GB2048688A - Programmable demand pacer - Google Patents

Abstract

An implantable R-wave inhibited cardiac pacer has integrated digital logic circuits including programmable registers 12, 16 (210, 262, Fig 4) to determine the pulse rate and width. A 5-3-4 code applied by an external pulsed electromagnet to a reed switch 20 (400, Fig.5) unlocks a digital lock 18 so that rate register 12 can be reprogrammed to any value between 30 and 120 BPM. A 5-3-2 code allows the width register 16 to be reprogrammed, and the circuitry ensures that the programmed width can only lie between set limits. If a main timing circuit failure gives rise to a rate above 130 BPM, a protection circuit 32 feeds pulses to the heart 36, at half the rate from generator 28. A refractory circuit 42 prevents the generator being reset by a heart signal for 256 or 320 msec following a sensed natural beat or a pacing pulse respectively, noise sensing circuits operating from the last 64 msec of the refractory period, and the paceroperating uninhibited at its programmed rate in the presence of noise. Closure of switch 20 by a permanent magnet causes uninhibited operation at 100, 90 or 85 BPM depending on whether the battery is new, half-discharged or should be replaced, the pacer giving six uninhibited pulses at the programmed rate on removal of the magnet. The disclosure is the same as that in G.B. Specification No. 2,014,858A. <IMAGE>

Description

1

GB 2 048 688 A

1

SPECIFICATION

Programmable demand pacer

5 This invention relates generally to electronic pacer devices for providing artificial electrical stimulation to the heart of a patient and more specifically to an improved pacer of the demand-inhibit type.

In the normal heart, electrical signals are gener-10 ated and appear in the atrium at a rate of approximately 60 to 120 times per minute, depending upon the physical activity of the individual. About 0.1 seconds following the generation of each atrial signal, a signal is transmitted to the ventricle, causing it to 15 contract and force blood from the ventricle through the body. Following the ventricular contraction, the heart experiences a refractory delay period which persists approximately 0.4 seconds and during which the heart is un-responsive to electrical stimu-20 lation.

An often observed abnormality of the heart is its failure to regularly or periodically generate atrial signals. Atrial signals may be generated with perfect regularity but then, suddenly, cease altogether or 25 occur at an abnormally low rate. To compensate for such a defect the so-called "demand-inhibit" pacer is a practical solution. Such a device applies stimulating pulses to the ventricle by means of surgically implanted electrodes only when natural pulses do 30 not occur spontaneously. As long as natural pulses occur at a desired rate, the implanted pacer provides no stimulation. However, when the natural or spontaneous ventricular pulses fail to occur or occur at irregular intervals the pacer comes into play to pro-35 vide artificial stimulation at a desired rate.

Prior art heart pacers as well as the pacer described hereafter include a simulated refractory delay period. The reason for including a simulated refractory delay in the pacer circuitry is to ignore 40 spontaneous signals developed by the heart which occur within a prescribed time period following application of an artificial stimulating pulse.

The pacer described hereafter operates asynchronously at a predetermined frequency (e.g., 100 45 beats per minute) when a permanent magnet is applied over the implanted pacer. This is the so-called "magnetic rate". The unit will remain in the magnetic rate mode as long as the permanent magnet is in proximity to the implanted pacer device and 50 for a maximum of 2,256 milliseconds after removal of the magnet. Immediately after the magnetic rate interval, the pacer will operate asynchronously at the programmed nominal rate for six pacer pulses. This allows the pulse generator to be checked for both 55 magnetic rate and nominal rate with the use of an externally applied magnetic field.

The pacer also incorporates an Elective Replacement Time indicator feature. The ERT indicator is tied in with the magnetic rate of 100 beats per 60 minute, therefore, a magnet is also used to check the ERT. When the output of the implanted energy source (battery) has decreased 1/2 of the way to ERT,

the magnetic rate will drop to 90 beats per minute. When the ERT is reached, the magnetic rate will drop 65 to 85 beats per minute. Therefore, a technician can readily determine by sensing the frequency or rate of pacer pulses when in the magnetic rate mode just when further surgery may be required to replace the energy source.

70 The pacer circuit also includes noise protecting circuitry. Specifically, a bandpass amplifier is used as a signal discriminator which attenuates unwanted signals, i.e., EMI, T-waves, 60 cycle pickup, etc., but passes signals containing the fundamental fre-75 quency component of R-waves. Thus, all unwanted signals are filtered and only desirable signals are passed. Noise protection is afforded by the fact that the pacer of the present invention includes apparatus for monitoring the pacer inputs during 80 refractory intervals. If for any reason, electrical noise should pass through the bandpass amplifier, a noise detection circuit, including a chain of bistable flip-flops, takes over and switches the pulse generator into its asynchronous mode in which state it remains 85 until the noise source is no longer present.

Another feature of the pacer is the incorporation of a dual refractory system. More specifically, the circuitry incorporated in the pacer provides a 256 millisecond refractory period following the occurrence 90 of a natural heartbeat and a 320 millisecond refractory period following a heartbeat that is a result of an artificial stimulating pacer pulse. During the final 64 milliseconds of the refractory periods, the pulse generator will be monitoring the heart electrodes for 95 electrical signal activity, but will not allow the resetting of the pulse generatortiming. This too is important in rendering the device immune from external noise.

The pacer also permits reprograrriming of the rate 100 of generation of pacer pulses as well as the width of such pulses. Programming is accomplished by means of an external console into which an operator may enter a code combination which functions to unlock circuitry in the implanted unit to allow new 105 pulse rate and pulse width information to be entered. This is accomplished by means of an electromagnet which is positioned outside of the body but in proximity to the implanted pacer unit. The electromagnet is pulsed by the control circuits in the 110 console and operates a magnetic reed switch which is found in the implanted circuitry. A three digit pulse train is employed to allow programming of the pulse generator. Before a new pulse rate may be entered, the external programming unit must first present a 115 certain preselected code, such as 5-3-4- code to the reed switch. Similarly, before the width of the pacer ' pulse may be changed, a second preselected code, such as a 5-3-2 code must be entered.

Upon application of the first pulse of the initial 120 digit (5) the circuitry will revert to its magnetic rate 'mode (100 BPM) and will remain there until the programming operation has been completed. When programming in a new pulse rate, the operator initially presents a 5-3-4 address code to the pulse

The drawings originally filed were informal and the print here reproduced is taken from a later filed formal copy.

2

GB 2 048 688 A

2

generator. Once the pulse rate generator is addressed, a single pulse from the programmer initiates the pulse period. A second pulse from the programmer terminates the pulse period at the desired pulse 5 interval. This interval, in conjunction with the pulse generator internal clock, allows the programming of any rate between 30 and 120 BPM. The programming time for any rate between 30 and 120 BPM is 2,256 milliseconds. Regardless of the rate program-10 med, i.e. between 30 and 120 BPM, the pulse generator will remain in the magnetic rate for the full 2,256 milliseconds.

Should an abnormally long or short pulse rate be supplied to the pulse generator, the following will 15 occur. In the case of an excessively long rate, e.g., greaterthan 2,048 milliseconds, the pulse generator will remain in the paced magnetic rate mode (100 BPM) until an acceptable rate is inserted. If the rate is less than 496 milliseconds, the pulse generator will 20 remain in the paced magnetic rate mode until an acceptable rate is programmed.

An alternative width for the pacer pulse is programmed in a somewhat different manner. Once the pulse generator is addressed, a pulse train corres-25 ponding to a particular pulse width is presented to the pulse generator. The minimum width is assured by a hard wired connection within the pacer unit and is 0.05 milliseconds. This is incorporated to guard against failure to enter any width count after suc-30 cessfully addressing the pulse generator by the 5-3-2 code. In the event that an excessive number of pulses are included in the pulse train, the pulse generator will reach its maximum width of 3.2 milliseconds and recycle back to 0.10 milliseconds. It 35 will continue to cycle until no more pulses are received. Following the reprogramming operation, the pulse generator switches to its asynchronous mode for six pacer pulses and then to the demand mode.

40 The three digits of the unlocking codes are transmitted in three consecutive 68 milliseconds time frames. The first digit time frame overlaps the pulse generator's first digit time frame by 4 milliseconds to ensure entry of the digit. The second digit time frame 45 overlaps the pulse generator's second digit time frame by 8 milliseconds and the third digit overlaps by 12 milliseconds. This overlap will ensure that the unlocking code is entered even if the clock frequency in the implanted unit has drifted somewhat from its 50 design value.

In accordance with this invention there is provided an implantable heart pacer for applying articifial stimulating pulses of a predetermined pulse width and repetition rate to the heart of a patient, compris-55 ing;

(a) a first binary register for storing a first predetermined binary number value corresponding to a desired stimulating pulse repetition rate;

(b) a second binary register for storing a second 60 predetermined binary number value corresponding to a desired stimulating pulse width;

(c) a first presettable counter having a clock input terminal, a load enable input terminal, a carry output terminal and a plurality of input terminals opera-

65 tively coupled to the outputs of said first binary register for periodically receiving said first binary number value each time its load enable input terminal is stimulated;

(d) a second presettable counter having a clock input terminal, a carry output terminal, a load enable input terminal and a plurality of input terminals operatively coupled to the outputs of said second binary register for periodically receiving said binary number value each time its load enable input terminal is stimulated.

