Advancements in the development of a proportional
counter measuring system
Ondrˇej Kola´rˇ
Department of Radio Electronics
Brno University of Technology
Brno, Czech Republic
ORCID:0000-0003-4187-4890
Michal Kubı´cˇek
Department of Radio Electronics
Brno University of Technology
Brno, Czech Republic
ORCID:0000-0001-8247-8013
Abstract—This paper is an interim report of theory and con-
tinuous work on the proportional counter measuring system. The
first part of the article covers general principles and properties
of several types of electronic detectors for measuring the ionizing
radiation fields including the proportional counters, to highlight
their positives and negatives. The following part discusses the
processing of proportional counter output signal. Next, the
conducted measurement of a neutron source is described, and its
results are shown. Based on the gained experience, the simulation
and design of a new preamplifier for the measuring system was
carried out.
Index Terms—ionizing radiation, proportional counters, mea-
surement system, low noise amplifier, transimpedance amplifier
I. INTRODUCTION
Ionizing radiation is utilized in many fields of modern
science and research, but it is also used in power generation or
commercially. It is a set of physical phenomena which can, by
definition, ionize a material or medium that has been exposed
to it. Depending on many factors, it can have potentially
harmful effects on life organisms or inanimate objects. There-
fore, the need for trustworthy and precise characterization of
ionizing radiation fields is crucial.
The ionizing radiation can be described as a stream of
moving particles. The number of particles traveling through
a given area per unit of time is called particle flux, and it
indicates the intensity of a given radiation field. The next
important parameter is the energy carried by every individual
particle since it indicates the amount of ionization that can
be caused by the said particle. Last but not least, the type of
particle is also very important. Based on all of these parameters
combined, we can estimate the radiation behavior, its degree
of danger, and also its source can be identified or studied.
II. DETECTORS
Various detector types are used for the measurement of
ionizing radiation. The choice of type depends on many factors
and will vary for different applications. This article covers only
some of the electronic detectors, but other types exist as well
[1].
Research described in this paper was supported by the Internal Grant
Agency of the Brno University of Technology under project no. FEKT-S-23-
8191.
The most important aspect of every detector is its mea-
surement capabilities. They differ in spectral and particle-type
sensitivity and resolution, the maximum detectable intensity,
or the output signal strength [2].
A. Scintillation counters
Scintillators are a group of materials which emit visible light
when irradiated with ionizing particles. Every incident particle
transfers its energy to the material which then re-emits it in
the form of a visible light impulse. These impulses are very
short and dim, so for their conversion to an electric signal,
photomultiplier tubes (PMTs) are often used [2]. The more
energy the incident particle dissipates, the more photons of
visible light are generated. This results in the output electric
impulse being proportional to the energy of the originating
particle. Thanks to this, scintillation counters are widely used
for ionizing radiation spectrometry [2].
The scintillator could be of various materials [3]. Solid
scintillators often come in the form of cylindrical crystals
with a protective cover. Liquid scintillators must be contained
in specialized sealed containers. Both of them must be made
in such a way as to block all of the light from surrounding
space, as any of it would overwhelm the scintillation photons
and could potentially damage the PMT if turned on. They
must also have an optical window for coupling with the PMT.
This optical coupling has to be well-made to be as transparent
as possible. The PMT serves two purposes. It transforms the
optical signal into an electric signal, which then gets amplified
for further processing in subsequent electronics [1] [2].
B. Gaseous ionization detectors
Gaseous ionization detectors are a group of ionizing radia-
tion detectors. All of them share the same general mechanical
design, which is depicted in Fig. 1. Their electrically conduc-
tive shell is connected to the negative terminal of a voltage
source. The positive terminal is connected to the anode, which
usually consists of a thin wire in the center of the chamber.
The inside of the chamber is filled with a gas or a gas mixture
[2].
