[email protected]

RF power & LLRF 20171017 Outlook

Combiners Preamble Transmission Lines Power sources Fundamental Power Couplers Tetrodes Range of HPRF at a glance Klystrons Overhead IOT Solid State Power Amplifiers (SSPA) True life for HPRF and LLRF Conclusion

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 2 Preamble

W → kW → MW HPRF € → k€ → M€ Cavity Very important for all projects Rule of thumbs, HPRF acquisition costs LLRF 5 € / W (MW amplifiers) to 10 € / W (kW amplifiers)

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 3 Preamble

W → kW → MW Class of operation A B C € → k€ → M€ Advantage Linear Almost Non linear linear Very important for all projects DC to RF efficiency 30 % 78.5 % 90 % Wall plug efficiency 20 % 40 % 50 % Rule of thumbs, HPRF Cost per 10 year [k€] operational costs 100 kW RF CW 3750 1870 1500 5000 hours 0.15 € / 푘푊푕 Acquisition 500 500 500 푒푓푓푖푐푖푒푛푐푦

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 4 RF power sources classification

Vacuum Tubes Transistors

Grid Tubes Linear Beam Tubes Bipolar Junction Transistor (BJT) Triodes Klystrons Field Effect Transistor (FET) Tetrodes Travelling Wave Tubes Junction Gate FET (JFET) Pentodes (TWT) Metal Oxide Semiconductor FET Diacrodes Gyrotrons (MOSFET) Inductive Output Tube power MOSFET (IOT) Vertically Diffused Metal Oxide Semiconductor (VDMOS) Laterally Diffused Metal Oxide Crossed-field Tubes Semiconductor (LDMOS)

Magnetrons

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 5 Grid tubes

1904 Diode, John Ambrose Fleming 1906 Audion (first triode), Lee de Forest The first diode prototype 1912 Triode as amplifier, Fritz Lowenstein Fleming Diode, 1904 1913 Triode ‘higher vacuum’, Harold Arnold 1915 first transcontinental telephone line, Bell 1916 Tetrode, Walter Schottky 1926 Pentode, Bernardus Tellegen 1994 Diacrode, Thales Electron Devices Thales TH 628 diacrode, 1998

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 6 Essentials of grid tube

Vacuum tube Ua e- Anode e- e- Heater + Cathode e- Heated cathode e- e- Coated metal, carbides, borides,… e- e- e- e- e- thermionic emission Electron cloud Cathode & Anode Filament Diode

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 7 Essentials of grid tube

Vacuum tube Ua Anode Heater + Cathode Heated cathode Coated metal, carbides, borides,… e- e- e- thermionic emission Electron cloud Cathode & Anode Filament Diode

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 8 Essentials of grid tube

Triode Ua e- Anode e- Modulating the grid voltage proportionally e- e- modulates the anode current Transconductance e- Ug1 e- Control Grid e- e- e- Voltage at the grid Current at the anode

Cathode Limitations & Filament Parasitic capacitor Anode/g1 Tendency to oscillate

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 9 Essentials of grid tube

Ua Tetrode e- Anode e- e- e- Screen grid Positive (lower anode)

e- Ug2 Ug1 e- Decouple anode and g1 Screen Grid Control Grid e- e- e- Higher gain Limitations Cathode & Secondary electrons Filament Anode treated to reduce secondary emission

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 10 Essentials of grid tube

DC power is Pdc = Vdc Idc

Assuming the tube is linear whilst it is conducting, Ipk the dc anode current is found by Fourier analysis of the current waveform and is Idc = Ipk/π Class B Irf = Ipk/2 = Idc π/2, and ideal class B, Vrf = Vdc So, RF power is Prf = ½ Vrf Irf Prf = ½ Vdc Idc π/2 = π/4 Vdc Idc

Theoretical efficiency η = Prf/Pdc = ¼ Vdc Ipk / Vdc Idc η = 78.5 % Vdc

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 11 Essentials of grid tube

Two reasons for not achieving this impressive number Ipk tube is not fully linear whilst it is conducting