(e) a source of regularly occuring timing signals coupled to said clock input terminals of said first and second presettable counters for changing the numbers stored therin at a fixed rate;

(f) means coupling said carry output terminal of said first presettable counter to said load enable terminals of said first and second presettable counter; and

(g) output terminal means for coupling said carry output terminals of said first and second presettable counters to the heart of the patient, the arrangement being such that when said first binary number value initially in said first presettable counter is changed by said timing signals to a given value, a heart stimulating pulse is initiated and when said second binary number value entered into said second presettable counter is changed by said timing signals a given value, said heart stimulating pulse is terminated.

The pacer described hereafter is also described in U.K. patent specification No. 2014858, which claims different features of the pacer.

Reference is now made to the accompanying drawings, wherein:-

Figure 1 is a functional block diagram of the cardiac pacer according to the present invention;

Figure 2 is a schematic electrical diagram of the voltage doubler, bandpass amplifier, and level detector circuitry incorporated in the pacer of the present invention;

Figures 3a and 3b are schematic electrical diagrams of a portion of the circuitry used to implement the ERT check feature;

Figure 4 shows the way that Figures 4a and 4b should be positioned to yield a composite schematic diagram.

Figures 4a and 4b when oriented as indicated in Figure 4 illustrate the digital logic circuitry for determining the pulse rate and pulse width of the pacer pulses as well as certain additional logic implementing features described in the introductory portion of this specification;

Figure 5 shows the way that Figures 5a and 5b should be positioned to yield a composite schematic diagram.

Figures 5a and 5b when arranged as shown in Figure 5 illustrate the logic circuitry for implementing the combination lock feature and certain control circuitry utilized therein; and

Figure 6 illustrates by means of a digital logic diagram the circuitry utilized to establish the dual refractory periods and to provide noise detection control.

Referring first to Figure 1 there is shown a functional block diagram of the preferred embodiment of the present invention. Identified by numeral 10 is a crystal controlled oscillator which serves as the main

70

75

80

85

90

95

100

105

110

115

120

125

130

3

GB 2 048 688 A 3

timing element. The crystal controlled oscillator may be designed to produce regularly occurring square wave pulses at a rate of 10 KHz, for example, where pulses of a lower frequency are required, a conven-5 tional frequency dividing network may be employed. The output from the oscillator is connected to a block 12 labelled "New Rate Storage" as well as to a block 14 labelled "New Rate Check" The main timing oscillator 10 also provides clock pulses to the block 16 10 labelled "New Width Storage". Both the New Rate Storage element 12 and the New Width Storage element 16 are underthe control of a digital combination lock element 18. Thus, as will be fully explained hereinbelow, before a new pulse rate 15 value or a new pulse width value can be programmed into the apparatus, the digital combination lock 18 must have a predetermined code applied thereto to enable the new entries to be made. The digital combination code can only be applied to the element 20 18 when a magnetically operable reed switch is actuated. This reed switch 20 also controls the element 22 labelled "Mag Rate", the purpose of which will become more apparent from discussions which follow. The functional block labelled 24 symbolically 25 represents the circuitry whereby a physican may determine the voltage of the implanted energy source for the purpose of determining whether a surgical replacement of an energy source should be made. The manner in which this is accomplished will 30 also be described when the detailed circuitry forming the preferred embodiment is discussed. For now, however, it is sufficient to say that the Elective Replacement Time (ERT) is determined when the pacer is operating in its so-called "magnetic rate 35 mode" and this is the reason that the functional block 24 is shown as being coupled to the Magnetic Rate block 22.

In a similar fashion, once a new pulse repitition has been entered into the device, it is desirable to 40 check this new rate to see that it is properly entered. This is also accomplished when the apparatus is in its mag rate mode and hence, the functional block 14 is also shown as being coupled to the Mag Rate block 22.

45 The block 26 entitled "Nominal or Mag Rate

Select" receives as an input the output from the New Rate Storage block 12 and the Mag Rate block 22 and the functions to select one or the other for applying or steering the output therefrom to the Resettable 50 Oscillator 28 which is the device which produces the pacer pulses upon demand for application to the heart of the patient. The Resettable Oscillator 28 is also controlled by the Crystal Controlled Oscillator 10 and by a block labelled "Pulse Width Select" 30. 55 This latter device functions to apply signals either from the Crystal Controlled Oscillator 10 or from the New Width Storage element 16 to the Resettable Oscillator 28.

The block labelled "Runaway Protecting" 32 is 60 shown as being connected to the output of the Resettable oscillator 28. As will be fully explained, the Runaway Protecting network 32 serves to insure that the pulse rate will never exceed 130 beats per minute (BPM) irrespective of the output from the 65 Resettable Oscillator 28. In a reprogrammable rate type demand pacer of the type described herein,

such a protecting network is deemed necessary to protect the patient in the event that a malfunction of the main timing circuit causes the Resettable Oscillator to run at too high a pulse rate.

As shown in Figure 1, the output of the Runaway Protecting block 32 is coupled to a Voltage Doubler network 34. This circuit operates in a conventional fashion to approximately double the amplitude of the pulses applied to its input and insured that for a desired battery potential, the stimulating signals applied to the heart load 36 byway of the surgically implanted electrodes labelled H+ and H- will be sufficient to provide the degree of stimulation required.

As in all demand type pacers/means are provided for sensing normal heart activity. As is shown in the functional block diagram of Figure 1, the same electrodes H+ and H- that couple the heart stimulating pulses from the output of Voltage Doubler 34 are used to pick-up or sense the electrical output from the heart occurring during normal activity. These signals are fed through a Bandpass Amplifier 38, and a Level Detect network 40 to a so-called "Refactory Generator" 42. In the pacer of the present invention, a dual refractory arrangement is provided. More specifically, a refractory period of 256 milliseconds is provided for a sensed heart signal and a 320 millisecond refractory period is provided for an artifi-cally stimulated beat. During the last 64 milliseconds of these refractory periods, the Amplifier 38 will be monitoring the heart electrodes for electrical activity, but will not allow resetting of the Resettable Oscillator 28.

Afunctional block44 labelled "Noise Detect" is shown as being associated with the refractory generator 42. During the last 64 milliseconds of the two refractory intervals mentioned above, the electrical activity of the paper inputs is being monitored. If for any reason, electrical noise signals should pass through the bandpass amplifier 38, the noise detection circuit 44 takes over and switches the Resettable Oscillator 28 into its "asynchronous mode". The pulse generator 28 will then be operating asynchronously at its nominal rate until the noise signals disappear.

Now that the overall organization of the implantable demand type pacer of the present invention has been described with the aid of the functional block diagram of Figure 1, the details of the various electronic circuits used to implement the system will be described. Following that, a description will be given of the mode of operation.

Referring to Figure 2, there is shown by means of a schematic electrical diagram the circuitry used to implement the Voltage Doubler network 34, the Bandpass Amplifier 38, the Level Detect 40 and a portion of the Refractory Generator 42 shown in the block diagram of Figure 1. In this figure there is shown a positive bus 50 and a negative bus 52 which are adapted to be connected to the B+ and B- terminals 54 and 56 to a source of direct current which, in the preferred embodiment, may be Lithium Iodide cell. As is well known in the art. Lithium Iodide batteries are especially well suited for use in implanted pacemakers because of their relatively long life and

70

75

80

85

90

95

100

105

110

115

120

125

130

4

GB 2 048 688 A

4

the absence of gas generation during discharge thereof.

A resistor 58 couples the output from the pacer rate gate 330 connected to terminal 57 (Figures 4b) 5 to the base electrode of a PNP transistor 60. The emitter electrode of transistor 60 is connected to the positive bus while the collector electrode thereof is coupled by a resistor 62 to the base electrode of an NPN transistor 64. The emitter electrode of transistor 10 64 is connected to the B- bus 52. The collector electrode of transistor 64 is coupled to the positive bus by means of a resistor 66 and to the negative heart terminal 70 by a conductor 68. A bidirectional diode 72, connects the collector electrode of the transistor 15 64 (via conductor 68) to the positive heart electrode terminal 74. Electrode terminals 70 and 74 are adapted to be connected to surgically implanted electrodes affixed to the heart of the patient. They serve to apply artificial stimulating pulses to the 20 heart and also act as the pickup means for sensing electrical output from the heart occurring during normal contractions thereof.

A resistor 76 couples the collector electrode of transistor 64 to the base or control electrode of a 25 PNP transistor 78. The emitter electrode of transistor 78 is connected directly to the positive bus 50. Its collector electrode is coupled through a resistor 80 to the negative bus 52. A voltage doubling capacitor 82 is connected between the collector electrode of 30 the PNP transistor 78 and the positive heart terminal 74.

Coupling capacitors 84 and 86 couple the bipolar signals developed during normal heart contractions to the input terminals of a differential bandpass amp-35 lifier 87. A voltage divider including the series connected resistors 92 and 94 is connected directly between the positive bus 50 and the negative bus 52 and the junction between these two resistors is connected to the positive input terminal of the differen-40 tial bandpass amplier 87. This arrangement determines the operating point for the differential amplier. A resistor 96 in series with coupling capacitor 86 is connected to the minus terminal of differential amplifier 87. Further, a feedback resistor 97 is con-45 nected between the output terminal of the differential amplifier 87 and the minus input terminal thereof. The combination of resistors 96 and 97 establish the voltage gain of the amplifier 87. The resistor 98 coupled between the positive bus 50 and 50 an auxiliary terminal of the amplifier 87 establishes the bias current for the amplifier. Thus, there will be developed at the output of the differential amplifier 87 an amplifier signal which is proportional to the differential signal developed across the implanted 55 heart electrode terminals 70 and 74.

