Pulsed passive charge recovery circuitry for an implantable medical device

Abstract

The problem of a potentially high amount of supra-threshold charge passing through the patient's tissue at the end of an Implantable Pulse Generator (IPG) program is addressed by circuitry that periodically dissipates only small amount of the charge stored on capacitances (e.g., DC-blocking capacitors) during a pulsed post-program recovery period. This occurs by periodically activating control signals to turn on passive recovery switches to form a series of discharge pulses each dissipating a sub-threshold amount of charge. Such periodic pulsed dissipation may extend the duration of post-program recovery, but is not likely to be noticeable by the patient when the programming in the IPG changes from a first to a second program. Periodic pulsed dissipation of charge may also be used during a program, such as between stimulation pulses.

Claims

1. A pulse generator, comprising: a plurality of electrode nodes, wherein each electrode node is configured to be coupled to one of a plurality of electrodes in contact with a patient's tissue; a reference voltage; a plurality of switches, wherein each of the switches is configured to couple the reference voltage to a different one of the electrode nodes; stimulation circuitry configured to provide stimulation pulses to two or more of the electrode nodes; and control circuitry configured to repeatedly close and open at least some of the plurality of switches during a time period when the stimulation circuitry is not providing stimulation pulses, wherein the time period occurs between successive stimulation pulses provided to the two or more of the electrode nodes.

2. The pulse generator of claim 1, wherein the control circuitry is configured to select and repeatedly close and open only the switches coupled to the two or more electrode nodes during the time period.

3. The pulse generator of claim 1, wherein the control circuitry is configured to repeatedly close and open all of the switches during the time period.

4. The pulse generator of claim 1, wherein the control circuitry is configured to repeatedly close and open the at least some of the plurality of switches at a constant frequency during the time period.

5. The pulse generator of claim 1, wherein the control circuitry is configured to repeatedly close the at least some of the plurality of switches for a duration that varies over the time period.

6. The pulse generator of claim 1, wherein the control circuitry is configured to repeatedly close and open the at least some of the plurality of switches at a variable frequency during the time period, but wherein the control circuitry is configured to repeatedly close the at least some of the plurality of switches for a duration that is constant over the time period.

7. The pulse generator of claim 1, wherein the control circuitry is configured to repeatedly open the at least some of the plurality of switches for a duration that varies over the time period.

8. The pulse generator of claim 1, wherein the stimulation circuitry is configured to provide the stimulation pulses to two or more of the electrode nodes during a program, and wherein the control circuitry is configured to repeatedly close and open the at least some of the plurality of switches after an end of the program.

9. The pulse generator of claim 1, further comprising a plurality of direct current (DC) blocking capacitors, wherein each electrode node is configured to be coupled to one of the electrodes via one of the DC blocking capacitors.

10. The pulse generator of claim 1, further comprising a plurality of resistors, wherein each of the resistors is serially connected to one of the switches between one of the electrodes nodes and the reference voltage.

11. The pulse generator of claim 10, wherein the control circuitry is further configured to vary a resistance of the plurality of resistors.

12. The pulse generator of claim 1, further comprising a battery, wherein the reference voltage comprises a voltage of the battery.

13. The pulse generator of claim 1, wherein the stimulation circuitry is powered by a compliance voltage, and wherein the reference voltage is a function of the compliance voltage.

14. The pulse generator of claim 13, wherein the compliance voltage is variable.

15. The pulse generator of claim 13, wherein the reference voltage is one half of the compliance voltage.

16. The pulse generator of claim 1, further comprising a conductive case, wherein at least one of the electrode nodes comprises a case electrode node configured to be coupled to the conductive case.

17. The pulse generator of claim 1, further comprising at least one implantable lead, wherein the electrodes are located on the at least one implantable lead.

18. The pulse generator of claim 1, wherein the control circuitry configured to repeatedly close and open at least some of the plurality of switches during each of a plurality of time periods when the stimulation circuitry is not providing stimulation pulses, wherein the time periods occur between different successive stimulation pulses provided to the two or more of the electrode nodes.

