PFC/JA-82-19
MULTI-WIRE PROPORTIONAL COUNTER
FOR SOFT X-RAY DETECTION
J. K~ilne, E. K~ilne, L. G. Atencio,
C. L. Morris, A. C. Thompson
Plasma Fusion Center
Massachusetts Institute of Technology
Cambridge, MA
02139
September 1982
This work was supported by the U.S. Department of Energy Contract
No. DE-AC02-78ET51013. Reproduction, translation, publication, use
and disposal, in whole or in part by or for the United States government is permitted.
By acceptance of this article, the publisher and/or recipient acknowledges the U.S. Government's right to retain a non-exclusive,
royalty-free license in and to any copyright covering this paper.
Multi-Wire Proportional Counter for Soft X-Ray Detection
J. Kallne and E. Killne
Smithsonian Astrophysical Observatory
Cambridge, MA
L.G.
02138, USA
Atencio and C.L.
Morris
Los Alamos National Laboratory,
Los Alamos, New Mexico
A.C.
87545, USA
Thompson
Lawrence Berkeley Laboratory
University of California
Berkeley, California
The design and
counter
performance
of
94720, USA
a
multi-wire
for 2-6 keV X-rays are presented.
use of this detector in
application
of
tokamak at MIT.
a
Bragg
crystal
proportional
We also report on the
spectrometer
in
the
X-ray diagnostics of the plasma of the Alcator C
Page 2
1.
Introduction
In this report we describe the design, construction, and use
of a multi-wire proportional counter for detection of soft X-rays
in the energy range 2 to 6 keV.
detect
photons
The purpose of the device is
in a Bragg crystal spectrometer.
the spectrometer are wavelength dispersed along
while
The photons in
the
x-direction
the image of the source is in the y-direction;
the source
in this case is the hot plasma of the Alcator C tokamak
In
the
application
we
require
good
spatial
to
at
MIT.
x-resolution of
100 pm, moderate y-resolution of the order of millimeters and a
count
rate
capability _of
better
than
100
KHz.
This can be
achieved with the detector described here.
The detector constructed is
a
fast
line proportional counter
.
It has a cathode plane which
delay
is a printed circuit board delay line.
spaced
wires
gives
pulse,
low
The anode plane
impedance
of
2 mm
y-information using an external delay line.
The aim is to build a 300 x 100 mm
2
(x x y) detector to cover the
whole dispersion plane of the X-ray spectrometer.
we report on the' design and use of a prototype
2
dimensions 35 x 35 mm.
Here, however,
detector
of
the
Page 3
2.
Design
entrance window (35 x 9 mm2) is screwed to the box and
the
with
sealed with an 0-ring
relieved
appropriately
was
foil
The area of the Be
Al plate.
0.076 mm
The entrance window is
(Fig. 1).
(51 x 51 mm2 ) which is glued with epoxy to the
beryllium
thick
plate
The
box.
aluminium
The detector is enclosed in an
into the aluminimum plate to provide a flat inside surface.
This
planes
(see
surface served
The
Fig. 1).
detector
the
of
one
as
cathode plane is a helical delay line made
other
used
fabrication convenience, we
the
where
conductor
1 mm
ribbons
conductor
wide
ribbons
two
the sake of
For
apart.
circuit
parallel
give
to
The boards are etched
from two printed circuit boards.
0.2 mm
cathode
boards
are soldered at the two ends
(see
Fig. 1).
To avoid charge accumulation on the circuit boards, the
surfaces
were coated with a thin layer graphite deposited in the
form of micro-graphite in aqueous base (aqua dag).
of
circuit
the
anode plane at a
consisted
of
(G-10).
distance
20 pm
spacing of 2 mm.
board
boards
and
of
diameter
this
one
the Al-Be cathode surface, was the
1.6 mm
from
each
cathode.
