectangular configuration it should be understood that the energy tr~n~mi~ion device may take any desired shape. Power to the TET device 50 is provided via cord 3 from an external power source such as a 120 VAC outlet on batteries (Figure 2). An indicator 131 illnmin~tPc when TET device 50 is correctly aligned with the implanted device 14 for maximum charging efficiency.
Referring now to Figure 2, TET device 50 of the tr~n~c lt~nP.ous recha~ ,g system generally Colll~ CS a AC/DC converter 5 and an inverter 20 connecting via conductors 6. An alignment ~ indicator 40 also cormects to the inverter 20 to receive signals from inverter 20 when the TET device 50 is properly aligned with respect to the implanted medical device 14 for m~imnm efficiency. A
pulse width modulation (PWM) controller 200 controls the output power level of inverter 20 (and thus the charging current). The PWM controller 200 can also periodically hlLe~ L the inverter's charging current, thus producing a duty-cycled charging current. The TET device 50 COlll~uOIIelll~
preferably are housed in a single enclosure 51. In the preferred embodiment, an alternating current CA 0226~984 1999-03-17 is converted substantially to a DC voltage by AC/DC converter 5 which also regulates the DC
voltage at a level appropriate for tr~n~cu~:ln~ous energy ll,~n~",i.c~ion. AC/DC converter 5 also provides operational power to PWM controller 200 over conductor 201~ One of ordinary skill in the art will recognize that a mllltinlfle of Icnown circuit implem~ont~tions are possible for AC/DC
converter 5, and thus the present inveMion should not be limited to any particular embodiment for AC/DC converter 5.
The regulated DC voltage output signal of the AC/DC converter 5 is tr~n~mi~ted to the inverter 20, which converts the regulated DC voltage output to a sinusoidal current that flows through primary coil 9. Electrical current in primary coil 9 electromagnetically induces a corresponding current in a secondary coil 10 which preferably is located in the implanted medical device 14. Alternatively, secondary coil 10 may be external to housing 15 of the implanted device 14, but electrically coupled to the implantable device 14. In either case, the electrical energy of primary coil 9 couples tr~n~cut:-n~ously between primary and secondary coils through the patient's skin 100.
Referring still to Figure 2, the implanted device 14 preferably includes a rectifier 12, battery 13 and a resonant circuit comprising secondary coil 10, parallel capacitor 410, and series capacitor 420. The resonant circuit in the implanted device 14 preferably is tuned to the frequency of the AC
current in the primary coil 9 Capacitors 410 and 420 are both removable from the circuit by switches 411 and 421. respectively. A switch drive logic 300 controls the state (open or closed) of switches 411 and 421 via lines 301 and 302 depending on the amount of curreM through battery 13, as indicated by current sensor 270.
The rectifier 12 converts the sinusoidal voltage received by the secondary coil 10 to a DC
voltage for charging the battery 13. The rechargeable battery 13 preferably comprises any of a number of different lithium chemistries, as disclosed in detail in commomy assigned U. S . Patent No.
5,411,537. One of ordinary sKill in the art, however, will recognize that the presen~ invention may also be used to recharge other types of batteries, as desired.

When an alternating magnetic flux passes through a metal plate, eddy current is generated in the metal plate. The m~gni~lde of eddy current is a function of frequency and m:lgnitl-dP of the m~gn~ti~ flux. Eddy currents cause temperature increases. For implantable devices such as defibrillators, eddy currents are induced on the can and metal casings of components internal to the implantable device. A high charging current, therefore, creates large ~ ,lp.,.dture rises. The present invention advantageously manages the level of charging current over time so as to minimi7e peak REPll~CEMENT

.

