The electron beam diagnostic requirements for the
FEL driver accelerator are similar in many respects to those for
other electron accelerators. There are, however, a few parameters
that require additional emphasis, such as the short bunch lengths
(<1 psec), low average beam loss (1 nA), dependence on
the angle and position of beam entering the wiggler, and the 1/4
megawatt of electron beam power.
Beam Position Monitors
The beam position monitoring (BPM) system will consist
of a detector mounted to the beamline and a set of hybrid time
domain/frequency domain electronics. This arrangement will enable
the operations crew to select and track an individual group of
electrons through the accelerator. This type of a system is desired
since there are one or two different electron beams in the various
parts of the accelerator at any given time.
The power of the machine is controlled by varying
the drive laser frequency, and the machine relies on "90%
energy recovery of the megawatt of power in the beam. Through
the linac there are two different electron beams: the beam on
the way to the wiggler and the beam going through energy recovery.
The two beams in the linac are separated by 125 nsec. This allows
ample time to capture the position of an individual bunch of beam.
The detector will consist of four 1 cm buttons mounted
to a vacuum flange. The bipolar signal from each of the electrodes
will shock-excite a resonant circuit. An automatic gain control
(AGC) amplifier will amplify this signal and pass it to both a
synchronized track-and-hold amplifier and a log ratio amplifier.
During low-power operation (2 MHz) this voltage will be captured
using a track-and-hold amplifier synchronized to the beam bunch
selected. When the accelerator is running at full power (40 MHz)
the log ratio circuit will provide the position signal of the
Bunch Length Monitor
The bunch length in the injector will be measured
at injection into the accelerator. The nominal bunch length at
this point is 1 psec. Thus, it is desired to have bunch length
measured to within 0.1 psec (10%). Since the bunch length will
be adjusted only during setup, the range of the measurement should
be 0.5 to 5 psec. In the injector, the bunch length will be measured
using coherent synchrotron radiation generated by the final bend
of the injection staircase. The device uses coherent properties
in the far IR to produce measurable power levels. An interferometer
and detector (bolometer or Golay cell) will be used to measure
the power spectrum of the synchrotron light. The bunch profile
can be derived from this spectrum. Using synchrotron light allows
the meter to monitor the bunch while running at full beam power.
If it is necessary to measure bunch length in the straight section
prior to the staircase, coherent transition radiation can be analyzed
using the same type of detector. However, the foil that produces
the transition radiation cannot remain in the beam path during
full-power operation. Measurement of bunch length using coherent
transition or synchrotron radiation has been successfully performed
[Kung et al.., Phys. Rev. Lett. 73 (7) Aug. 15,
1994, p. 967, and Shibata et al ., Phys. Rev. E
49 (1) Jan. 1994, p. 785]. At CEBAF, tests of a bunch length monitoring
system using coherent synchrotron radiation are currently under
way. Bunch length in the FEL driver accelerator can be measured
using a similar technique.
The beam current will be measured in two different
ways: one method to provide information about fast variations
("10 kHz) in the charge per bunch, and a slower ("1
sec), more accurate method to show variation in average current.
These systems would be located in the 10 MeV injection line and
used to monitor the performance and stability of the photocathode
and drive laser.
The fast system will consist of a modified BPM connected
to an integrator. The sum signal provided will be proportionate
to the charge passing per microsecond. This system will be optimized
for relative changes in the charge per bunch. The slow system
will use a current transformer with a 1 Hz bandwidth to provide
the average current with an absolute accuracy of 10 µA.
The nominal value for normalized rms emittance at
the injection point is 4p
mm-mrad. The required value is 6p
mm-mrad. Hence, it will be necessary to measure emittance to within
30%. at this level. Since space dominates the beam optics in this
region, optical transition radiation (OTR) methods will be used
[Fiorito and Rule, AIP Conf. Proc No. 319, Beam Instrumentation
Workshop 1994, 21]. This type of device uses the OTR interference
pattern generated by the 10 MeV electron beam passing through
two parallel foils. The rms beam divergence can be determined
from the interference rings and the rms beam size can be determined
from the OTR image itself.
In the main accelerator, the emittance will be measured
by varying a quadrupole magnet and observing the beam profile
on a downstream OTR foil.
Table 3 summarizes FEL driver accelerator electron
|500 keV||Beam profile||1 wire HARP|
|Setup position||Beam viewer|
|Beam position||1 stripline BPM|
|10 MeV||Beam position||7 BPMs|
|Bunch Length||SR /interferometer|
|Energy||Beam viewer @ dump|
|Energy spread||Harp @ dump|
|Beam current||Current transformer|
|Beam current||Resonant cavity|
|Linac||Beam position||59 BPMs|
|Wigglers||Position & angle||8 BPMs|
Personnel Safety System
Any safety system requires hardware, procedures, and knowledge to ensure the well-being of personnel. To meet this goal, designs for three systems have been developed to provide protection from the dangers of accelerator operations:
Electrical safety systems to provide protection from high voltage and/or high-current devices.