19. A method for operating a pulse generator comprising a plurality of electrode nodes each configured to be coupled to one of a plurality of electrodes in contact with a patient's tissue, comprising: actively driving stimulation circuitry in the pulse generator to form a succession of stimulation pulses to two or more of the electrode nodes; and without actively driving the stimulation circuitry, providing passive charge recovery during time periods between successive stimulation pulses, wherein during each time period charge is passively recovered by repeatedly closing and opening switches connected at least to the two or more electrode nodes.

20. The method of claim 19, wherein the switches are configured to couple the two or more electrode nodes to a reference voltage.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIGS. 1A-1C show an Implantable Pulse Generator (IPG), and the manner in which an electrode array is coupled to the IPG, in accordance with the prior art.

(2) FIG. 2A shows an architecture for an IPG utilizing at least one Application Specific Integrated Circuit (ASIC), and FIG. 2B shows circuitry blocks within the ASIC and connection to off-chip components.

(3) FIG. 3A shows aspects of the Digital-to-Analog converters within the stimulation circuitry of the ASIC, and stimulation pulses formable thereby, while FIG. 3B discloses passive recovery switches and related circuitry used during passive charge recovery phases after each stimulation pulse.

(4) FIGS. 4A and 4B show examples of stimulation programs, and the accumulation of charge on DC-blocking capacitors despite use of passive charge recovery phases following pulses of the stimulation program.

(5) FIG. 5 shows use of a post-program passive recovery period, and the equivalent circuit formed thereby.

(6) FIGS. 6A and 6B show improved passive charge recovery circuitry for the IPG, and passive recovery logic for closing passive recovery switches coupled to each of the electrode nodes.

(7) FIG. 7 shows different examples of pulsed post-program recovery periods useable after a program operating in the IPG has ended to discharge capacitance in the IPG.

(8) FIG. 8 shows generation of control signals usable to implement a pulsed post-program recovery period.

(9) FIG. 9 shows use of pulsed post-program recovery period between programs operating in an IPG, which program may comprise stimulation programs or other operational IPG programs.

(10) FIG. 10 shows use of pulsed discharging during a program and in particular between stimulation pulses of a program.

DETAILED DESCRIPTION

(11) As noted earlier, an IPG 10 may include circuitry and techniques designed to remove the charge from DC-blocking capacitors 55 in the electrode output paths that provide stimulation to a patient's tissue, Rt. An IPG 10 may issue biphasic pulses, with the second pulse phase 94b (FIG. 3A) designed to actively recover charge stored on the DC-blocking capacitors 55. Beyond this, a passive charge recovery phase 98 may be implemented to recover any remaining charge that was not recovered during active charge recovery (at the end of second pulse phase 94b).

(12) Using both active (94b) and passive (98) charge recovery techniques, the charge Q across DC-blocking capacitors 55 in any electrode output path involved in providing stimulation is ideally zero before those electrode output paths need to issue a next pulse. In other words, the voltage across such capacitors (V=Q/C) is ideally zero at the end of each passive charge recovery phase 98.

(13) However, the inventors realize this ideal goal is not always achievable, particularly when the DAC circuit 172 of the IPG 10 is instructed to form pulses of a high frequency, f, such as, but not limited to, ≥5 kHz. Such high frequency pulses as issued pursuant to a stimulation program (SP1) are shown in FIG. 4A, which again comprise biphasic pulses with first 94a and second 94b pulses phases. Also shown in FIG. 4A is passive charge recovery phase 98 which follows each second pulse phase 94b. Notice that because the frequency f is high, the duration of each passive charge recovery phase 98—from the end of a second pulse phase 94b of one pulse to a first pulse phase 94a of a subsequent pulse—must be relatively short. The inventors note that the duration of this phase 98 may be too short to passively recovery all charge remaining on the DC-blocking capacitors 55.

(14) This is shown at the bottom of FIG. 4A, which shows the voltage across the DC-blocking capacitors 55 involved in providing stimulation, e.g., capacitors C1 and C2 associated with electrodes E1 and E2. After a first phase 94a of a first pulse, the voltage across each capacitor C1 and C2 (VC1 and VC2) is high (X), but is reduced to near zero (Y) after the second active-charge-recovery pulse phase 94b. A passive charge recovery phase 98 begins thereafter. However, as shown in FIG. 4A, the end of the passive charge recovery phase 98—although reducing VC1 and VC2 still further (Z)—does not bring these voltages exactly to zero.