CA 0226~984 1999-03-17 WO 98/11942 PCT/US97tl6319 invention advantageously manages the level of charging current over time so as to ~ i7~ peak le~ elalul~ rises without reducing the amount of charging energy delivered to the implantable device.
For typical prior art charging systems, which often implement constant current charging S protocols, the temperature rise ~ les~ d by curve I in Figure 3 is exemplary. The present invention is capable of delivering the same amount of energy to the battery in the same amount of time as prior systems, but can do so in such a manner that the peak L~ lalule of the battery is less than it would be for prior systems, Curve II exemplifies the l~ e~aLul~ rise response of the present invention. Significantly, the peak temperature produced by the present invention is less than the peak temperature of the prior art charging system represented by curve I, as reflected at point P.
Prior art recharging systems often charge the battery at a constant current as exemplified in Figure 4A. As shown in Figure 4A, a prior art constant current IPA is delivered to the battery for the entire charging period. To deliver the same amount of energy to the battery in the same period of time, the recharging system of the present invention preferably initially delivers to the battery a charging current I, (which is higher than IPA) for a first predetermined period of time, followed by a lower current I2, which is lower than IPA for the rem~inder of the charge cycle as shown in Figure 4B. The energy delivered and time required using the recharging system of the present invention embodied in the preferred protocols of Figures 4B and 4C are s~-bst~nti~lly the same as the energy delivery and time requirement for the prior art charging protocol of Figure 4A. However, the peak temperature rise of the implantable device can 15 is less using the protocol of Figure 4B, referred to in this description as the "current step" protocol.
An alternative method for achieving the same result is shown in Figure 4C. Using this protocol, a charging current I3, which is higher than IPA~ is delivered to the battery with hllellllillenl periods with no charging current. Because of the periodic intervals without current, the effective charging current per unit time is substantially less than the peak current I3. Moreover, the average current delivered to the battery is a direct function of I3 and the duty cycle of the charge current waveform. The average current, therefore, is higher during the charging period in which a higher duty cycle is used as in~ t~d in interval H in Figure 4C. As the duty cycle is reduced during internal L, the result is a lower average charging current. The charging protocol of Figure 4C is hele.llarler referred to as the "duty cycle" charging protocol. The resulting effect of this duty cycle charging protocol is a step-down in average charging current.
Consi~lelll with the pl~rel~cd embo-lim~nt a duty cycle of 100% may be implent~nt~d during the H interval of the charging protocol of Figure 4C. In other words, the initial high charging current phase of the protocol of Figures 4C may be identical to that of Figure 4B, but where the CA 0226~984 1999-03-17 protocol of Figure 4B reduces the current m~gnit~ P to 12, the protocol of Figure 4C m:~int lin.C the same high current m~gni~lde while duty cycling the current waveform to achieve a lower average charging current.
Figures 5-8 show ~l~r~llcd h~ ions of the circuitry employed to achieve the charging protocols shown in Figures 4B and 4C. Referring initially to Figure 5, the PWM controller 200, inverter 20, and ~lignmPnt indicator 40 of TET 50 are shown in more detail. In accordance with the preferred embodiment, the inverter 20 includes a pair of switches 21, 22, a pair of capacitors 26, 29, and a tuning capacitor 25. The PWM controller 200 preferably includes a pulse controller 231, a pulse generator 232, an RC oscillator 233, resistor 233R, and capacitor 233C. One of ordinary skill in the art will recognize other circuit impl.omPn~:ltions are possible for PWM
controller 200. The PWM controller 200 preferably includes functions such as dual output capabilities and high source and sink current.
High frequency h~rmc ni~ content in the current con~l~lrtPd through primary coil 9 will induce eddy currents in the housing or can 15 of the implantable medical device 14 causing a detrimental increase in can te~ dt~re. In addition to the benefits provided by employing the preferred charging protocols of Figures 4B and 4C, the present invention advantageously minimizes the increase in can temperature ~\T by geneldli-lg a charging current signal having a sllhst~nli~lly full sinusoidal waveform with little harmonic content. This sinusoidal charging current signal is transcutaneously I~A~ d to the implanted medical device 14 to charge the battery 13. To generate the desired ~ t~ical sinusoidal waveform, the inverter 20 includes two switches, 21 and 22 (SWl and SW2, respectively). Preferably these switches 21, 22 are solid state devices and, as shown in Figure 8, may be irnplPmPnt~ d with metal oxide field effect transistors (MOSFET's) 21 and 22. As shown in Figures 5 and 6, the output of pulse ~ lol 232 turns switches 21 and 22 on and off alternately such that only one switch is "on" (i.e., con~ cting electricity) at any given time. As shown in Figure 6, a short time period D (for example, 2 mieloseconds) is provided after one switch turns off and before the other switch turns on. This "dead time" between activation of switches 21, 22 insures that the switches are not on simlllt~n~ously which may cause a short circuit condition between the voltage input terminal V", and ground. The dead time between swit~;lling off one switch and turning on the other preferably is modified to control the charging current applied to the batteries, COnSIDI~III with the current step protocol of Figure 4B, and as described more fully below.
Increasing the time when both switches 21, 22 are off results in a decrease in the power supplied to the primary charging coil 9.
