Copyright ©1996, American Institute of Aeronautics and Astronautics, Inc.

AIAA Meeting Papers on Disc, July 1996A9636949, AIAA Paper 96-2720

NSTAR ion thruster and Breadboard Power Processor functional integrationtest results

John A. HamleyNASA, Lewis Research Center, Cleveland, OH

Luis R. PineroNASA, Lewis Research Center, Cleveland, OH

Vincent K. RawlinNASA, Lewis Research Center, Cleveland, OH

John R. MillerNASA, Lewis Research Center, Cleveland, OH

Roger M. MyersNYMA, Inc., Brook Park, OH

Glen E. BowersGilcrest Electric, Elyria, OH

AIAA, ASME, SAE, and ASEE, Joint Propulsion Conference and Exhibit, 32nd, Lake

Buena Vista, FL, July 1-3, 1996

A 2.3-kW Breadboard Power Processing Unit (BBPPU) was developed as part of the NASA Solar ElectricPropulsion Technology Application Readiness (NSTAR) program. The NSTAR program will deliver an electricpropulsion system based on a 30-cm xenon ion thruster to the New Millennium (NM) program for use as theprimary propulsion system for the initial NM flight. The final development test for the BBPPU, the FunctionalIntegration Test, was carried out to demonstrate all aspects of BBPPU operation with an Engineering ModelThruster. Test objectives included (1) demonstration and validation of automated thruster start procedures, (2)demonstration of stable closed loop control of the thruster beam current, (3) successful response and recovery tothruster faults, and (4) successful safing of the system during simulated spacecraft faults. These objectives weremet over the specified 80-120 VDC input voltage range and 0.5-2.3 output power capability of the BBPPU. Twominor anomalies were noted in discharge and neutralizer keeper current. These anomalies did not affect thestability of the system and were successfully corrected. (Author)

Page 1

NSTAR ION THRUSTER AND BREADBOARD POWER PROCESSOR FUNCTIONAL INTEGRATION TESTRESULTS

John A. Hamley,* Luis R. Pinero,* Vincent K. Rawlin.t and John R. Miller**NASA Lewis Research Center

Cleveland, Ohio 44135

Roger M. MyersttNYMAInc.

Cleveland, Ohio 44135

Glen E. Bowers***Gilcrest Electric

Elyria, Ohio 44035

Abstract

A 2.3 kW Breadboard Power Processing Unit (BBPPU)was developed as part of the NASA Solar ElectricPropulsion Technology Application Readiness(NSTAR) Program. The NSTAR program will deliveran electric propulsion system based on a 30 cm xenonion thruster to the New Millennium (MM) program foruse as the primary propulsion system for the initial NMflight. The final development test for the BBPPU, theFunctional Integration Test, was carried out todemonstrate all aspects of BBPPU operation with anEngineering Model Thruster. Test objectives included 1)demonstration and validation of automated thruster startprocedures, 2) demonstration of stable closed loopcontrol of the thruster beam current, 3) successfulresponse and recovery to thruster faults, and 4)successful safing of the system during simulatedspacecraft faults. These objectives were met over thespecified 80-120 VDC input voltage range and 0.5-2.3output power capability of the BBPPU. Two minoranomalies were noted in discharge and neutralizer keepercurrent. These anomalies did not affect the stability ofthe system and were successfully corrected.

Nomenclature

JA Accelerator Current, mA

JBJDINKVAVBVp

Beam Current, ADischarge Current, ANeutrali/er Keeper Current, AAccelerator Voltage, VBeam Voltage, VDischarge Voltage, VNeutralizer Keeper Voltage, V

Introduction

A flight electric propulsion system based on 30 cmxenon ion system technology is being developed underthe NASA Solar Electric Propulsion TechnologyApplications Readiness (NSTAR) program. TheNSTAR program will deliver a 30 cm thruster, PowerProcessing Unit (PPU), Digital Control and InterfaceUnit (DCIU), and Xenon Feed System (XFS) forintegration on the first New Millennium Spacecraft1'2As part of the NSTAR program, a Breadboard PowerProcessing Unit (BBPPU) and several EngineeringModel Thrusters (EMTs) were developed prior to theaward of a contract for flight hardware.

NSTAR engineering model thrusters incorporate severalinnovations with respect to the state of the art in iontechnology. These innovations include a conical, spunaluminum, integral thruster body/anode, elimination of

* Electrical Engineer, On-Board Propulsion Branch, 21000 Brookpark Rd., MS 301-3t Electrical Engineer, On-Board Propulsion Branch, 21000 Brookpark Rd., MS 301-3, Member AIAA** Electronics Technician, Energy and Spacecraft Branch, 21000 Brookpark Rd., MS 301-1tt Deputy Director, Aerospace Technology Department, 2001 Aerospace Pkwy., 44142, Member AIAA*** Electronic Systems Mechanic, 570 Ternes Ave