(15) As a result, a next pulse (with first and second phases 94a′ and 94b′) will not start with capacitors C1 and C2 that are completely discharged, i.e., VC1 and VC2 don't equal zero. This increases the voltage across the capacitors: X′ after the first pulse phase 94a′ is higher than after first pulse phase 94a X; Y′ after the second pulse phase 94b′ is higher than after second pulse phase 94b Y; and Z′ at the end of the passive charge recovery phase 98′ is higher than at the end of passive charge recovery phase 98 Z. As subsequent pulses are issued, these voltages begin to climb over time period t1, as shown by the dotted line in FIG. 4A.

(16) Eventually, the voltages across the capacitors will increase to a steady-state condition as shown during time period t2 in FIG. 4A. In the depicted example, the voltages across the capacitors C1 and C2 are established around a new steady-state baseline voltage, such as 3V, as shown in the dotted line of t2. This baseline voltage could differ depending on the particulars. In any event, from this baseline, the voltages across the capacitors increase after a first pulse phase 94a″ (to X″), then decrease after a second pulse phase 94b″ (to Y″), and then decrease back to the baseline (to Z″=3V) at the end of a passive charge recovery phase 98″. The steady-state baseline voltage across the capacitors is reached at Z″ because—as discussed further below—the passive charge recovery phases 98 decrease the voltage across each capacitor exponentially. When the baseline voltage is high enough during time period t2, the exponential drop experienced during the passive charge recovery phases 98 are significant enough to keep the voltages from increasing, and thus allowing a steady-state to be reached.

(17) While is it desirable that the voltages across the capacitors VC1 and VC2 be zero at the end of each second pulse phase 94b, or at least at the end of each passive charge recovery phase 98, it is not necessarily a problem that this doesn't occur, particularly if the compliance voltage VH is adjustable as it is in IPG 10. As described earlier, IPG 10 includes on ASIC 160 compliance voltage generation circuitry 76 (FIG. 2B) that produces a compliance voltage HV. Compliance voltage VH can vary during the provision of stimulation pulses, preferably in conjunction with the feedback provided by measurements such as the voltage drops across the PDAC 172p (Vp) and NDAC 172n (Vn) used in providing stimulation, as described earlier. An increase in VC1 and VC2 will cause Vp and Vn to drop for a given VH (FIG. 3A), such that Vp and Vn would now indicate that VH is too small. Thus, compliance voltage generation circuitry 76 would raise VH to compensate.

(18) This is shown in FIG. 4A. Compliance voltage generation circuitry 76 may initially during time period t1 set VH to 8V as an example. As the voltages VC1 and VC2 across the capacitors C1 and C2 grow, compliance voltage generation circuitry 76 increases VH. Thus, during time period t2, when both VC1 and VC2 are established around a steady-state baseline of 3V, VH has been increased from 8V to 14V (to compensate for VC1+VC2=6V). To summarize, even though the voltages VC1 and VC2 across the capacitors C1 and C2 increased over time, DAC circuitry 172 is still able to output high-frequency pulses with the proper amplitude (+A, −A) by increasing the compliance voltage, VH.

(19) FIG. 4B shows another example of stimulation program SP1 in which charge builds on the DC-blocking capacitors 55 over time. In this example, monophasic pulses are used consisting of only a first pulse phase 94a. In between pulse phases 94a are passive recovery phases 98. This stimulation program, which lacks active recovery second pulse phases 94b (FIG. 4A), is again logical to use at high frequencies f where pulse periods T are necessarily short. Having said this, monophasic pulses can be used at lower frequencies as well. Note that the duration of the pulse phases 94a (D1) and the passive recovery phases 98 (D2) may be equal (e.g., 100 microseconds each), although they may also have different durations.

(20) Similarly to FIG. 4A, and as shown at the bottom of FIG. 4B, the voltages (VC1, VC2) across the capacitors (C1, C2) associated with the active electrodes (E1, E2) increase during monophasic pulse phase 94a (to X), and then will exponentially decay during the passive recovery phase 98, although it does not do so perfectly as a residual voltage remains at Y. The voltages thus increase during time period t1 for subsequent pulses (e.g., X′, Y′), and eventually establish at a baseline (X″ and Y″) during time period t2 as before, and with the compliance voltage increasing to compensate for the additional voltages dropped across the capacitors.

