Design of an Improved Ion Chamber for the SNS
R. L. Witkovert and D. Gassner
Brookhaven National Laboratory, fTechSource, Inc., Santa Fe, NM
Abstract. Ion chambers are in common use as beam loss monitors at many accelerators. A unit
designed and used at FNAL and later at BNL was proposed for the SNS. Concerns about the ion
collection times and low collection efficiency at high loss rates led to improvements to this unit
and the design of an alternate chamber with better characteristics. Prototypes have been tested
with pulsed beams. The design and test results for both detectors will be presented.
BACKGROUND
The ion chambers (IC's) designed for the FNAL Tevatron by Shafer in 19821 and
built by Troy-Onics2, have been used at both FNAL and BNL. These are simple in
design, consisting of a hollow nickel inner electrode and a nickel foil cylinder outer
electrode in a glass enclosure filled with argon. During testing of the chambers for use
in RHIC at BNL3 it was found that detector response varied unacceptably using the
preferred bias voltage polarity. Normally, from space charge considerations, electrons
are collected on the inner electrode. The IC's were tested with a Cs-137 source which
produced 1 Rad/hr at the chamber. The signal current, measured as the bias voltage is
varied, is characterized by a rapid increase which quickly rolls over ("knee") into a flat
region ("plateau") as the field becomes high enough that all charge produced is
collected. Ion chambers operate in this plateau region. At higher voltage, the electric
field is sufficient to cause multiplication and the chamber enters the proportional
region. Results for the first 40-50 RHIC production IC's varied widely with many
having extended "knees" and early onset of the proportional region. Test results
obtained from FNAL for the original detectors did not show this behavior. It is
possible that the original FNAL detectors, built 15 years earlier, did not exhibit this
problem but subsequent vendor technicians may have unknowingly changed details of
the construction. However the detectors were not usable as delivered.
At BNL, conditioning using a Tesla coil and "spot-knocking" showed some success
but required significant time per unit. It was suggested that the opposite bias polarity
would improve uniformity. Tests showed excellent unit-to-unit reproducibility when
ions, rather than electrons, were collected on the inner electrode, so it was decided to
use the "unconventional" field polarity for RHIC. The major consequences were lower
collection efficiency (saturation) for high dose rate losses, and slower signal risetimes,
which were not expected to be problems in RHIC. For SNS, however, this drop in
efficiency would be significant for the worst-case 1% local loss, and a rise time
comparable to the pulse width would be unacceptable.
CP648, Beam Instrumentation Workshop 2002: Tenth Workshop, edited by G. A. Smith and T. Russo
© 2002 American Institute of Physics 0-7354-0103-9/02/$19.00
337
SIGNAL RISE TIME CONSIDERATIONS
With a 1 msec pulse width, signal rise time is important in the SNS. Electrons are
collected in a few microseconds with either bias polarity, but the heavier ions take
much longer moving to the inner electrode. At 2000 V the ion collection time would
be close to 700 usec. If the losses were constant the signal would continue to rise over
the 1 msec Linac pulse as the slow ions arrived, followed by a tail for another 700
usec after the pulse. While the electron signal would allow rapid beam abort for large
fast losses, the waveform during the pulse would require unfolding.
Electron and ion charges are generated equally, but will produce equal voltages
only for a parallel plate geometry. For cylindrical geometry, using the relative
capacitance of a line charge to the 2 electrodes, Shafer4 showed that the fraction of
current due to electrons at the inner electrode (anode) is:
a
\F2(K)RdR
<F2>=—-—————— = ———-———————
2(x2-l)Ln(x)
where x = a/b
(1)
For the FNAL ion chambers, a = 0.75" and b= 0.125" and <F2>= 0.749 . That is,
the fast electron signal will be 3 times the slow ion current with the conventional
polarity but only 1/3 for the reverse-bias polarity. This was observed in tests in the
RHIC transfer line. Clearly, the FNAL chambers would be unsuitable for SNS unless
the conventional bias polarity could be used.