It
gold plated tungsten wires with a
The wires were suspended on a
in
Between
prototype
potential spatial information from the
frame
of
epoxy
detector we disregarded the
anode
plane
and
simply
shorted all anode wires.
The delay line was designed to have a low impedance of 100Q
which
was chosen for the sake of time resolution.
unit length of the delay line was 0.66 ns/mm.
The delay per
Page 4
3.
Operation
X-rays
The detector was tested with 5.9 keV
from
a
5 5 Fe
The.25mm line source was collimated to a width of about
source.
Gas mixtures
50 pm (FWHM).
of both argon-ethane and xenon-ethane were tried with pressure
2
This pressure range was safely
and 2.25 kg/cm .
between 0.85
point
the
below
consequently,
we
found
a
in
up
range
and,
no change in the electronic performance
The applied high
due to cathode plane deformation.
varied
bulging
window
noticeable
any
of
point of breakdown for the gas
the
to
was
voltage
conditions tested.
The
positive
cathode
and
inverted
were
signals
then
amplified a factor of 100 using Hewlett-Packard HP8947A amplifier
was
noise
level
Unfortunately, these modules
6 mV measured into 500.
about
amplifier
The
followed by a LeCroy 612 amplifier.
presented an impedance mismatch to the detector delay line
(50Q
compared to 100Q) which resulted in some loss in signal amplitude
and signal reflections.
of
Because of the fast time characteristics
the signals (the rise time was < 3 ns and the width was about
8 ns, FWHM) the reflections
not
did
the
affect
leading
edge
signal shape;
the detector was connected to the amplifiers using
2 m long 100Q
cables.
constant
fraction
provided the
start
The amplified signals were fed
discriminators
and
stop
(EG&G
pulses
for
into
ORTEC Model 934) which
a
time-to-amplitude
converter .(Tennelel model TC 861) which in turn was read by a
multi-channel analyzer (see Fig. 2).
two
Page 5
4.
Performance
The detector performance was investigated with the
X-ray
source.
the
notice
that
above 2.4 kV
observed
conditions.
In
Fig. 3
resolution as function of applied high voltage
Using argon-ethane gas
the range 2.3 to 2.7 kV.
we
keV
Of particular interest was the spatial resolution
achieveable for different operating
shown
5.9
the timing resolution
while
there
resolution
is
a
(V) in
2.5 kg/cm 2
at
(r) changes only slightly
significant
below 2.4 kV.
is
decrease
The signal amplitude
in
the
(S) has
then decreased to 60 mV so that the noise at the level of 6 mV is
no
longer
neglible
at
V
2.3 kV;
we note that the rate of
change in resolution (Ar/AV) approaches that of
variation
region
in
can
S(V).
therefore
noise-to-signal
exponential
The
decrease
in
time resolution in this
be
ascribed
to
the
ratio.
increase
in
timing
peak
broading
should
be
(FWHM) of the observed 8 channels shown in Fig. 3;
peak width of 7 channels corresponds to a broadening in the
of
the
At a ratio of 1:10 and a pulse rise-time
of 3 ns, we estimate that the
7 channels
the
0.25 ns
which
resolution.
electronic
is
equivalent
to
0.18 mm
Above 2.5 kV, or pulse amplitudes
noise
is
no
a
line
spatial detector
of
>200 mV,
the
longer what limits the accuracy in the
position determination.
In the amplitude range >200 mV, one limiting factor for
spatial
resolution
photoelectron
significance
from
is
the
the
finite
stopping
photo-absorption
range
of
2)
process2.
the
the
The
of this range effect is illustrated in Fig. 4 where
the observed resolution for Ar and Xe gas mixtures at
1.1,
1.6,
Page 6
and
2.3 kg/cm2
pressures
are shown.
