for Microdosimetric Measurements
I. Almási (a), E. Anachkova (b,d),
T. Bartha (a),
Dr. Katalin Erdélyi (a), A.M. Kellerer (b,c)
and H. Roos (b)
(a) MicroVacuum Ltd., H-1147
Kerékgyártó u.10. Budapest, Hungary
Institut der Universitä t München, Schillerstrasse 42, D-80336 München, Germany
(c) Institute für
Strahlenbiologie, GSF Forschungszentrum für Umwelt und Gesundheit, Postfach 1129, D-85758
(d) Institut für
Strahlenhygiene, Bundesamt für Strahlenschutz, Postfach 1108, D-85758
Various measurement systems based on tissue equivalent
proportional counters (TEPC) are used in radiation protection practice ( 1 , 2 ), two of these being commercially available ( 3 , 4 ). Most of the systems measure pulses from single events,
proportional to the energy imparted in the counter gas, convert the pulse height
distribution to a distribution of dose in lineal energy, and evaluate dose rate, or dose
equivalent rate by different approximations of the dependence of the quality factor on
An alternative technique for determination of the
microdosimetric parameters is the variance method. In this case the electric charge
proportional to the energy imparted over a specified time interval (including multiple
events) is measured by an electrometer connected to the TEPC. The fluctuations of the
energy imparted in the counter are used to determine the dose average lineal energy or the
dose average specific energy.
While for the single event measurements the conventional
and widely developed pulse height technique is applicable, the variance method requires
high precision current measurements. For the variance-covariance method5 simultaneous measurements in
two detectors and two independent channels of signal processing are needed. They must
contain low noise electronics with high resolution to allow the precise determination of
the fluctuations of the energy deposition. Furthermore, in the measurements at high dose
rate high sampling frequencies are desirable. These requirements are not readily met and -
for lack of a generally applicable multi-purpose design - considerable work is usually
invested in the signal processing electronics, when the variance or the
variance-covariance method is applied. In spite of the efforts that are involved, the
resulting instrumentations tend to be bulky and to lack versatility, a notable exception
being a portable system for performing variance and variance-covariance measurements (the Sievert
instrument) that has recently been designed by Lindborg et al6 and was applied for measurements of the
ambient dose equivalent and the average quality factor on board of aircrafts.
In view of the potential of the variance-covariance
technique it was felt desirable to create a suitable multi-purpose device for
microdosimetric measurements in terms of the variance-covariance method. The resulting
design that is here reported is portable and fully battery operated. It provides simple
communication with computers via standard RS 232 interface, and it equally provides the
possibility to connect different TEPCs or ionisation chambers. The large range of sampling
frequencies permits broad applicability, including measurements at very high dose rate,
pulsed fields of different intensities, or fields with low dose rate. The evaluation
program can be readily modified and adapted for special applications.
The operational amplifier AD549 is applied as low-noise
switched electrometer with guarded input lines. The offset voltage of each channel can be
adjusted by a potentiometer. Each electrometer converts the input current to an output
voltage by integration over a high precision (270 pF ± 1%) integration capacitor. Two
fully symmetrical measuring electrometer units are built into a separately guarded box.
Each electrometer output is connected directly to an
analog-to-digital converter DDC101 (Burr Brown). The DDC101 units are programmed in CDS
(Correlated Double Sampling Mode) which allows to compensate for internal errors related
to factors such as steady state, charge injections, and thermal noise. The DDC101s operate
in integrating unipolar mode with 20-bit resolution: the sampling time is programmed by
the control unit. The resolution of the voltage measurements is 5.96 µV.
The controlling unit includes 80C537 Microcontroller, 32
kbyte EPROM, 32 kbyte RAM, 8 kbyte EEPROM, an 8 character LCD display and four control
buttons. The evaluation algorithm is put on the EPROM. The operating parameters (sampling
frequency, high voltage, number of samples) and the detector parameters (gas
multiplication factor for each detector, mean chord length, and air-kerma calibration
factor for photons) are put on the EEPROM. The configuration can be readily read or can be
promptly changed via the RS 232 interface. The electrometer readings are stored on the RAM
(maximum 2000 in each channel) and can be read out via the RS 232 port. The operating
parameters need to be set through the control buttons before each measurement.
