Gridded Tubes Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016 Paul Scherrer Institute

Preamble Speaker guidelines for Session C (RF sources) The goal of this session is to compare the efficiency of different RF sources to assess today’s performance to evaluate their potential performance if present R&D programs are successful Please prepare your talk for a non-expert audience Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI RF source type Gain [dB] Maximum output power pulsed [kW] Rise time [us] Pulse length range [ms] Rep rate range [Hz] Max. output power CW [kW] Efficiency at working point [%] High voltage needs [kV] Frequenc y range [MHz] Typical performance Performance potential

Outline Tetrodes Basics Today’s performances R&D program, Diacrodes IOT Basics Today’s performances R&D program, MB-IOT Summary Efficiency Frequency and Power range General considerations Overall efficiency Overhead Experienced team Case study for medical application Gridded tubes specificities What R&D program ? Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

Tetrodes 1904Diode, John Ambrose Fleming 1906Audion (first triode), Lee de Forest 1912Triode as amplifier, Fritz Lowenstein 1913Triode ‘higher vacuum’, Harold Arnold 1915first transcontinental telephone line, Bell 1916Tetrode, Walter Schottky 1926Pentode, Bernardus Tellegen 1998Diacrode, Thales Electron Devices Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI The first diode prototype Fleming Diode, 1904 Thales TH 628 diacrode, 1998

Essentials of tetrodes Vacuum tube Heater + Cathode Heated cathode Coated metal, carbides, borides,… thermionic emission Electron cloud Anode Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Cathode & Filament e- Ua Anode e-

Essentials of tetrodes Vacuum tube Heater + Cathode Heated cathode Coated metal, carbides, borides,… thermionic emission Electron cloud Anode Diode Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Ua Anode e- Cathode & Filament

Essentials of tetrodes Triode Modulating the grid voltage proportionally modulates the anode current Transconductance Voltage at the grid Current at the anode Limitations Parasitic capacitor Anode/g1 Tendency to oscillate Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Ug1 Control Grid Ua Anode e- Cathode & Filament

Essentials of tetrodes Tetrode Screen grid Positive (lower anode) Decouple anode and g1 Higher gain Limitations Secondary electron Anode treated to reduce secondary emission Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI e- Ug1 Control Grid Ug2 Screen Grid Ua Anode e- Cathode & Filament

Frequency & Power range of tetrodes Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

Frequency & Power range of tetrodes Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

Frequency & Power range of tetrodes Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

Tetrode amplifier Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI CERN SPS, RS 2004 Tetrode, Trolley (single amplifier), and transmitter (combination of amplifiers) Two transmitters of eight tubes delivering 2 x MHz, into operation since 1976

Theoretical efficiency Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI RF power in DC power in Heat out RF power out

C B AB A EnEfficient RF Sources workshop, June 2014, Cockcroft Institue Eric Montesinos, CERN-RF14 Amplifier class Input Signal Output Signal Operative curve Class A Input Signal Output Signal Operative curve Class B Unsused area Input Signal Output Signal Less than 180⁰ Operative curve Class C Unsused area 90⁰ Conduction Angle 0⁰ 0 AABBC 0% 25% 50% 75% 100% Efficiency Amplifier ClassDescription Class-AFull cycle 360⁰ of conduction Class-ABMore than 180⁰ of conduction Class-BHalf cycle 180⁰ of conduction Class-CLess than 180⁰ of conduction

Theoretical Class B efficiency DC power is Pdc = Vdc Idc Assuming the tube is linear whilst it is conducting, the dc anode current is found by Fourier analysis of the current waveform and is Idc = Ipk/π 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 % EnEfficient RF Sources workshop, June 2014, Cockcroft Institue Eric Montesinos, CERN-RF15 Vdc Class B Ipk

Tetrode RS 2004 CERN SPS example 26 May - 5 June, 2015, CAS, Accelerators for Medical Applications, Vösendorf, Austria RF Powering, CERN-BE-RF16 Ug1 Control Grid Ug2 Screen Grid Ua Anode Grounded Screen Grid RF in RF out CERN SPS, RS 2004 Tetrode (very) simplified bloc diagram Cathode & Filament

Class B efficiency in practice Two reasons for not achieving this impressive number 1.tube is not fully linear whilst it is conducting 2.Anode voltage must be higher than G2 voltage, VG2 being ~ 10% Vdc 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 % EnEfficient RF Sources workshop, June 2014, Cockcroft Institue Eric Montesinos, CERN-RF17 Class B UG2 Ipk Vdc

Measurement on a YL1530 tube EnEfficient RF Sources workshop, June 2014, Cockcroft Institue Eric Montesinos, CERN-RF18 Operating point RF/DC efficiency = 66.4 %

R&D program, Diacrode Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Example of calculated RF losses on the screen grid for the same cathode length at an output power of 1.4 MW 120 MHz — Diacrode — Tetrode The main difference is in the position of the active zones of the tubes in the resonant coaxial circuits, resulting in improved reactive current distributing in the tube’s electrodes. The basic Diacrode design limits electrical losses and electrodes heating by minimizing the reactive currents in the cathode and grids meshes. This means that compared with conventional tetrodes, Diacrodes can either double the output power at a given operating frequency or double the frequency for a given power output. Diacrodes provide the same gain and efficiency as conventional tetrodes - but at frequencies which are out of reach for tetrodes at an equivalent output power. (already available from Thales)