The positive ion transit time is given by:
t=
where:
d = Effective electrode separation [cm] for cylindrical geometry5
d=
(3)
PO = Atmospheric
ricpi
pressure
P = Working pressure
|io = Ion mobility at STP [ cm2/(V-sec)]
V = Applied Voltage [V]
The transit time can be reduced by decreasing the electrode gap but with increased
gradient corona may occur. A design was proposed6 in which the radii were closer
together and the chamber lengthened to keep the same volume. Keeping the original
outer electrode radius, a = 0 .75" but increasing the inner radius to b= 0.5", the ion
collection time at 3 kV bias would be reduced from 560 usec to 72 usec. The
338
calculated results
results are
calculated
are shown
shown in
in Figure
Figure 1.
1. “Negative
"Negative bias”
bias" refers
refers to
to negative
negativevoltage
voltageon
on
the outer
outer electrode.
electrode. Clearly
Clearly the
the
the “positive
"positive bias”
bias" case
case isis not
not suitable
suitablefor
forSNS.
SNS.The
TheFNAL
FNAL
chamber with
with negative
negative bias
bias might
might be
chamber
be acceptable
acceptable since
since the
the electrons
electrons contribute
contribute75%
75%of
of
the signal
signal within
within aa few
few microseconds.
microseconds. The
the
The new
new design
design provides
providesthe
thebest
bestresponse.
response.
1.2
V=
V= 3000
3000 V,
V, Argon
Araon
FNAL Design Neg bias
Current [Normalized]
1
New Design Neg bias
0.8
FNAL Design Pos bias
0.6
0.4
0.2
0
0
500
500
1000
1000
Time [microseconds]
1500
1500
2000
Time [microseconds]
Figure 1.
1. Calculated
Calculated signal
Figure
signal for
for SNS
SNS Linac
Linac pulse
pulse on
on original
original FNAL
FNALand
andnew
newBLM.
BLM.
COLLECTION
COLLECTION EFFICIENCY
EFFICIENCY
While the
the equations
equations describing
While
describing the
the collection
collection efficiency
efficiency have
have been
been known
known since
since
Thomson
described
them
in
1899,
they
have
never
been
explicitly
Thomson described them in 1899, they have never been explicitly solved.
solved.
Approximate solutions
solutions for
the simpler
applied to
Approximate
for the
simpler parallel
parallel plate
plate geometry
geometry can
can be
be
7 applied to
cylindrical
if
an
“equivalent”
gap
(eq.
3)
is
used.
Boag
and
Wilson
7 solved the
cylindrical if an "equivalent" gap (eq. 3) is used. Boag and Wilson solved the
equations assuming the ionization distributions were second and third order
equations
assuming the ionization distributions were second and third order
polynomials, and that the radiation was constant. Their result, which gives a good fit
polynomials, and that the radiation was constant. Their result, which gives a good fit
for efficiency > 0.7 is given by:
for efficiency > 0.7 is given by:
1
(4)4
f =
ξ2
f=-^
1+
6
<>
where:
where:
α
P d 4q
(5)
ek1k 2 P0 V 2
(5)
Pn V
e is the electron charge,
k1 is the electron mobility,
eα is
charge,recombination
is the
the ion
electron
mobility,
is the
the electron
first Townsend
kki2 is
mobility,
acoefficient,
is the first Townsend recombination
k
is
the
ion
mobility,
q 2is the ionization change density:
coefficient,
q is the ionization change density:
ξ2 =
q=
a P d42q
I signal
× 3×109
v
where Isignal is the signal current generated in the ion chamber by the beam loss and
where
signal
in the
chamber
bysensitivity
the beam loss
gnai is the
v is theIsiactive
volume
ofcurrent
the ion generated
chamber. For
theion
FNAL
IC, the
is 70and
3
vnA/rad/sec,
is the active
volume
of
the
ion
chamber.
For
the
FNAL
1C,
the
sensitivity
70
and the volume is 110 cm 3 so for the accepted upper limit of 1 %islocal
nA/rad/sec,
and
the
volume
is
110
cm
so
for
the
accepted
upper
limit
of
I
%
local
loss, the dose rate is 9.2 krad/sec during the linac pulse.
loss, the dose rate is 9.2 krad/sec during the linac pulse.