Copyright © by the American Institute of Aeronautics and Astronautics, Inc. No copyright is asserted in the United Statesunder Title 17, U.S. Code. The U.S. Government has a royalty-free license to exercise all rights under the copyright claimedherein for Governmental Purposes. All other rights are reserved by the copyright owner.

magnetic metals in the thruster body, and biasing thecathode keeper electrode through a 1 kQ resistorconnected between the keeper and anode. This method ofkeeper biasing eliminated the need for a dedicatedcathode keeper power supply. Technologies pertainingto cathode ignition and life were also transferred fromthe Space Station Plasma contactor program. The resultwas a 2.3 kW, 30 cm thruster with a mass of only 7kg. NSTAR thruster and plasma contactor cathodetechnology have been described in detail elsewhere.3-6

The BBPPU incorporates "dual use" power supplies forthe neutralizer and main cathodes. These power suppliesprovided two switchable outputs tailored to power boththe thruster heaters and anode discharges. Dischargeignition was accomplished with a high voltage pulsetransformer integral with the output filter inductor inthese power supplies. This form of pulse ignition wasfirst demonstrated successfully in arcjet powerelectronics.7 A microcontroller was utilized to sequencethe operation of ah" thruster power supplies and provideclosed loop control of the thruster beam current viaadjustment of the thruster discharge current. TheBBPPU controller automates the thruster cathodeconditioning, ignition, transition to steady state, andthrottling between sixteen user-programmable throttlepoints. Control of the xenon feed system was notincluded in the BBPPU control software, and is residentin the DCIU for the flight system. The NSTARBBPPU has been described in detail elsewhere.8.9

As has been demonstrated in several previous programs,one of the most critical tests in the development of anion propulsion system is the initial integration test ofthe thruster and BBPPU.10.11 In these tests the thrustercontrol laws, power supply functionality, overallstability, and system performance are evaluated. In thecase of the NSTAR system, several power suppliesmust operate in unison, performing discrete regulationof thruster parameters without crosstalk. Further,several procedures which are under the control of thesystem microcontroller including the automatic start ofthe thruster and closed loop control of the thruster beamcurrent must be validated. This paper presents the testobjectives, procedure, and results of the NSTARFunctional Integration Test (FIT), which was the firstintegration of the BBPPU with an EMT. The results ofthis test were immediately provided to the flighthardware developer for incorporation into the final PPUdesign.

Test Objectives

The fundamental objective of the FIT was to

demonstrate the specified functional performance of theintegrated EMT and BBPPU combination. To fulfillthis objective, the operation of the integrated EMT andBBPPU was evaluated over the required power throttleand BBPPU input voltage range by characterizing theintegrated EMT and BBPPU response to startup,shutdown, and power setpoint commands. Based on theresults of the test, required design changes to the EMTand BBPPU would be identified and implemented toachieve the specified system functional performance.

FIT Hardware

Test ArticlesEngineering Model Thruster #2 (EMT 2), was used asthe test thruster. EMT 2 incorporated all engineeringchanges made to the NSTAR thruster based on datafrom 2000 and 1000 hour lifetests of EMT 1,3,12conducted as precursors to a planned 8000 hour lifetimevalidation test of EMT 2. A simplified schematic of thethruster appears in Figure 1. Power was brought to thethruster via seven wires. A single wire from the thrusterconnected the discharge anode to the positive output ofthe beam and discharge power supplies. A new featurefor EMT 2 was a cathode keeper electrode, which wasbiased from cathode common by a connection to thedischarge anode through a Ikfl resistor. Potential leads,not shown in the figure, were used to measure thedischarge voltage. Xenon propellant was providedthrough three feed lines, one for each cathode and onefor the discharge chamber. Detailed descriptions of thethruster are available elsewhere.3-4

A block diagram of the NSTAR BBPPU appears inFigure 2. Power was input to the PPU via two separatepower busses, a 28 VDC regulated bus, which poweredall housekeeping functions including themicrocontroller, and an 80 - 120 VDC high power buswhich supplied power to the four power converterswhich operated the thruster. The discharge andneutralizer power supplies employed a dual usetopology to allow the cathode heater power to be derivedfrom the same converter which supplied power to thecorresponding anode. These power supplies were parallelconverters with a switching frequency of 50 KHz. Highvoltage ignition pulses which were used to initiate thethruster discharges were generated by a pulse transformerwhich also acted as the output filter for the supplies.7The beam and accelerator power supplies were separatephysically, but were controlled in unison during thrusteroperation. The beam power supply was a full bridge,zero-voltage switched converter with a switchingfrequency of 50 KHz. Four secondaries with dedicatedbridge rectifier circuits were placed in series to generate

the 1100 V beam power output. The low output powerof the accelerator power supply enabled a 50 kHzflyback topology to be utilized. A microcontroller (jiG)was used to automate the cathode conditioning, thrusterstart, and steady state operation of the thruster. Sixteenpre-programmed power levels were available for thrusteroperation. These were stored internal to the \iC. Asingle command was available to start the thruster andmaintain steady state operation at any of these powerlevels. The data in the throttling table was modifiableby another command. A sample throttling table isshown in Table 1. Cathode and main flow rates rangedfrom 2.1 seem to 3.0 seem and 6.0 to 23.5 seem,respectively. The \iC also digitized all analog powersupply outputs and power supply status bits andtransmitted the telemetry data over an RS-232 data linkto a command and telemetry display terminal. Detaileddescriptions of the NSTAR BBPPU, its functions, andoperation are available elsewhere.8.9

