Transcript Document

Front End Test
Stand Klystron
Alan Letchford
UKNF Plenary Meeting
22nd April 2009
• The Front End Test Stand (FETS) aims to demonstrate key technologies
for the front end of the next generation of high power pulsed proton
• Applications include:
• UK Neutrino Factory
• ISIS upgrades
• Future Spallation Neutron Sources
• Waste Transmutation
• ADS etc
• The key components of FETS are:
• High intensity, high duty factor, H- ion source
• Magnetic Low Energy Beam Transport (LEBT)
• 324 MHz 4-vane Radio Frequency Quadrupole (RFQ)
• Medium Energy Beam Transport (MEBT)
• Very high speed beam chopper
• Comprehensive diagnostics
FETS is a collaboration between ISIS, ASTeC, Imperial College London,
Warwick University and the University of the Basque Country in Spain.
A big challenge for all future MW scale proton accelerators is controlling
beam loss. Beam loss leads to component activation – component
activation hinders hands-on maintenance.
Absolute loss levels in the future machines (1 – 10 MW beam power) must
be similar to that on ISIS (160 kW beam power). Fractional loss must
therefore be reduced by orders of magnitude.
A significant reduction in beam loss in circular accelerators can be
achieved by chopping the beam in the injector linac so as to precisely fill
the ring RF bucket – no trapping loss and much reduced extraction loss.
‘Perfect’ chopping is extremely challenging. One of the main focuses for
FETS is to demonstrate perfect chopping on a full power beam.
Chopping will be demonstrated on a 3 MeV H beam delivered by the
A frequency of 324 MHz has been chosen due to the availability of RF
power at this frequency. The alternative possibility of 405 MHz makes
perfect chopping more difficult.
The 4m long RFQ will need 6-700 kW of structure power and about
175 kW for the beam.
Total RF requirement is ~1MW, up to 50pps and 2ms pulses.
The chosen RF power source is a Toshiba E3740A klystron as used on the
J-PARC linac.
Due to the low frequency and high power it is a very large device.
~5m long, ~3 tonnes.
3MW max peak power, 80kW max average window power
Modulating anode
110kV cathode potential, 50A peak beam current
A fully solid state power supply and modulator has been purchased from
Diversified Technologies Inc of Boston, USA
Due to the high power and voltage this is also quite large.
The system consists of an autotransformer, 2 switching power supplies, a
capacitor bank , solid state switch, floating heater power supply and
modulating anode power supply.
Total weight ~9 tonnes – before being filled with ~7000l of oil.
A dummy load and waveguide system including directional coupler and
arc detector was supplied by Mega Industries.
Temperature controlled ethylene glycol load.
This proved to be the least reliable
part of the system.
But it was cheap!
Due to the klystron being the only suitable load for the power supply, the
system had to be commissioned as a whole. All components had
previously undergone full power factory tests.
One week was scheduled for commissioning with engineers from Toshiba
and diversified.
3 days were spent making all the interconnects and high voltage
terminations, filling both klystron and modulator with oil and testing
During the first day of running 70kV, 10pps, 250µs pulses were delivered
to the klystron.
There were no signs of breakdowns or vacuum rises so RF was gradually
applied to the input.
~600kW @ 10pps & 250µs was achieved.
On Good Friday the voltage was increased to 95kV and 1MW @ 30pps &
1ms pulses was achieved before problems with the dummy load halted
Not a single HV breakdown or waveguide arc was detected.
Some low level x-rays were detected around the collector and window.
Multiple problems with the dummy load limited progress towards full
FETS operating conditions:
Overheating motor
Leaking pump seal
Faulty flow switch
Blocked heat exchanger
Although frustrating these are relatively trivial problems to fix but there
was no time before the teams from Toshiba and Diversified returned
The load will be fixed and some temporary aspects of the installation
made permanent before running again.
95kV, 2ms pulse at low prf
Pulse rising edge
50pps, 500µs pulse
Wire test (crawbar)
Over 1 week in early April the FETS klystron and PSU/modulator were
successfully commissioned to close to full power.
1MW, 30pps, 1ms were achieved with remarkable ease.
Limiting factor was the poorly designed cooling circuit on the high power
glycol load.
Both the Toshiba klystron and the Diversified PSU/modulator appear to
be high quality pieces of equipment.
The installation will now be made permanent and the x-ray shielding
improved before further long term running is attempted.