339
Consider a chamber with argon gas:
a = 1 x 10'6 [cm3/sec]
ki = 1.8[cm2/(V-sec)]
k2=1.3[cm2/(V-sec)]
V = 3000 [Volts] (Negative bias)
P = 725 [mTorr]
Then
Outer Radius [cm]
Inner Radius [cm]
Active Length [cm]
Volume [cm3]
Effective Gap [cm]
lonization Density [esu/cm3-sec]
Ion Transit Time [usec]
Electron Fraction (neg on outer)
Collection Efficiency
FNAL
1.905
.3175
10
110.84
1.991
16850
564
.7495
.357
New Design
1.905
1.27
17.4
110.21
0.6436
16850
72.2
.5669
.971
While the efficiency calculation for values <0.7 may not be accurate, clearly the
new chamber would be superior to the original, both in ion transit time and collection
efficiency, but, since the electrode gap is smaller, great care must be taken with the
high voltage design. Guard electrodes would be required to decouple the signal
electrode from leakage. The design would be more difficult to manufacture than the
simple FNAL detector, with higher detector cost. Note that the electron fraction is
0.567, with 43% of the signal coming from the now ions.
Boag8 also found a solution for the pulsed radiation case when the loss occurs in a
time short compared with the ion transit time. For this case and the above parameters
the FNAL chamber efficiency is 21% and the new design is 60%. This might apply to
the first 5-10 mini-pulses for SNS but the static case solution might better describe the
efficiency late in the pulse.
THE NEW ION CHAMBER DESIGN
The design of the new ion chamber is shown in Figure 2. Voltage gradients have
been reduced by rounding the electrode ends. Guard electrodes divert leakage from the
HV electrode to ground and assure the signal collector is in a uniform electric field
region. They also prevent ions at the chamber ends from contributing to the signal
resulting in longer transit times. Three prototypes units have been made. The first was
limited to 3 kV when filled with argon due to HV feedthrough breakdown, but when
filled with nitrogen was able to reach 5 kV. It was then installed in the transport line
between BNL Linac and Booster. Units 2 and 3 were built with feedthroughs rated for
higher voltage and with improved ceramic design. The radius of curvature of the ends
of the guard rings and signal electrodes were also increased. These units have been
340
tested
installed in
in the
the BNL
BNL Booster
Booster
tested to
to 4.5
4.5 kV
kV with
with argon. Prototype 2 has been installed
tested to 4.5 kV with argon. Prototype 2 has been installed in the BNL Booster
extraction
similar to
to SNS
SNSRing
Ringextraction.
extraction.
extraction line
line for
for tests
tests with
with short pulse beams similar
extraction line for tests with short pulse beams similar to SNS Ring extraction.
Macor
Macor
Ceramic
Ceramic
Signal
Signal
Electrode
Electrode
Guard
Guard
electrode
electrode
HV
HV
electrode
electrode
Figure 2.
2. Drawing of
of new design
BLM
Figure
BLM prototype.
prototype.
Figure 2. Drawing
Drawing of new
new design
design BLM
prototype.
IMPROVEMENTS TO
TO THE FNAL
DETECTOR
IMPROVEMENTS
DETECTOR
IMPROVEMENTS TO THE
THE FNAL
FNAL DETECTOR
As an alternative
alternative to designing
designing a new chamber,
chamber, the original
was
As
was studied
studied to
to see
see ififif
As an
an alternative to
to designing aa new
new chamber, the
the original
original was
studied
to
see
modifications
could
be
made
to
improve
its
performance.
The
difficulties
with
modifications
The difficulties
difficulties with
with the
the
modifications could
could be
be made
made to
to improve
improve its
its performance.
performance. The
the
conventional
(negative)
bias
polarity
in
the
FNAL
IC
were
believed
to
be
due
to
high
conventional
(negative)
bias
polarity
in
the
FNAL
1C
were
believed
to
be
due
to
high
conventional (negative) bias polarity in the FNAL IC were believed to be due to high
field points at
at the tip
tip of the
the inner electrode
electrode and at
field
at the
the supports
supports for
for the
the outer
outer electrode,
electrode,
fieldpoints
points at the
the tip of
of the inner
inner electrode and
and at
the
supports
for
the
outer
electrode,
since
these
are
where
breakdown
occurs
when
the
voltage
is
raised.
Changes
were
since
these
are
where
breakdown
occurs
when
the
voltage
is
raised.