A laboratory feed system with commercial xenon flowcontrollers and meters was used. The test procedure waswritten such that the behavior of the laboratory XFSwas similar to that of the flight system. The responsetime of the commercial flow controllers, which was onthe order of a few seconds, was much faster than thoseexpected from the anticipated flight system. The flightsystem, which regulates flow rate by maintaining aconstant pressure in a plenum that is connected to aflow restrictor, has a response time on the order ofseveral minutes. Thus, the BBPPU beam currentregulator control loop was given a conservative case testfor stability when encountering time varying xenonflow to the thruster.

Data AcquisitionThree methods of data acquisition were utilized in theFIT. Digital Multimeters (DMMs) were used to directlymeasure all thruster voltages and currents. To eliminatethe effects of cable losses, potential leads were attachedto the thruster anode and main cathode to read thedischarge voltage. The digital thruster data acquired bythe BBPPU |iC was stored at regular intervals on a harddisk drive connected to the BBPPU command and dataterminal. In addition, the BBPPU also provided driversfor a strip chart recorder. The strip chart data arepresented in the Results and Discussion section of thisreport.

Test FacilityThe EMT was mounted on the centerline of the largespace propulsion testbed at NASA LeRC (VF 5). VF 5is a 4.6m x 19.5m cryopumped facility with a xenonpumping speed on the order of 300,000 l/s.13 A block

diagram of the test setup appears hi Figure 3.

BBPPU Operation

Of primary concern in the FIT was the operation andstability of the automated procedures in the BBPPUover the required input voltage and output power range.Prior to the FIT, the individual power supplies hi theBBPPU had already been integrated with a FunctionalModel Thruster (FMT).9 The individual power supplieswere also integrated into a single unit and operated withan FMT under manual control in previousdevelopmental tests. Thus, the stability of theindividual power supply control loops were proven tobe stable. However, integration of the breadboard withthe microcontroller resulted in some modifications tothe power supplies to improve the noise immunity ofthe analog control voltages in the breadboard. Thesemodifications were to later prove troublesome, andcorrective actions were needed to return the BBPPU tospecifications.

Thruster Start ProcedurePrior to thruster start, the flow rates are set to thevalues commensurate with full power operation, and theneutralizer and discharge power supplies are set to theheater output mode. The cathode heaters are energizedfor six minutes at full heater power. After six minutes,the neutralizer power supply is shut down, switched toanode mode and restarted with the pulse ignitor active. Ifthe neutralizer cathode plasma discharge starts, thedischarge power supply is shut down and restarted in asimilar fashion. After discharge ignition, the thruster isallowed to operate without beam extraction for twentyseconds to allow the discharge plasma to stabilize. Afterthe twenty second warm-up period, the high voltage isapplied and the throttling algorithm adjusts the beamcurrent to the proper value. Power supply setpoints arerestricted to those assigned to power level 1 in thethrottling table.

Each cathode is given 20 seconds to start with the highvoltage ignitor. If the cathode fails to start, the BBPPUrecycles back to the beginning of the six minute pre-heat period and attempts to start the cathodes again. If acathode does not start in three attempts, the BBPPUcontroller shuts down the BBPPU, displays a warningmessage on the telemetry terminal, and waits foroperator intervention.

Beam Current RegulatorThe beam current regulator loop relies on theproportionality of the beam current to the dischargecurrent at a fixed flow rate. The beam current is

measured at a rate of one sample per second and iscompared to the desired value. If the beam current is inerror, the discharge current is adjusted in proportion tothe error. The maximum permissible discharge currentadjustment is limited to 100 mA. In the event that thebeam current is unstable, or below 250 mA, thedischarge current is not adjusted.

Throttling AlgorithmThe throttling algorithm for the FIT was identical tothat planned for the flight unit. Because the BBPPUcontroller has no control authority over the feed system,flow at the appropriate rate to the thruster is assumed.

To throttle to a higher power level, the feed system isfirst commanded to the flow rate for the new operatingpoint. When the flow rate stabilizes, the throttle upcommand is then sent to the BBPPU. The BBPPUresponds by first increasing the beam and acceleratorvoltages to the new values in the throttling table. Thebeam current is then increased by the closed loop beamcurrent regulation algorithm, which begins regulation atthe new setpoint. The operation of the beam currentregulator is described in detail in the following section.

Throttle down reverses the process. The discharge powersupply output current is commanded to the new value inthe throttling table, and the beam current regulatorbegins to regulate the current at the new value. Whenthe beam current is stable within ± 0.015 A of the finalvalue, the beam and accelerator voltages are reduced totheir new values, if necessary. When the thruster iselectrically stable, the flow rates are adjusted to theirnew levels.