An NPN transistor 100 has its emitter-collector path connected directly in parallel with the feedback resistor 97. The base electrode thereof is coupled by way of a conductor 101 to the output of flip-flop 550 60 (Figure 6). As will be explained more fully when the details of Figure 6 are set forth herein, the flip-flop 550 provides a signal to the transistor 100 for a predetermined time interval during which the gain of amplifier 87 is turned down to zero. Thus, the amp-65 litude of the output signal from the differential amplifier 87 is decreased during the time that the flip-flop 550 is set.

The resistors 92,94,96 and 97 along with the cou-

i"

pling capacitors 84 and 86 form a low pass filter network. The feedback capacitor 99 associated with the amplifier 87 creates a high pass filter network. Thus, the combination of these components with the differential amplifier 87 can be condidered as creating a bandpass amplifier. The component values compris- , ing this bandpass amplifier are chosen to give a centre frequency of approximately 31 Hz with a lower 3db point of 12 Hz and an upper 3db point of 62 Hz. As such, this bandpass amplifier arrangement acts as a discriminator, attenuating unwanted signals which may be generated by extraneous sources outside of the body of the patient. However, it passes signals containing the fundamental frequency components of R-wave signals produced by the heart muscle. The output from the bandpass amplifier stage is applied via coupling capacitor 107 to the base electrode of a PNP transistor 101 whose emitter electrode is tied to the positive bus 50 and whose collector electrode is coupled through a bias resistor

102 to the negative bus 52. The output from the bandpass amplifier is also applied through a coupling capacitor 108 to the base of an NPN transistor

103 having its collector electrode coupled through a bias resistor 104 to the positive bus 50 and its emitter electrode connected directly to the negative bus 52.

The switching threshold for the PNP transistor 101 is established by a voltage divider including the series connected resistors 109,110,111,112 and 113 which are connected between the positive bus 50 and the negative bus 52. A transistor 105 having its collector open has its emitter tied directly to the positive bus 50 and its base electrode tied to the junction point formed between series connected resistors 110 and 111. Similarly, a transistor 106 of opposite conductivity type has its collector electrode open and its emitter electrode tied to the negative bus 52 and its base electrode connected to the junction point formed between the series connected resistors 111 and 112. Thus, the transistors 105 and 106 function as semiconductor diodes. As was inferred above, the resistors 109,110,111 112, and 113 along with the diode connected transistors 105 and 106 establish . the switching point forthe level detect transistors 101 and 103. The resistors 109 and 113 can be considered as trimming devices which are used to precisely set the exact switching point forthe level detecting network.

It is sufficient at this point merely to mention that the level detector network senses the output from the bandpass amplifier and produces a pulse on the output line 114 or 115 associated respectively with the level detect transistor 101 and 103, whenever the output from the bandpass amplifier exceeds a predetermined threshold in either a positive or a negative sense. The output signals which appear on lines 114 or 115 are fed directly into the buffer amplifier 504 and inverter 506 in Figure 6.

Resistor 109 is trimmed so that for the desired negative excursion of the amplifier 87 output, transistor 101 will turn on, causing line 114, to go positive. This signal propagates through buffer 504 causing

70

75

80

85

90

95

100

105

110

115

120

125

130

5

GB 2 048 688 A 5

the output of gate 508 to go positive. In a like manner, resistor 113 is trimmed such that for the desired positive excursion, transistor 103 will turn on, causing line 115 to go low and the output of buffer 504 to 5 go low, thus forcing the output from gate 508 to go high. Thus, no matter whatthe polarity of the amplifier output may be, gate 508 produces a high output.

Figures 3a and 3b illustrate a portion of the cir-10 cuitry used in the ERT detect network 24 in the functional block diagram of Figure 1. These two circuits monitor the battery voltage and produce logic signal outputs for use by the remainder of the ERT detect circuitry, yet to be described. The terminals 158 and 15 160 in Figure 3a are adapted to be connected respectively to the positive bus 50 and the negative bus 52 in Figure 2. A voltage divider including a resistor 162 and an adjustable resistor 164 causes a potential to exist at the mid-tap point 166 which is proportional 20 to the battery voltage. This signal is applied to the base electrode of a transistor 168 to thereby control the conductivity state thereof. The potentiometer 164 is set so that when the potential of the battery exceeds the predetermined voltage selected as the 25 ERT halfway point, the transistor 168 will be conductive and the output appearing on line 170 will be relatively "low" in a binary sense. However, when the voltage of the battery drops below the predetermined halfway point voltage, the transistor 168 will 30 be turned off and the binary signal appearing on terminal 170 will go "high".

The circuit of Figure 3b is substantially identical to that of 3a except that the potentiometer 172 is initially set so that the transistor 174 will be conductive 35 whenever the battery voltage applied between the terminals 176 and 178 exceeds the predetermined ERT voltage. Thus, when the battery voltage drops below the ERT voltage, transistor 174 will be rendered nonconductive and the binary signal 40 appearing on output line 180 will go from a "low" condition to a "high" condition. More will be said about the mode of operation of the level sensing circuit of Figures 3a and 3b when the remainder of the ERT detect network 24 is described hereinbelow. 45 Figures 4a and 4b when arranged as shown in Figure 4 comprise a logic diagram of Crystal Controlled Oscillator 10, the New Rate Storage unit 12, the New Rate Check circuitry 14, the New Width Storage unit 16, and the Nominal or Mag Rate Select unit 26. 50 Referring to Figure 4a, a stable square wave oscillator 200 which is crystal controlled, produces pulses at a desired rate, e.g., 10 KHz at its output terminal 202. Connected to the oscillator output terminal 202 is a frequency divider network 204 which operates 55 upon receipt of the leading edges of the pulses from the oscillator 200 to achieve a predetermined frequency division. In the preferred embodiment, the network 204 divides the input frequency by a factor of 10. Thus, for a 10 KHz oscillator output, the output 60 from the frequency divider appearing at junction 206 (Figure 4b) would be a signal having a pulse repetition rate of 1 KHz. The output from the frequency divider 204 is applied as a first input to a NAND gate 208 whose second input is obtained from the Q out-65 put of a toggle flip-flop 211. The output from the

N AND gate 208 is coupled to the input terminal of a 12-stage binary, ripple-up counter 210 which is advanced on the trailing edge of the input signals applied thereto. The binary counter 210 comprises the New Rate Storage device 12 shown functionally in the block diagram of Figure 1. The true output terminal from the 12 stages of the counter 210 are labelled, respectively, Q1 through Q12 and are individually connected to the Y-input terminals of a multiplexer or selector network 212. The selector 212 corresponds to the "Nominal or Mag Rate Select" unit 26 in Figure 1. In addition to the Y-input terminals labelled Ay, BY LY, the multiplexer 212 also has a plurality of X-input terminals labelled Ax, Bx... Lx. Whenever the input applied to the select input terminal 214 is low, the input data appearing on the X-input terminals will be transferred to the corresponding output terminals, i.e., signal on line Ax will be transferred to the A-output terminal, the signal on the input terminal, Bx will be transferred to the B-output terminal etc. Similarly, when the input signal on the select terminal 214 is high, the signals appearing on the input terminals AY, BY, etc., will be transferred to the output terminals A, B etc.

As shown in Figure 4b, the output terminals of the multiplexer 212 are coupled to the input terminals labelled J1 through J12 of a 12-stage, pre-settable, binary, ripple-down counter 216. The clock input terminal 218 of the ripple-down counter 216 is connected by a conductor 220 to the output of the frequency divider 204 at junction point 206. Hence, the counter 216 will be decremented at a 1 millisecond rate.

The logic circuit shown enclosed by the dashed line box 222 along with the ERT sensing transistor circuits of Figures 3a and 3b comprise the major portion of the "Elective Replacement Time Indicator. Output line 170 of Figure 3a is connected to the input line 224 and the output line 180 of Figure 3b is connected to the input line 226. A FET switching mechanism indicated generally by box 228 is adapted to connect the input lines 224 and 226 to the logic network 222.

Specifically, the input line 224 will be connected through the FET switch to a first input of a NOR gate 230 while the input line 226 will be connected through the FET switch to the input of a NOR gate 232. The second input to each of the gates 230 and 232 comes by way of a conductor 233 from the second stage of a 12-stage binary upcounter462 in Figure 5a by way of inverter 471 in Figure 5b. The output from the NOR gate 232 is applied to the clock input terminal of a D-type flip-flop 235. Similarly, the output from the NOR gate 230 is applied to the clock input terminal of D-type flip-flop 237. The Q output of the flip-flop 235 serves as a first input to an NAND gate 234. The Q output from flip-flop 237 is applied as the other input to NAND gate 234.

A conductor 236 connects the output from NAND gate 234 to the input terminal Gx of the multiplexer 212. Conductors 238 and 240 connect the output from the flip-flop 237 to the input terminals Bx and Hx ofthe multiplexer212. Finally, conductor 242 connects the output from flip-flop 235 to the input terminals Dx and Ex ofthe multiplexer 212. It is to be

70

75

80

85

90

95

100

105

110

115

120

125

130

6

GB 2 048 688 A 6

noted that terminal Ax, Cx, Fx, Ix, Kx, and Lx are wired to ground and thus continually carry binary "low" signals.