Diacrode Pulse duration [µs] Repetition rate [pps] Anode Voltage [kV] Anode current [A] Grid2 voltage [kV] Pout [MW] η RF/DC [%] Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI John Lyles, Los Alamos National Laboratory, Design, test and implementation of new MHz RF power amplifier for LANSCE Linac* LA-UR Los Alamos already successfully operate several Diacrodes since 2015 Within the Test Infrastructure and Accelerator Research Area (TIARA) program, CERN and Los Alamos tested a Diacorde for the Ionisation Cooling Test Facility at the Rutherford Appleton Laboratory Novel pulsed RF power amplifier design, Milestone MS pdf pdf Design report of a 3 MW power amplifier, Deliverable pdf pdf

IOT (Inductive Output Tube – Klystrode) Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI 1937Klystron, Russell & Sigurd Variant 1938IOT, Andrew V. Haeff 1939Reflex klystron, Robert Sutton 1940Few commercial IOT 1941Magnetron, Randall & Boot 1945Helix Travelling Wave Tube (TWT), Kompfner 1948Multi MW klystron 1959Gyrotron, Twiss & Schneider 1963Multi Beam Klystron, Zusmanovsky and Korolyov 1980High efficiency IOT

Essentials of IOT IOT density modulation converts the kinetic energy into radio frequency power Vacuum tube Triode input Thermionic cathode Grid modulates e- emission Klystron output Anode accelerates e- buckets Short drift tube & magnets Catcher cavity Collector Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI U grid e- Cathode & Filament e- U anode

IOT available from industry Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Possible frequency range even if never requested yet

IOT amplifiers 26 May - 5 June, 2015, CAS, Accelerators for Medical Applications, Vösendorf, Austria RF Powering, CERN-BE-RF24 CERN SPS, TH 795 IOT, Trolley (single amplifier), and transmitter (combination of amplifiers) Two transmitters of four tubes delivering 2 x MHz, into operation since 2014

Measurement on a TH795 tube EnEfficient RF Sources workshop, June 2014, Cockcroft Institue Eric Montesinos, CERN-RF25 Operating point RF/DC efficiency = 68 %

R&D program, MB-IOT In order to provide an alternative to klystrons, ESS launched a R&D program for Multi Beam IOT Two prototypes will be delivered in 2016 The goal is MHz pulsing up to 3.5 ms – 14 Hz CERN contributes by building a test place Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

R&D program, MB-IOT Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Operating point RF/DC efficiency = 70 % HVPS

High Voltage With gridded tubes, HVPS is very simple No RF -> idle current (can be zero in class B or class C) Even if HV is drooping, the LLRF will impose output power, and tetrode remains able to deliver requested Power Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI RF Pulse time Pdc HV % tolerant HV less tolerant Gridded tubes klystrons

Efficiency summary Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI RF source typeGain [dB] Maximum output power pulsed [kW] Rise time [µs] Pulse length range [ms] Repetition rate range [Hz] Maximum output power CW [kW] Efficiency at working point [%] High voltage needs [kV] Frequency range [MHz] Tetrode ns Almost whatever requested (depends on HVPS design) – 2530 – 400 Diacrode ns – 3030 – 400 IOT ns – 38? – 1300 MB-IOT ns150 (tbc?) Reminder : RF modulating grid voltage -> No RF means No current, direct impact onto efficiency

Frequency & Power range of gridded tubes 26 May - 5 June, 2015, CAS, Accelerators for Medical Applications, Vösendorf, Austria RF Powering, CERN-BE-RF30

Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Good Machine = (when my boss does not hear about it) Efficiency * Reliability * AVAILABILITY * Experienced team * k Overhead Preventive maintenance Operation costs Oversize Acquisition cost Quantity of spares Obsolescence RF/DC efficiency Overall efficiency Including peripherals & building HVAC Manpower management Age profile Training Everything else I missed

Overall Efficiency Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Building HVAC (Heating, Ventilation, Air Conditioning) RF power in DC power in Heat out RF power out AC/DC AC power in Amplifier Cooler Building Cooler DUT (Device Under Test)

Reliability and Availability Overhead and maximum power Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Operating point RF/DC efficiency = 70 % Operating point Including overhead for LLRF and margin for availability Possibility of short power excursions for LLRF (up to 1.5 x operating point)

Reliability and availability Overhead and maximum power Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI Operating point Including overhead for LLRF and margin for availability Operating point RF/DC efficiency = 70 % Possibility of short power excursions for LLRF only for gridded tubes (possible for SSPA at a cost of an overdesigned system)

Experienced team Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

Case study for medical application Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI ** Tubes need highly qualified HV specialists for maintenance* Construction of the infrastructure not included SSPA option requires more volume

Gridded tubes specificities Advantages Not sensible to radiation Can be close to equipment Power supplies can be in an other building Power level sizable for one system per cavity Easier LLRF Better granularity, better Availability RF modulated grid No RF, no current, no need of HV modulation Drawbacks HV & tube Need specialists (not easy to find) Lower Gain Need powerful drivers (SSPA can do the job) Obsolescence Tubes market is decreasing (science applications become more important for suppliers) Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

What R&D program ? Continue Diacrode tests Launch R&D for IOT at low frequencies Train young people to gridded tubes Improve cooling system Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI

Proton Driver Efficiency Workshop 29 February 2016 – 2 March 2016, PSI