Anode voltage must be higher than G2 voltage, Class B

VG2 being ~ 10% Vdc UG2 This leads into Pdc = Vdc Idc = Vdc 1.05 Ipk/π Prf = ½ Vrf Irf = ¼ 0.9 Vdc Ipk Theoretical efficiency in practice η = Prf/Pdc = ¼ 0.9 Vdc Ipk / 1.05 Vdc Ipk/π

η = 67 % Vdc

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 12 Tetrode RS 2004 CERN SPS amplifier @ 200 MHz

CERN SPS, RS 2004 Tetrode, Trolley (single amplifier), and transmitter (combination of amplifiers) Two transmitters of eight tubes delivering 2 x 1 MW @ 200 MHz, into operation since 1976

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 13 Tetrodes & Diacrodes available from industry

peak < 1 ms CW

10000

1000

100 Power per singleper tube kW Power

10 0 100 200 300 400 500 Frequency MHz

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 14 Linear beam tubes

1937 Klystron, Russell & Sigurd Variant 1938 IOT, Andrew V. Haeff 1939 Reflex klystron, Robert Sutton 1940 Few commercial IOT Russell & Sigurd Varian klystron, 1937 1941 Magnetron, Randall & Boot 1945 Helix Travelling Wave Tube (TWT), Kompfner 1948 Multi MW klystron 1959 Gyrotron, Twiss & Schneider 1963 Multi Beam Klystron, Zusmanovsky and Korolyov

1980 High efficiency IOT Thales TH 1802, 2002

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 15 Essentials of klystron

U beam Klystrons velocity modulation Collector converts the kinetic energy into radio frequency power Vacuum tube Drift Electron gun Space Thermionic cathode Anode

Electron Electron beam Gun Drift space Uanode Collector Cathode & e- constant speed until the collector Filament

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 16 Essentials of klystron

U beam Cavity resonators RF input cavity (Buncher) modulates e- velocity Some are accelerated Some are neutral Some are decelerated Bunching the e- RF output cavity (Catcher)

Cavity Coupling Resonating at the same frequency as the input loop cavity U anode At the place with the numerous number of e- Kinetic energy converted into voltage and Cathode & extracted Beam Accelerating Filament line gap

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 17 Essentials of klystron

Cavity resonators RF input cavity (Buncher) modulates e- velocity Some are accelerated Some are neutral Some are decelerated Bunching the e- RF output cavity (Catcher) Resonating at the same frequency as the input cavity At the place with the numerous number of e- Kinetic energy converted into voltage and extracted

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 18 Essentials of klystron

Cavity resonators RF input cavity (Buncher) modulates e- velocity Some are accelerated Some are neutral Some are decelerated Bunching the e- RF output cavity (Catcher) Resonating at the same frequency as the input cavity At the place with the numerous number of e- Kinetic energy converted into voltage and extracted

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 19 Essentials of klystron

Cavity resonators RF input cavity (Buncher) modulates e- velocity Some are accelerated Some are neutral Some are decelerated Bunching the e- RF output cavity (Catcher) Resonating at the same frequency as the input cavity At the place with the numerous number of e- Kinetic energy converted into voltage and extracted

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 20 Essentials of klystron

Voltage in e- density at e- density at Cavity resonators cavities input cavity output cavity vs time RF input cavity (Buncher) time modulates e- velocity Some are accelerated Some are neutral accelerated Some are decelerated neutral Bunching the e-

decelerated RF output cavity (Catcher) Resonating at the same frequency as the input cavity At the place with the numerous number of e- Kinetic energy converted into voltage and Distance (drift space) extracted Bunching of e- beam in a klystron

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 21 Essentials of klystron

Collector Additional bunching cavities Resonate with the pre-bunched electrons beam Generate an additional accelerating/decelerating field Better bunching Gain 10 dB per cavity Focusing magnets Anode To maintain the e- beam as expected and Cathode & where expected Filament