When an ionizing particle enters the detector, it ionizes
the gas fill along its trajectory. The more energy the incident
particle dissipates, the more of the gas gets ionized. These ions
198DOI 10.13164/eeict.2023.198
=
+
–
R
V
Cathode
Anode
Gas filling
Output
Fig. 1. Diagram of a simple cylindrical gaseous ionization detector [2]
start to move toward the corresponding electrodes as they are
accelerated by the electric field. The acceleration is directly
proportional to the intensity of the electric field, which is
determined by the shape of both electrodes and the voltage
supplied to them. When any ions finally reach the electrodes,
they manifest as a current impulse at the output of the detector
[2].
The optimal physical shape of the detector is chosen with
respect to the intended region of operation, which then de-
termines properties of the output signal. For a given detector
shape, the region of operation depends on the supplied voltage
as seen in Fig. 2. The chart shows the relation between the
supply voltage and the output signal strength for two incident
particles of different energies.
1) Geiger-Mu¨ller tubes: are a well-known group of gaseous
ionization detectors designed to work in the region of opera-
tion of the same name. They are usually of a thinner cylindrical
shape. Due to this, a high electric field intensity can be
achieved without the need of a very high supply voltage. This
strong electric field accelerates the primary free ions, created
by an incident ionizing particle, so much that they hit other
gas particles. This causes secondary ionization and causes
a chain reaction that forms multiple Townsend avalanches
a) b) c) d) e)
E1
E2 > E1
Detector supply voltage
O
ut
pu
t
si
gn
al
am
pl
itu
de
(l
og
ar
ith
m
ic
sc
al
e)
Fig. 2. Regions of operation of gaseous ionization detectors (a: ion recombi-
nation; b: ion saturation; c: proportional; d: limited proportionality; e: Geiger-
Mu¨ller) [2]
throughout most of the detector interior. This means that
the amplitude of the output current pulse is relatively high,
therefore the subsequent processing of the signal is easy and
does not require very sophisticated electronics. However, since
the detector gas fill becomes saturated with the avalanches
by any incident particle regardless of its properties, only the
intensity of ionizing radiation fields can be measured with the
Geiger-Mu¨ller tubes by counting the number of pulses per unit
of time [2].
2) Ionization chambers: are a group of gaseous ionization
detectors that work in the ion saturation region. On the
contrary of the Geiger-Mu¨ller tubes, in this region, the primary
ions are accelerated sufficiently so that neither their recombi-
nation is possible nor they could cause secondary ionization.
This means that their amount is the same as the number of
ions reaching the detector electrodes. Therefore, the resulting
output current pulse is directly proportional to the energy
dissipated by the incident ionizing particle. The downside is
the small amplitude of such signal, so the requirements for
a very high amplification and a low noise properties of the
subsequent processing electronics are inevitable [2].
3) Proportional counters: are a group of detectors working
in the proportional region, which combines the benefits of
both previously mentioned regions. As the plot in Fig. 2
shows, the output signal amplitude remains proportional to the
energy of an incident particle. Additionally, the amplitude also
increases with higher detector supply voltages. This is caused
by the electric field intensity near the anode in the detector
center, which is higher than the minimum intensity needed for
Townsend avalanche formation. The avalanche region is kept
only within a small radius around the anode, as seen in Fig. 3,
to ensure the formation of small nonoverlapping avalanches
from primary electrons. This process multiplicates the number
of secondary ion pairs and is called gas amplification [2]. As
the final number of ions is much larger, the output signal am-
plitude is higher, and so is the signal-to-noise ratio. Therefore,
the low-noise requirements of processing electronics are less
demanding than those of ionization chambers.
III. PROPORTIONAL COUNTER SIGNAL PROCESSING
An output signal of a proportional counter comes directly in
the form of an electric current pulse. Although its amplitude is
higher, relative to an ionization chamber, it is still too weak for
any direct processing. Therefore, it has to be amplified first.
The specialized proportional counter preamplifiers often serve
multiple purposes. In addition to amplification, they inject
the supply voltage into the detector. They also transform the
current signal to a voltage signal.
A. Particle energy measurement
Commercial preamplifiers for ionizing radiation spectrom-
etry often come as charge sensitive amplifiers, which are
basically integrating transimpedance amplifiers. Therefore,
their amplification factor (gain) is given in units of mV/pC
[4]. The integration of the signal comes with two benefits.