(21) The fact that the DC-blocking capacitors 55 cannot be completely discharged raises concern noticed by the inventors, which are illustrated in FIG. 5. At the end of a program—such as either of the stimulation programs SP1 illustrated in FIG. 4A or 4B—the inventors find it desirable to close all of the passive recovery switches 96(x) during a post-program recovery period 99 before allowing the IPG 10 to execute a new stimulation program SP2. New program SP2 may involve wholly different stimulation parameters and electrodes, or may involve changes to the stimulation parameters and electrodes involved during program SP1, such as a change in pulse amplitude A. By way of review, closing all of passive recovery switches 96(x) can be affected by the assertion of all of control signals RCVx (FIG. 3B).

(22) Closing all passive recovery switches 96(x) during post-program recovery period 99 at the end of a program such as SP1 is desirable in the inventors' view. First, this will recover any charge that had built up on the DC-blocking capacitors 55 (e.g., C1 and C2) associated with the electrodes (e.g., E1 and E2) selected to provide stimulation during program SP1, which as just explained may be significant. Other DC-blocking capacitors 55 not associated with electrodes involved in providing stimulation during program SP1 (e.g., C3-C17) would normally have no charge stored on them, and thus no voltage across them. But this may not always be the case, and thus the inventors' preference for closing all passive recovery switches 96(x) during the post-program recovery period 99. For example, such other electrodes may have charge stored on them by virtue of being active in another timing channel different from that running stimulation program SP1. Further, an electrode may have experienced a fault, such as an open circuit, and its DC-blocking capacitor 55 may have stored charge prior to the fault.

(23) When the passive recovery switches 96(x) are closed during post-program recovery period 99 (RCVx=‘1’), the equivalent circuit 95 of FIG. 5 results. Equivalent circuit 95 includes the series connection of passive recovery resistor 97 Rx and DC-blocking capacitor 55 through each switch 96x for each electrode Ex. These series connections are coupled in parallel between the common reference voltage (e.g., Vbat) and the patient's tissue (e.g., Rt). Because the capacitors may be charged to different values (e.g., VC1=VC2=3V, while other capacitors are charged to 0V), a discharge current, I.sub.D, will flow through the patient's tissue, Rt, as the equivalent circuit 95 seeks to equilibrate the amount of charge across each of the DC-blocking capacitors 55. Preferably, at the end of the post-program recovery period 99, each capacitor stores little or no charge, and hence will have little or no voltage across them.

(24) Because the equivalent circuit 95 is an RC circuit, the discharge current I.sub.D will start at a maximum (I.sub.D(max)) and will exponentially decay from this value down to zero at a rate dictated by its time constant, τ, as shown in FIG. 5. If we assume typical values for the components in the equivalent circuit 95 (e.g., R1=R2=30052; Rt=500Ω; C1=C2=5 μF) and that capacitors C1 and C2 are significantly charged as they are end the end of SP1 (e.g., VC1=VC2=3V), then the maximum discharge current I.sub.D(max) may equal several mA, and the time constant τ may equal a few milliseconds. Note that to equate and/or remove all of the charge remaining on the DC-blocking capacitors 55 during the post-program recovery period 99, the duration of this period 99 is preferably significantly longer that the time constant τ. For example, the duration of period 99 may be 10τ, or a few tens of milliseconds in the depicted example.

(25) How the discharge current I.sub.D will flow through the patient's tissue Rt during the post-program recovery period 99—i.e., the extent to which the discharge current will flow to or from any particular electrode Ex—will depend on different factors, such as the location of the electrodes Ex relative to one another in the patient's tissue, Rt, which electrodes are charged during time period t2 and to what extent, etc. Nonetheless, the discharge current I.sub.D can occur with a significant magnitude (I.sub.D(max)) and—because τ is relatively large—for a significant amount of time. In total, a significant amount of charge may pass through the patient's tissue during the post-program recovery period 99. This discharge current I.sub.D therefore runs the risk of being supra-threshold—that is, felt by the patient. Moreover, the discharge current may be significant enough to actually cause discomfort to the patient, who may perceive the discharge current as an unwanted “zap” to their tissue.