The circuitry for accomplishing the "New Rate 5 Check" represented by Block 14 in Figure 1 is shown as being enclosed by the dashed line box 244 (Figure 4b). It includes an OR gate 246 an AND gate 248 and a D-type flip-flop 250, the latter being a leading edge-triggered device. The OR circuit 246 receives as 10 its input a "Rate Reset" signal on conductor 252 and the output from the highest order stage ofthe counter 210. Specifically, stage Q11 of counter 210 is connected by a conductor 254 to the second input terminal ofthe OR gate 246. The AND gate 248 15 receives at its inputs the true output from stages 5 through 9 ofthe 12-stage ripple-up counter 210. The output from AND gate 248 is applied to the "clock" input terminal of a D-type flip-flop 250. The output from the OR gate 246 is connected to the "reset" 20 input terminal ofthe D-type flip-flop 250. A conductor 256 connects the complement output (Q) to a first input terminal of a NOR gate 258 (Figure 4a). The second input to NOR gate 258 connects to the Mag Rate out ofthe flip-flop 414 shown in Figure 5a. It is 25 the output of gate 258 which determines whether the X or Y inputs to the multiplexer 212 will be applied to the input lines J1 through J12 ofthe ripple-down counter 216.

The circuitry for determining the pulse width ofthe 30 stimulating pulses applied to the heart is determined primarily by the circuitry enclosed by dashed line box 260. Included are a binary up-counter 262, a presettable down-counter 264 and an inverter 266. The counter 262 has its output lines Q1 through Q5 con-35 nected to the respective input terminals J1 through J5 ofthe width counter 264. The counter 262 and the counter 264 are each five stages and the count overflow terminal of counter 264 is connected as an input to inverter 266. The "counter over-flow" terminal 40 (C.O.) ofthe rate counter 216 is coupled through an inverter 268 and is applied via conductor 270 to the "presetenable" terminal (P. E.) 272 ofthe 5-stage counter 264. The output from inverter 268 is also coupled by a conductor 274 to the "Reset" terminal 45 of a D-type flip-flop 276. The output from the inverter 266 is applied byway of conductor 278 to the Clock input terminal of D-type flip-flop 276.

Signals comprising a new width count are applied by way of input line 280 to the Clock input terminal of 50 the binary up-counter 262 which serves as the width register. As will be explained more fully hereinbelow, the new width count is obtained at the output of NOR circuit 454 shown in Figure 5b. The width register 262 is reset by the output from an Inverter 55 282. Inverter 282 receives as its input a Width Reset signal on line 284 originating at the output ofthe gate 458 in Figure 5b.

In order to insure at least a given minimum width to the pacer pulses in the event that the newly prog-60 ranrimed width count is erroneously entered, the network shown enclosed by dashed line box 286 is provided. Included are a D-type flip-flop 288 who Reset input terminal is connected to the input conductor 280 and whose Clock input is connected to 65 the output of inverter 282. The true output terminal,

Q, of flip-flop 288 is connected to a first input of a NAND gate 290 and of an inverter 292. The second input to NAND gate 290 comes directly from the output ofthe 10 KHz oscillator byway of conductor 294.

This same 10 KHz signal on conductor 294 is inverted by network 296 and applied to a first input terminal of a NAND gate 298. The second input to NAND gate 298 comes from the output of inverter 292. The outputs from gates 290 and 298 are logically combined " in a NAND gate 300 whose output is connected by way of conductor 302 to the Clock input terminal of the width counter 264. The combination of logic •

components, 290,292,296,298 and 300 merely provide the Exclusive OR function on the output signals from the master clock 200 and the flip-flop 288.

Hence, this circuitry can be replaced with a conventional Exclusive OR device.

In order to test the unit after a new rate or pulse width is entered, additional circuitry is provided to permit six cycles of paced pulse generation even in the presence of normal heart activity This test circuitry is shown as enclosed by dashed line box 304 (Figure 4b). Included are a three stage binary counter 306, an AND gate 308 and a D-type flip-flop 310. The resetterminai ofthe counter 306 is connected to the output of a NOR gate 312 (Figure 4a) by way of inverter 311 and conductor 313. A first input to this last mentioned gate comes from the output of flip-flop 414 in Figure 5a and carries the signal Mag Rate The second input to NOR gate 312 is applied byway of conductor 314 and originates at the complementary output terminal, Q, ofthe D-type flip-flop 250. The Clock input terminal of counter 306 comes from the complementary output terminal Q ofthe D-type flip-flop 276 by way of a conductor 316. The true output terminals of stages 2 and 3 ofthe counter 306 are applied as inputs to an AND gate 308 and the output terminal of gate 308 is connected to the Clock input terminal ofthe D-type flip-flop 310. The flip-flop 310 is triggered on the leading edge ofthe input clock pulse. The true output terminal, Q, of flip-flop 310 is connected to a first input of NAND gate 318. The second input to this last mentioned gate is applied by way of conductor 320 which comprises the Refractory Reset signal originating at the output of flip-flop 568 in Figure 6. The output from inverter 311 * (Figure 4a) is connected by conductors 313 and 322 to the reset terminal ofthe D-type flip-flop 310. The Q output of flip-flop 310 is also coupled to NAND gate 524 (Figure 6). The output from NAND gate 318 is connected to a first input to a further NAND gate 324. The second input to NAND gate 324 originates at the Q output terminal ofthe D-type flip-flop 276. The output from NAND gate 324 is applied by way of conductor 326 to the preset enable (P. E.) terminal of the rate counter 216.

The Q output from D-type flip-flop 276 is also connected by a conductor 332 to a first terminal of NAND gate 330 (Fig. 6). The complementary output, Q, from D-type flip-flop 276 is also applied by way of conductor 332 to the trigger input terminal of a negative edge triggered, non-retriggerable, monostable multivibrator 336 (Fig. 6). The output from the one-shot circuit 336 is applied as a second input to NAND gate 330 (Fig. 6). It is the output from the NAND gate

70

75

80

85

90

95

100

105

110

115

120

125

130

7

GB 2 048 688 A 7

330 which serves as the "Pacer Rate" input to the voltage doubler circuit of Figure 2.

Next to be described is the digital logic networks comprising the digital combination lock 18 in the 5 functional block diagram of Figure 1. This explanation will be facilitated by reference to the logic diagram of Figures 5a and 5b.

There is shown at 400 a magnetic reed type switch ofthe single-pole double-throw type. This switch 10 controls the state of a bistable flip-flop 402 which includes the cross-coupled OR inverters 404 and 406. The flip-flop 402 acts in a conventional fashion to eliminate the effects of contact bounce occurring upon operation ofthe magnetic reed switch 400. The 15 output from the "set" side ofthe flip-flop 402 is connected by conductor 408 to the Clock input terminal of a three-stage input pulse counter 410. It is also connected by a conductor412tothe Resetterminai of a D-type flip-flop 414. A set of four NOR gates 416, 20 418,420 and 422 are provided fortranslating the true and inverted versions ofthe output signals from the binary counter 410. More specifically, the output from stage Q! of counter 410 is applied as an input directly to NOR gates 420 and 422 while the inverted 25 version thereof is applied as an input to the gates 416 and 418. Similarly, the output from stage Q2 of counter 410 is connected directly to an input of gates 416 and 420 and through an inverter to the inputs of gates 418 and 422. Finally, the true output Q3 from 30 the counter 410 is connected directly to an input of NOR gates 418 and 422 and through an inverter to the inputs of gates 416 and 420. The respective output lines from gates 416 through 422 are applied as inputs to an OR gate 424, the output of which is cou-35 pled by a conductor 426 to the reset side of a flip-flop 425. The output from the flip-flop 425 is connected to an OR gate 427 whose output is then connected to the Reset input terminal ofthe binary up-counter 410 by means of a conductor 429. The output from the 40 gates 416,418,420 and 422 are also individually connected to a Clock input terminal of a corresponding set of D-type flip-flops 428,430,432 and 434. The Reset terminals of each of these last-mentioned flip-flops are connected by a conductor 436 to the Q out-45 put terminal ofthe D-type flip flop 414. It is to be further noted that the true output terminal, Q, of flip-flop 428 is connected by a conductor 438 to the "Data" input terminal (D) ofthe flip-flop 430, and likewise, the true output terminal, Q, of flip-flop 430 50 is connected to the D input terminals ofthe flip-flops 432 and 434. A D-type flip-flop herein termed the "Pulse Code Sample Flip-Flop" 448 is provided and when set, provides an enabling signal, via conductor 450, to the gates 416,418,420 and 422. The condi-55 tions under which flip-flop 448 will be set will be presented later in this specification.

A further set of four NOR gates 452,454,456 and 458 is also included in the combination lock circuitry. NOR gate 452 receives as its input, the complemen-60 tary output, Q, ofthe flip-flops 428,430 and 432. In addition, the output from the NOR gate 406 ofthe flip-flop 402 is applied by way of a conductor 460 to an additional input terminal ofthe NOR gate 452. NOR gate 454 receives as its input the complemen-65 tary output, Q, from flip-flops 428,430 and 434 as well as the output from the set side of flip-flop 402. NOR gate 456 has as its inputs the EToutput from flip-flop 430 and the output from NOR gate 420 via inverter 421. Finally, NAND gate 458 received as its inputs the true output from the flip-flop 430 as well as the output from the NOR gate 422.

The output from NOR gate 452 comprises the input signal "New Rate Count" applied to the toggle inputterminal (T) of flip-flop 211 (Figure 4a). The output from NOR gate 454 comprises the input signal "New Width Count" applied to the input line 280 (Figure 4a). The output from NOR gate 456 is the "Rate Reset" applied to the input line 252 in Figure 4a and the output from NAND 458 comprises the signal Width Reset applied to the input line 284 in Figure 4a. The complementary output from the flip-flop 432 comprises the input signal "Clear Rate Enable" applied to the Reset input terminal ofthe toggle flop-flop 211 in Figure 4a.