The first is noise filtering, since the integrator behaves like
199
- +
Cathode
Anode Incident
particleAvalanche
region
Avalanche
Fig. 3. Example of a single ion pair interaction in a cylindrical proportional
counter [2] [5]
∝ E
Time
C
ur
re
nt
w
av
ef
or
m
/
Ti
m
e
in
te
gr
al
i(t)
k
∫
i(t) dt
Fig. 4. Example of an artificial noisy current pulse and its numerical
integration
a low-pass filter. The second benefit comes with the fact
that the current i(t) through the detector terminals in any
given moment is proportional to the rate of ions reaching the
electrodes. By integrating this current pulse, it results in the
total amount of charge generated by the incident particle which
is proportional to the energy E dissipated by it. This simplifies
further processing of the signal because the particle energy,
which is of interest when using the proportional counter for
spectroscopy, can be determined by measuring the height of
the voltage step. To show both of these properties of charge
sensitive preamplifiers, an example noisy pulse was generated
and its integral (scaled by an appropriate constant k) computed
using MATLAB and plotted to Fig. 4.
B. Pulse shape discrimination
When measuring ionizing radiation field, it can be composed
of multiple types of particles simultaneously. One example of
Custom data
acquisition unit
PC Ethernet
230 V AC
High-voltage
power supply Preamplifier
Proportional
counter
Amp. power Signal
Fig. 5. Diagram of the measuring system
these mixed fields can be neutron radiation, which is often
accompanied by gamma radiation, as the nuclear reactions
usually produce both of them. For a proper characterization
of such a field, the ability to separate them is important.
Pulse shape discrimination are various methods of signal
analysis for classification of an incident particle type. For
the scintillation counters, described in II-A, several of these
methods are established and commonly used [6] [7]. How-
ever, the usual combination of charge sensitive preamplifiers
and shaping amplifiers used with the proportional counters
are deteriorating the original pulse shape. This makes them
unusable for subsequent pulse shape analysis. The use of a
non-integrating transimpedance amplifier is more suitable for
the application of a pulse shape discrimination method [8].
IV. MEASUREMENT
The custom developed measuring system used for data
acquisition from proportional counters consists of several parts
shown in Fig. 5. Its detailed description is covered in [9].
Using the system, the first real measurement of a radi-
ation field was conducted at the Department of Electrical
Power Engineering, FEEC, BUT. As a source of radiation the
Americium-beryllium neutron source AS010/15, manufactured
by Eckert & Ziegler Cesio s.r.o. was used. The commercial
charge sensitive preamplifier used was the CSP10 manufac-
tured by FAST ComTec GmbH. Its gain is 1.4 V·pC−1. The
spherical proportional counter used for the measurement was
manufactured by LND, Inc. with the serial number 270133
and it was supplied with 2500 V.
The detector was placed in the shielded room on top of a
fixture in which the neutron source is securely kept. However,
the number of detected events was too small to acquire a
reasonable amount of data. To increase the number of detected
particles, the neutron source was taken out of the fixture
and placed at a distance of 20 cm from the detector. This
drastically increased the number of detected events. A photo
of the detector and source placement is in Fig. 6.
200
Fig. 6. Placement of the detector (left) and the neutron source (right) during
the measurement
The measurement was run for 10 minutes and 5 seconds,
during which the number of captured events was 15 019.
The average detection rate was approximately 24.8 events per
second. The saved binary data was then loaded into MATLAB
for processing.
Since the system saves the signal in raw form, a simple
moving average filter was applied to filter out the noise first.
The difference between the raw and the filtered signal can
be seen in Fig. 7. The use of the integrating preamplifier
facilitates easy computation of a measured energy spectrum
by calculating the height of every impulse as described in
Section III-A and plotting the values to a histogram (Fig. 8).