Switches 21 and 22 preferably are turned on for the same amount of time each cycle to produce a symmetrical voltage waveform across junctions 30 and 31. Capacitors 26 and 29, which CA 0226~984 1999-03-17 preferably have i~lPnric~l values, form a voltage divider network. Tuning capacitor 25 connects between the common connection point for capacitors 26 and 29 (junction 30) and terminal 32 of l~dl~rulmer 9. It should be noted that this duty cycle is not the same duty cycle described in Figure 4C. The implempn~ on of the duty cycle protocol of Figure 4C is described below.In order to ~ the eddy current induced in the housing or can 15 of implantable device 14, the operational (or carrier) frequency of the PWM controller 200 preferably is set at 5 KHz, but it may be set to operate within a range of 1 KHz and 40 KHz. Tuning capacitor 25 is selected to generate the desired current amplitude with the primary coil 9 leakage inrlllr.t~nre so that a sinusoidal all~ d~hlg current waveform flows through the primary coil 9 with little high frequency harmonics.
Through proper selectiûn of the value of capacitor 25, the natural resonant frequency of the resunalll circuit formed by primary coil 9 and capacitor 25 can be controlled to be slightly less than the operational frequency in order to achieve the zero-voltage turn-on of both switches 21 and 22.
In general, the inverter 20 produces a purely sinusoidal transfer current waveform between coils 9 and 10 using a resonant circuit comprising the leakage in~ ct~nre of primary coil 9 and tuning capacitor 25. Resonance is continuously m~int~inP.d by alternately activating switches 21 and 22. The present invention can provide a wide range of charging current from 0 to 1 amperes and charging voltage from 0 to 20 V. The distance between coil 9 and coil 10 preferably is less than 2.5 inches. Although a purely sinusoidal current waveform is preferred to reduce eddy currents (and thus temperature elevation) which in part are created by higher frequency harmonics, the present invention further reduces 1e~ el~lu.~ rises by using the charging protocols of Figures 4B and 4C
as described below.
Referring now to Figures 5 and 7, PWM controller 200 controls the time when switches 21, 22 (SW1 and SW2, respectively) turn on, as well as the time period during which the switches are activated. Resistor 233R and capacitor 233C connect between RC oscillator 233 and ground. RC
oscillator 233 provides a periodic signal. As one of ordinary skill in the art will recognize, the voltage across capacitor 233C, as inrlir~ted by Vc in Figure 5, is substantially a saw-tooth voltage waveform as shown in sub-part (a) of Figure 7. Superimposed on the saw-tooth Vc waveform is an exemplary current control signal provided by the pulse controller 231. As shown in Figure 5, the current control signal is ~ lPd to the VREF input terminal of pulse gene.~lol 232 via conductor 250. Figure 7 demonstrates how the current control signal on line 250 can be used to achieve different levels of charging current and the curreM step protocol of Figure 4B. The GATE signals (GATE1 and GATE2) from the pulse genelaLol 232 are coupled to the gate inputs G of switches 21, 22. These GATE signals dictate when each of the switches 21, 22 are turned on. Alternating cycles of the sawtooth Vc waveform of Figure 7 are used to generate the GATE signals. The first Vc cycle CA 0226~984 1999-03-17 shown in Figure 7 produces the first GATE1 pulse in sub-part (b) of Figure 7. The second Vc cycle produces the first GATE2 pulse in sub-part (c) of Figure 7. The third Vc cycle produces the second GATE1 pulse, and so on. Each pulse is generated when the Vc voltage rises above the voltage of the current control signal (VREF). The width of each GATE pulse is dictated by the point when Vc rises above VREF and the point where Vc drops below VREF, exemplified in Figure 7 as the pulse PW1. Accordingly, by hlcleasillg the level of the current control voltage on the VREF terminal of pulse generator 232, the width of the GATE pulses are reduced as shown by pulse PW2 in Figure 7. As described above, increasing the time when both switches 21, 22 are "off" reduces the m~gnitll(le of primary coil current, and thus charging current, as intlic~t~d in sub-part (d) of Figure 7 at D1. Thus, the current control signal on line 250 is used by the pulse generator 232 as a reference value to vary the duty cycle of switches 21, 22 and thus control the charging current.
The pulse generator 232 preferably includes an enable input terrninal (EN) which controls the status of PWM controller 200. When an enable signal is provided to the enable input via line 255, the controller is enabled. Conversely, if a disable signal appears on line 255, controller 200 is disabled. More specifically, the pulse generator 232 is enabled and disabled. According to the preferred embodiment, the enable/disable signal on line 255 is used to define a duty cycle for the current delivered to primary coil 9, according to the charging protocol of Figure 4C.
The enable signal preferably is generated by the pulse controller 231 and ~ n~ Pd via line 255 to the enable (EN) input terminal of the pulse generator 232 to turn the pulse geneldtor on and off. Thus, for example, a logic "0" on the enable input line 255 may turn the PWM controller on, while a logic " 1 " turns it off. Once the pulse generator 232 is turned off through the enable signal, no current flows through the primary coil 9 and thus, the battery is not charged during the period when controller 23 is disabled. The pulse controller 231 preferably produces an enable signal sirnilar to the duty cycle waveform of Figure 4C. As ~ cllc~ed above, the resulting charging current is determined by the amplitude of the charging current when the pulse generator 232 is enabled (on) and the duty cycle of the enable input signal as one of ordinary skill in the art would understand.
The amplitude of the charging current when the PWM controller 23 is on can be preset through a specified current control voltage (VREF) on line 250. Thus, the PWM controller 200 can direct the inverter 20 to produce a step down average charging current by geneldlillg an enable signal with a variable duty cycle, thereby interrupting the gate signals to switches SW1, SW2. The average charging current is directly related to the duty cycle of the enable signal. ~t should be noted that for the step down current protocol of Figure 4B, the PWM controller's enable input is contimloll~ly asserted to constantly m~int~in the pulse generator 232 in the on condition.