The complementary output terminal, Q, of flip-flop 414 is applied as a first input to a NAND gate 461. The second input to this last mentioned NAND gate comes from the output ofthe frequency divider 204 in Figure 4a and the output from the NAND gate 461 is applied to the count input terminal, 0, of a 12-stage, binary, ripple up counter 462. The true output from stages Q5, Q7, Qs and Q12 ofthe counter 462 are combined in AND circuit 464 and the output thereof is applied by a conductor 466 to the Clock input terminal ofthe D-type flip-flop 414 and the D input of a flip-flop 467. The flip-flop 467 has its Clock input terminal connected to the output ofthe frequency divider 204 and is therefore energized at a 1 KHz rate. The output from flip-flop 467 is connected to the Resetterminai ofthe counter 462. The output from stage 06 ofthe counter 462 is applied by way of an inverter 468 to the Clock input terminal ofthe D-type flip-flop 448. This flip-flop is adapted to be reset by the output from the NAND gate 461. The signal "Mag Rate" applied to the NOR gate 312 in Figure 4a comes from the true output, Q, ofthe D-type flip-flop 414. The output stage Q2 ofthe program interval counter 462 is connected by a conductor 469, an inverter 471 and a conductor 233 to the input terminals ofthe NOR gate 230 and 232 in Figure 4a. The outputs from stages Q5, Q7 and Qe are combined in AND gate 473 and the resulting logic signal output therefrom is applied to the clock input terminal of a D-type flip-flop 475. A Reset terminal of this last-mentioned flip-flop is connected to the Q output terminal ofthe flip-flop 414. The Q output terminal of flip-flop 475 is connected to the OR gate 427. The complementary output, Q, of flip-flop 475 is connected to the D-input ofthe Pulse Code Sample flip-flop 448.

This completes a description ofthe circuits for implementing the digital combination lock network. Next to be described will be the construction ofthe Refractory Generator 42 and the Noise Detect network 44 illustrated in the functional block diagram of Figure 1. This will be accomplished with the aid of Figure 6, which shows by means of a digital logic diagram, the structure implementing same.

Shown in Figure 6 are first and second flip-flops 500 and 502. The flip-flops 500 may be referred to as

70

75

80

85

90

95

100

105

110

115

120

125

130

8

GB 2 048 688 A

8

the "Sense Refractory Flip-Flop" while the flop-flop 502 may be referred to as the "Paced Refractory Flip-Flop." As was mentioned earlier in this specification, the cardiac pacer ofthe present invention has 5 a dual refractory system incorporated in it. A 256 millisecond refractory period is provided following a naturally occurring heartbeat whereas a 320 millisecond refractory period is established for a heartbeat resulting from the application of a pacer pulse. 10 It will be recalled that a signal is produced on line 114 or line 115 (Figure 2) whenever the heart is producing a natural electrical output, upon contraction, which exceeds a given threshold in the absence of extraneous noise. These signals are applied byway 15 of buffer amplifiers 504 and 506 to the inputs of a NAND gate 508. The output from NAND gate 508 connected to the Clock input terminal ofthe Sensed Refractory Flip-Flop 500. The Paced Refractory Flip-Flop 502 has the "Paced Refractory Enable" signal 20 from Q output terminal of flip-flop 276 (Figure 4b) applied to its Clock input terminal. The complementary output, Q, of flip-flop 502 is connected to the data input terminal ofthe flip-flop 500. The complementary output signals from both flip-flops 500 25 and 502 are applied as inputs to a NAND gate 512. The output from this gate is connected to the D input terminal of a noise flip-flop 514. It is also applied by way of an inverter 516 to the reset input terminal of the noise flip-flop 514. It is also applied byway of an 30 inverter 516 to the reset input terminal ofthe noise flip-flop 514.

The output from NAND gate 508 is also connected by a conductor 518 to a first input terminal of an AND gate 520. The output from this last-mentioned gate is 35 applied to the clock input terminal ofthe D-type flip-flop 514. The output of AND gate 520 is also connected to an input of a NAND gate 522. A second input to gate 522 comes from the Q output terminal ofthe noise flip-flop 514. NAND gate 522 has its out-40 put connected first to one input of a four-input NAND gate 524; a second input to this last-mentioned gate comes from the output of flip-flop 310 (Figure 4b) and another input comes from the flip-flop 414 in Figure 5a which is the "Mag Rate" signal. The final 45 input to the NAND gate 524 comes from the output ofthe NAND gate 512.

The output from NAND gate 524 is connected to the resetterminai of a nine-stage binary up/down counter 526 which is referred to herein as the 50 "Refractory Interval Counter". This counter is advanced by pulses appearing on the input line 528, each time the NAND gate 530 is enabled. Gate 530 has two inputs, one being connected to the output from NAND gate 512 and the other coming from the 55 output ofthe frequency divider 204 in Figure 4a by way of line 207.

In this particular application, the signals appearing only at the output of stages 7,8 and 9 ofthe counter 526 are of interest. The output from stage Q7 is 60 applied to a number of gates. Specifically, gate 532 has a first input connected to stage Q7 and a second input connected to the output from stage Qg ofthe counter 526. The stage Q7 output is also connected to an input to a NAND gate 534 whose output is cou-65 pled by a conductor 536 to a first input of a NAND

circuit 538. The output ofthe NAND circuit 538 is connected to the resetterminai ofthe Sensed Refractory Flip-Flop 500. The output from NAND gate 534 is also coupled through a NAND circuit 540 to the reset terminal ofthe Paced Refractory Flip-Flop 502 by way of a conductor 542. The second input to NAND gate 540 is provided by the output from the NAND gate 532. The second input for the NAND circuit 538 comes from stage 9 ofthe counter 526 by way of an inverter 544. The output from stage Q7 of counter 526 is applied to an AND gate 546 along with the output from stage Qa ofthe counter. The output from gate 546 is connected by conductor 548 to the resetterminai of a D-type flip-flop 550. The clock input terminal of this flip-flop 550 is connected by conductors 552 and 554 to the output from the flip-flop 276 (Figured) which provides the "Paced Refractory Enable".

A signal from the Q output terminal of flip-flop 500 is logically combined with the output from AND circuit 546 in the NAND gate 556. The output from NAND gate 556 is applied as a first input to a still further NAND gate 558 whose other input is connected to the output ofthe inverter 544. The output from gate 558 is connected to the resetterminai of a D-type flip-flop 560. The complementary output from the flip-flop 560 is connected by way of conductor 562 to a second input terminal ofthe AND circuit 520.

The output from NAND gate 522 is connected to a Clock input terminal of a D-type flip-flop 564. The Q output terminal of flip-flop 564 is applied as a second input to the NAND gate 534. The resetterminai of flip-flop 564 is connected to the output from the inverter 516.

A pair of additional D-type flip-flops 566 and 568 along with a NAND gate 570 are also included in the dual refractory period network circuitry. The clock input terminal of flip-flop 566 is connected to the Q output terminal ofthe flip-flop 500. The Q output terminal of flip-flop 566 is connected to the Data (D) input terminal ofthe flip-flop 568.The clock terminal of flip-flop 568 is connected byway of a conductor 572 to the 10KHz clock line which connects to oscillator 200 in Figure 4a. The Q output terminal of flip-flop 568 is connected to the resetterminai ofthe the flip-flop 566. The signal emanating from the Q output terminal of flip-flop 568 is the "Rate Counter Reset" signal utilized by the circuitry of Figure 4a.

This completes a description ofthe construction of the preferred embodiment ofthe invention. To follow will be a discussion of its operation in its various modes.

Operation Paced Mode With Programmed Rate

When operating in the paced mode with a programmed rate the following initial conditions apply;

1. The binary counter 210 (Figure 4b) contains a count which corresponds to the programmed rate interval. .

2. The flip-flop 250 is set and the Magnetic Rate line (Figure 5b) is low.

Under the asumed condition gate 258 is fully enabled and produces a high output signal to the input select line 214 ofthe multiplexer 212. As is indicated by the legend on the drawing, this selects

70

75

80

85

90

95

100

105

110

115

120

125

130

9

GB 2 048 688 A

9

the Y inputs ofthe multiplexer 212 such that each of the Qt through Q12 outputs from the binary register 210 are presented as the J, through J12 inputs to the binary rate counter 216.

5 Starting with a high pulse from the NAND gate 324 (Figure 4b), the preset enable input (PE) ofthe rate counter 216 is high causing the count from the rate register210to be entered into the rate counter 216. The 1 KHz clock signals emanating from the fre-10 quency divider 204 are applied to the clock inputter-minal 218 ofthe rate counter 216 such that at each 1 millisecond interval, the counter 216 is decremented until it reaches zero. Then, on the next clock pulse, the counter 216 reverts to its all 1 's state. This causes 15 the carry output (C.O.) signal to go low. This low signal is inverted by circuit 268 such that the signal on conductor 270 is high. This high signal is applied to the "Reset" terminal of flip-flop 276 by means of conductor 274 and the flip-flop 276 is thereby reset. 20 This results in the removal ofthe preset enable from the counter 216 (by way of gate 3^4) which causes the output from the inverter 268 to return to its low state.

