OUTCOME OF R&D PROGRAM FOR ITER ICRF POWER SOURCE SYSTEM

Size: px

Start display at page:

Download "OUTCOME OF R&D PROGRAM FOR ITER ICRF POWER SOURCE SYSTEM"

Eustacia Lucas
5 years ago
Views:

1 OUTCOME OF R&D PROGRAM FOR ITER ICRF POWER SOURCE SYSTEM a Rajesh Trivedi 1, Aparajita Mukherjee 1, Raghuraj Singh 1, Kumar Rajnish 1, Dipal Soni 1, Sriprakash Verma 1, Gajendra Suthar 1, Akhil Jha 1, P. Ajesh 1, Manoj Patel 1, Rohit Anand 1, Rohit Agarwal 1, Kartik Mohan 1, JVS Hari 1, Harsha Machchhar 1, Pareshkumar Vasava 1, Hriday Patel 1, Hrushikesh Dalicha 1, Ujjwal Baruah 1, Amit Patel 1, N.P. singh 1, Niranjanpuri S. Goswani 1, Kush R Mehta 1, Dishang V Upadhyay 1, Hitesh Dhola 1, Bhavin Raval 1, Sandip Gajjar 1, Alan White 2, David Francois 2, Jose Sainz 2, Kerry Kozard 2 1 ITER-India, Institute for Plasma Research, Bhat, Gandhinagar , India 2 Continental Electronics, 4212 S. Buckner Blvd., Dallas, Texas , USA a rajesh.trivedi@iter-india.org Abstract As per agreed procurement arrangement with IO, INDA has to supply nine (1 Prototype-unit + 8 subsequent supply) ICRF Source to IO, which is capable to deliver 2.5 MW/2000s RF power at VSWR 2:1 in the frequency band MHz or 3.0 MW/3600s at VSWR 1.5:1 in the frequency band MHz, accompanied by other strict requirements. As there is no such single RF active device for amplifier available to get the RF power of 2.5MW for ITER, therefore, ITER RF source consists of two parallel three-stage amplifiers chains and a combiner to achieve 2.5 MW power by each RF source for ITER. Such RF source is unique in terms of its challenging requirements. Research & Development drive is started by INDA to confirm the use of available high power vacuum tube before start of activities related to ITER deliverables, using Diacrode and Tetrode tubes. During R&D drive, only one chain experiment at 1.5MW/2000s is performed for the frequency range MHz up to VSWR 2:1, with any phase of reflection coefficient. The main objective for the R&D test is to confirm the system performance for the power, duration and frequency range as per ITER need and to check the reliability of both the tube and the amplifier with matched as well as with mismatched load (up to VSWR 2:1), which essentially simulates the plasma load condition. To upkeep the R&D experiment, required infrastructure to test RF source is established at INDA for both the type of vacuum tubes. For Diacrode based system, high power ITER relevant tests are completed in 2016 and reported elsewhere [1]. Since 2-3 years, installation and commissioning of R&D RF chain using Tetrode tubes at INDA test facility is completed with validation of all the relevant sub-systems/systems as standalone mode. The high power RF test of Tetrode based RF amplifiers chain is also conducted at 1.7MW of power for 3600s duration at 36 MHz. For other ITER operating frequencies, the system was operated at 1.5MW/2000s successfully. This paper reports commissioning of RF amplifier using Tetrode technology with various operating scenarios, dissipation limit, safety system, challenges faced during high power operation and outcome of R&D activity at INDA test facility. 1. BRIEF INTRODUCTION ITER require up to 20 MW of ICRF power to a variety of plasmas like at low power level, assisted plasma startup & wall conditioning in between the plasma shots and at full power level, heating and driving plasma current with Ion Cyclotron (IC) frequency band [1]. There will be 2 ICRH antennas, main TX lines, matching units, RF Sources, High Voltage Power Supplies (HVPS) for anode and real time local controls in the IC system. 4 number of RF sources will feed each antenna connected via 8 TX lines and matching units. The complete system will be fabricated & supplied by 3 Domestic Agencies (DAs) and ITER Organization (IO). Under RF source package, INDA is responsible for supplying total 9 (1 Prototype + 8 series production) of ICRF sources along with auxiliary power supplies and local control units to IO. Each RF source shall capable to handle 2.5 MW, 2000s RF power at VSWR 2:1 in the frequency band MHz or 3.0 MW, 3600s at VSWR 1.5:1 in the frequency band MHz and with other necessities. For qualifying end stage high power tetrode-tube & cavity for ITER need, dedicated R&D program has been initiated by INDA. To establish the Tetrode technology related to ICRF amplifiers, ITER-India has signed contract with Continental Electronics (USA) in The contract is to design & develop driver and final stage amplifiers using commercially available CPI make Tetrode tubes (4CW150000E & 4CM2500KG) and to test the system at 1.5MW/2000s/35-65 MHz. Once successful, such two identical chains can be combined to meet the ITER specifications. As reported earlier, the dedicated test facility at ITER-India laboratory mainly consists of low power RF section, wideband solid state pre-driver amplifier, common high voltage power supply for anodes of driver and final stage amplifiers, high power transmission line system, local control unit for control & monitoring, key safety system, water & air cooling system and 3 MW test load for testing the RF source on match and mismatch load. 1