CA 0226~984 1999-03-17 WO 98/11942 PCT/US97tl6319 Referring now to Figures 5 and 8, switches 21 and 22 preferably are impl.om~nt.od with MOSFET devices. The pulse generator 232 (Figure 5) COllllfCI~ to the gate G and source S terminals of MOSFET switch 21 through conductors 98 and 97 (Figures 5 and 8). The source terrninal S of MOSFET 21 and the drain terminal D of MOSFET 22 connect to terminal 31 of primary coil 9. The S drain terrninal D of MOSFET 21 c~-nn~ctC to capacitor 26 and also receives the input regulated DC
voltage Vjn on conductor 8 from AC/DC converter 5 (Figure 2). The gate terminal G of MOSFET
22 connects to the GATE2 terminal of controller 23 via conductor 27. The source terminal S of MOSFET 22 provides a path for current from the primary coil 9 to ground through the current sensing resistor 42 of the :~lignmPnt indicator 40. When switch 21 (SW1) turns on, a current path is completed from Vjn, through switch 21, node 31, coil 9, capacitor 25, node 30, and capacitor 29 to ground. When switch 22 (SW2) turns on, a current path is completed from Vjn, through capacitor 26, node 30, capacitor 25, coil 9, node 31, switch 22, and resistor 42 to ground.
Referring now to Figure 5, an additional advantageous feature of the present invention is an indicator for providing a visual indication of when the TET device 50 is properly positioned with respect to implanted device 14 for m~xim~1m charging efficiency. When switch 22 is turned on by controller 23, current flows from primary coil 9 through switch 22 and to resistor 42 in ~lignm~nt indicator 40. Due to the symrnetric AC current on the prirnary coil 9, the current through the switch 22 comprises half of the coil current during one-half the time of each cycle of the AC waveform.
Thus, only half of the primary coil current is received by resistor 42. In the preferred embodiment, the DC component of the voltage across the resistor 42 is used as an indication of DC input current from the voltage source Vjn.
Alignment indicator 40 provides a light emitting diode (LED) in LED circuit 48 or other output device to indicate proper positioning of the TET device 50 with respect to implanted device 14. The TET system can be tuned so that the amplitude of the AC current through the primary coil 9 decreases when the primary coil 9 is not properly aligned with secondary coil 10. The input DC
current, therefore, depends on the power draw of the load on the secondary coil and the proximity and orientation of the primary coil 9 to the secondary or receiving coil 10. Therefore, a mca~ulcl.lent of the m~gnitu(le of the input current preferably is used in the present invention to d~ ,lli--e if the TET device 50 is positioned properly for m~ximllm energy ~ .";.c.~ion efficiency.
The following discussion details the construction and operation of the ~lignm~nt indicator 40 which uses the correlation between the input current and ~lignmPnt to provide an output signal which in(li(~.~t~s when the energy ~ ion device 50 is sufficiently aligned with the receiving coil 10 of the implanted device 14.