Radiation systems to protect personnel from exposure to particle beams.
Laser safety systems to prevent exposure to dangerous levels of IR or UV light.
The personnel safety system will be modeled after
the existing CEBAF systems.
Inherent in each of these systems is the concept
of redundancy. All hardware must be designed in such a way that
no single failure will result in the loss of protection. To accomplish
this, two separate circuits are used to detect specific conditions.
For example, each door that is monitored uses two separate switches
to detect the status of the door. Each of these switches in turn
is connected to a separate control circuit. Thus if one switch
were to fail, the other would still operate, thereby providing
the necessary protection. An extension of the redundancy concept
is used in controlling radiation safety system critical devices.
A critical device is one that prevents beam from entering an area.
Two critical devices will be controlled by a radiation safety
system. Thus when a possibility exists for personnel to be in
an area, two devices are inhibited to provide protection of "fail-safe"
circuits. All circuits are designed in such a way that if a circuit
fails, the failure would result in a safe condition. For example,
if the cable that controls a device were cut, the device could
not be enabled. In this way personnel are still safe. Perhaps
the most important part of the Personnel Safety System is the
searching and securing of an area. This is done to ensure that
the area is not occupied before beam or power supplies are enabled.
The securing of an area requires that a minimum of two qualified
persons thoroughly search an area in a predefined sequence.
The search sequence will be programmed into a Programmable
Logic Controller for the FEL Personnel Safety System. The order
in which the interlocks are reset has been designed to ensure
the area is searched in a logical order, thereby ensuring personnel
are not missed by the search team.
The FEL Personnel Safety System is segmented into
seven areas: the accelerator hall and the six optical labs. Once
the accelerator hall is secured, each of the optical labs must
have its two critical devices in place or also be secured. The
normal procedure is to have all of the optical labs locked out
when the accelerator is in open access, then to secure each lab
and retract its critical devices as required once the accelerator
has been locked up.
Once an area has been searched and secured, status
displays on the outside of each controlled access door and each
section gate indicate to individuals that the area is interlocked
and that access is forbidden to unauthorized personnel. Immediately
before the beam or laser light is brought into an area or power
supplies are enabled, a prerecorded message consisting of a siren
and verbal announcement will be played so that any personnel who
have somehow been missed in the search-a highly unlikely search
outcome, but nonetheless necessary to address-can push an emergency
crash button and/or leave the area. In addition to the audible
warning, strobe lights will flash, giving a visual indication
the area is transitioning into a hazardous state while the audio
message is played. Emergency crash buttons, audio warning speakers,
and visual warning strobes will be located at approximately 50
m increments in the accelerator enclosure and in plain view in
each of the six optical labs. Pushing one of these buttons will
inhibit beam and power supplies and require that the actuated
switch be locally reset before the area can be enabled.
All doors to an area are locked and the keys to open
these doors are interlocked and guarded in the FEL Control Room
key tree. Distribution of these keys is not taken lightly. Only
authorized personnel are allowed access. The type of access determines
the authorization level required for the individual.
Control Computers and Software
The function of the control system is to keep the
operations crew informed of the current state of the accelerator
hardware and to provide an efficient means of starting up the
machine, making required minor corrections, and archiving and
retrieving data. This collection of hardware and software will
be patterned closely after the CEBAF control system.
The EPICS operating system provides a reliable connection
from operators on work stations in the control room to the hardware
devices in the accelerator. The FEL will take advantage of the
software development that has been done at CEBAF and the other
labs in the EPICS collaboration. This will save man-years of software
The FEL control system will consist of 12 workstations
which will be available for both development of the required additional
software and for operation of the accelerator. There will be a
server (HP 730 for example) which will interface the workstations
to the hardware in the field. This server contains the bootable
code for the remote intelligent local input/output controllers
(IOCs). The IOCs will contain between 4 and 16 megabytes of memory
depending on the complexity of the task.
Each IOC also contains four serial links which can
be configured to control the drive laser and the bulk power supplies.
This serial link is based on the single ended RS232 format which
requires a simple adapter to convert it to the differential RS485
link. The subsystem needs are shown in Table 4.
|RF control||1 systems can be operated from as single IOC; the 1 1/4 + buncher is = 1 1/2|
|Diagnostics||There are 40 BPMs, 28 beam viewers, 7 synchrotron light monitors, and a double slit|
|DC power||41 dipoles, 86 quads, 36 sextupoles, 35 horizontal trims, 35 vertical trims, plus 6 injector trims|
|Cryogenics||One to interface to CAMAC|
|Vacuum system||37 ion pumps, 35 gate valves|
|Machine protection||Monitor only|
|Personnel safety||Monitor only|