scispace - formally typeset
Open AccessJournal ArticleDOI

A Proposed RF System for the Fusion Materials Irradiation Test Facility

Reads0
Chats0
TLDR
In this paper, the phase and amplitude control of the EIMAC 8973 tetrode was evaluated for the deuteron beam at the FMIT facility and it was shown that it is the most suitable final amplifier tube for each of a series of 15 amplifier chains operating at 0.5-MW output.
Abstract
Preliminary rf system design for the accelerator portion of the Fusion Materials Irradiation Test (FMIT) Facility is in progress. The 35-MeV, 100-mA, cw deuteron beam will require 6.3 MW rf power at 80 MHz. Initial testing indicates the EIMAC 8973 tetrode is the most suitable final amplifier tube for each of a series of 15 amplifier chains operating at 0.5-MW output. To satisfy the beam dynamics requirements for particle acceleration and to minimize beam spill, each amplifier output must be controlled to ±1° in phase and the field amplitude in the tanks must be held within a 1% tolerance. These tolerances put stringent demands on the rf phase and amplitude control system.

read more

Content maybe subject to copyright    Report

IEEE Transactions on Nuclear Science, Vol. NS-26, No. 3, June 1979
A PROPOSED RF SYSTEM FOR THE FUSION MATERIALS IRRADIATION TEST FACILITY*
M. V. Fazio, H. P. Johnson,** W. J. Hoffert,* and T. 3. Boyd
Los Alamos Scientific Laboratory,
LOS
Alamos, NM 87545
Abstract
Prelimi.nary rf system design for the accelerator
portion of the Fusion Materials Irradiation Test
(FMIT) Facility is in progress. The 35-MeV, 100~mA,
cw deuteron beam will require 6.3 MW rf power at 80
MHz.
Initial testing indicates the EIMAC 8973 tetrode
is the most suitable final amplifier tube for each of
a series of 15 amplifier chains operating at 0.5-MW
output.
To satisfy the beam dynamics requirements for
particle acceleration and to minimize beam spill, each
amplifier output must he controlled to +l” in phase
and the field amplitude in the tanks must be held
within a 1% tolerance.
These tolerances put stringent
demands on the rf phase and amplitude control system.
General Description
Prelimin‘lry rf system design for the linear
accelerator (linac) portion of the FMIT facility is in
progress. 1
The 35-MeV accelerating structure will
consist of a low-beta radio-frequency quadrupole (RFQ)
accelerator/huncher*
up to 2 MeV and two post-
coupled Alvarez tanks in series, with an intertank
spacer at the 20-MeV point. The 35-MeV, 100-mA cw
deuteron beam will require approximately 6.3~MW rf
power at 80 MHz.
*Work supported by the U.S. Department of Energy.
*Hanford Engineering Development Laboratory (HEDL),
Westinghouse-Hanford Company, Richland, WA 99352, HEDL
employee wor’king at the Los Alamos Scientific
Laboratory.
?Consultant, Los Alamos Scientific Laboratory, Los
Alamos, NM 87545
The basic layout of the FMIT rf system is illus-
trated in Fig. 1. Each rf amplifier chain will
generate at least 500-kW of power. The RFQ will
require two amplifier chains and each linac tank will
need at least six or seven.
A multiplicity of drive
loops will be used for inductively coupling the rf
into the accelerator tanks.
Each power amplifier (PA)
will have its own coupling loop.
To obtain optimum
particle acceleration and to minimize beam spill, beam
dynamics requires that the accelerator fields be con-
trolled to +l” in phase.
Also, the field amplitude
in the tank: must be held within a 1% tolerance.
These requirements place stringent demands on the
phase and amplitude control systems.
High Power Amplifiers
Amplification from 100 W to 500 kW will be accom-
plished by several stages of vacuum tube amplifiers.
The tube selected for the final amplifier is the EIMAC
8973 (formerly X-2170) power tetrode. The Los Alamos
Scientific Laboratory (LASL), in cooperation with
EIMAC, has embarked on a test program to determine the
capabilities of the 8973 tetrode.
The tube has been
operated in the grounded-grid grounded-screen config-
uration at 80 MHz for over 6-l/2 hours at a 525-kW rf
output * The tube had a 14-db power gain, which was
slightly higher than the calculated gain of 12 db,
because in this configuration it was behaving more
like a high-p triode than a tetrode. The most vulner-
able tube element is believed to be the screen grid.
Because of the high plate to screen capacitance
(-140pF), rf displacement current heating of the
screen structure becomes significant at frequencies
above 50 MHz. Because of the very low thermal capac-
ity of the screen structure, this element should reach
9 l
100
2
DRIFT TUBE 20
DRIFT TUBE
35 M*V
INJECTOR 2 RF QUADRUPOLE ---+
MtV LINAC - TANK *I
LINAC -TANK ‘2 100mo
.
RF
IOOW
‘* AMPTD.4 LOW
CONT.
+-’ POWER
RF
RF
--, AMPL.
+ PHASE
A -
CONT.
A
RF RF
oow oow
I+ AMPTD __) LOW I+ AMPTD __) LOW
CONT. 7 POWER CONT. 7 POWER
RF RF
RF RF
-+ AMPL -+ AMPL
4 PHASE 4 PHASE
A A
CONT. CONT.
L L
4 4
0’
CENTRAL
CONTROL
SYSTEM
FREO.
I
CONT.
SYSTEM
TYP
OF
3018
Fig. 1. The FMIT rf system.
OOlg-9499/79/0600-3018$00.75 @ 1979 IEEE
© 1979 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material
for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers
or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