Changes
were
since these are where breakdown occurs when the voltage is raised. Changes were
made
to
the
outer
electrode
to
eliminate
crimping
at
the
flared
end
of
the
outer
made
end of
of the
the outer
outer
made to
to the
the outer
outer electrode
electrode to
to eliminate
eliminate crimping at the flared end
electrode at the
the support pin
pin attachment locations.
Figure 3a shows
electrode
shows the
the original
original outer
outer
electrode at
at the support
support pin attachment
attachment locations. Figure 3a shows
the
original
outer
electrode
crimped
inward
by
the
support
rods.
Figure
3b
shows
a
modified
chamber
electrode
chamber
electrode crimped
crimped inward
inward by
by the
the support rods. Figure 3b shows a modified chamber
with the outer
outer electrode slotted
slotted to allow the support to pass the flared end
with
end without
without
with the
the outer electrode
electrode slotted to allow the support to pass the flared end
without
bending
it
inward.
bending
bending itit inward.
inward.
Figure 3a. Original FNAL Outer electrode
Figure3a.
3a. Original
Original FNAL
FNAL Outer
Outer electrode
electrode
Figure
Figure 3b. Slotted outer electrode
Figure 3b.
Figure
3b. Slotted
Slotted outer
outer electrode
electrode
So far 5 chambers have been modified in this way. Tests indicate that this has
So
far 55 chambers
chambers
have
been
modified
in this
this Figure
way.
Tests
indicate
that
has
So
far
have
in
way. 4a
Tests
indicate
that this
this
has
significantly
improved
thebeen
unit modified
to
unit response.
shows
the cesium
source
significantly
improved
the
unit
to
unit
response.
Figure
4a
shows
the
cesium
source
significantly
improved
the
unit
to
unit
response.
Figure
4a
shows
the
cesium
source
tests of 5 of the first RHIC chambers with conventional bias polarity. Figure 4b shows
testsof
of 55 of
of the
the first
first RHIC
RHIC chambers
chambers with
with conventional
tests
conventional bias
bias polarity.
polarity. Figure
Figure 4b
4b shows
shows
341
results
electrodes. The
The uniformity
uniformity isis clearly
clearly
results for
for 55 units
units with
with modified
modified (slotted)
(slotted) outer
outer electrodes.
improved
with
the
slotted
electrodes.
improved with the slotted electrodes.
ItItwas
may have
have been
been due
due to
to sharp
sharp edges
edgesatat
wasfelt
felt that
that some
some of
of the
the voltage
voltage limitation
limitation may
the
open
end
of
the
hollow
center
electrode.
Several
chambers
were
built
with
the
the open end of the hollow center electrode. Several chambers were built with the
slotted
outer
and
a
rounded
solid
tip
inserted
in
the
end
of
the
electrode.
Tests
of
these
slotted outer and a rounded solid tip inserted in the end of the electrode. Tests of these
units
from the
the slotted
slotted outer
outer electrode
electrode alone.
alone.
units did
did not
not show
show aa significant
significant improvement
improvement from
While
the production
production uniformity,
uniformity, the
theion
ion
While these
these modifications
modifications seem
seem to
to have
have improved
improved the
transit
are still
still aa problem
problem for
for SNS.
SNS. IfIf the
the design
design
transit time
time and
and high
high dose
dose rate
rate saturation
saturation are
could
voltage then
then the
the ion
ion transit
transit time
time would
would
could be
be improved
improved to
to allow
allow higher
higher operating
operating voltage
decrease
and
collection
efficiency
increase.
The
improvement
is
somewhat
less
than
decrease and collection efficiency increase. The improvement is somewhat less than
linear
with
voltage,
so
this
might
not
be
very
productive.
linear with voltage, so this
productive.
2.50E-11
35.00
30.00
2.00E-11
Notched 1
Notched 2
Notched 3
Notched 4
Notched 5
1.50E-11
20.00
Current
Signal Current [pA]
25.00
r
15.00
1.00E-11
BLM001
BLM002
BLM003
BLM004
BLM008
10.00
5.00
5.00E-12
0.00
00
500
500
1000
1000
1500
1500
2000
2000
2500
2500
0.00E+00
3000
3000
0
BiasVoltage
Voltage [V]
[V]
Bias
500
1000
1000
1500
2000
1500
2000
Voltage
Voltage
2500
2500
3000
3000
Figure4a.