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 22 Klystron TH 2167 CERN LHC @ 400 MHz

CERN LHC, TH 2167 klystron in lab and in UX45 cavern 16 klystrons delivering 330 kW CW @ 400 MHz, into operation since 2008

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 23 Klystrons available from industry

peak < 50 µs CW

100000

10000

1000

100 Power per single tube kW tube single per Power

10 0 2000 4000 6000 8000 10000 12000 14000 Frequency MHz

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 24 Essentials of IOT

U anode IOT density modulation converts the kinetic energy into radio frequency power Vacuum tube Triode input Thermionic cathode Grid modulates e- emission e- e- Klystron output e- e- e- Anode accelerates e- buckets Ugrid Short drift tube & magnets Cathode Catcher cavity & Filament Collector

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 25 IOT TH 795 CERN SPS @ 800 MHz

CERN SPS, TH 795 IOT, Trolley (single amplifier), and transmitter (combination of amplifiers) Two transmitters of four tubes delivering 2 x 240 kW @ 801 MHz, into operation since 2014

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 26 IOT available from industry

peak < 10 ms CW

1000

100 Power per singleper tube kW Power

10 0 200 400 600 800 1000 1200 1400 Frequency MHz

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 27 Transistor for RF power

1925 theory, Julius Edgar Lilienfeld 1947 Germanium US first transistor, John Bardeen, Walter Brattain, William Shockley 1948 Germanium European first transistor, Herbert Mataré and Heinrich Welker First transistor invented at 1953 first high-frequency transistor, Philco BELL labs in 1947 1954 Silicon transistor, Morris Tanenbaum 1960 MOS, Kahng and Atalla 1966 Gallium arsenide (GaAs) 1980 VDMOS 1989 Silicon-Germanium (SiGe) 1997 Silicon carbide (SiC) XXI century LDMOS 2004 Carbon graphene

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 28 Essentials of RF transistor

Vdc

In a push-pull circuit the RF signal is NPN BJT applied to two devices One of the devices is active on the positive

Vin Vout voltage swing and off during the negative voltage swing The other device works in the opposite manner so that the two devices conduct half the time PNP BJT The full RF signal is then amplified

Two different type of devices

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 29 Essentials of RF transistor

Vdc

0 ⁰ NPN BJT Another push-pull configuration is to use a balun (balanced-unbalanced) 0 ⁰ it acts as a power splitter, equally dividing the input Input balun Output balun power between the two transistors (Unbalanced-Balanced) (Balanced-Unbalanced) the balun keeps one port in phase and inverts the second port in phase Vin Vout

Since the signals are out of phase only one 180 ⁰ device is On at a time 180 ⁰ NPN BJT This configuration is easier to manufacture since only one type of device is required

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 30 Essentials of RF transistor

CERN SPS 200 MHz, 1250 Wp – 700 W CW solid state amplifiers from TTi Norte (Spain) that replace discontinued YL 1440 tube

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 31 Transistors SOLEIL @ 352 MHz and ESRF @ 352 MHz

SOLEIL 45 kW @ 352 MHz tower ESRF four 150 kW @ 352 MHz tower solid state amplifiers (2004 & 2007) solid state amplifiers (2012)

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 32 Transistors available from industry

CW

10000

1000

100 Power per single transistor single W per Power 10 0 500 1000 1500 2000 2500 3000 Frequency MHz

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 33 Combiners

Resistive power Combiners Cheap and easy to build Use of resistors to maintain the impedance Power limitation and losses induce by the resistors (→ not used in high power)

Hybrid power Combiners Use RF lines Low level of loss Limitation by the size of the lines

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 34 Combiners

Low loss T-Junction

Z 푁 c λ/4 Z c Zc

Zc λ/4

Zc 푁

With 푍 = 푍푐 푁 λ/4 We have a N-ways combiner 160 to 1 @ 352 MHz T-junction combiner [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 35 Combiners