a stable temperature in
just
a few minutes of
operation.
Consequently,
if the screen doesn’t fail
within several minutes of operation, it will probably
survive indefinitely.
On the other hand, the ceramic
tube seals are more prone to deterioration over a long
period of time but seal integrity is more dependent on
the larger dissipations of the filament and the anode,
which in all cases were running within or well below
maximum ratings.
The maximum output of the 8973 has
not been determined yet because of output cavity
arcing at the 550-kW level and a lack of sufficient
drive power.
Drive modulation tests up to 300 kW exhibited a
reasonably linear response,
There was no evidence of
discontinuities in gain and no spurious components
could be detected in the 8973 output under several
conditions of drive modulation.
Low Power rf
The low-power rf stage will supply an rf signal of
the proper frequency, phase, and amplitude (O-100 W)
to drive the PA such that the PA output is also at the
required frequency, phase, and amplitude (O-500 kW)
for optimum accelerator performance.
The accelerating field is the vector sum of the rf
fields produced by the energy delivered to each tank
by several rf power amplifiers. Feedback control
loops around each PA and around each accelerator tank
insure that the field in the tank remains at the
correct phase and amplitude with or without beam,
The
phase and amplitude controlling takes place at the 1-W
level,
The rf drives for all three tanks are essen-
tially the same so only one amplifier chain will be
described in detail.
Figure 2 contains the block
diagram for one complete rf amplifier chain including
feedback control loops,
FREQUENCY CONTROL 8.
PHASE FIEF: LINE
The tank resonant frequency is critically tempera-
ture dependent.
When the tank is cold, as in start
up, the resonant frequency could be as far as 1.5 MHz
from 80 MHz and the bandwidth of the tank is so narrow
(-3 kHz) that the 80-MHz amplifiers cannot drive
energy into it.
The frequency control system is
designed to find, lock on, and track the linac reso-
nance so that
the tank can be heated with rf energy.
In the search mode the rf source is switched to a
voltage controlled oscillator (VCO), which generates
1-ms pulses with a pulse repetition frequency of 120.
The VCO shifts about one tank bandwidth on each
pulse.
When a feedback signal from a pick-up loop in
the tank indicates power flow into the tank, the VCO
locks
onto that
frequency and tracks it as the tank
warms up to normal operating temperature.
At this
time the VCO is switched out and is replaced with a
crystal oscillator with an 80-MHz output for normal
operation.
The central control system (CC’S) generates an
amplitude set-point signal proportional to the desired
field level in the accelerator tank. This set point
is also proportional to the average output power
required of each rf amplifier chain to achieve the
desired tank field. This set point is distributed to
the amplitude control circuit of each PA.
Coupling
loops in the accelerator sample the field amplitude in
several places.
These sample signals are detected and
summed to generate a tank feedback signal that is pro-
portional to the average rf field in the tank. The
tank amplitude set-point signal is subtracted from the
tank feedback signal to generate a control signal pro-
portional to their difference (error). This tank
amplitude control signal is distributed to the ampli-
tude con.rol circuit of each PA.
mm--
hi POWEl$
PS -POWER SPLITTER
0 -PHASE SHIFTER
~/~~IGITAL TO ANALOQ
CONVERSION
DS-DIODE SWITCH
L!!
Fig. 2. A single rf amplifier chain.
3019