4a. Original
Original FNAL
FNAL 1C
IC test
test data
data Figure
Figure 4b.
Figure
4b. Modified
Modified (slotted)
(slotted)FNAL
FNALdata
data
TESTS WITH
WITH BEAM
TESTS
BEAM
non-modified FNAL
FNAL 1C
IC and
and Prototype
Prototype 11 were
AA non-modified
were installed
installed together
together inin the
the BNL
BNL
Linac-to-Booster
(LTB)
line,
approximated
1
foot
from
the
beam
line,
where
Linac-to-Booster (LTB) line, approximated 1 foot from the beam line, where high
high
doserate
ratelosses
losses from
from the
the 200
200 MeV
MeV H-minus
H-minus beam
dose
beam could
could be
be observed.
observed.Two
Twosets
setsof
ofdata
data
runs were
were made.
made. In
In the
the first
first aa high
high sensitivity
sensitivity multi-channel
multi-channel integrator
runs
integrator module
moduleof
ofthe
the
type used
used in
in the
the source
source testing
testing of
of the
the chambers
chambers was
type
was used
used to
to condition
condition the
the signals.
signals.
They were
were modified
modified to
to have
have aa 470
470 Ohm
Ohm input
They
input so
so rise
rise times
times of
of several
several µsec
usec could
could be
be
observed on
on aa LeCroy
LeCroy digital
digital scope.
scope. The
The bias
observed
bias voltage
voltage was
was varied
varied to
to determine
determine the
the
collection efficiency
efficiency curves
curves for
for both
both bias
bias polarities.
polarities. The
collection
The beam
beam current
current waveform
waveform was
was
also
captured.
Figure
5
shows
the
results
of
this
run.
Clearly
the
positive
also captured. Figure 5 shows the results of this run. Clearly the positive bias
bias (non(nonpreferred)FNAL
FNALresponse
response shows
shows low
low efficiency,
efficiency, especially
preferred)
especially at
at lower
lower bias
biasvoltage.
voltage. The
The
data
for
the
negative
bias
shows
large
excursions,
while
summation
of
data for the negative bias shows large excursions, while summation of the
the beam
beam
current data
data does
does not
not show
show more
more than
than aa few
few percent
percent variation
current
variation between
between data
data points,
points,
implying beam motion. The data for Prototype 1 appears fairly flat throughout the
implying beam motion. The data for Prototype 1 appears fairly flat throughout the
entire voltage scan. Calculations indicate that the FNAL chamber should be 77%
entire voltage scan. Calculations indicate that the FNAL chamber should be 77%
342
efficient while
while Prototype
Prototype 111 should
should be
92%
at
efficient
be 92%
92% at
at 333 kV
kV for
for negative
negative bias
bias at
at this
this loss
loss level.
level.
efficient
loss
level.
Dedicated runs
runs are
are planned
planned to
to study
study
the
efficiency
at
various
dose
rate
Dedicated
study the
the efficiency
efficiency at
atvarious
variousdose
doserate
ratelevels.
levels.
levels.
Current
vs Bias
Bias Voltage
Voltage for
for FNAL
FNAL and
and New
Prototype
BNL
Currer
it vs
vs
New Prototype
Prototype BNL
BNL
Current
Bias Voltage
for FNAL
and New
200 MeV
Linac
April
3,
2002
200
MeV Linac
Linac April
April 3,
3, 2002
2002
200
MeV
6.E-05
6.E-05 -,
6.E-05
Signal
Current
[A]
Signal
Current
[A]
5.E-05
5.E-05
J
4.E-05
4.E-05
3.E-05
3.E-05
2.E-05
2.E-05
/^•^x—
^——
/^
-^——
FNAL - Neg Bias on Outer
-•- FNAL
FNAL --Neg
Neg
Bias
on
Outer
Bias
on
Outer
Prototype
- Neg
Bias
1.E-05
1.E-05
0.E+00
0.E+00
Prototype
- Neg
FNAL - Pos
BiasBias
on Outer
-"&— FNAL
FNAL -- Pos
Pos
Biasbias
on Outer
Outer
Bias
on
Prototype
- Pos
•'-¥?"• Prototype
Prototype -- Pos
Pos bias
bias
0
0)
500
500
500
1000
1000
1000
1500
2000
2500
1500
2000
2500
1500
2000
2500
Bias Voltage [V]
Bias
Bias Voltage
Voltage [V]
[V]
3000
3000
3000
3500
3500
3500
4000
4000
4000
Figure 5.