CA 0226~9X4 1999-03-17 Referring to Figure 5, the re~ict~n~e value of resistor 42 preferably is small to minimi7~ the loading effect on the inverter 20 that would otherwise result. In the preferred embodiment, resistor 42 is selected as approximately 0.5 ohms. It will be understood that the purpose of resistor 42 is to sense current in the primary coil 9 and provide an output signal indicative of the current amplitude and phase shift. Accordingly, although a resistor is preferable, any current sensing device can be used in place of resistor 42.
Referring still to Figure 5, the ~lignment indicator 40 preferably includes a low-pass amplifier 43, a peak detector 45 to detect the peak DC current amplitude through switch 22, a differential amplifier 46 to amplify the difference between the peak current amplitude and the amplitude of the output current signal from low-pass amplifier 43 on line 108, a comparator 47 to compare the amplified difference with ground voltage, and an LED or other output circuit 48. In the preferred embodiment, the LED circuit 48 (or other output device) only provides.an output signal intli~ ing ~lignm~ont if the present sensed current amplitude is within a predetermined range of the peak value.
Current flow through resistor 42 from switch 22 generates a voltage Vs across resistor 42 which is amplified and filtered by low-pass amplifier 43 to effectively obtain the DC component of the waveform through resistor 42, and to filter out the AC portion of the waveform. The peak detector 45 senses the peak amplitude value of the output signal on conductor 108, which connects to the output terminal of the low-pass amplifier 43. The peak detector 45 stores the peak value, unless a higher amplitude is snhsPq-lPntly sensed. If a higher value is subsequently sensed, the peak detector 45 replaces the stored peak value with the new peak value. The output signal of the peak detector 45 on conductor 116 c~ pollds to the peak positive voltage sensed by the peak detector 45. This peak voltage (which is scaled to provide a threshold value that is somewhat less than the peak value), is provided as an input to the dirre-enlial amplifier 46. The other input to the differential amplifier colll~lise5 the current sensed output of the low-pass amplifier 43 (conductor 108). The dir~.~.-lial amplifier 46 amplifies the dirrere.lce between the scaled peak value, and the present sensed value, and provides an output signal to co---l)al~ol 47. Co--lpal~lul 47 compares the difference with ground voltage, and turns on the LED circuit 48 when the current sensed value is greater than the scaled peak value. This condition will occur when the TET device 50 is positioned properly over the implanted device. In order to capture the O~ti~ llll Iocation, the primary coil 9 has to pass the optimal location at least once to let the peak detector 45 record the peak DC current value. Thereafter, the LED circuit 48 will not be turned on unless the primary coil 9 stays at the optimum location and orientation. If the lateral pl~ç~m~nr of the TET device is mi.c~lignPd with respect to the receiving coil, or if the TET device 50 is positioned at a nonoptimal angle with respect CA 0226~984 1999-03-17 to the implanted device for peak lla~ ion ~rriciency, the scaled peak value will be greater than the present output voltage at the output terminals of filter 44, and the comparator 47 will produce an output signal de-activating the LED circuit 48.
One of ordinary skill in the art will recognize that a plurality of circuit implp-mpnr~ions are possible for the low-pass amplifier 43, peak detector 45, dirrelelllial amplifier 46, comparator 47, and LED circuit 48 of ~lignmP.nt indicator 40. In addition, the functions of two or more of these co~ on~ may be pelrolllltd by a single device. The circuit Srl,r~ 'ir,s of Figures 9-13 are shown as the preferred embodiment of the alignmPnt indicator of the present invention.As noted above, the voltage waveform across resistor 42 includes both AC and DC
0 CUIIIIJOnel1~. In the preferred embodiment, the AC component is filtered to permit ex:~min~ion of the DC component. Referring now to Figures 5 and 9, low-pass amplifier 43 is configured as an inverting amplifier, with an operational amplifier 103, an input resistor 102, a feedback resistor 105, a feedback capacitor 106, and an output resistor 107. The negative ratio of the resistance of feedback resistor 105 to the 1~ t,~l~re of resistor 102 (lrlf~ Ps the DC voltage gain of the amplifier 43. Preferably, the gain is set at 100. Therefore, the re.Ci.~t~nre of resistor 105 should be one hundred times greater than that of resistor 102. R~.si.ct~nce values of 44.9 Kohms for resistor 105 and 449 ohms for resistor 102 are pler.,.led, but numerous other values are possible. Capacitor 106, together with resistor 102, provide low-pass filter capabilities to amplifier 43. A resistor 104 connects the non-inverting input terminal of operational amplifier 103 to ground. The output terminal of operational amplifier 103 connects to feedback resistor 105, capacitor 106, and output resistor 107. The output of amplifier 43 (which preferably in-lir:l~Ps a negative voltage value) is provided on conductor 108 to peak detector 45.
Referring now to Figures 5 and 10, the preferred construction and operation of the peak detector 45 will now be described. ln the preferred embodiment, peak detector 45 colllylises an operational amplifier 114, peak storage capacitor 120, and voltage follower 46. The low-pass amplifier 43 connects through conductor 108 to the non-inverting input terminal of operational amplifier 114. The output terminal of operational amplifier 114 connects to the cathode of diode 117, the anode of which co~ e-;L~ to the cathode of diode 119. Current from operational amplifier 114 (with a negative amplitude) flows through diodes 117 and 119, charging storage capacitor 120 to a voltage indicative of the peak value at the non-inverting input of operational amplifier 114.
Diode 115 prevents operational amplifier 114 from saturating in the absence of peak values, and resistor 216 provides a path through which the current from diode 115 can flow. Switch 122 resets the peak detector output signal to 0 V upon closure of that switch.