(26) Note that the equivalent circuit 95 of FIG. 5 may also be formed during each passive charge recovery phase 98 during program SP1 (e.g., FIGS. 4A, 4B), but this circuit is unlikely to cause patient discomfort in this context. The pulses 94a/94b formed during SP1 are high frequency (e.g., f≥5 kHz), and therefore the duration of each passive charge recovery phase 98 would be of short duration (e.g., ≤0.2 ms, or even less when duration D of the pulse phases 94a and 94b are considered). As such, only a small amount of charge Q (Q=I*t) would be removed from the DC-blocking capacitors 55 during each passive charge recovery phase 98. The inventors reason that removal of such a small amount of total charge from the DC-blocking capacitors 55 during such a short-duration phase 98 is sub-threshold—that is, undetectable by the patient. By contrast, the significantly long duration of the post-program recovery period 99 together with the relatively high amount of current I.sub.D passing through the patient's tissue Rt during this period 99 may be supra-threshold as already mentioned.

(27) The inventors address the problem of a potentially high amount of supra-threshold charge passing through the patient's tissue post-program by periodically dissipating only small amount of the charge stored on the DC-blocking capacitors 55 during use of a pulsed post-program recovery period 100. This occurs by periodically activating the control signals RCVx during a post-program recovery period to form a series of discharge pulses after the end of the stimulation program, i.e., during a time period when the stimulation circuitry is not providing stimulation pulses. This will extend the duration of post-program recovery because gaps are included between the discharge pulses during which no charge is recovered from the DC-blocking capacitors 55. Nonetheless, the extended duration of pulsed post-program recovery period 100 should be tolerable, as it is not so long as to likely be noticed by the patient when the programming in the IPG 10 changes from a first to a second program.

(28) IPG architecture 140 as introduced previously includes passive recovery circuitry used to remove charge on the DC-blocking capacitors 55 during each passive charge recovery phase 98 between pulses issued during a program. Such passive recovery circuitry has been modified to remove charge post-program—i.e., after the program has ended—during a pulsed post-program recovery period 100. FIGS. 6A and 6B show specific circuitry details. FIG. 6A shows an improved stimulation circuitry block 170 within ASIC 160, which can include the DAC circuitry 172 used to source or sink stimulation currents at the electrode nodes 61a via a PDAC 172p or NDAC 172n respectively (FIG. 3A). It is again assumed that ASIC 160 supports sixteen electrode nodes 61a E1′-E16′ coupleable to sixteen electrodes 16 E1-E16 used to stimulate a patient's tissue (e.g., on lead(s) 18; see FIG. 1A), as well as a case electrode node (Ec′) coupleable to a case electrode 12 (Ec). PDAC and NDAC circuitry 172p and 172n can be as described earlier, or may take other forms, or may be as disclosed in the '520 Publication referenced above. Note that PDAC 172p and NDAC 172n can comprise constant current sources or constant voltage sources.

(29) Within improved stimulation circuitry block 170 is a recovery control block 174. Recovery control block includes a mode/resistance select control module 175 that stores data used to select different passive recovery modes and different resistances Rx for the passive recovery resistors 97 (FIG. 3B). As mode and resistance adjustment is described in detail in the above-incorporated '937 Patent, it is not further discussed here, although such adjustments could be used in disclosed embodiments as well.

(30) Recovery control block 174 further includes a pulsed discharge block 180 configured to initiate and control operation of a pulsed post-program recovery period following the end of a program. The digital logic of the pulsed discharge block 180 can be made aware when the end of a program occurs in different ways. For example, pulsed discharge block 180 may receive a control signal P(end) at the end of a program. Control signal P(end) may issue from the ASIC 160—such as from the microcontroller 150 via internal bus 192 (FIG. 2B). The pulsed discharge block 180 may also understand when a program operating in the stimulation circuitry has ended, and in this regard may monitor the currently running program (SP) as shown in FIG. 6A.

(31) The recovery control block 174—under control of blocks 175 and 180—issues control signals to recovery logic circuitry 176. These control signals include Rec[17:1], which indicate when passive recovery switches 96(x) (FIG. 3B) for the case electrode (Rec 17) and electrodes E16-E1 (Rec16-Rec1) are to be turned on (i.e., closed). Control signal EnRec comprises a passive recovery enable signal, which as described below will enable control of the passive recovery switches 96. Control signal AllOn comprises a control signal that turns on all of the passive recovery switches 96 as described in detail later. As described in the above-incorporated '937 Patent, the recovery logic 176 ultimately produce control signals RCV[17:1] each of which controls a passive recovery switch 96 (FIG. 3B) associated with the case 12 (Ec) or one of the electrodes E16-E1. FIG. 6B shows one example of logic circuitry 176, which is discussed in detail later with reference to FIG. 8.