High Voltage RecyclesThe following thruster faults require the BBPPU to takecorrective action: 1) Jt > 3A, 2) neutralizer extinction,or 3) discharge extinction. Upon detection of a fault, thebeam and accelerator power supplies are immediatelyshut down and the discharge current is commanded to 4A. Following a 500 ms delay, the high voltage isreapplied only if both thruster discharges are lit andanother 500 ms delay follows. The discharge current isthen ramped back to its original value. In the event thatthe recycle was caused by a discharge extinction,BBPPU shuts down and awaits further operatorintervention.

FIT Procedure

System power-up was initiated with the data terminal.The 28 VDC housekeeping power was applied to the

BBPPU and communication between the BBPPU anddata terminal was verified. The 80-120 VDC inputpower supply was set to 100 VDC output and poweredon. The main and cathode flow rates were set to thelevels required for the 2.3 kW power level. Theoperating points selected for the FTT are listed in Table1. For the FIT, the thruster was started at power level 1and throttled up in power in the discrete steps indicatedin Table 1.

The command to operate the thruster at power level 1was sent. This started the thruster at the lowestavailable power level of 0.5 kW. After the thruster wasoperating, flow rates were adjusted to levels appropriatefor low power operation. No flow adjustments werenecessary for the next operating point so the BBPPUwas commanded to increase power to power level 2.Following stabilization at this power level, data weretaken and the BBPPU was commanded to throttle up topower level 3, and the data acquisition process wasrepeated.

Prior to throttling to power level 5, a flow adjustmentwas made to ensure adequate flow to the thruster duringthe power increase. The process was repeated and thethruster was operated at power levels 10 and 15. Whilethe thruster was operating at 2.3 kW input, severalthruster faults requiring high voltage recycles weremanually simulated including anode to tank ground,anode to accelerator grid, accelerator grid to ground, andcathode common to ground. BBPPU behavior duringthese faults was recorded on a high speed digitaloscilloscope.

With the steady state performance evaluation at 100 Vthus completed, the thruster was commanded off. Theinput voltage to the BBPPU was reduced to 80 VDCand the process repeated to evaluate BBPPU performanceat the lower input voltage limit. One more set of datawas taken with the input voltage set to the upper inputlimit of 120 VDC.

Following steady state operation demonstration, severalfaults were induced to characterize the BBPPUmicrocontrollers ability to detect and recover from thesefaults. The input voltage to the BBPPU was set at 100VDC, and the neutralizer keeper lead was disconnectedto preclude thruster ignition. The command was sent tooperate the thruster at power level 1 and the BBPPUreaction was recorded. The neutralizer keeper lead wasreconnected, and the thruster started. Thruster power wasthrottled up to power level 10 and a shorting bar washeld across the anode terminal and facility ground for tenseconds. The shorting bar was then released and the

recovery to steady state operation characterized. Finally,with the thruster operating at power level 15, the 80 -120 VDC bus power was removed, and the DCIUmicrocontroller reaction to this fault was observed.

Finally, the throttle up tune from startup to steady stateoperation at high power was characterized to determinethe total amount of time necessary to start a "cold"thruster.

Results and Discussion

A large amount of data was generated while the FIT wasbeing conducted. For brevity, only the data at 100 VDCBBPPU input and selected fault recovery data will bepresented here. The data at 100 VDC input were typicalof all of the data collected.

Figure 4 is the strip chart record of a typical thrusterstart. The data from the cathode preheat is not shown.Cathode heater power was simultaneously applied toboth cathodes for six minutes. After the six minutepreheat, the BBPPU (J.C turned off the neutralizer powersupply, switched the output to anode mode and turnedthe supply back on with the pulse ignitor. Neutralizerignition was immediate, with some overshoot in theneutralizer current. The peak neutralizer current was 4A, but lasted only 60 ms. This short duration overshootwas determined to be inconsequential to the cathode.The neutralizer current quickly stabilized at the 1.5 Asetpoint. Five seconds later, the main cathode heaterpower was removed, and the discharge power supply re-energized in the anode mode. A similar overshoot wasobserved in the discharge current, but the power supplysettled at the 4 ADC setpoint within 60 ms. A twentysecond discharge stabilization period followed prior tothe application of the high voltage. Both beam andaccelerator power supplies were powered onsimultaneously, and the accelerator supply was allowedto overshoot by design. There was no overshoot in thebeam power supply. Closed loop control of the beamcurrent was immediately engaged and the regulatoradjusted the current to the 0.54 A setpoint within 30seconds. A reduction in neutralizer keeper voltage wasobserved when beam extraction began. This wasexpected due to the higher neutralizer emission currentwith beam extraction.