V. PREAMPLIFIER DESIGN
Following the measurement, it was decided that a custom
non-integrating amplifier for the measuring system will be
designed and built. The reason for this is that the manufacturer
prevents any modifications to the commercial CSP10, as the
printed circuit board of the amplifier module is obfuscated by
encasement in a black epoxy resin.
0 2 4 6 8 10 12 14 16 18 20
−10
0
10
20
30
40
50
Time [µs]
Si
gn
al
[a
rb
.u
.]
Fig. 7. Raw waveform of a captured event (blue) and its filtered form (red)
0 20 40 60 80 100 120 140
0
200
400
600
800
1 000
Event energy [arb. u.]
In
te
ns
ity
[a
rb
.u
.]
Fig. 8. The measured energy spectrum
As a core for the preamplifier, the AD8099 integrated circuit
was chosen. It features an ultra-low noise operational amplifier
with high-speed capabilities given by its gain-bandwidth prod-
uct of 3.8 GHz [10]. The manufacturer Analog Devices, Inc.
provides useful designer tools [11], including the Photodiode
Circuit Design Wizard and SPICE model of the AD8099.
A. First stage design
The Photodiode Circuit Design Wizard is an online tool for
a simple and quick design of a photodiode transimpedance
amplifier. The photodiode is simulated as a square waveform
current source with its capacitance and shunt resistance in
parallel. The values of all parameters can be set according
to the required design. To model a proportional counter, the
peak current was set at 100 µA and the capacitance to 50 pF.
The shunt resistance was set to 100MΩ as this is a value of
a resistor cascade through which the proportional counter is
supplied with high voltage.
In the second tab of the tool, the AD8099 was selected.
Various values of the required output signal parameters were
then tried, and their influence on the estimated circuit char-
acteristics was observed. The reasonable balance was found
by setting the required bandwidth of 100 MHz, peak output
voltage 100 mV, and peaking capacitor value 2.4 pF. The tool
then estimated the output signal-to-noise ratio of 37.1 dB.
B. Simulation
With the SPICE model of the AD8099 available, the circuit
was transferred to the LTspice simulator. For better utilization
of the maximum ±1V range of the ADC in the acquisition
unit, a second inverting amplifier stage with the voltage gain
of 10 was added to the design. To ensure the modularity of the
measuring system, both stages will be constructed separately,
each of them in its own casing. This requires proper impedance
matching to 50Ω coaxial cables used for interconnection.
The final simulation schematic is shown in Fig. 9. Running
the noise analysis of the circuit in the range of 1 Hz to
1 GHz revealed its frequency characteristics, which are plotted
in Fig. 10. The resulting bandwidth ranges approximately from
201
V1
5V
V2
5V
C2
2p4
I1
PULSE(0 100u 100n 100n 1u 100n 10u)
C1
1n
C3
10µ
C4
3p
T1
Td=15n Z0=50
+in
-in
VCC
VSS
out
FB
PD
CC
U1
AD8099
+in
-in
VCC
VSS
out
FB
PD
CC
U2
AD8099
R6
510R
R7
51R R8
51R
R5
51R
R4
51R
R3
51R
R2
1kR1
100Meg
C5
50p
Out
Out2
Vcc
Vss VccVssVcc
Vss
.noise V(Out2) I1 dec 100 1 1000MEG
 ---  C:\Users\xkolar74\Documents\LTspiceXVII\projects\ProPreAmpV1.asc  --- 
Fig. 9. The designed preamplifier simulation schematic
100 102 104 106 108
0
500
1 000
1 500
2 000
Frequency [Hz]
G
ai
n
[V
/A
]
0
20
40
60
80
N
oi
se
sp
ec
tr
al
de
ns
ity
[
n
V
√
H
z
]
Fig. 10. The simulated frequency characteristics - gain (blue), noise spectral
density (red)
246 Hz to 96.8 MHz with a gain of 1.59V/mA. The total
output RMS noise is 653.4 µV.
C. Physical design
The physical design of the preamplifier followed. To lower
the cost of manufacturing, only a single universal version
of a 4-layer printed circuit board (PCB) design was created
using KiCad software. The board contains only one amplifier
stage with all of the passive component footprints needed for
populating either of the stages.