(32) FIG. 7 shows different manners in which a post-program recovery period can be controlled by the pulsed discharge block 180 in the recovery control block 174 to dissipate the charge on the DC-blocking capacitors 55 at the end of a program (e.g., SP1). The top shows the approach described earlier (FIG. 5) in which all passive recovery switches 96 are continually closed during a post-program recovery period 99. As noted earlier, this may lead to a significant and supra-threshold amount of charge being passed through the patient's tissue, Rt.

(33) Periods 100, 100′, and 100″ comprise examples of pulsed post-program recovery periods, which are preferable as they periodically permit small sub-threshold amounts of charge to pass through the patient's tissue. In each of periods 100, 100′ and 100″, the dissipation of charge occurs periodically during discharge pulses 101, during which all passive recovery switch control signals RCVx are preferably asserted to close their associated passive recovery control switches 96. Less preferably, only the individual passive recovery switch control signals RCV associated with the previously-active electrodes during program SP1 (e.g., RCV1 and RCV2 for electrodes E1 and E2) are asserted, but this option isn't further discussed, as it would not assist in dissipating charge stored on other non-active capacitors arising from parasitic coupling from the tissue Rt as explained earlier. Between discharge pulses 101 are gaps 102 during which none of the passive recovery switch control signals RCVx are asserted, and thus all passive recovery switches 96 are opened. In other words, DC-blocking capacitors 55 are not discharged during gaps 102.

(34) In a first example of a pulsed post-program recovery period 100, the duration of the discharge pulses 101, tx, are fixed, and the duration of the gaps 102, ty, between them are also fixed. In one example, the duration tx of the discharge pulses 101 comprises 20 microseconds or less, and the duration ty of gaps 102 comprise 500 microseconds or less. These values are preferred because it is believed that discharge through the tissue Rt of 20 microseconds or less will be sub-threshold and not felt by the patient, regardless of the magnitude of the discharge current I.sub.D. In other words, the total charge passing through the patient's tissue during a discharge pulse 101 should be insignificant from a patient perception standpoint. Further rendering such stimulation imperceptible is the significant duration ty of the gaps 102 between such discharge pulses 101, which allow the tissue ample time to recover.

(35) Note that periodic dissipation of the stored charge may lengthen the post-program recovery period. For example, if a continuous dissipation of the stored charge occurs over a period of tens of milliseconds (99), then fractionalizing this dissipation (100) will extend the length of this period by a factor of (tx+ty)/tx, which may be several hundred milliseconds for the values of tx and ty provided earlier. In other words, pulsed post-program recovery period 100 might last hundreds of milliseconds before a new program is run by the IPG 10 to ensure that the DC-blocking capacitors 55 have been discharged. While this is significantly longer than a continuous post-program recovery period 99, pulsed post-program recovery period 100 from the patient's standpoint comprises only a short delay during which the IPG 10 will transition from a first program SP1 to a new program SP2, which may not even be noticeable by the patient as a practical matter.

(36) In second and third examples of pulsed post-program recovery periods in FIG. 7, modifications are made to reduce the time needed for recovery. In pulsed post-program recovery period 100′, subsequent discharge pulses 101 are increased in duration, with duration tx2 being longer than tx1; tx3 being longer than tx2, etc. Because the discharge current I.sub.D decreases over time and hence during subsequent discharge pulses 101, increasing the duration of the discharge pulses can keep the total amount of charge (Q=I*t) relatively constant for each discharge pulse. This is especially helpful for discharge pulses 101 nearer to the end of the pulsed post-program recovery period 100′, where discharge pulses 101 are low enough in current that it is significantly less risky to the patient to issue these pulses with greater durations. In the example shown, the frequency f.sub.D at which the discharge pulses 101 issue is constant; therefore, as discharge durations tx increase over time, the gap durations ty decrease over time (with ty1>ty2>ty3, etc.). Although not shown, eventually the duration tx of the discharge pulses 101 will equal that set by the frequency (1/f.sub.D), at which point the passive recovery switches 96 may be held constantly on.