2 The entire RF chain can be operated from the control room with slow and fast safety interlocks. Adequate water and air cooling safeguards high power amplifiers by removing dissipated heat for long pulse operation. 2. REQUIRED SPECIFICATION FOR ITER R&D RF POWER SOURCE The required specification is shown in Table I for tetrode based R&D RF source which is similar as Diacrode based R&D RF source. TABLE I: Tunable frequency range Output power Harmonic level MAJOR SPECIFICATION FOR R&D RF SOURCE 1dB Bandwidth at any centre frequency Overall electrical efficiency Emergency shutdown MHz 1.5MW/2000s/VSWR up to 2:1 with 25% duty cycle at 36 MHz and 60 MHz < -20 dbc ±1MHz 50-65% (Mismatched to Matched condition) <10 sec RF exposure limit < 1.0 mw/cm 2 3. DESCRIPTION ON TETRODE BASED R&D RF POWER SOURCE R&D RF source consist of cascaded chain of RF amplifiers along with two line stretchers in between amplifier stages, water & air cooling systems, auxiliary & anode high voltage power supplies, local control unit and high RF power test rig Cascaded RF chain of amplifiers RF chain is made of low power RF section followed by a 10 kw level wide band solid state power amplifier (SSPA) and tetrode based tuned, 100kW driver stage (HPA2) and 1.5MW final stage (HPA3) amplifiers. A signal generator with -10 dbm to +10 dbm levels drives 10kW. The selected tetrodes for HPA2 and HPA3 are CPI make tetrodes 4CW150000E and 4CM2500KG respectively. The operating frequency is set by adjusting the position of tuning elements. A common grid configuration of the tetrodes is used in the RF path to provide stable operation. The nominal load impedances and operating parameters for the Tetrode tubes are selected according to operation with VSWR 2:1 with all the phases of reflection angles. Input and output cavity circuits are tunable to provide impedance matching for the entire frequency band. Coaxially bounded circuit ensures negligible spurious, low losses and strong EMI shield. Since amplifier can self-oscillate with higher frequencies, absorbing ferrite tiles are used to suppress higher order modes Driver Stage Amplifier (HPA2) Input circuit matches the specified 50 Ω impedance to the calculated cathode impedance ZK which is approximately 21 Ω in parallel with the 175 pf input capacitance of the tetrode. A1 and A2 motors are used for input cavity tuning and loading adjustment respectively. A5 is to adjust the line stretcher between HPA2 and HPA1. Output tuning network consists of variable length transmission lines to transform the nominal operating anode impedance from approximately 400Ω to the required 50Ω impedance. A3 and A4 motors are used for output cavity tuning and loading adjustment respectively. Schematic circuit diagram for the HPA2 is shown in Fig.1. 2

3 FIG.1. Schematic circuit for driver stage (HPA2) amplifier (A1: input cavity tuning, A2: input cavity loading, A3: output cavity tuning, A4: Output cavity loading) Final Stage Amplifier (HPA3) Input circuit matches the specified 50 Ω input impedance to the calculated cathode impedance ZK which is approximately 3.5 Ω in parallel with the 385 pf input capacitance of the tetrode. A1 and A2 motors are used for input cavity tuning and loading adjustment respectively. A6 is to adjust the line stretcher between HPA2 and HPA3. The output tuning network of the HPA-3 amplifier is a coaxial structure consisting of an integral DC blocking capacitor and two variable position shunt capacitors that match the nominal 92.5Ω anode impedance to the 50Ω load impedance. A3 and A4 motors are used for output cavity loading and tuning adjustment respectively. A11 is used to adjust the cavity for harmonic suppression and A12 is used to supress the parasitic oscillation. Schematic circuit for HPA3 is shown in Fig.2. Seven tuners are used for desired tuning and matching purpose. 3

4 FIG.2. Schematic circuit for final stage (HPA3) amplifier (A1: input cavity tuning, A2: input cavity loading, A3: output cavity loading, A4: output cavity tuning, A11: output parallel tuning for harmonic suppression, A12: Ferrite positioning for parasitic suppression Active water and forced air cooling are used to take away dissipated heat from tetrodes and related components during long RF pulse and interlock is provided to turn off the amplifier, if nonconformity of cooling circuit is 4