5. Measured
Measured bias
bias curve
curve (collection
(collection
efficiency)
for
200
MeV
Linac
beam.
Figure
(collection efficiency)
efficiency) for
for 200
200 MeV
MeVLinac
Linacbeam.
beam.
Aproximately
1000
W/m
loss.
Aproximately 1000 W/m loss.
A second
second run
run was
was made using
using the prototype
prototype analog front
front end (AFE)
circuitry
A
A
second run
was made
made using the
the prototype analog
analog front end
end (AFE)
(AFE) circuitry
circuitry
designed for SNS999. This circuit has a current amplifier first stage using a Burr-Brown
designed for
for SNS
SNS .. This
This circuit
designed
circuit has
has aa current
current amplifier
amplifier first
first stage
stage using
using aa Burr-Brown
Burr-Brown
OPA627BM opamp with outputs from a unity gain (50 kHz BW) buffer, a “leaky
OPA627BM opamp
OPA627BM
opamp with
with outputs
outputs from
from aa unity
unity gain
gain (50
(50 kHz
kHz BW)
BW) buffer,
buffer, aa “leaky
"leaky
integrator” and a filtered (1 kHz) slow amplifier. Data from the fast output and the
integrator” and
and aa filtered
integrator"
filtered (1
(1 kHz)
kHz) slow
slow amplifier.
amplifier. Data
Data from
from the
the fast
fast output
output and
and the
the
beam current transformer is shown in Figures 6a and 6b.
beam current
current transformer
beam
transformer is
is shown
shown in
in Figures
Figures 6a
6a and
and 6b.
6b.
FNAL Ion Chamber
FNAL
Ion
Chamber
FNAL
Ionthrough
Chamber
BNL 200 MeV
beam
SNS Protoype
BNL
200
MeV
beam
through
SNS
BNL 200 MeV
beam
through
SNS Protoype
Protoype
AFE
circuit
2 kV bias
AFE
circuit
2
kV
bias
AFE circuit 2 kV bias
lufl
1
Time [sec]
Time
[sec]
Voltage
[V][V]
Voltage
Signal
[V][V]
Signal
2
2
1
1
0
\.
.
^._ . . , _
0 j
0
-1
1
^r^"**™
1
£" --1-2
1 -2
-2- 1 1
/^
-3
.2> -3
FNAL Neg Bias
FNAL
Bias
-4
BeamNeg
Current
llnllllllllill
——FNAL Neg Bias
-4
Beam Current
-5
-5
-6
0.00E+00
2.00E-04 4.00E-04 6.00E-04 8.00E-04 1.00E-03 1.20E-03
-6
-6
0.00E+00
2.00E-04 4.00E-04
4.00E-04 Time
6.00E-04
8.00E-04
[sec]
0.00
E+OO 2.00E-04
6.00E-04
8.00E-04 1.00E-03
1.00E-03 1.20E-03
1.20E-03
2
2
1.5
1.5
1
1
0.5
E
0.5
0
Prototype 1 (Nitrogen) withBNL 200 MeV
Prototype
11(Nitrogen)
withBNL
MeV
Prototype
(Nitrogen)
withBNL
200
MeV
Linac Beam
SNS Prototype
AFE200
Circuit
Linac
Prototype
AFE
Linac Beam
Beam SNS
SNS
Prototype
AFECircuit
Circuit
2 kV
bias
22kV
kVbias
bias
I
j
^
0 *
-0.5
.
• t^**«,n ,.»*,*,*,*«
-0.5
-1
* "°
-1
-1.5
-1.5
-1-2
-2
-2.5
lilllliiil Mull
||P||P|||lPl'
-2.5
-3
0.00E+00
-3
2.00E-04
0.00E+00 2.00E-04
0.00 E+OO 2.00E-04
4.00E-04
Prototype 1 Neg Bias
Prototype 1 Neg Bias
Beam Current
Beam Current
Bedl,,Cu,,el,l
~~ Prot°type 1 Ne9 B'as
6.00E-04
8.00E-04
1.00E-03
1.20E-03
4.00E-04 Time
[sec] 8.00E-04
4.00E-04 6.00E-04
6.00E-04
8.00E-04 1.00E-03
1.00E-03 1.20E-03
1.20E-03
Time
Time[sec]
[sec]
Figure 6a. FNAL IC Beam Test.