3 dB phase combiner

P1 Σ λ/4

3 푑퐵

P Δ 2 Correctly adjusting the phase and the gain, With correct input phases P1 = P2 = P

푃 + 푃 푃 + 푃 Σ = 1 2 + 푃 푃 Σ = + 푃푃 = 2 푃 2 1 2 2

푃 + 푃 푃 + 푃 Δ = 1 2 − 푃 푃 Δ = − 푃푃 = 0 2 1 2 2

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 36 Combiners 200 MHz CW combiners

2 * 40 kW to 80 kW 10 * 4 kW to 40 kW

3 dB : 550 kW & 160 kW

3 dB : 1.1 MW

CERN SPS 64 to 1 combiner @ 200 MHz

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 37 Combiners

Since some years now we are moving to Cavity combiners and Radial combiners The CaCo from Alba is one of the first example CERN 144:1 cavity combiner 144 kW CW @ 200 MHz Then with ESRF with funding from the EU as WP7 in calculations by ESRF within CRISP, construction by CERN the framework of the FP7/ESFRI/CRISP program we launched a first Cavity Combiner (I invite you to CERN 16:1 radial cavity combiner 2.4 MW CW @ 200 MHz read the mopc005 IPAC 11 and wepfi004 IPAC13 reports from ESRF) Recently CERN developed a new VHPCC (Very High Power Cavity Combiner) 16 * 150 kW : 1 * 2.4 MW CW @ 200 MHz

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 38 Combiners

Advantages of Cavity Combiners Compact No dummy load Less losses

Drawbacks Reduced bandwidth (still large enough) Reflection in case of wrong phase

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 39 Coaxial cables are Coaxial Lines often with PTFE foam to keep concentricity

Characteristic impedance is

60 퐷 Flexible lines have Zc = ln ε푟 푑 spacer helicoidally With placed all along the D = inner dimension of the outer conductor line d = outer dimension of the inner conductor

εr = dielectric characteristic of the medium

Outer conductor Inner conductor Size Outer diameter Inner diameter Outer diameter Inner diameter Rigid lines are made 7/8" 22.2 mm 20 mm 8.7 mm 7.4 mm of two rigid tubes 1 5/8" 41.3 mm 38.8 mm 16.9 mm 15.0 mm maintained 3 1/8" 79.4 mm 76.9 mm 33.4 mm 31.3 mm concentric with 4 1/2" 106 mm 103 mm 44.8 mm 42.8 mm supports 6 1/8" 155.6 mm 151.9 mm 66.0 mm 64.0 mm

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 40 Coaxial lines Maximum Power handling

Power handling of an air coaxial line is related to breakdown field E

푑 퐷 푉 = 퐸 푙푛 푝푒푎푘푚푎푥 2 푑 D 푉 2 푃 = 푝푒푎푘푚푎푥 d 푝푒푎푘푚푎푥 2푍푐

퐸2 푑2 휀 퐷 푃 = 푟 푙푛 푝푒푎푘푚푎푥 480 푑 With E = breakdown strength of air (‘dry air’ E = 3 kV/mm, commonly used value is E = 1 kV/mm for ambient air) D = inside electrical diameter of outer conductor in mm d = outside electrical diameter of inner conductor in mm Zc= characteristic impedance in Ω ε = relative permittivity of dielectric r Curves valid for straight lines, in case of elbow, f = frequency in MHz reduce it by 30 %

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 41 Coaxial lines

CERN SPS 200 MHz, 4 * 150 meters of 345 mm coaxial lines from the surface building to the cavities

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 42 Rectangular waveguides

The main advantage of waveguides is that b waveguides support propagation with low loss a

λ λ푔 = Wavelength λ 1 − ( )2 2푎

c Cutoff frequency dominant mode f = c 2푎 c Cutoff frequency next higher mode f c2 = 4푎

