Components & Subsystems New generations of RF Amplifiers : from system requirements to technologies Michel Caplot, Christian Robert, Michel Grezaud, Bernard Darges and Pascal Ponard cw RF Workshop – CERN – March 2008

1 Components & Subsystems System requirements (1) From System requirements to Technologies….and not the opposite way System requirements do include technical performance but are not limited to these performance Systems requirements should cover system life time Cost of a system is not limited to purchasing cost: it clearly include Life Cycle Costs (LCC) Purchasing cost Maintenance cost System evolution cost Supplier support cost Spare part cost ……

2 Components & Subsystems System requirements (2) Systems Requirements  Technical performance  LCC including availability of spare parts over 2 or 3 decades  System architecture  System reliability  System availability (fault tolerance, graceful degradation, maintainability, testability, …)

3 Components & Subsystems Architectures (1) Large klystron based RF amplifiers Multi-IOT based RF amplifiers High power Few items to maintain Centralized architecture Lower power at each stage Easy replacement Decentralized architecture with local control

4 Components & Subsystems Multi-coupled IOT based RF amplifiers Architectures (2) Multi-coupled Solid State based RF amplifiers Coupling losses Graceful degradation Easy replacement Coupling losses Low voltage Obsolescence management

5 Components & Subsystems Architectures (3) Architecture should not take into account technologies Architecture is strongly dependent on requirements Solid State vs. Electron Device architectures : really a debate ? Solid StateElectron Devices Low voltageHigh voltage Graceful degradation but coupling lossesSome possible graceful degradation Obsolescence managementFew suppliers but existing (!!!) Components not driven by science market Ability to produce components some decades after system start LCC (design evolution due to new components)LCC In both cases, experienced high power RF designers are mandatory !!!

6 Components & Subsystems Electron device evolution - Klystrons & IOTs (1) Klystrons 1 GHz10 GHz 100 GHz 10 MW 1 MW 100 kW 10 kW IOT CW Klystrons Pulsed Klystrons MBK EIK µs Klystrons 1 kW Radiotherapy Security Industry Radiotherapy Security Industry Scientific Accelerators & Fusion Scientific Accelerators & Fusion Broadcast Defense f P peak IOTs

7 Components & Subsystems Electron device evolution – Triodes & Tetrodes (2) Diacrodes 1 MW 100 kW 10 kW 10 MHz100 MHz 1 GHz 1 MHz TV Diacrodes Induction Laser Radio Triodes & tetrodes Scientific Accelerators & Fusion Scientific Accelerators & Fusion Scientific Accelerators Scientific Accelerators Triodes & Tetrodes

8 Components & Subsystems Electron device evolution – Solid State Device (3) Solid State amplifiers Solid State amplifiers 1 W 1 kW 1 MW Si/GaAs PWM (S-class) E class AB class F class PWM (S class) C (D) class AB class C class A class Si/SiC/GaN 1 MHz 1 GHz

9 Components & Subsystems Diacrode - A major evolution for tetrode technology (1) Cathode Control Grid Screen Grid Anode The end capacitance is mainly done in this area Fig 18 TH 628 Diacrode K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A Two tetrodes K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A K G 1 G 2 A One Diacrode

10 Components & Subsystems Diacrode - A major evolution for tetrode technology (2) Frequency - MHz cw Power - MW VSWR = 2.0:1 Both tubes with the same cathode

11 Components & Subsystems Solid State – Amplifier Class of Operation Impact (1) 100% A AB B C D 98% 0.1 Frequency, MHz 75% 65% 55% 25% % Thales new design Si MOSFET Technology Amplifier Class of operation Efficiency, %

12 Components & Subsystems Solid State – Amplifier structure Impact (2) Si LDMOS Technology (low frequency) No required tuning or adjustment 90 % SMT technology Last MOSFET technology RoHS compliance In case of component technology evolution, Pallets are changing but not Case Interface and not Racks Pallet Case Racks 10 kW Complete rack 600 W Complete rack

13 Components & Subsystems Solid State – GaN device for high frequency (3) KORRIGAN : a GaN foundry for Europe GaN TECHNOLOGY RF power densityx 10 Output powerx 10 Power added efficiencyx 2 Junction temperature> 250°C Voltage supply > 28V

14 Components & Subsystems Solid State – GaN device for high frequency (4) Substrates Epitaxy Process Die Bank Assembly Test Final Product Delivery CTH, ISOM, CIDA, PTO U.Roma TV, CNR IFN GaN HEMT Foundries SELEX, UMS (<-Tiger), QinetiQ TNO, IRCOM, CTH System Houses ELT, SELEX, Thales, Indra, EMW, SAAB, BAE ISOM TNO GaN epi-wafers Picogiga, QinetiQ LiU, Tiger, ULE U.Padova, ISOM, CIDA NORSTEL KORRIGAN : a GaN foundry for Europe

15 Components & Subsystems IOTs – Modularity based design (1) Diamond TH793 IOTs MHz 80 kW cw TH713 IOTs L band 16 kW cw

16 Components & Subsystems IOTs – Modularity based design (2) Combined RF Output Coupling cavity Modularity allows for Fault tolerance through graceful degradation, due to the operation of each RF source at a lower power: in case of one source failure, two remaining can supply full power

17 Components & Subsystems Conclusion New generations of RF amplifiers should be defined by System Requirements that include technical performance but also system operation requirements and Life Cycle Costs (and not only Purchasing Cost) There is not a debate between Solid State and Electron devices: based on system requirements, technological answers may be based on a unique type or on a combination of technologies Both Solid State and Electron devices have evolutions. SSD design based amplifier must incorporate an architecture that allows to mix different solid state technologies in order to take into account that science is not the market driver