Figure 6b. Prototype 1 IC beam test
Figure 6a.
6a. FNAL
FNAL 1C
IC Beam
Beam Test.
Figure
IC beam
Figure
Test.
Figure 6b.
6b. Prototype
Prototype 111C
beam test
test
The signal from the FNAL chamber (Figure 6a) can be seen to have a very long tail
The signal
from the
FNAL
chamber
(Figure 6a)
to
aavery
long
The
signal
FNAL
chamber
6a) can
can be
be seen
seen
to have
have
veryThey
long tail
tail
following
thefrom
pulse,thedue
to the
slower (Figure
ions being
collected
after
the beam.
are
following
the
pulse,
due
to
the
slower
ions
being
collected
after
the
beam.
They
following
the during
pulse, the
due pulse,
to the causing
slower ions
being collected
beam.following
They are
are
also arriving
the gradual
rise ratherafter
thanthe
directly
also arriving
arriving during
during the
the pulse,
pulse, causing
the
gradual
rather
than
directly following
also
causing
the apparent
gradual rise
rise
rather
than
the beam current
(also see
Figure
1). The
noise
during
thedirectly
pulse isfollowing
actually
the
beam
current
(also
see
Figure
1).
The
apparent
noise
during
the
the beam current (also see Figure 1). The apparent noise during the pulse
pulse is
is actually
actually
modulation of the beam current to tailor the stacking in the Booster. The signal from
the Prototype ion chamber has a much faster ion collection time, as expected.
CONCLUSIONS
Improvements have been made to the original FNAL detector design which result
in more uniform characteristics of the production units. These changes allow the use
of the conventional bias polarity (center electrode is anode) without conditioning of
the chambers. However, concern over collection efficiency (saturation) resulting in
departure from linearity at high dose rates may preclude its use in SNS. The slow ion
transit times are comparable to the SNS pulse width, requiring unfolding to obtain the
true time behavior.
The new chamber design appears to yield significantly faster ion transit time and
higher collection efficiency at high dose rate, as predicted. However, higher
fabrication costs and the uncertainties of bringing a new design to production need to
be weighed against the benefits to the SNS application. A firm specification of
linearity at high dose rate is needed before the detector choice can be made.
ACKNOWLEDGMENTS
The authors wish to acknowledge the contributions of C. J. Liaw in the mechanical
design, Dave Kipp is assembling the prototype chambers and Paul Ziminski in testing
the chambers with the cesium source. The prototype circuit used in the beam tests was
built and tested by Chaofeng Mi. Anthony Curcio (BNL) was invaluable in the
installation and setup of the beam tests.
REFERENCES
1
R. E. Shafer, et al., The Tevatron Beam Position and Beam Loss Monitoring Systems", Proc. 12th Int'l
Conf on High Energy Accel., FNAL (1983) p609
2
Troy-Ionics, Inc., Kenvil, NJ
3
Witkover, R. L., Michnoff, R. J. and Geller, J. M., "RHIC Beam Loss Monitor System Initial
Operation", Proceedings of the 1999 PAC, NY, 1999, 2247
4
Shafer, R. E., Private communication
5
Boag, J. W. in "Radiation Dosimetry, Vol. II", F. Attix, et al, eds.,Academic Press (1966), p22
6
R. Witkover," Proposal for a Quasi-Parallel Plate, Cylindrical Geometry Ion Chamber for SNS",
November 7, 2001, unpublished
7
Boag, J. W. and Wilson, T., "The saturation curve at high radiation intensity", Brit. J. Phys., 3 222-9,
(1952)andLoc.Cit(5),pl6
8
Boag, J. W., "the Dosimetry of Ionizing Radiation", Vol. 2, Ed. By K. R. Kase, B. E. Bjamgard and F.
H. Attix, Academic Press, (1987), p 191
9
Witkover, R. L. and Gassner, D., "Preliminary Design of the SNS Beam Loss Monitoring System", To
be published in these proceedings.
344