Usable frequency range 1.3 fc to 0.9 fc2

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 43 Rectangular waveguides

Waveguides are usable over certain frequency ranges For very lower frequencies the waveguide dimensions become b impractically large a For very high frequencies the dimensions become impractically small & the manufacturing tolerance becomes a significant portion of the waveguide size

Waveguide name Recommended Cutoff frequency Cutoff frequency Inner dimensions of frequency band of lowest order of next waveguide opening EIA RCSC IEC of operation (GHz) mode (GHz) mode (GHz) (inch) WR2300 WG0.0 R3 0.32 — 0.45 0.257 0.513 23.000 × 11.500 WR1150 WG3 R8 0.63 — 0.97 0.513 1.026 11.500 × 5.750 WR340 WG9A R26 2.20 — 3.30 1.736 3.471 3.400 × 1.700 WR75 WG17 R120 10.00 — 15.00 7.869 15.737 0.750 × 0.375 WR10 WG27 R900 75.00 — 110.00 59.015 118.03 0.100 × 0.050 WR3 WG32 R2600 220.00 — 330.00 173.571 347.143 0.0340 × 0.0170

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 44 Rectangular waveguides Maximum Power handling Peak Power vs Frequency

1000 WR2300 λ2 WR2100 −4 2 2 2 푃 = 6.63 10 퐸푚푎푥 푏 (푎 − ) WR1800 4

WR1500

With WR1150

P = Power in watts Peak power [MW] power Peak a = width of waveguide in cm WR975 b = height of waveguide in cm λ = free space wavelength in cm 100

Emax = breakdown voltage gradient of the dielectric filling

350 800 200 250 300 400 450 500 550 600 650 700 750 850 900 950

the waveguide in Volt/cm (for dry air 30 kV/cm, for 1.000 ambient air 10 kV/cm) Frequency MHz [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 45 Rectangular waveguides

CERN SPS 800 MHz, 2 * 150 meters of WR1150 waveguides from the surface building to the cavities

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 46 Fundamental Power Couplers (FPC)

SRF2017 – Tutorial on FPC http://ddl.escience.cn/f/M2wh LHC

LHC 400 MHz, 500 kW SPS 2.0 200 MHz, 750 kW SPL 1.0 704 MHz, 900 kWp Crab SPL 2.0 704 MHz, 1000 kWp ESRF & APS DQW & Linac4 352 MHz, 1000 kWp RFD Crab DQW 400 MHz, 100 kW Crab RFD 400 MHz, 100 kW ESRF 352 MHz, 250 kW SOLEIL 352 MHz, 250 kW APS 1.0 352 MHz, 250 kW SPS LIU 200 MHz, 800 kW SPS LIU HG (SPL 3.0) 704 MHz, 1200 kWp LHC 2.0 400 MHz, 500 kW APS 2.0 352 MHz, 250 kW HG Crab 2.0 400 MHz, 100 kW Linac4

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 47 Fundamental Power Couplers (FPC)

Even if we can imagine cavities able to receive a large amount of power FPC power ratings In the frequency range, Fundamental Power 1500 Coupler will limit the maximum power that can be delivered to the cavity 1000

Currently the limit is around y = 200000x-1

200000 CW Power[kW] 푃 [푘푊] = 500 푚푎푥 푓 [푀퐻푧]

0 0 200 400 600 800 1000

(for a stable operation 10-20 years lifetime of a Frequency [MHz] Fundamental Power Coupler) [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 48 Fundamental Power Couplers (FPC)

LLRF07 A Fully Integrated Controller for RF Conditioning of the LHC Superconducting cavities, John C. Molendijk

CWRF08 LHC RF Embedded SC Cavity Conditioning System, John C. Molendijk

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 49 Tetrode peak Available RF sources at a glance Tetrode CW Klystron peak 100000 Klystron CW IOT peak IOT CW

10000 SSPA*100

1000 Power [kW] Power

100

10 0 500 1000 1500 2000 Frequency MHz [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 50 Tetrode peak Available RF sources at a glance Tetrode CW Klystron peak 100000 Klystron CW IOT peak IOT CW