The preset enable applied to the width counter 264 25 causes the contents ofthe width register 262 to be transferred into the width counter 264. The 10KHz pulses occurring at the output from the crystal controlled oscillator 200 are applied by way of conductor 294 and through the minimum width control logic, 30 shown enclosed by dashed line box 286, such that 0.1 millisecond clock pulses are applied by way of conductor 302 to the width counter 264. These clock pulses cause the count in counter 264 to be decremented until, upon reaching all 1's stage the 35 carry output terminal (C.O.) goes low. This low signal is inverted by circuit 266 causing a high clock signal to be applied to the flip-flop 276 by way of conductor 278. The clock signal applied to the flip-flop 276 returns it to its "set" state. It can be seen, 40 then, that a pulse whose width has been determined by the contents of the width register 262 has been generated at the Q output terminal ofthe flip-flop 276 at a rate determined by the contents ofthe rate register 210. This pulse from the flip-flop 276 passes 45 through the gate 330 (Figure 6) to generate the "Pacer Rate" signal.

The Pacer Rate signal from the gate 330 is coupled through the resistor 58 (Figure 2) to turn on transistors 60 and 64. With transistor 64 conducting, the 50 potential at its collector drops and the transistor78 is also turned on. With transistor 78 conducting, a pulse approximately equal to the sum ofthe battery voltage plus the voltage across the capacitor 82 is applied to the heart electrode terminals 70 and 74 55 and to the heart ofthe patient via implanted electrodes (Not shown).

Paced Mode With Magnetic Rate

When a permanent magnet is placed in proximity to the implanted pacer, the reed switch 400 (Figure 5) 60 closes, causing an "O" signal to be applied at the input of NOR circuit 404 ofthe flip-flop 402. This results in a setting ofthe flip-flop 402 and the logical "1" output from gate 404, when applied by way of conductor 412 causes the flip-flop 414 to be reset, 65 causing the "Magnetic Rate" line (Figure 5b) to go high. This, in turn, causes the output ofthe NOR gate 258 (Figure 4a) to assume a low state which has the effect of selecting the X inputs ofthe multiplexer 212. Thus, the pulse rate will now be determined by the combined outputs ofthe gate 234 and flip-flops 235 and 237. That is, rather than having the contents ofthe binary counter 210 transferred to the Rate Counter 216 as was the situation in the "Paced Mode With Programmed Rate" described above, the new rate will be a function ofthe logic circuits enclosed by dashed line box 222 along with the prewired low inputs of the multiplexer 212.

The placing ofthe permanent magnet in proximity to the pacer also causes the output of gate 312 (Figure 4a) to go low. The P channel FET'S (in box 228) which are normally off to conserve power now turn on while the N channel FET'S turn off. Resistor 164 (Figure 3a) is adjusted such that transistor 168 will be at the threshold of conduction when the battery is at the predetermined voltage selected as the ERT halfway point (e.g., 2.3 volts). Thus for a battery voltage above this value, line 170 will remain low while for a battery voltage below below this halfway point, it will go high. Transistor 174 is similarly set at the predetermined ERT value (e.g., 1.8 volts) After the Magnetic Rate Mode has been initiated, a 2 millisecond delay allows time for lines 170 and 180 to settle to their correct state. Subsequently, the Q2 output of counter 462 (Figure 5a) goes high causing the output of inverter 471 to go low. This results in flip-flop 237 being clocked to the set state, if transistor 168 is conducting and also causing flip-flop 235 to be clocked to the set state, if transistor 174 is conducting.

Let it first be assumed that the battery voltage is at its beginning of life (BOL) value. Under this assumed condition the ERT 90 line 224 and the ERT 85 line 226 (lines 170 and 180 in Figures 3a and 3b) are both low. Thus, the Q output of flip-flop 235 will be high (a binary 1), the Q output of flip-flop 237 will be low (a binary 0) and the output of gate 234 will be high (a binary 1). It can be seen that this causes the X inputs to the multiplexer 212 to assume the binary state 001001011000 (60010). (The A input is the least significant bit). This corresponds to an interval of 600 milliseconds and, hence, the pacer rate will be 100 beats per minute.

As the battery discharges during its life and falls halfway to the elective replacement time (ERT)

value, transistor 168 (Figure 3a) becomes nonconducting, causing the ERT 90 line to go high. The Q output of flip-flop 235 will now be a 1, the Q output of flip-flop 237 will be a 1 and the output of gate 234 will be a 0. This generates a binary count of 001010011010 (66610), at the X inputs ofthe multiplexer 212. This number corresponds to a 666 millisecond interval or 90 beats per minute pacer rate.

As the battery voltage decays lower and falls to the ERT voltage, both transistors 168 and 174 in Figures 3a and 3b will be nonconducting and both the ERT 90 and ERT 85 lines 224 and 226 assume a high condition. Under these circumstances, the Q output from flip-flop 235 is a zero, the Q output of flip-flop 237 is a 1 and the output of gate 234 is a 1, generating a binary count of 001011000010 (70610). This corres70

75

80

85

90

95

100

105

110

115

120

125

130

10

GB 2 048 688 A

10

ponds to an interval of 706 milliseconds or 85 beats per minute. Thus, it can be seen thatthe condition of the battery can be inferred by monitoring the pacer rate when the system is in its magnetic mode. Note 5 that flip-flop 235 and 237 cannot be reset while in the Magnetic Rate Mode. This feature eliminates the possibility of erratic fluctuations in rate when either of transistors 168 and 174 are exactly atthe conduction threshold.

10 Refractory Timing Generation

Figure 6 contains all ofthe circuitry associated with the generation ofthe refractory interval for the pacer of the present invention. The refractory intervals are controlled by the 9-stage, binary, ripple-up 15 counter 526. In the normal mode, it will be recalled thatthe Magnetic Rate line is high. Also, the flip-flops 500 and 502 and the counter 526 are initially reset. If a normal heart signal is picked up at the heart electrode terminals 70 and 74 (Figure 3) it will 20 be amplified, filtered and level detected by the circuitry of Figure 2 and if the amplitude ofthe electrical signal from the heart exceeds the requisite threshold, the transistor 101 or 103 (Figure 2) will be turned on. This will cause a positive or negative 25 pulse to be produced atthe output ofthe buffer amplifier 504 or inverter 506 (Figure 6) which clocks the flip-flop 500 into its "set" condition. The positive going edge ofthe Q output from the flip-flop 500 clocks flip-flop 566 to the set state. The next leading 30 edge ofthe 10KHz clock sets flip-flop 568 which resets flip-flop 566 causing a reset of flip-flop 568 on the next clock cycle. Thus, a 0.1 millisecond positive going Rate Counter Reset pulse is generated on line 517. This pulse is applied byway of line 320 (Figure 35 4a) to an input of gate 318 (Figure 4b). The output of gate 318 will therefore go low, causing the gate 324 to produce a high output signal, which acts as a preset enable forthe Rate Counter 216 and allows the contents ofthe Rate Register 210 to be re-entered 40 into the Rate Counter 216. At the same time, NAND gate 512 (Figure 6) produces a high output signal which, when applied to an input ofthe NAND gate 524 causes the output of this last-mentioned gate to go low, thereby releasing the reset on the Refractory 45 Interval Counter 526.

The output from gate 512 also enables gate 530 such that the 1 KHz clock signal can be propagated to the counter input terminal 528. Upon each trailing edge of a clock signal applied to the counter input 50 terminal 528, the contents ofthe counter will be incremented until the output Q9 (the highest order stage) assumes its 1 state. When this happens, the signal applied byway of inverter 544 and NAND gate 538 to the reset input terminal ofthe flip-flop 500 55 causes the flip-flop 500 to be reset. This causes the NAND gate 512 to output a low signal and the gate 524 to output a high signal which resets the Refractory Interval Counter 526 to a value of zero. With a 1 KHz clock rate, it will take 256 milliseconds for stage 60 9 ofthe counter to become a binary 1. This corresponds to the basic 256 millisecond refractory period associated with a sensed natural heart beat. During this time, additional heart output signals picked up by the electrodes 70 and 74 are inhibited from 65 generating a rate counter reset pulse. As long as there is a sensed natural heart signal following the refractory period which generates a rate counter reset prior to the countdown of counter 216 (Figure 4b) the device will run in this normal sensed mode with a 256 millisecond refractory period. Whenever, a natural heart signal is missing such that the rate counter reset is not generated, the counter 216 will complete its countdown, which generates a Pacer Pulse and a Paced Refractory Enable which clocks the flip-flop 502 to its set condition. The complementary output, Q, ofthe flip-flop 502 causes a low or zero signal to be applied to the data terminal, D, of flip-flop 500, thereby preventing it from being set. In this manner, the state of flip-flop 502 now controls the refractory period. The pacer pulse also sets flip-flop 550 causing transistor 100 to turn on and Differential Bandpass Amplifier 87 to be switched to a zero gain state. As such, currents flow to charge capacitors 86,107 and 108 to hold the bases of transistors 101 and 103 at their quiescent level independent of any input signals. In this mode, the Amplifier 87 can rapidly recover from large noise pulses induced by a pacer pulse when transistor 100 is turned off. Amplifier 87 assumes a high gain state and input signals then will cause excursions from this fixed quiescent level. By incorporating this controlled gain feature, problems related to amplifier recovery following a natural heart signal are minimized.

As was true before, the output from NAND gate 512 goes high, causing the counter 526 to begin counting clock pulses arriving at a 1 KHz rate via gate 530. At a count of 192, AND gate 546 goes high which resets flip-flop 550 thus restoring Amplifier 87 to its normal gain. This time, when the count reaches 25610, the signal applied to the resetterminai of flip-flop 500 has no effect since it was never set. Counting therefore proceeds to a binary count of 101000000 which corresponds to a 32010 millisecond interval. When the aforementioned count is reached, the gate 532 will be fully enabled and will produce an output signal that passes through NAND gate 540 and on line 542 which resets the flip-flop 502. The D input ofthe flip-flop 500 now assumes a "1" state and any subsequent natural heart signals will generate a refractory reset. Thus it can be seen that a 320 millisecond refractory period is established after the occurrence of a paced heart beat.