Following the thruster start, the flow rate to the thrusterwas reduced from the full power setpoint to the powerlevel 1 setpoint. The flow adjustment was completed inapproximately 38 seconds. Figure 5 shows a strip chartrecord of the thruster parameters during the flowadjustment. As expected, the discharge voltage rose with

the flow rate reduction, and the accelerator impingementcurrent was reduced. The beam current showed a modestdeviation of 50 mA during this period. This deviationwas largely due to the rapid change in thruster flow rate,which was at least one order of magnitude faster thanwould be expected with the flight system.

Throttling from power level 1 to 2, shown in Figure 6,did not involve any adjustments to the thruster flow.The increase in power at this level was accomplished byan adjustment in beam voltage only. Upon receipt ofthe throttle up command, the BBPPU increased thebeam voltage from 650 to 850 VDC in approximately20 seconds. During this transition, the thruster beamcurrent remained constant. A slight decrease in dischargecurrent was commanded by the beam current regulator tocompensate for the increase in total accelerating voltageand subsequent increase in ion extraction efficiency. Thetransition to power level 3 from power level 2 requiredan increase in beam voltage to 1100 VDC and wasaccomplished without a change in the xenon flow. Thereaction of the thruster parameters were identical tothose shown in Figure 6. Subsequent throttling tohigher power levels required increases in both flow rateand discharge current to allow the beam current toincrease to the higher levels.

Figure 7 shows the effects of both the flow rate increaseand the throttle up in engine power from level 3 to 5.Approximately 15 seconds were required to adjust theflow rate to the new level. During this time period, aslight increase in beam current was noted, but the closedloop regulator returned the beam current to the originalsetpoint within 10 seconds. This was illustrated by thereduction in discharge current. Upon receipt of thethrottle up command, the discharge current began toramp up to increase the beam current to the new value.The throttle up was complete within 30 seconds, and noovershoot in beam current was noted.

A xenon flow rate increase was again needed for thethrottle up to power level 10. This was accomplished in10 seconds, and this rapid change caused a perturbationin the beam current, as shown in Figure 8. Upon receiptof the throttle up command, the accelerator voltage wasdecreased in a stepwise fashion to -180 V prior to theincrease in beam current. This was to ensure that theincrease in beam current did not approach the electronbackstreaming limit of the grids. After receipt of thethrottle up command, the discharge current began toramp up until the beam current reached the targetedvalue. Shortly after the throttle up was complete, a highvoltage fault and recycle occurred. The BBPPU recoveredcorrectly, and the beam current was restored to its

original value.

The throttle up to power level 15 was identical to thepower level 5 to 10 procedure, with the exception of thechange hi accelerator voltage. After all data were takenat power level 15, a thruster shutdown command wasissued. The thruster shutdown was as expected. The datafor the 80 and 120 VDC BBPPU input cases wereidentical to the data presented here.

After completing the 120 VDC input cases, the thrusterwas shut down, and the electrical lead to the neutralizerkeeper lead was removed. The BBPPU was thencommanded to start the thruster. The neutralizer failed toignite and the BBPPU recycled back and re-entered thecathode heat period. After three attempts to light thethruster, the BBPPU sent the appropriate failuremessage to the data terminal and shut down. Followingthe shutdown, the neutralizer keeper lead was attached.

The thruster was restarted and throttled up to powerlevel 10. While the thruster was operating in the steadystate mode, the input voltage to the BBPPU was variedfrom 80 - 120 VDC in 20 seconds. No discernableeffects on the thruster were noted. The input voltagewas returned to 100 VDC and a shorting bar was appliedfrom the thruster anode to the facility ground. Thereaction of the BBPPU shown in Figure 9. While theanode was shorted to ground, the BBPPU recycled andattempted to return to the steady state. Each tune thehigh voltage was applied to the thruster, large beam andaccelerator currents resulted due to the short, forcing theBBPPU to recycle again. Approximately twentyrecycles occurred. During this tune period, the dischargecurrent remained at the 4A recycle cutback level and wasstable. Similarly, the neutralizer keeper current wasunaffected by the thruster fault. When the fault wascleared, the beam current returned to its value prior tothe fault without overshoot.

Finally, the thruster was restarted to determine thelength of time needed to start the thruster and throttle upto power level 15. This would be the likely condition ofinterest for an earth orbital application. The results ofthis test are shown in Figure 10. Following the sixminute preheat, the thruster was operating at full powerin under three minutes. Throttle up was smooth and freeof recycles, with no overshoot in beam current. Itshould be noted here that the throttling time could bedecreased significantly by increasing the maximumallowable step in discharge current from 100 mA to ahigher value. The maximum allowable step was notdetermined in the execution of this test.

Two anomalies were noted in the FIT, both having todo with oscillations in the neutralizer keeper anddischarge current. A 300 kHz, nearly sinusoidaloscillation of 1 Ap.p amplitude was noted in theneutralizer keeper current. The AC component of theneutralizer keeper current is shown in Figure 11. Themagnitude of oscillation was somewhat proportional tothe beam current. The normal neutralizer keeper currentripple is a 100 kHz sawtooth of approximately 100mAp.p amplitude. The cause of this oscillation wasfound to be a failed diode in the high voltage pulseignitor circuit. The diode failure allowed the pulseshaping capacitor to resonate with the output filterinductor of the neutralizer power supply. The result wasthe 300 kHz oscillation. Further investigation showedthat the diode was of marginal rating and was replacedwith a more appropriate device. Following the diodereplacement, the neutralizer keeper current returned tonormal as shown in Figure 12.