(37) It is not strictly necessary that duration ty of the gaps 102 decrease as the duration tx of the discharge pulses 101 increase. Instead, although not shown, the duration of the gaps 102 may be kept constant between the discharge pulses 101 in pulsed post-program recovery period 100′. In effect then, the discharge pulses 101 would not issue with a constant frequency, but with a frequency that decreases over the duration of the period 100′. Still, in any of these variations, because the duration tx of the discharge pulses 101 increase, the duration of the total period 100′ can be smaller than for period 100.

(38) Pulsed post-program recovery period 100″ at the bottom of FIG. 7 is also of a smaller duration than period 100. However, in this example, the duration of the discharge pulses 101 are kept constant (tx), while the duration of the gaps 102 decreases over time (e.g., ty1>ty2>ty3, etc.). Thus, notice in this example that the frequency f.sub.D of the discharge pulses 101 is not constant, but increases over time. Reducing the duration ty of the gaps 102 is sensible in this example, because smaller amount of charge are discharged from the DC-blocking capacitors 55 over time. Hence, gaps 102 can be made smaller, as the patient's tissue will need less time to recover to decrease the amount of charge.

(39) All of the above examples 100, 100′ and 100″ by which discharge pulses 101 are issued during a pulsed post-program recovery period can be modified or combined. What is important is that a plurality of discharge pulses issue during the pulsed post-program recovery period, using constant or variable discharge pulse durations tx, gap durations ty, and/or pulse frequencies f.sub.D, such that the entirety of the charge stored on the DC-blocking capacitors 55 is not continuously discharged through the patient's tissue, Rt, as occurs in post-program recovery period 99.

(40) As noted earlier, pulsed discharge block 180 (FIG. 6A) can control the passive recovery circuitry to affect pulsed post-program recovery periods. This occurs by block 180 issuing various control signals at appropriate times corresponding to timing of the discharge pulses 101, which pulses as explained earlier can differ in their duration tx and in the duration ty of gaps 102 between them. As noted earlier, it is preferable that all passive recovery switches 96 be closed during each discharge pulse 101. This can be affected by pulsed discharge block 180 and recovery logic 176 of FIG. 6B in different ways, two of which are explained in FIG. 8.

(41) In a first option (Op1) shown in FIG. 8, operation of the passive recovery switches 96 is enabled (EnRec=‘1’), and pulsed discharge block 180 asserts all of control signals Rec1-Rec17 during the discharge pulse 101 durations tx. In recovery logic 176 (FIG. 6B), this causes OR gate 182 to output a ‘1’, and AND gate 184 to output a ‘1’. Because Rec1-Rec17 are asserted, all of OR gates 185 would also output a ‘1’, and so each of AND gates 186 would output a ‘1’ as control signals RCV1-RCV17. As noted earlier (see FIG. 3B), this closes all of the passive recovery switches 96 during the discharge pulse 101 durations tx. During gap 102 durations ty, Rec1-Rec17 are not asserted, and hence OR gates 185 would output a ‘0’. This would cause each of AND gates 186 to output a ‘0’ as control signals RCV1-RCV17, thus opening all of the passive recovery switches 96 during the gaps 102.

(42) In a second option (Op2) shown in FIG. 8, operation of the passive recovery switches 96 is again enabled (EnRec=‘1’), and pulsed discharge block 180 asserts AllOn during the discharge pulse 101 durations tx. In recovery logic 176 (FIG. 6B), this again this causes OR gate 182 to output a ‘1’, and AND gate 184 to output a ‘1’. Moreover, because AllOn=‘1’, all of OR gates 185 output a ‘1’ regardless of the state of Rec1-Rec17. Thus, each of AND gates 186 would output a ‘1’ as control signals RCV1-RCV17, which again closes all of the passive recovery switches 96 during the discharge pulse 101 durations tx. During gap 102 durations ty, Rec1-Rec17 are not asserted, and hence OR gates 185 would output a ‘0’. This would cause each of AND gates 186 to output a ‘0’ as control signals RCV1-RCV17, thus opening all of the passive recovery switches 96 during the gaps 102.