Noise Detection.

The noise detection circuitry operates to assure thatthe rate counter will only be reset on a sensed heart beat and not on spurious noise. When a Paced Refractory period is initiated by setting flip-flop 502, the output ofthe inverter 516 goes low, releasing the reset on flip-flops 514 and 564. Flip-flop 560 is also set by the negative going pacer pulse applied to the input of NAND gate 570. With the Q output of flip-flop 560 low, noise pulses are inhibited by AND gate 520. Two hundred and fifty-six milliseconds into the refractory period, stage Q9 ofthe refractory interval counter 526 sets, causing the output of inverter 544 to go low and the output of NAND gate 558 to go high, thus resetting flip-flop 560. A subsequent noise pulse will now propogate through AND gate 520 setting flip-flop 514 and causing the output of NAND

70

75

80

85

90

95

100

105

110

115

120

125

130

11

GB 2 048 688 A 11

gate 522 to go low, thus resetting counter 526 via NAND gate 524. On the trailing edge ofthe noise pulse the output of NAND gate 522 goes high, setting flip-flop 564 and thus enabling NAND gate 534. 5 if counter 526 reaches a count of 64,0, NAND gate 534 goes high, generating a reset of flip-flop 502 via NAND gate 538. However, another noise pulse will reset counter 526 and thus extend the refractory period an additional 64 milliseconds.

10 When flip-flop 502 is reset, flip-flops 514 and 564 are also reset via NAND gate 512 and inverter 516. Should the noise condition persist, the rate counter 216 will eventually count down through zero and generate a pacer pulse. The noise detection circuits 15 function in the same manner in the "sensed mode" except that noise detection is started earlier in the refractory period. The Q output from flip-flop 568 sets flip-flop 560 via NAND gate 570 so, as before, noise pulses are initially inhibited. With flip-flop 500 20 set NAND gate 556 is enabled thus when Q7 and Qg of counter 526 go high at a count of 192i0 the output of AND gate 546 goes high causing flip-flop 560 to be reset via NAND gates 556 and 558, thus initiating noise detection 64 milliseconds earlier than the 25 "paced mode".

To summarize this section ofthe specification dealing with the refractory period generation, without rate counter resets, the unit goes into a paced mode with pacer pulses being generated at a rate 30 determined by the contents ofthe rate register210. For a rate counter reset to be generated, flip-flop 500 must be reset. Once flip-flop 500 or 502 has been set, it can only be reset through achieving the appropriate count in the refractory interval counter 526. Any 35 noise pulses during the noise detection period will cause the counter 526 to repeatedly reset and to extend the refractory period until a 64 millisecond noise free interval has passed. Furthermore, the "Magnetic Rate" mode causes the counter 526 to be 40 continually reset by virtue ofthe output from the NAND gate 524.

Combination Lock Circuits

In order to reprogram a new pulse rate into the implanted pacer from an external source, it is first 45 necessary to unlock the combination lock circuitry by presenting to it a predetermined code word. A three digit pulse train has been adopted in the preferred embodiment. In order to effect a change in the pulse rate, a code 5-3-4 must be presented. Similarly, in 50 order to change the pulse width ofthe pacer signal, the code 5-3-2 must be presented.

Referring to Figures 5a and 5b, initially all ofthe flip flops 428 through 434 and the counters 462 and 410 are reset. Flip-flop 414, however, remains set. On 55 the first cycle ofthe magnetic reed switch 400

caused by the pulsing of an external electromagnet, the output from the set side ofthe flip-flop 402 goes high which resets the flip-flop 414 via a pulse on conductor 412. This enables the IKHz clock signals to 60 propagate through NAND gate 461 and to begin incrementing the Program Interval Counter 462. The flip-flop 414 remains reset throughout the programming interval, causing the pacer to run in the "Magnetic Rate" mode previously described. The 65 external electromagnet positioned in proximity to the implanted pacer operates at a modulation frequency of 100 Hz. Each cycle ofthe magnetic reed switch 400 causes the input counter 410 to increment. When the program interval counter 462 70 reaches a count of 6410, inverter 468 outputs positive going clocking signals forthe Pulse Code Sample flop 448. The complementary output, Q, of flip-flop 448 goes low and then is immediately reset to the high state by the next clock pulse appearing at the 75 output of NAND gate 461. The output from the flip-flop 448 on conductor 450 therefore serves to sample the state ofthe input pulse counter 410 by enabling the gates 416 through 422. If five pulses have been registered in counter 410 during this first 64 80 millisecond interval, the count in counter 410 will be 1012 and the outputs of inverters 411 and413 will be low as will the output of stage Q2 of counter 410.

With all four inputs ofthe gate 416 low, its output goes high, causing the output ofthe OR gate 424 to 85 go high, thus setting flip-flip 425 which then resets counter 410. At the same time, flip-flop 428 is clocked to its "set" state. For any count otherthan five, the output of gate 416 remains low and counter 410 and flip-flop 428 are unchanged. The counter 410 90 then accumulates counts during the next sampling interval and 64 milliseconds later, the Q output of flip-flop 448 again goes low. If a count of three has been entered into counter 410, the output of gate 418 will go high, clocking flip-flop 430 and also resetting 95 the counter 410 by way ofthe output from OR circuit 424, flip-flop 425, and OR circuit 427. In a like manner, a count of four will enable gate 420 and a count of two will enable the gate 422.

It should be noted at this point that the output 100 from the flip-flop 428 feeds the flip-flop 430 and the output of flip-flop 430 feeds flip-flop 432, such that flip-flop 430 cannot be set until flip-flop 428 has first been set, and neither flip-flops 432 and 434 can be set until flip-flop 430 has been set. This, then is the 105 basis forthe combination lock feature. In the first 64 milliseconds time period, there must be a count of five, in the second such time period there must be a count of three and in the third such time period there must be either a count of 4 or 2 to set flip-flops 432 110 and 434 respectively. Any other pulse pattern is selected without altering the counters 210 or 262 (Figures 4b and 4a). Since the combination lock requires a specifically different count for each ofthe sampling intervals it is impossible for noise of any 115 single frequency to open the combination. Furthermore, it is very improbable that any random pattern of noise would match the combination.

After three sampling pulses have been presented from the Q output of flip-flop 448, flip-flop 475 gets 120 set causing the D input of flip-flop 448 to go low, thus inhibiting the production of further sampling pulses. Counter 410 is also maintained in a reset condition via gate 427. This assures that gates 416,418,420 and 422 will also be inhibited through the remainder 125 ofthe programming cycle.

For a valid 5-3-4 code, the output from gate 420 goes high causing the inverter 421 to output a low signal. Since flip-flop 430 had previously been set, the output from NOR circuit 456 goes high producing 130 the "Rate Reset" signal which is applied byway of

12

GB 2 048 688 A

12

conductor 252 to the Reset input terminal ofthe Rate Register 210 in Figure 4b. Atthe same time, the flip-flop 250, which is the Program Rate Check flip-flop, is reset by the output from OR circuit 246.

5 When flip-flop 432 assumes its "set" condition, the "Clear Rate Enable" line goes low releasing the reset on the toggle flip-flop 211. The next cycle ofthe reed switch will cause the output from gate 452 to go high, toggling flip-flop 211 to its "set" state. This 10 allows the 1 KHz clock signal to propagate through gate 208 and begin incrementing the counter 210 at a 1 millisecond rate. This incrementing operation continues until a second cycle ofthe reed switch 400 generates a second pulse at the output of gate 452 15 which toggles the flip-flop 211 to the "reset" state and terminates the incrementation ofthe rate register 210. In this manner, the resulting count in the rate register 210 is determined by the time interval between the first rate count pulse and the second rate 20 count pulse. Use of this gated clock method is considerably faster than counting input pulses. The use of this latter method might require up to one minute to complete the programming function.

If a valid 5-3-2 code is entered by way ofthe 25 magnetic reed switch 400, then flip-flop 434 is set. At this time, gate 422 is producing a high output, causing the output from NAND gate 458 to go low. This low signal is applied by way of conductor 284 (Figure 4a) and causes the output from inverter 282 to go 30 high, thereby resetting the Width Register counter 262. The output from inverter 282 is also effective to set the flip-flop 288. For each subsequent programmer pulse, the output of gate 454 goes high, producing the "New Width Count" signal which is applied 35 by way of conductor 280 to the clock input terminal ofthe Width Register 262. Hence, the contents ofthe Width Register are incremented each time a programmer pulse causes the output from gate 454 to go high.

40 In the event that there are no width count signals emanating from NOR gate 454, the flip-flop 288 (Figure 4a) will remain in its set condition. Should an excessive number of width counts be applied to the counter 262, it will overflow and automatically recy-45 cle to its lowest count and then increment again. The effect of this action is that atthe end ofthe programming operation, flip-flop 288 will be cleared unless counter 262 is in its zercrstate. With flip-flop 288 set, the gates 290,292,296,298 and 300 cause 50 the phase ofthe clock input to the counter 264 to be inverted. This has the effect of delaying the time for the counter to change state by 1/2 clock pulse or .50 milliseconds. Thus the circuitry contained in the dashed line box 386 comprises a hard-wired connec-55 tion assuring at least a minimum width to the pacer pulses. This feature is provided to guard against a failure to enter any width count after successfully addressing the pulse generator.