Since the excited states
created by the photo-absorption process
electron
emission
mostly
decay
by
Auger
and the energies of these electrons are lower
than the energy of the photo-electron 2),
we may assume
that
the
photo-electron energy characterizes the range differences between
Ar and Xe;
for
Ar
fit
the
i.e.,
the range 2)()is about a factor of
than for Xe
data
[(R/P)2 + C 2
on
(RAr
the
11/ 2
=
1.5 RXe).
time
which
by
the
P
is
the
function
expresses the resolution as the
sum of two terms in quadrature namely C (fixed) and
dependent);
greater
We use this assumption to
resolution
simply
1.5
pressure
expressed
in
R/P
(density
kg/cm .
This
determines C to be in the range 2.4-2.9 channels or 60
to 75 Pm
2 -1
for R-~ 11.0 in units of [kg/cm ] .
A significant portion of the
value of C stems from the finite source
width
of
about
50 Pm.
Accounting for the source contributions, we are still left with a
residue of some 30
to
50 Pm
which
could
diffusion
electrons
while
drifting
of
the
be
due
to
thermal
from the point of
initial ionization to the anode wire or other subtle effects.
make
no
attempt to study these effects here, partly because the
contributions from the source width and
resolution
below
100 vm.
It
spectrum (Fig. 5) recorded after
optimizing
We
may
alignment
suffice
careful
to
source
the resolution using Xe at 2.25 kg/cm 2.
resolution was 80 vm (FWHM).
dominate
show
the
the time
alignment
for
The observed
Page 7
5.
Applications
The detector was used as the position sensitive element in a
Bragg
crystal
region
spectrometer
2.4-2.7 keV.
detector
operates
In
in
+
through
detector.
the
measurements
this
The
ethane(40%)
is
for
soft
X-rays
spectrometer
in the energy
application
the
vacuum with a gas pressure of 1 atm and a
high voltage of 2.0 kV.
krypton(60%)
3)
detector
which
The
gas
was
was
a
allowed
purpose
of
mixture
of
slowly to flow
the
spectrometer
to analyze the characteristic X-ray emission of
impurity ions in the hot plasma of the Alcator C tokamak at MIT.
A typical spectrum obtained with
shown
in
Fig. 6.
present
n=2
and
n=l
electron
made
a
line
fit
to
prominent lines (see Fig. 6).
resolution,
two
of
electrons.
(FWHM) for the four
We
most
We then determine the average line
width to be about 11.8 channels or 0.73 mm
detector
orbits
the observed spectrum and find line
widths of between 9.9 and 13.8 channels
the
is
At the high plasma temperatures of about 1.2 keV,
these ions are stripped of all but the last
have
detector
The four prominent peaks are each identified
with transitions between the
chlorine ions.
the
(FWHM).
Apart
from
there are three major contributors to
the line broadening of these X-ray lines;
namely,
the
entrance
slit of the spectrometer (0.4 mm), the Bragg diffraction width of
the pentaerythritol (PET) crystal (AX = 0.89 mA, from the
specification resolution for the 002-planes of
AX/X = 1/5000 at
X - 4.444
£hA.
crystal of
.A) and the
Doppler broadening
( AX = 1.9 mA ) due to the temperature of
radiating
( estimated
ions
conversions are 1 mA =
0.27 mm
to
=
Tion z1.2
keV ).
4.49 channels.
The
These
the
unit
three
Page 8
components,
when
added
observed line width.
with
a
in
quadrature, account for most of the
In particular,
the
result
is
consistent
neglible (<0.2 mm) contribution from the detector.
this detector, we can thus approach the resolution limit
spectrometer
decreased
which
with
is
the
set
by
practical
the
crystal;
lower
limit
design
and
With
of
the
the slit can be
set
by
intensity
requirements.
6.
Conclusion
We have
multi-wire
reported
proportional
the region 2-6 keV.
demonstrated
that
on
the
detector
in
of
a
counter for detection of soft X-rays in
In tests with a 5.9 keV X-ray source we have
(FWHM)
a spatial resolution of 80 Jim can be achieved
successfully with this design.
this
construction
We have reported on
the
use
of
a spectrometer application for analyzing soft
X-rays in the energy region
2.4-2.7 keV.