(43) To this point, pulsed post-program recovery periods 100, 100′, 100″ or like post-program periodic discharge periods, have been illustrated as being executed by pulsed discharge block 180 at the end of a stimulation program (such as SP1), and/or after an adjustment to such stimulation program. However, use of the disclosed pulsed post-program recovery periods are not limited to transitions in such stimulation programs, but may also be used more generally at the end of any generic IPG 10 program. This is shown in FIG. 9, in which a pulsed post-program recovery period 100, 100′, 100″, etc., is implemented at the end of a general IPG 10 program (such as P1 or P2) or between such general IPG 10 programs (such as between P1 and P2, or between P2 and P3). General IPG programs P1, P2, or P3 may comprise different operating modes of the IPG 10. For example, programs P1, P2, or P3 may comprise IPG maintenance programs or IPG test programs, which programs may or may not involve providing useful stimulation therapy to the patient. In one example, programs P1, P2, or P3 may comprise a program to test the resistance between various electrodes pairs 16 in the IPG 10, which as noted earlier is valuable to discern for different reasons. Such a resistance test program would normally comprise providing stimulation to the patient, but at levels that are preferably sub-threshold—i.e., not detectable by the patient.

(44) To this point, it has been assumed that use of recovery control block 174 and its pulsed discharge block 180 will provide pulsed discharging of the DC-blocking capacitors 55 only after the end of a program, or between transitioning from one program to another. However, this technique is not so limited, and instead pulsed discharging may be used in other contexts, and during other time periods when the stimulation circuitry is not providing stimulation pulses.

(45) For example, pulsed discharge may occur in between stimulation pulses within a program, as shown in FIG. 10. FIG. 10 essentially mimics FIG. 3 discussed earlier, which shows a stimulation program of biphasic pulses being issued from electrodes E1 and E2. As explained earlier, a passive recovery period 98 follows each pulse (more specifically in this example, the second pulse phase 94b). However, pulse discharge block 180 (FIG. 6A) in this example controls the passive recovery switches 96x during the passive recovery periods 98 to periodically issue discharge pulses 101 with gaps 102 in between, similarly to what occurs during the pulsed post-program recovery periods illustrated earlier (FIG. 7). In other words, the passive recovery switches 96 are not held continually on during the passive recovery periods 98 as typically occurs, but are only on when passive recovery control signals RCV(x) are on during the discharge pulses 101. In the example shown, all recovery control signals RCV(x) are asserted in pulsed fashion during the passive recovery periods 98, and thus all passive recovery switches 96 are closed during those times. However, although not illustrated, the asserted control signals RCVx can be limited by pulsed discharge block 180 to those corresponding to the electrodes being stimulated (e.g., RCV1 and RCV2 for E1 and E2).

(46) Using periodic discharge pulses 101 could be beneficial during passive recovery periods 98 between stimulation pulses for a number of different contexts. For one, although not illustrated, the stimulation pulses may not be biphasic, but instead may monophasic comprising just the first pulse phases 94a as illustrated in FIG. 4B. As the end of such a singular pulse phase, the charge on the DC-blocking capacitors may be significant, and hence risky to discharge using a continuously-on passive recovery period 98: a supra-threshold current may be generated in the patient tissue Rt for the reasons explained earlier. Using a pulsed discharge during period 98 would alleviate this concern.

(47) While the improved passive recovery circuitry has been described as useful to recover charge from DC-blocking capacitors 55, this is not strictly necessary. Some IPG architectures may not use DC-blocking capacitors, yet may still have inherent capacitances that will charge as a stimulation current is provided. Such inherent capacitances may for example occur at various boundaries, such as the boundary between the electrodes and the patient's tissue. The improved recovery circuitry can be used to recover charge in such architectures, even though they lack intentionally-placed capacitances like the DC-blocking capacitors 55.

(48) While disclosed in the context of an implantable pulse generator, it should be noted that the improved passive recovery circuitry could also be implemented in a non-implantable pulse generator, such as an External Trial Stimulator (ETS). See, e.g., U.S. Pat. No. 9,259,574 (describing an ETS).

(49) Although particular embodiments of the present invention have been shown and described, it should be understood that the above discussion is not intended to limit the present invention to these embodiments. It will be obvious to those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the present invention. Thus, the present invention is intended to cover alternatives, modifications, and equivalents that may fall within the spirit and scope of the present invention as defined by the claims.