The power for the AD8099 amplifier is provided by the
measuring unit which outputs a symmetric voltage of ±15V
from its internal switching power supply. Since the voltage is
quite noisy and too high for the amplifier, it is filtered by an
LC filter and lowered to the ±5V by a dual symmetric low
noise LDO regulator LT3032-5.
To ensure the low noise properties, the PCB was designed
to fit inside an extruded aluminum enclosure, and on top of
that, the AD8099 circuitry is placed under an additional SMT
EMC shield to ensure suppression of any noise which could
be brought inside of the amplifier enclosure with the power
supply voltage.
VI. CONCLUSION
Taking into account the theory and experience gained from
the conducted measurement, a custom modular preamplifier
for the measuring system was designed. The design was
verified using the LTspice simulator, which showed promis-
ing results. However, the simulation did not account for all
real-world inaccuracies, such as parasitic parameters of the
components used or especially the parasitic capacitance of the
PCB, therefore the results might not be accurate and will need
to be verified.
So, as a next st p, a physical prototype of the preamplifier
is about to be built, tested, and fine-tuned if necessary. At
the time of writing this paper, the prototype is already in the
process of production. The experimental results are therefore
expected to be known soon.
REFERENCES
[1] V. Ullmann, “Detekce a spektrometrie ionizujı´cı´ho za´rˇenı´“, AstroNukl-
Fyzika.cz.
[2] G. F. Knoll, “Radiation Detection and Measurement,“ 3rd ed. Michigan:
John Wiley & Sons, 2000.
[3] B. D. Milbrath, A. J. Peurrung, M. Bliss, and W. J. Weber, “Radiation
detector materials: An overview“, Journal of Materials Research, vol.
23, no. 10, pp. 2561-2581, 2008.
[4] FAST ComTec. (2023). CSP10 Charge Sensitive Preamplifier [Online].
Available: https://www.fastcomtec.com/amp/csp10-charge-sens-preamp
[5] T. Neep, I. Katsioulas, P. Knights, J. Matthews, K. Nikolopoulos, and R.
J. Ward, “A simulation framework for Spherical Proportional Counters“,
in Proceedings of 40th International Conference on High Energy physics
— PoS(ICHEP2020), 2021, p. 917.
[6] G. Liu, M. D. Aspinall, X. Ma, and M. J. Joyce, “An investigation
of the digital discrimination of neutrons and gamma rays with organic
scintillation detectors using an artificial neural network“, Nucl. Instrum.
Methods Phys. Res. A: Accel. Spectrom. Detect. Assoc. Equip., vol. 607,
no. 3, pp. 620-628, 2009.
[7] I. A. Pawełczak, S. A. Ouedraogo, A. M. Glenn, R. E. Wurtz, and
L. F. Nakae, “Studies of neutron– pulse shape discrimination in EJ-
309 liquid scintillator using charge integration method“, Nucl. Instrum.
Methods Phys. Res. A: Accel. Spectrom. Detect. Assoc. Equip., vol. 711,
pp. 21-26, 2013.
[8] T. J. Langford, C. D. Bass, E. J. Beise, H. Breuer, D. K. Erwin, C.
R. Heimbach, and J. S. Nico, “Event identification in 3He proportional
counters using risetime discrimination“, Nucl. Instrum. Methods Phys.
Res. A: Accel. Spectrom. Detect. Assoc. Equip., vol. 717, pp. 51-57,
2013.
[9] O. Kola´rˇ, “Syste´m pro meˇrˇenı´ s proporciona´lnı´mi detektory“, Diplomova´
pra´ce, Brno, 2022.
[10] Analog Devices, “Ultralow Distortion, High Speed, 0.95 nV/Hz Voltage
Noise Op Amp“, AD8099 datasheet, 2016 [Rev. E].
[11] Analog Devices. (2023). AD8099 Tools & Simulations [Online]. Avail-
able: https://www.analog.com/en/products/ad8099.html#product-tools
202