5 found. A safety key lock system is implied on all power supplies and HV box of RF amplifiers to avoid hazardous access Auxiliary Power Supplies for tetrodes Tetrodes for HPA2 and HPA3 require power supplies for filament, screen grid and control grid for biasing. The major specifications for different power supplies are shown in Table II. TABLE II: MAJOR SPECIFICATION FOR AUXILIARY POWER SUPPLIES Power supply for Filament-FPS Control Grid-CGPS Screen Grid-SGPS HPA2 15.5V DC/215A 0 to -500V DC/1.5A 0-1.1kV/1A HPA3 15.5V DC/640A 0 to -550V DC/12A 0-1.3kV/12A FPS permits automatic ramp-up and ramp-down by 8 minutes to limit inrush current and extend tube life. The remote control variability is provided to the SGPS of HPA High Voltage Power Supply for anode of tetrodes To meet the performance specifications of HVPS i.e. fast dynamics, very low ripple and vital restriction on stored energy, a Pulse Step Modulation (PSM) based HVPS is preferred solution. Normally for HPA-2 and HPA-3, individual HVPSs are required with power requirement of 250 kw and 2.8 MW, correspondingly. As per configuration, two nos. of thoroughly synchronized PSM based HVPSs are essential to upkeep the driver & end stage amplifiers. For R&D drive, 2 HVPSs are combined into a one dual output arrangement with a local controller to run two self-regulating control loops. The dual output arrangement seems more beneficial over normal scheme as use of one HVPS for driver & end stage amplifiers rises power density moderately. Major specifications for dual output HVPS are mentioned in Table III 1[2]-[4]. TABLE III: MAJOR SPECIFICATION FOR AUXILIARY POWER SUPPLIES Input Parameter Value Input voltage : 22 kv + 10 %, 50 Hz Output HVPS for HPA kv, 250 kw HVPS for HPA kv, Amp,2.8 MW Δ V Maximum = 15 kv Accuracy Resolution Ripple Rise Time + 1 % of maximum value of specific stage 100 V + 1 % of maximum value of specific stage 100 µs to 5 ms (Programmable) Energy dumped during fault condition in the load should be less than 10 J. The HVPS is undergone to acceptance tests with resistive load banks (Pulsed Duty) and simulated fault tests well before connecting to RF amplifiers. To verify energy dump <10 Joules during fault, wire burn test was conducted. A fuse wire of specified length (30 AWG, 6 Inch, Copper Wire) is used in test setup in sequence with a short circuit switch, across HVPS output. Once the short circuit switch operates, the controller senses fault and stops all IGBT pulses. Thus, it is proven that cumulative energy dumped to load must remain less than 10 J, observed by visual examination of wire condition whether blown or remain intact Local Control Unit For trustworthy and harmless operation of RF system, local control unit (LCU) is used, which is having sequence control using PLC and Real Time (RT) control using PXI, interlocks, acquisition and display modules. 5

Feedback control loops used in LCU: (1) Anode voltage control loop to adjust anode voltage for controlling anode and screen grid dissipations, (2) VSWR loop to keep constant RF power at load.

Data acquisitions used in the LCU: (1) 40 channel continuous at 1 khz sampling frequency using PXI-6255 I/O board; (2) Event driven with 100 ms pre-post data at 1MHz sampling frequency using PXI-6133