It should be noticed that for any other code than 60 the 5-3-4 or the 5-3-2 codes, gates 452,454,456 or 458 are not enabled and the pre-existing width value and rate value remain unchanged. After a period of 2,256 milliseconds has elapsed following the start of a reprogramming operation, the output from gate 65 464 goes high which clocks flip-flop 414 to its "set"

state thereby inhibiting clock pulses to counter 462. On the next clock pulse, flip-flop 467 sets, resetting counter 462, which causes the output of gate 464 to clear flip-flop 467 on the following clock pulse. The 70 output from the Q side ofthe flip-flop 414 is applied byway of conductor 436 to the Reset terminals of each ofthe flip-flops 428,430,432 and 434, resetting same and thereby conditioning them, or the combination lock circuitry for receipt of subsequent 75 reprogramming codes.

Programming Rate Check and Runaway Protect

The "Magnetic Rate" mode enables NOR gate 312 (Fig 4a) causing the three stage counter 306 to be reset by way ofthe output from inverter 311 on con-80 ductor 313. Similarly, the Magnetic Rate signal on line 322 resets the flip-flop 310. This inhibits the "Rate Count Reset" pulse on line 320 from propagating through the NAND gate 318. Hence, the unit is always operating in the paced mode when the 85 magnetic rate mode is operative. After removal of the permanent magnet used to establish the magnetic rate, the counter 306 is incremented each time that flip-flop 276 is clocked. This continues until a binary count of 110 (decimal 6) is reached at which 90 time the output from gate 308 goes high, thus clocking flip-flop 310 to the "set" condition. This allows Rate Counter Reset pulses to propagate on through the gate 318. This feature provides for six paced intervals running atthe programmed rate even in the 95 presence of naturally occurring electrical heart activity which is operative to produce refractory resets. When the flip-flop 276 is clocked to its "1" state, the nonretriggerable monostable multivibrator 336 (Figure 6) is triggered to its active state causing its Q 100 output to go positive for a predetermined time period. Specifically, the output from the one-shot circuit 336 will remain positive, independent of any further input trigger signals, for a period of 462 milliseconds. If during this interval a second pacer pulse 105 generated, it will be inhibited by the gate 330. Thus, the paced rate can never exceed 130 beats per minute. If a rate in excess of 130 beats per minute is attempted, the monostable multivibrator suppresses every other pulse to thereby halve the output rate. 110 During rate reprogramming, the flip-flop 250 (Figure 4b) serves as a check on the rate count. When counter 210 reaches a binary value of 11111000 (49610), the output from gate 248 goes high, clocking flip-flop 250 to the "set" state. If the counter runs 115 beyond 204710, bit Q12 of counter 210 goes high causing flip-flop 250 to reset by way of a pulse from OR _ gate 246. If at the end of a reprogamming cycle the Q output of flip-flop is high, then counter 306 will be maintained in the reset state by way of NOR gate 312 120 and inverter 311 and the Mag Rate inputs to multiplexer 212 will be selected by way of NOR gate 258. The pacer will therefore run in the "Paced Magnetic Rate" mode, unless the rate count is between 49610 and 204710. This corresponds to a programmed rate 125 between 30 beats per minute and 120 beats per minute.

GENERAL

The logic devices shown in Figures 4,5 and 6 are all commercially available components. For exam-130 pie, the D-type flip-flops such as flip-flops 500,502,

13

GB 2 048 688 A

13

508,510 and 512,276 288,448,414., etc., may be the type CD 4013M/CD 4013C dual D-type flip-flops manufactured and sold by the National Semiconductor Corporation. As such, they comprise monolithic 5 complementary MOS integrated circuit constructed with N and P channel enhancement transistors. The logic level present atthe "D" input is transferred to the Q output during the positive going transition of the clock pulse. Setting or resetting is independent 10 ofthe clock and is accomplished by a high level on the set or reset line respectively. The rate and width counters 210,216,262 and 264 may comprise type CD4029M/CD4029C presettable, binary/decade, up/down counters also manufactured by the 15 National emi-conductor Corporation. A logical "1" preset enable signal to such a counter allows information atthe "jam" inputs to preset the counter to any state asynchronously with the clock. The counter is advanced one count atthe positive-going edge of 20 the clock if the carry-in and preset enable inputs are at logical "0" levels. Advancement is inhibited when either or both of these two inputs is at a logical "1" level. The "carry out" signal is normally at its logical "1" state and goes to the logical "0" state when the 25 counter reaches its maximum count in the "up" mode or the minimum count in the "down" mode, provided the carry input is at logical "0" state. The program interval counter 462 may be implemented with the National Semiconductor Corporation type 30 CD 4040M/CD4040C counter. This is a 12 stage ripple-carry binary counter having buffered outputs from each stage externally available. The counter is reset to its logical "0" state by a logical "1" applied to its reset input terminal. The counter is advanced 1 35 count on the negative transition of each clock pulse.

NAND gates, NOR gates, AND gates and OR gates are also commercially available in integrated circuit technology.

From the above description, it is apparent that the 40 apparatus ofthe present invention provides a pulse generator capable of providing artificial stimulating electrical pulses in the absence of normal heart activity, with the pulse rate and pulse width of such stimulating signals being programmable by an 45 externally located mechanism. The digital implementation lends itself to microminiaturization which is, of course, a desired feature in implanted stimulator devices. The device ofthe present invention also incorporates requisite safety features to 50 prevent inadvertent or spurious operation thereof to the detrement of the user.

Claims (7)

1. An implantable heart pacer for applying artificial heart stimulating pulses of a predetermined 55 pulse width and repetition rate to the heart of a patient, comprising:

(a) a first binary register for storing a first predetermined binary number value corresponding to a desired stimulating pulse repetition rate; 60 (b) a second binary register for storing a second predetermined binary number value corresponding to a desired stimulating pulse width;

(c) a first presettable counter having a clock input terminal, a load enable input terminal, a carry output 65 terminal and a plurality of input terminals opera-

tively coupled to the outputs of said first binary register for periodically receiving said first binary number value each time its load enable input terminal is stimulated.

(d) a second presettable counter having a clock input terminal, a carry output terminal, a load enable input terminal and a plurality of input terminals operatively coupled to the outputs of said second binary register for periodically receiving said second binary number value each time its load enable input terminal is stimulated.

(e) a source of regularly occurring timing signals coupled to said clock input terminals of said first and second presettable counters for changing the numbers stored therein at a fixed rate;

(f) means coupling said carry output terminal of said first presettable counter to said load enable terminals of said first and second presettable counter; and

(g) output terminal means for coupling said carry output terminals of said first and second presettable counters to the heart ofthe patient, the arrangement being such that when said first binary number value initially in said first presettable counter is changed by said timing signals to a given value, a heart stimulating pulse is initiated and when said second binary number value entered into said second presettable counter is changed by said timing signals a given value, said heart stimulating pulse is terminated.

2. An implantable heart pacer according to Claim 1 and further including;

digital logic means responsive to binary pulse code patterns generated external to the body of a patient for producing an enabling signal only when pulse code patterns of predetermined binary significance are received during a preset time interval; and means coupling said digital logic means to said first and second binary registers and responsive to said enabling signal for enabling the storing of a new digital value in said first and second binary registers.

3. An implantable heart pacer according to Claims 1 or 2, and further including;

digital multiplexer means having first and second sets of input data terminals and a single set of output terminals, said multiplexer means having said first set of input terminals connected to the output of said first binary number register and said single set of output terminals connected to the input terminal of said first presettable counter;

magnetically actuable means responsive to the presence of an externally applied magnetic field for generating digital codes of predetermined significance indicative of a voltage magnitude; and means coupling said magnetically actuable means to said second set of input data terminals or said digital multiplexer.

4. An implantable heart pacer according to any one of Claims 1,2 or3 and further including counter reset means responsive to the occurrence of a natural heartbeat for stimulating said load enable input terminal of said first presettable counter only if said natural heartbeat occurs subsequent to a predetermined period following the next preceding one of said natural heartbeats.

70

75

80

85

90

95

100

105

110

115

120

125

130

14

GB 2 048 688 A 14

5. An implantable heart pacer according to Claim 4 wherein said counter reset means comprises; electrode means coupled to the heart ofthe patient; amplifier means having input terminals coupled to

5 said electrode means and an output terminal;

digital timing means including a third digital counter having a clock input terminal coupled through first gating means to said source of regularly occurring timing signals, a resetterminai and a plurality of 10 output terminals;

a bistable flip-flop having an input terminal coupled to said output terminal of said amplifier and an output coupled through second gating means to said resetterminai of said third digital counter and to said 15 first gating means; and third gating means coupled to said plurality of output terminals of said third digital counter for inhibiting stimulation of said load enable input terminal of said first presettable counter until the number of tim-20 ing signals accumulated in said third digital counter following the setting of said flip-flop reaches a predetermined value.

6. An implantable heart pacer according to Claim 5, further including noise sensing means coupled to

25 the output of said amplifier means, and to said second and third gating means for resetting said third digital counter upon the detection of noise during the period that said third digital counter is accumulating timing pulses and prior to the time 30 when said third digital counter reaches said predetermined value.

7. An implantable heart pacer according to Claim 1, in which said magnetically actuable means is responsive to the application of a specific applied

35 magnetic field to cause said pacer to operate at a first defined rate in the presence of said field and to operate at its asynchronous rate for a set number of pulses immediately after removal of said specific field.

Printed for Her Majesty's Stationery Office by The Tweeddale Press Ltd., Berwick-upon-Tweed, 1980.

Published atthe Patent Office, 25 Southampton Buildings, London, WC2A1 AY, from which copies may be obtained.