CA 0226~984 1999-03-17 WO 98tll942 PCT/US97/16319 When a new peak arrives at the non-inverting input of operational amplifier 114, the output of op amp 114 swings in the negative direction, turning diode 115 off (preventing current flow through resistor 216) and turning diodes 117 and 119 on, pc~ lg capacitor 120 to charge. As the input voltage on conductor 108 drops, the output of operational amplifier 114 swings in the positive direction, turning off diode 117 and diode 119. As a result, c~citcr 120 m~int~in.c its peak voltage charge, with diode 119 and resistor 118 limiting the leakage of capacitor 120. As the output voltage continues in the positive direction, diode 115 turns on to prevent saturation of the op amp 114.
The voltage follower buffer 121 not only provides a high input impedance to ,,,il,i..,i,~.
loading on other stages of the circuitry, but also scales down the peak detected voltage through the use of a manually adjustable potentiometer 118. Potentiometer 118 connects between the output of operational amplifier 121 and ground to provide an adjustable voltage divider in which conductor 116 carries the scaled down peak voltage to an input of differential amplifier 46. The output of operational amplifier 121 is fed back to the inverting input of amplifier 121 and is provided via conductor 115 to peak detector 45.
Referring now to Figure 11, the dirr~lellLial amplifier 46 preferably cul~ ises an operational amplifier 123, a feedback resistor 122b, and input resistors 120, 122a. The output signal from peak detector 45 couples to the inverting input terminal of operational amplifier 123 through resistor 122a.
The output signal from the low-pass amplifier 43 couples through resistor 120 to the non-inverting input terminal of operational amplifier 123. Operational amplifier 123 amplifies the difference between the scaled peak value on conductor 116, and the present sensed value on conductor 108, and provides the amplified difference as its output 224. In the pref~lled embodiment of Figure 11, the resistance of resistor 122b is equal to the r~si~t~nre of resistor 119, and the resistance of resistor 122a is equal to the reSict~n~e of resistor 120, to provide a gain for dirr~ellce amplifier 46 that equals the ratio of resistor 122b to resistor 122a.
Referring now to Figures 5 and 12, the comparator 47 and LED circuit 48 are shown in detail. The comparator circuit 47 pler~ilably comprises a comparator 126, a pull-up resistor 127, input resistor 124, and capacitor 125. The LBD circuit 48 includes an LED 131, l.dnsislur 130, current limiting resistors 128 and 129. The output of dirr~ Lial amplifier 46 preferably connects via conductor 224 to the inverting terminal of cu~ aldlol 126, through input resistor 124. The non-inverting input terminal of comparator 126 connects to ground, and to the inverting input terminal of comparator 126 through capacitor 125. The output of colllpal~lur 126 provides an input signal to the LED circuit 48 to turn on LED 131, or an alternative output device. Resistor 127 collll,l;ses a pull-up resistor which may be necessary if comparator 126 has an open-collector output stage. In CA 0226~984 1999-03-17 WO 98/11942 PCTtUS97/16319 the pl~felled embodiment, the output terminal of comparator 126 connects to the base termin:ll B
of transistor 130 through current limiting resistor 128. Power from the voltage source +Vcc is provided to LED 131 through resistor 129 when transistor 130 is turned on by the supply of sufficient base current from the comparator 126 to the base terminal B of the transistor 130.
Although an NPN lldll~ Ol iS shown in Figure 12, one of ordinary skill in the art will recognize that other types of LED driver circuits are possible, including the use of PNP ~Idnsi~lul~, and the present invention should not be construed as limited by the particular circuit embodiment shown in Figure 12. Similarly, although an LED 131 is shown as the output device, it will also be understand that other output devices, such as audible indications, may be used as an alternative, or in addition to LED 131.
An important parameter associated with a recharging system is charging efficiency.
Referring to Figure 2, recharging efficiency can be defined as the ratio of the electrical energy input to the charging system, Ejn~ in TET device 50 that is to the energy delivered to the battery, Ebatl in implantable device 14. Part of the energy coupled to the implanted device that is not delivered to the battery 13 is converted to thermal energy detrimentally heating the implanted device. Thus, it is desirable to maximize recharging efficiency in order to minimi7e lel~lpel~t~lre increases of the implantable device.
Referring now to Figure 13, it has been de~ il,ed that charging efficiency is a function of the m~gnih~ . of the charging current. Curve I exemplifies the efficiency when both capacitors 410 and 420 of implantable device 14 in Figure 2 are used in the charging circuit simlllt~n~.ously Improved efficiency can be achieved if either capacitor 410 or 420 is used, but not both simlllt~n~ ously. As shown in curve II, higher efficiency can be achieved using just capacitor 410 than using both capacitors together, but the ~rr,ciell~y still drops off at higher charging current levels.
If just capacitor 420 is used (curve III), higher efficiency is obtained at higher current levels, but lower efficiency at lower current levels than if just capacitor 410 is used. To m~ximi7e charging t;rrl~iel~y, therefore, it is preferable to use only capacitor 410 when charging at lower current levels and only capacitor 420 when charging at higher current levels, but not both capacitors sim~llt~n~oucly. Current level Io may be defined as the charging current above which capacitor 420 is used and below which capacitor 410 is used.