Following the diagnosis and repair of the neutralizerkeeper power supply, it was suspected that the dischargepower supply had suffered a similar fault The dischargecurrent waveform was expected to be similar to that ofthe neutralizer keeper current, with a peak-to-peakamplitude of 500mA at full power. However, as shownin Figure 12, the discharge current was sinusoidal inshape. Further, the magnitude of the oscillation was2 Ap.p at full power. Inspection of the start circuitrevealed no failures. Subsequently, a circulating currentwas discovered in bypass capacitors which were installedfrom the outputs of the BBPPU to ground to eliminatecommon mode noise from the telemetry andmicrocontroller interface circuits. When these capacitorswere appropriately isolated from the thruster withresistors, the circulating current no longer flowed andthe discharge current returned to normal.

Conclusions

The Functional Integration Test was the finaldevelopment test of the NSTAR BBPPU andEngineering Model Thruster prior to their delivery toNASA JPL for an 8000 hour wear test. The BBPPUcontained all power supplies necessary to operate thethruster and a microcontroller which sequenced thepower supplies to automate the thruster ignition andcathode conditioning procedures. The controller alsoregulated the beam current about a setpoint and throttledthe engine over a 0.5 - 2.3 kW power range. Thestability of these routines and the BBPPU powersupplies was verified over the full input voltage range

of the breadboard.

The automated sequences were conducted withoutincident. The thruster started reliably and transitionedsmoothly to steady state at all input power conditions.The beam current regulation loop maintained stabilityunder all conditions tested. Throttling was accomplishedwithout overshoot. Maximum time for a throttle upfrom low power to full power was under three minutes.

The BBPPU also correctly detected and reacted to severaldifferent thruster faults. All high voltage transient faultconditions resulted in nominal BBPPU responses, and amanually induced, continuous thruster fault did notdamage the BBPPU. When the fault was removed, theBBPPU restored the thruster to the steady statecondition without overshoot in beam current. Failedstart attempts and input bus failures resulted hi a proper,orderly shutdown of the BBPPU and safing of thethruster.

Two anomalies, manifested as neutralizer keeper anddischarge current oscillations, were noted during theFIT. These were traced to failed components andimproper telemetry circuit common mode filter designs.Repairs and design changes eliminated these anomaliesand restored the BBPPU to nominal operation.

The FIT successfully concluded, and the BBPPU/EMTdelivered for integration into a planned 8000 hr test Alldata and "lessons learned" were entered into the NSTARthruster development database, and also transferred to theflight hardware contractor.

Acknowledgements

The authors wish to thank Messrs. Stanley Krauthamerand David Rogers of JPL for their continued assistancein the development of the NSTAR BBPPU. The authorsalso extend special thanks to Mr. G.I. "Hap" Cardwellof Cardwell Associates Inc. for his assistance in thediagnosis of the discharge current anomaly.

References

1. Kakuda, R., Sercel, ]., and Lee, W., "Small BodyRendezvous Mission Using Solar Electric IonPropulsion: Low Cost Mission Approach andTechnology Requirements," IAA Paper L-0710, April1994.2. Janson, S.W., "The On-Orbit Role of ElectricPropulsion," AIAA Paper 93-2220, June 1993.3. Patterson M.J., et al., "2.3 kW Ion Thruster WearTest," AIAA Paper 95-2516, July 1995.

4. Patterson, M.J., et al., "NASA 30 cm Ion ThrusterDevelopment Status," AIAA Paper 94-2849, June1994, (Also NASA TM-106842).5. Soulas, G.C., "Multiple Hollow Cathode WearTesting for the Space Station Plasma Contactor," AIAAPaper 94-3310, June 1994.6. Sarver-Verhey, T.R., "Continuing Life Test of aXenon Hollow Cathode for a Space Station PlasmaContactor," AIAA Paper 94-3312, June 1994.7. Sarmiento, C.J., and Gruber, R.P., "Low PowerArcjet Pulse Ignition," NASA TM-100123, July, 1987.8. Hamley, J.A., et al., "A 2.5kW Power Processor forthe NSTAR Ion Propulsion Experiment," AIAA Paper94-3305, July 1994.9. Hamley, J.A., et al., "Development Status of theNSTAR Ion Propulsion System Power Processor,"AIAA Paper 95-2517, July 1995.10. Biess, J.J., Inouye, L.Y., and Schoenfeld, A.D.,"Electric Prototype Power Processor for a 30-cm IonThruster," NASA CR-135287, 1978.11. Herron, E.G., et al., 30-cm Ion Thruster PowerProcessor," NASA CR-135401, 1978.12. Polk, J.E., "A 1000 Hour Wear Test of the NASA30-cm Xenon Ion Thruster," AIAA Paper 96-2717, tobe published.13. Grisnik, S.P., and Parkes, J.E., "A Large, HighVacuum, High Pumping Speed Space SimulationChamber for Electric Propulsion," IEPC-93-151,September 1993.