At
the
level
of
a
measured line width of 0.7 mm, the detector contribution is found
to be neglible.
Acknowledgements
This work was supported by the U.S.
Department of Energy.
Page 9
References
1.
L.G. Atencio, J.F.
Nucl.
2.
F.
Inst.
Amann, R.L.
and Meth., 187
Boudrie, and
C.L.
Morris,
(1981) 381.
Sauli, "Principles of Operation of Multiwire Proportional
and Drift Chambers", CERN NP Internal Report 77-09 (1977).
3.
L.
van Hamos, Ann. der Physik 17
(1933) 716.
Page 10
Figure Captions
Fig. 1:
(a) Photograph of the detector
Visible
are
with
one
side
opened.
the anode wires and the cathode of printed
circuit boards.
(b) Schematics of detector.
Fig. 2:
Schematics of the electronic set up.
Fig. 3:
Time resolution and
applied
high
source with a
signal
amplitude
as
function
voltage using 5.9 keV X-rays of
detector
gas
mixture
of
an
of
55Fe
+
argon(60%)
ethane(40%) at a pressure of 2.25 kg/cm2
Fig. 4:
Time resolution as a function of the inverted
the
pressure
in
the
detector
for
gas
argon(60%) + ethane(40%) and xenon(60%)
+
value
of
mixtures
of
ethane(40%).
The curves present fits to the data for the parameters C
= 2.4 and 'X= 11.0
Fig. 5:
(see text).
Time spectrum for four source positions using a detector
gas
mixture of xenon(60%) + ethane(40%) at 2.25 kg/cm 2
pressure and a high voltage of 2850 V.
Fig. 6:
Example
of
crystal
spectrometer
X-ray
spectrum
obtained
and
a
the
Bragg
using the present detector with a
krypton(60%) + ethane(40%) gas mixture
pressure
with
at
high voltage of 2.0 kV.
1.0
kg/cm 2
The lines are
Page 11
due to characteristic n=2 to n=l transitions of
Cl
of
which
the
prominant
lines
(w,x,y, and z) are
identified with the transitions from the excited
ls2p
is2
1
Pi,
IS0
ls2p
3
P 2 ,-
ls2p
ground state with
3
P1 ,
and
He-like
Is2s
3
S
states
to
the
X'= 4.444, 4.465, 4.468, and
0
4.497 A.
lines.
The fitted line widths are indicated for these
The spectrum refers to the sum of
discharges of 100 ms duration each.
three
plasma
yY
4PF
9A
-)-o
0>
C
0~
*.
e 4-
0)
S
~0
C
00000
C
C
t
C
m0
+HV
F-mF-
inv
inv
Amp
Amp
Delc y
(25 is)
Disc
Disc
Start
Stop
TAC
MCA
I
I
I
I
i
I
50 F
500
/
"5
E
/
~0
/
U)
-r
U
(9
/
0
/
10k
-4100
/
0
LUi
D
a_
/
0
/
5 -
0
50
.
V)
Ld
.
--I J
i
LiU
I
2.4
*
I
2.6
VOLTAGE (k V)
I
0
2.8
I
I
I
I
8
LI)
z 61
z
z
Q
L-
4
Of)
O
Ar
A
Xe
2
i
0.5
I
1.0
I
i
1/p
(kg/cm
2
-12.0
COUNTS /25 pLm bin
0
0
a
0
0
0
CA
0
0
0
0
-N
0
0
I
I
I
I
I
I
I I
I.
tom
r) 3
x
3
1-
I-------------------- j-
I
I
-
150 -
w
x
y
z
100-
L-J
z
z
-z
+
11.6
0
00
*.-
*.
.
200
-
-11-*
.
0
300
POSITION (CHANNEL NO.)
9.
.
.
400