6 Feedback control loops used in LCU: (1) Anode voltage control loop to adjust anode voltage for controlling anode and screen grid dissipations, (2) VSWR loop to keep constant RF power at load. To safeguard the ICRF amplifiers against dangerous fault situation, safety logic is implemented using PXI- 7841R module, which ensures firm shut off (<10µs) of ICRF power and all electrode biasings. Data acquisitions used in the LCU: (1) 40 channel continuous at 1 khz sampling frequency using PXI-6255 I/O board; (2) Event driven with 100 ms pre-post data at 1MHz sampling frequency using PXI-6133 I/O board. Out of so many channels, any 8 channels data of 100 sec at 1 khz sampling frequency can be shown as online graph. Loggings as well as display of selectable channels for offline investigation of faults are also implanted High Power Test Rig For simulation of plasma type load condition, a high RF power test rig (1.5MW/3MW/3600s/35-65 MHz) is used, which includes 1.5MW/3MW soda water (Na 2 CO 3 ) based dummy load (DL) along with Mis-Match Transmission Line-MMTL (which consists of a stub and phase shifter with adjustable electrical length) system. The VSWR of the dummy load is set to less than 1.07 for the operating frequency. HPA3 is tested on matched load for VSWR 1.1 through 12 transmission line (figure 3a). Similarly, HPA3 is tested by inserting MMTL system in between matched load and HPA3 output for 2:1 with different phase angles (fig. 3b). 4. SITE ACCEPTANCE TEST RESULTS HPA2 and HPA3 are erected and connected with all sub-systems at INDA test laboratory, as shown in Fig. 3. Before the initiation of high RF power test, the amplifier was undergone to static test, i.e. bandwidth (BW) and Higher Order Mode (HOM) measurement to check absence of unwanted oscillation for a particular frequency tuning at low level (~ kW) RF power. HPA3 DL (a) (b) MMTL FIG.3. View of HPA2, phase shifter & HPA3 connected with (a) matched (b) mis-matched load TABLE V: TEST RESULTS OF GLOBAL CHAIN ON MATCHED LOAD (VSWR1:1) Freq. HPA3 HPA3 HPA2 HPA2 SSPA Va-3 Ia-3 Eff-3 Diss-3 Gain-3 Pf Pr Pf Pr Pf MHz kw kw kw kw kw kv A % kw db After successful testing of HPA2, complete RF chain test (SSPA+HPA2+HPA3) was conducted at MW level for 2000s on matched load, as per Table V. At lower edge side (36 MHz), source was also tested for 1.7 MW/3600s, to show the design capability of amplifiers (Fig.4). 6

7 Gain (db) Overall Gain (HPA3+HPA2) Time (Min.) (a) (b) FIG.4: 1.7MW/36MHz/3600s on matched load (a) Run test, (b) Harmonics on spectrum analyser BW of complete RF chain was measured for all the operating frequencies at 1.5MW level and found within satisfactory range. Typical BW curve at 36 MHz is shown in Fig. 5. FIG.5: Measurement of BW at 1.5MW output power level The system was also tested for 1MW, 2000s including measurement of 1dB BW at 40, 45, 50 and 55 and 60 MHz. One of the key specifications for ITER R&D source is to validate RF amplifier for constant output power for mis-match load (VSWR up to 2:1) at any reflection phase which is like ELMs in the plasma. RF chain was tested for mismatched load condition, at various reflection phase angles as per Table V. Forward power could be kept constant to 1.5MW with 11% power reflected power (equivalent to VSWR 2:1) during test,, thus generating an significant databank for ITER RF source to be delivered. TABLE V: TEST RESULTS OF COMPLETE RF CHAIN ON VSWR 2:1 AT 55 MHz Ref. angle (Deg.) P output (kw) V Anode (kv) I Anode (A) Anode dis (kw) SG dis (kw) SSPA HPA2 HPA3 HPA2 HPA3 HPA2 HPA3 HPA3 HPA

8 Run test was conducted for 2000s for five consecutive RF pulses with 25 % duty cycle to confirm the ruggedness & availability of the full system and to benchmark the equipment for fusion application. Electrical efficiency of complete RF chain is around 55% - 60%. RF exposure is measured at the vicinity of RF source and found less than 25 W/cm 2 which is well within the limit. 5. SUMMARY R&D RF source using tetrode tubes for driver and final stage of amplifiers are designed, manufactured, assembled and commissioned successfully at INDA test facility along with other subsystems. This RF source is tested first time worldwide on matched and mismatched load for more than 2000s at different ITER frequencies at MW power level. To check the overall performance of the system, 5 consecutive 2000s shots with 25% duty cycle at 1.5MW/55MHz tested successfully. 6. REFERENCES [1] APARAJITA MUKHERJEE et. al., Progress in High Power Test of R&D Source for ITER ICRF system, 26 th IAEA FEC 2016, Oct 2016, Kyoto, Japan [2] AMIT PATEL et. al., Initial Operation of 3 MW Dual Output High Voltage Power Supply with IC RF System, presented in Symposium on Fusion Technology (SOFT-2016), September [3] A. PATEL et. al., Development of 3 MW Dual Output High Voltage Power Supply for IC RF System, presented in IEEE Power Modulator and High Voltage Conference, San-Francisco, 2016 and under review at IEEE Transaction on Plasma Science. [4] H. DHOLA et. al., Data transfer methods in Real Time controller of Ion Cyclotron High Voltage Power Supply, presented in Real Time conference, IEEE NPSS, Padua, Italy, May

Lesson learnt from development of ITER MW level R&D RF source in MHz frequency range

Lesson learnt from development of ITER MW level R&D RF source in MHz frequency range Rajesh Trivedi, On behalf of ITER-India IC H & CD Group rajesh.trivedi@iter-india.org ITER-India, Institute for Plasma