It should be noted that the secondary circuit in implantable device 14 can be properly tuned to the frequency of the AC current through the primary coil 9 by using either capacitor 410 or 420 in conjunction with secondary coil 10. That is, battery charging is possible regardless of which capacitor is included in the preferred embodiment. However, to achieve maximum efficiency, capacitor 410 preferably is used for low current charging and capacitor 420 is used for high current CA 0226~984 1999-03-17 charging. Accordingly, a mPchqni~m for switching the capacitors into the circuitry is described below.
Referring now to Figure 14, switches 411 and 421 of implantable device 14 are shown COII~ ,lelll with the preferred embodiment as MOSFET devices. Switch 421 effectively provides a S short circuit across capacitor 420 when switch 421 is closed and an open circuit when the switch is opened. Thus, opening switch 421 causes capacitor 420 to be used, which is preferable during high current charging periods (i.e., for geneldtil,g current I, in Figure 4B). Similarly, switch 411 can be opened, removing capacitor 410 from the charging circuit during high current charging. When switch 411 is closed, capacitor 410 forms part of the charging circuit, which is preferable during low current charging (for generating current l2 in Figure 4B). Preferably, switches 411 and 421 are opened for the duty cycle protocol of Figure 4C because only a single current level is generated.
Referring to Figure 14, the current sensor 270 (Figure 2) preferably Co~ liscs a current sensing resistor 435. As one of ordinary skill in the art will recognize, current sensing components typically include low r~ict~nfe resistors, such as 0.5Q resistors or less. Because one side of current sensing resistor 435 is grounded, the voltage on the other side at point S1 is proportional to the current through the resistor. Because the current sensing resistor is connected in series with the battery 13, the voltage at point S1 is proportional to the charging current applied to the battery.
Referring again to Figure 14, the voltage at point S1 (i.e., battery current) is provided as an input to switch drive logic 300. After filtering voltage S1 to remove high frequency noise, including ringing and carrier frequency, by low pass filter 440, the filtered signal is then amplified by inverting amplifier 445. Comparator 450 compares the amplified battery current signal to a reference voltage, VR, which is provided on line 447. The reference voltage y~ can be generated in several ways co~ ,Lent with the l~refe~l~,d embodiment. For example, voltage VR may be derived from the battery 13 through a voltage regulator. Reference voltage VR is predetermined so as to leplesellt the charging current level at which it is desired to switch from capacitor 410 to 420, or vice versa.
Preferably, VR is set to correspond to current level Io from Figure 13. One of ordinary skill in the art will also recognize the benefit in incorporating hysteresis into the comparator circuit 450 to prevent the comparator from oscillating if the amplified current signal S1 hovers around the value of the reference voltage VR. Low-pass filter 440, inverting amplifier 445, and comparator 450 can be provided by any of a number of conventional devices, and the present invention is not limited to any specific circuit implem~nt~tion.
Referring still to Figure 14, when the amplified current signal provided to the positive terminal of the culll~d,~tor 450 is greater than the r~ferel1ce voltage VR, the output of the CUIII~'dldtOI
on line 452 will become a logic high level. This high logic signal on line 452 in turn drives switches CA 0226~984 1999-03-17 411 and 421 open through inverting gate drives 455 and 457 which are of a type known to those of ordinary skill in the art. The low output signal of gate driver 455 on line 301 is provided to the gate terminal G of switch 411, thereby opening switch 411, and thus removing capacitor 410 from the circuit. Further, the low output signal of gate driver 457 on line 302 is provided to the gate terminal G of switch 421, thereby opening switch 421, and placing capacitor 420 in the charging circuit.
Therefore, for high charging currents over the trip point set by voltage VR (Io)~ charging efficiency is maximized by using only capacitor 420, not capacitor 410, in the charging circuit.
When charging the battery at a low current level (less than the trip point deffined by VR), the output signal on line 452 from the comparator 450 is driven to a low logic level because VR exceeds the output signal from inverting amplifier 445. The low logic level from the output of co~ al~tor 450 activates the inverting gate drives 455 and 457 to high output levels. A high output signal on line 301 from gate drive 455 closes switch 411 thereby placing capacitor 410 i~ltO the charging circuit. A high output signal on line 302 from gate drive 457 closes switch 421 and effectively removes capacitor 420 from the charging circuit. Therefore, for low charging currents (less than the trip point set by voltage VR), ~hal~i,lg efficiency is m~ximi7ed by using only capacitor 410, not capacitor 420, in the charging circuit.
Finally, referring to Figures 2 and 14, rectifier 12 preferably comprises a full-wave bridge rectifier including diodes 12a, 12b, 12c, and 12d. Capacitor 430 is provided as an output filter so that a good approximation to a DC level current is provided to battery 13.