Table 1. Sample NSTAR power throttling table. Levels 0 through 15 shown in descending order.Asterisks indicate levels tested during the FIT. A f denotes a controlled parameter

PowerLevel15*1413121110*98765*43*2*1*0

BeamVolt., Vt11001100110011001100110011001100110011001100110011008506500

DISC ANODE

BeamCurr., At1.781.681.581.481.381.281.171.070.970.860.660.640.540.540.54

0

Dis.Volt., V28.028.028.029.029.030.030.030.031.031.031.532.032.032.032.0

0

DisCurr., A10.810.19.48.78.07.46.86.25.65.04.44.24.04.04.0

0

Neut.Volt., V141414141515151616161820222424

0

Neut.Curr.,1.51.51.51.51.51.51.51.51.51.51.51.51.51.51.50

Power,At W

2288215920291908178016601521139512731133901878766635527

0

CATHODE FLOW

PLASMA SCREEN

INSULATOR

ACCEL GRID

SCREEN GRID

NEUT. KEEPER

NEUT. HEATER

Figure 1. NSTAR thruster schematic diagram. The plasma screen was isloated from facility ground for the FIT.

60-120V

INPUT

° NPUTFUER

CMD/DATAINTERFACE

*1

2»VDCINPUJ ——

—— < >

-•--

1

HSKPSUPPLY

4 >!..,

! >

NEUTfWLQER

POWERSTAGE

1

M 3\EE

OUTPUTRB3T1FER

OUTPUTpecTi-ieR

J^————— /=* b-

fDISCHARGE

POWERSTAGE

\

PWMCONTHOl

1iEE

OUTPUTPECTF1ER

OUTPUT

f--J J

.^

m 9 —

BEAM/ACCEL"T5SER

STAGEAND PWM

| POW] STACL_

iPW

IEHX

Mmoi

irJHEt

ompjrRECnPEP.

OUTPUTRECTTFER

OUTPUTRECtiFER

OUTPUTRECTFER

OUTPUTRECTTRER

OUTPUTRLTER

OUTPUTFLTER

[ —————— i

OUTPUTFILTER

OUTPUTFCTER

OUTPUTRLTER

——— 1

]

]

[I

T_

^H

i

NEUTF-O KEEPER

L NEUTHEATER

NEUT

^Q DSCHAR3E

L CATHHEATER

CATHODE^^O COMMON

Figure 2. NSTAR breadboard PPU block diagram.

Hart Disk CommarKiTetemetryTarminal

Figure 3. FIT electrical setup

Neut. Ignitionisch. Ignition

HighVoltage OnV

ICUlrJNK

VNK

JD

VD

JA

VA

5VDIV

0

^v.&^U. ^* :ii^..;~;Xy >£^;;^^^^^ -. .v.:,-;~. ^yi? 4,'y.j. 4- v! ,-^^ •'-* wi- t!. - .. A<^i> £;*£.«4y.'i*»^^r-f^^^^Jr^^HtH^^r.^

' s • • -;. •.'••':•';'•'••;•';':'."." •; •. •* ; i '•.';"-'.-. '•'.'$ : ',"''-, ''- - '•-'•'• '• • • - . ' . ' ; !."...'.•-.••;.:.<;•'•:••.-> - i.-*- • ;• . -- j.....;'..'.-.-:-. • . ' • • • - :-*....;.-. . i . - . - . ; . ; / • -:• ; • : - . • • - * • ; • ; • • • ;.ir.n'.it..!1;.n-i.-(f-ii.ii.-i..ii.m».<t:-.-r.An---HvJiM——W-.-rf-t,-.-.^.--^;— •-•-;-: .>'•'"-:":- •"•-•• ^;^i:---',:'.-;. :'i:-.X-.,-'.• ;'^.:--. -,'vv • ,'• •.'•i'•..{.: V:,'Vi/:"vVS.-v-V .'•--* .I)'.-. •?:...:>.• <f.•;'-•'- < U'i;. ''.'..V .,•.',',.':;:;-;,W

. .

Q^^^^^^^^^

6mADIV

•f:tg>v%£.:ft.£f:£A*,-4.:'• <i *jv^-f^i-i^. J:•';;.•-•/-.:.• .;•• .'• --*,: '*• '\::i-.;'i:-.•:'•*.'•>.^-/' /^'/ r y;^.%' ^ ^y.fy'<T#7J''.?^= * X ^ '. ^v:- .V ' • • • 'V-'i '•: •' -*. ^- V

:;.:;i;;-;::'i;v.,'^::.;.!.:":.;- ^,^.:;;.:|,'fi. •;.;.:..;-V,. ; . . . ; - . ;• . : ;• ; - , . . : . : ; . .: .:.....;.;i: ;:;:.•;:.>:;•.- ;•:•;•••-;••.:;•:"-it":::- •<• -.•; ;•:,•••:••-*-: -v?^ '••'•:''-•' •' \ "

VB. . ......-,..r-__^_,_._,_,__.__._,_^__.,

..