In normal operation each PA will deliver its equal
share of the required power. Under some conditions a
PA will operate at a higher or lower output power
level. The CCS may generate on “offset” signal for
some drive chains to increase or decrease their output
power.
The amplitude control circuit of each PA receives
four control signals:
1.
Amplitude set point, proportional to the average
power output of each PA.
2.
Amplitude offset, proportional to the difference
between actual power output of a PA and average
power output of all PAS.
3.
Tank amplitude control, proportional to the
difference in the actual tank field and the tank
amplitude set point.
4. PA amplitude feedback, proportional to the PA
power output.
Signals 1, 2, and 3 are combined to generate an
“effective set point” for each PA. This signal is
compared with the PA amplitude feedback (4 above) to
generate a control signal for the voltage variable
attenuator that controls the overall gain of the rf
amplifier chain.
When beam is injected, the heavy beam loading
causes field droop in the accelerator. The tank
amplitude control signal increases, which changes the
“effective set point” for each PA.
This change
increases the power gain of each PA causing the
average field level in the accelerator to maintain its
desired value.
The phase control system is quite similar to the
amplitude control system. The phase control system
consists of four parts:
phase detector, phase con-
troller,
+45O varactor-tuned electronic phase
shifter, and a PIN diode digital phase shifter with a
360° range in 45O increments. These four parts
are shown in Fig. 2. The SO-MHz oscillator acts as
The CCS generates a phase set point that is pro-
portional to the desired rf field phase in the tank
and also proportional to the PA output with respect to
the phase reference .
This set point is compared with
the actual detected tank phase and a phase control
signal (error signal) is generated and distributed to
each phase controller. The phase set point, the phase
control signal, and CCS generated phase off-set signal
are all inputs to the phase controller of each PA. A
phase feedback signal that is proportional to the
difference between the reference phase and the PA out-
put phase is also a phase controller input. Because
the phase shift through the PA changes with output
power, this controller input assures a constant PA
output phase under all operating conditions.
The phase controller generates an output signal
that tunes the varactor phase shifter to produce the
required phase shift.
The digital phase shifter is
adjusted to keep the varactor phase shifter within its
operating range,
Prototype
A prototype FMIT accelerator capable of producing
5-MeV H$ particles is under development now at
LASL.
This accelerator will require four rf amplifier
chains to supply the necessary energy for beam accel-
eration.
The prototype rf system used at LASL will
prove the final design for the FMIT accelerator to be
built at the Hanford Engineering Development
Laboratory in Richland, Washington.
References
1.
Jameson, R. A., “High-Intensity Deuteron Linear
Accelerator (FMIT) ,”
presented at the 1979
Particle Accelerator Conference (paper B-l), San
Francisco, California, March 12-14, 1979.
2.
3. M. Potter, R. H. Stokes, G. W. Rodenz, and F.
J. Humphrey,
“RF Quadrupole Accelerating Structure
Research at Los Alamos ,‘I presented at the 1979
Particle Accelerator Conference (paper D-121, San
Francisco, California, March 12-14, 1979.
the phase reference for the phase control system.
3020
Citations
More filters
Journal ArticleDOI

High-Intensity Deuteron Linear Accelerator (FMIT)

TL;DR: The Fusion Materials Irradiation Test (FMIT) Facility at the Hanford Engineering Development Laboratory (HEDL), Richland, Washington, was designed to provide a high neutron flux and a neutron energy spectrum representative of fusion reactor conditions in volumes adequate to screen and qualify samples of candidate fusion reactor materials as mentioned in this paper.
Journal ArticleDOI

The FMIT Accelerator

TL;DR: A 35-MeV 100-mA cw linear accelerator is being designed by Los Alamos for use in the Fusion Materials Irradiation Test (FMIT) Facility as mentioned in this paper.
Patent

Electron propulsion unit

TL;DR: An electron acceleration device uses thermionic fission cells, an electromagnetic scoop coil, and/or microwaves for power as discussed by the authors, where a power control junction and electron injector control and feed free electrons in packets into the acceleration components that consist of a series of either induction module units, or radio-frequency linacs module units.
Journal ArticleDOI

Higher Frequency Operation of Linear Accelerator

TL;DR: In this paper, the high frequency operation of linacs is discussed and several trial designs for fusion material irradiction test facility in Japan are given, where the size of linac namely the construction cost may be reduced by using 1 MW CW klystrons with the frequency of about 1 GHz.
References
More filters
Journal ArticleDOI

Radio Frequency Quadrupole Accelerating Structure Research at los Alamos

TL;DR: In this paper, the results of studies of radio frequency quadrupole accelerating structures are presented, and the results indicate that this is a promising structure for transporting and accelerating ion beams in the energy region of 50 KeV to 2 MeV.
Journal ArticleDOI

High-Intensity Deuteron Linear Accelerator (FMIT)

TL;DR: The Fusion Materials Irradiation Test (FMIT) Facility at the Hanford Engineering Development Laboratory (HEDL), Richland, Washington, was designed to provide a high neutron flux and a neutron energy spectrum representative of fusion reactor conditions in volumes adequate to screen and qualify samples of candidate fusion reactor materials as mentioned in this paper.
Related Papers (5)