Claims (26)

WHAT IS CLAIMED IS:

1. A transcutaneous energy transmission system for transmitting electrical power from an external charging device (51) to an implantable medical device (14) to recharge a battery (18) associated with the implantable device, said transcutaneous energy transmission system comprising:
an external charging device (51) containing a primary coil (9) for transmitting the power transcutaneously to said implantable device;
a capacitor (25) coupled to said primary coil, wherein said primary coil and said capacitor form a resonant circuit;
a controller (200) coupled to said resonant circuit by a first switch (21) and a second switch (22), said first switch and said second switch being alternatively switched on and off by said controller to produce a sinusoidal current waveform to said primary coil;
characterized by means (231. 232. 233) for controlling power radiated by said primary coil at at least a high level and a low level.

2. The device of claim 1 wherein said first switch (21) and said second switch (22) comprise solid state devices.

3. The device of claim 2 wherein said solid state switches (21, 22) comprise field effect transistors.

4. The device of claim 1 wherein said duty cycle is variable.

5. The device of any of the foregoing claims further characterized by said means (231, 232, 233) for contolling power uses a first duty cycle during an initial stage of charging and uses a second duty cycle during a subsequent stage of charging in which the second duty cycle is lower than the first duty cycle.

6. The device of claim 5, wherein the first duty cycle is 100%.

7. The device of claim 5, wherein the controller includes a first output terminal that couples to said primary coil.

8. The device of claim 7, wherein the controller (200) includes second and third output terminals connected to said first switch (21) and said second switch (22), respectively.

9. The device of claim 5, wherein the controller includes an enable signal, and the controller turns power on and off to the primary coil based upon the value of the enable signal.

10. The device as in claim 9, further comprising a pulse controller (231) for generating the enable signal.

11.

The system according to claim 1 wherein said means (231, 232, 233) for controlling said power is further characterized by timing means (231, 232, 233) for emitting pulses at at least a high and a low rate.

12. The system according to claim 11 wherein said timing means comprises circuitry (233) for generating a voltage signal with a varying amplitude;
a pulse width modulation controller (200) for controlling the level of charging current to said battery, said pulse width modulation controller comprising:
a pulse controller (231) providing a reference voltage signal;
a pulse generator (232) receiving said reference voltage signal and varying amplitude voltage signal, and in response generating an output pulse said output pulse having a width defined by the period when the varying amplitude voltage signal exceeds the reference voltage signal.

13. A device as in claim 12, wherein the varying amplitude voltage signal (Vc) exhibits a constantly repeating pattern.

14. A device as in claim 13 wherein the varying amplitude voltage signal (Vc) comprises a saw-tooth wave form.

15. A device as in claim 12 wherein the reference voltage signal (Vref) comprises one of two voltage levels. causing said pulse width modulation controller (200) to generate either a high charging current or a low charging current in said battery,

16. A device as in claim 15 wherein said high charging current is generated first followed by said low charging current.

17. A device as in claim 12, wherein said pulse width modulation controller (200) couples to said primary coil via a first switch (21) and a second switch (22), and wherein said pulse width modulation controller (200) transmits a first and second signal to turn on and off said first and second switch.

18. A device as in claim 17, wherein said first and second signals are alternately emitted to said first and second switches (21, 22) to generate said output pulse.

19. A device as in claim 18, further comprising a capacitor (25) connected to said primary coil (9) to form a resonant circuit.

20. A device as in claim 18, wherein said first and second switches (21, 22) comprise solid state switches.

21. A device as in claim 20, wherein said first and second signals comprise gate signals for said solid state switches (21, 22).

22. The transcutaneous energy transmission system according to any of the foregoing claims wherein said implantable medical device contains a secondary coil (10) in said implantable medical device;
a battery (18) for receiving electrical current from said secondary coil (10) a first capacitor (410) connected in parallel across said secondary coil (10);
a secondary capacitor (420) connected in series with said secondary coil (10);
further characterized by said first capacitor (410) being connected in series to a first switch (411). the combination of the first capacitor and first switch connected in parallel across said secondary coil;
and a second switch (421) connected in parallel across said second capacitor.

23. The system of claim 22, wherein the implantable medical device (14) further comprises a current sensor (270) for sensing the magnitude of current charging the battery.

24. The system of claim 23. wherein the implantable medical device further comprises a switch drive (300) for turning said first switch and said second switch on and off depending on the magnitude of current charging the battery, and wherein the magnitude of current charging the battery is indicated to the switch drive by the current sensor (270).

25. The system of claim 24, wherein said first switch (411) is closed to allow current to flow through said first capacitor (410) and said second switch (421) is closed to short circuit said second capacitor (420), when a relatively low charging current is applied to said battery (18).

26. The system of claim 24 wherein said first switch (411) is opened to prevent current from flowing through said first capacitor (410) and said second switch (421) is opened to allow current to flow through said second capacitor (420), when a relatively high charging current is applied to said battery (18).