10000 SSPA*100

1000 Power [kW] Power

100

10 0 500 1000 1500 2000 Frequency MHz [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 51 Overhead YL1530, 35 kW @ 200 MHz 70

60

50

40

30

Output power kW power Output 20

10 Courtesy Wolfgang Hofle 0 0 1000 2000 3000 4000 5000

Input power W [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 52 Overhead

Tetrodes, Diacrode, IOT, Klystrons, SSPA

1,8 1,6 1,4 1,2 klystrons 1 Overhead needed Tetrode, Diacrode & IOT 0,8 for LLRF operation Klystron 0,6 SSPA 0,4 RF Pulse 0,2

Output power / / nominalOutputpower Output power 0 0 1 2 3 4 Input power / nominal Input power [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 53 Overhead needed for LLRF operation Klystrons RF Pulse Klystrons 1,4 100 90 Output power reduces if we go over 1,2 Power 80 Efficiency saturation (nominal) point of operation 1 70 Need to operate lower than nominal point 0,8 60 50 of operation 68 % at saturation 0,6 40 Loss of efficiency 62 % at -1.2 dB 0,4 30 Double cost (acquisition + operation) 20 0,2 10

Output power / / nominalOutputpower Output power 0 0 Phase stability is given by construction 0 1 2 3 from HV stability (very expensive) Input power / nominal Input power [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 54 Overhead needed for LLRF operation Tetrodes, Diacrodes, IOT RF Pulse

Tetrodes, Diacrode, IOT 1,8 A great advantage of gridded tubes is that 1,6 they allow overdrive without damage 1,4 Thanks to that, they can be operated very 1,2 close to their nominal point 1 0,8 Tetrodes & Diacrodes are limited in 0,6 frequency (max ~ 400 MHz), not IOT 0,4 0,2 Lower gain, some more stages, addition of 0

Output power / nominal / power Output power Output 0 1 2 3 4 limiting parameters Input power / nominal Input power

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 55 Gridded klystrons HVPS Pdc tubes HV HV 10-20 % tolerant less tolerant

For gridded tubes HVPS is very simple RF Pulse RF Pulse No RF -> idle current (can be zero in class B or time class C)

Even if HV is drooping, the LLRF will impose RS2004 tetrode output power, and tetrode remains able to 100 90 deliver requested Power 80 Unominal (8.5 kV) 70 - 10 % U nominal Stability of the klystron is much more 60 -20 % Unominal 50 dependant on stability of the HVPS as any Pout kW 40 30 drop will result on different acceleration, and 20 10 length of drift tube remains the same, it 0 0 2 4 6 8 10 12 means a phase variation Pin kW

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 56 Overhead needed for LLRF operation SSPA RF Pulse

SSPA 1,4 Destruction in case of large overdrive (+ 3 dB) longer than ~ 100 µs 1,2 Hard protection limit are needed, 1 could be built in, but then it is 0,8 complex to manage for LLRF, so we 0,6 try to have a good LLRF protection 0,4 system Overhead must be perfectly and 0,2

correctly defined 0 Output power / nominal / power Output power Output Overhead very costly (compare to 0 0,5 1 1,5 2 2,5 gridded tubes) Input power / nominal Input power [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 57 [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 58 [email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 59 True life for LLRF & HPRF

For any new HPRF system, LLRF colleagues should be involved from the very beginning in order to provide the boundaries needed for operation

Be aware that HPRF is very expensive and that we will have, what we will have…

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 60 True life for LLRF & HPRF

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 61 True life for LLRF & HPRF

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 62 Conclusion

Whatever characteristics you need, please keep in mind that HPRF is very HPRF long to build, is very expensive, and Cavity that LLRF will have to compensate all our HPRF systems defaults LLRF Corrections You are the ones that will enable our Protections very expensive HPRF systems to be operational and reliable

[email protected] 20171017, LLRF17 BARCELONA, RF POWER & LLRF 63 [email protected]

Thank you very much 20171017