2.5 DIV

Figure 4. Thruster Start at 100 VDC input

Flow Adjustment

JNK&SAlDIV

VNK 5VDIV

JD 2.5A '! ? : , . , . , , .•Jl— .{.wt-wj^iw** .»*+. **-'.;*^^A. -*V'*-'ssj . hi- Wti** !D !v • yr^w.,*,?. -.v-, . .?' ^

2.5DIV

Figure 5. Flow adjustment from augmented thruster start flow to the nominal operating flow for power level 1.

H 2 CommandThrottle Complete

VNK

0 -^n-!^mr^-^H :;-,:i.-JA

VADIV

JB Q-25ADIV

DIVFigure 6. Throttle up from power level 1 to power level 2.

Flow AdjustmentTH 5 Command

Throttle Complete

f

DIV-.

\ / - 5V -4;-:H' -U -'i' : : ^V jMrC •*"'" "'" i j .- ' h-" '- -:- -'-- ^^^^

DIV ^ ^.^^'-.-,-—^.————w——————-™rrvv-»«.'...;S.w*»

o

VA 45VDIV

0

DIV

Figure 7. Throttle up from power level 3 to power level 5

Flow AdjustmentTH 10 Command

Throttle Completei Recycle

VNK

DIV

5Y

jA6mADIV

\ /A 45V ^J':v/:rl:'.y OP:.-.';;:;,.;.:':'iV,,-^"Pi^-^-vx.iY^ • ' . • • : * ; : • : - : s ; ;:-^;:.:-K-i;.i-*,;feyV A TT * ^ .. ^^ . ^ . . i. . ^^ . . .. . ^^ . -. ^^^^ ., ; . ^^ ! .; ^ . .. ^^ - .. ^ ; ,.

I~M\/ '"i' : ; ''r .•'•r^;-K^^i^j^g^ui;':;:AA.-:Ll^-:.-:''-V'?'^J';''?;^':'^''':vJ'::''^^UIV ^"^;;:^vr^:TT:-^^ITSi-:nfl%-i-,^^"*:{3^S:^?ii:^S®i®Sw|;:S:

0 ''•^^^v;.''r:;::^:};':!::'f'K:;^;;?5*

VB 250VDIV

10DIV

Figure 8. Throttle up from power level 5 to power level 10.

knode Fault Appliedault Removedr

DIV

VNK . ,. , , • ,;'-44 ;U;- M<?*VW.lS~|!g * ^P\IW • . - • • •• •• - • ; • ""• ••••*•• •^.i.-.^s.^aii.v^.Si^tfiiKi&iifc^^UIV ..«•.»•.•.; -Tv;r^T^;-?yyy^a;^^ ' Liiail^**^^fc-/^ujj&!i*'ttislw^"i^^^;;^;A;;Sii-:;iiiil>Iiv(^iiSii:Kfefelsi - I SsisM l S^ ^ S^ ^ ^Wii SSP i i- gSBlMiipSSI IIIsrt M.-K^a?^ :• • ; . . • • - . .< , ; - • ••:::i-*;«i ';.-«!';!|t8l!!Si «sf*;Si:i Stft ii»iMifti »;?f *.Ki~»SiS!rJ:;=J;,;i4j3*a,(S,5ju4s.&!«!:is-S5;!«!S?>s?s\J .•...•. ^*^-.-. , rt .,./.,v..-.v^rfvi.Vv™:.™^-**a-^v . , ; . i\SUin«;™u; : ^K^5 * ^ WXiSS^A4^ -.' >Vi'

VRD ._ ,. ^^^u^l:^^i^..ii^.;LJ..ivJ:J;';.J•ji.lj;^;^,^•.a;,^^i;..u^j.^l.;.^jJ!:^i;;;gJ;ij; ^K^Kfi4 ii.ysyg'tss,F l l\/ ^v';^^;Vi-5:r . .ft-^^\>i^1J;^^:^5^^;.«.U-^;^?iS^ ZtZ&fft&f c^^f^f^\J IV ^f^^y^^^^gl^e&^^^^K^^&^&x-i ijijaftSK ;iss#ssipSs^SisftKisi^iKsss^^aiSK&sMfeaaMsiSSSSsa sssafes asfeiagfa

2.5DIV

Figure 9. Continuous recycles.

"hruster Ignition Throttle CompleteThruster Shutdown

VB

10DIV

Figure 10. Throttle up from power level 1 to 15

—il- -t- - i- -ire..!....I...gjti.t.1

10|is/Div

Figure 11. Neutralizer Keeper Current Oscillations

JNK

10ns/Div

Figure 12. Nominal neutralizer keeper current and discharge current with oscillations