The following operation parameters can be chosen through
the control unit :
- sampling frequency: 2, 10, 100, 1000. The additional
frequencies 23, 57, 230, and 568 Hz are implemented in a second variant of the instrument.
- high voltage: from 200 V to 1300 V in steps of 50 V
- number of sampling intervals: 500, 1000, and 2000.
The second variant provides the possibility of repeated
measurement, the number of cycles (1 to 63) being chosen from the control panel. This
allows better statistics in pulsed radiation fields.
The quality factor, Q, is approximated by the relation ( 6 )
with a = 0.8 and b = 0.17 ?m/keV.
This relation holds below 150 keV/µm, beyond this value it
provides an overestimate of Q on the ‘safe’ side.
At the end of each measurement the following information
can be read off the display:
- dose average lineal energy, yD (in
keV/µm), obtained as the mean from the two detectors
- dose equivalent rate (in µSv/h)
- the mean value of the signal in each channel (in relative
- the relative variance of the signals in each channel
- the relative covariance of the signals from the two
- the number of sampling pairs actually used in the
The last four readings facilitate the optimisation of the
operation parameters. There is also a possibility to check the electrometer readings for
each sampling interval.
The instrument has been tested with cylindrical tissue
equivalent proportional twin-counter7 and a 137Cs-source. Although the geometrical shape and dimensions of the two single
detectors were identical, the gas multiplication was seen to differ, at the highest
possible voltage, up to 30 %; but this difference did remain constant during the
measurements. The gas multiplication was determined with the 37Ar calibration technique ( 8 ) and with a built-in Am ?-source (
7 ). The tests of the new
device were mainly concerned with its function under different operating conditions and
with its range of applicability. The dose average lineal energy, yD, was
determined in measurement series with different numbers of sampling pairs, N, and at
different sampling frequencies between 2 Hz and 4000 Hz. Figure 1 illustrates the measurements at a
dose rate of about 200 µGy/h (simulated dose rate : about 6 Gy/s).
Measured values of yD are represented in Figure 2 as
ratio of the yD determined at specified frequency to the average value yD
(mean) from all measurements shown in the picture.
The new algorithm, proposed recently by Kellerer ( 9 ) for the case of changing dose-rate ratio in two detectors, was also applied.
The results were generally identical, as is to be expected in a constant radiation field.
But even under these simple conditions the new algorithm was found to provide more stable
results in the presence of strong electronic noise. The implementation of the new
algorithm and an increase of accuracy in the numerical calculation will be the next steps
in the development of the device.
1.) L. Lindborg, D. Bartlett, H. Klein,
Th. Schmitz and M. Tichy, Radiat. Prot. Dosim.,1995, 61, 89.
2.) H. G. Menzel, L. Lindborg, Th.
Schmitz, H. Schumacher and A. Waker, Radiat. Prot. Dosim., 1989, 29(1-2), 55.
3.) A. W. Kunz, P. Pihet, E. Arendt and
H. G. Menzel, Nucl. Instrum. Methods, 1990, A299, 696.
4.) A. Aroua, M. Höfert and A. V.
Sannikov, Radiat. Prot. Dosim., 1995. Radiat. Prot. Dosim., 1995, 59, 49.
5.) A. M. Kellerer and H. H. Rossi,
Radiat. Res., 1984, 97, 237.
6.) L. Lindborg, J. E. Grindborg, O.
Gulleberg, U. Nilsson, G. Samuelson and P. Uotila , Radiat. Prot. Dosim., 1995, 61(1-3),
7.) J. Chen, J. Breckow, H. Roos and
A.M. Kellerer, Radiat. Prot. Dosim.,1990, 31, 171.
8.) E. Anachkova, A. M. Kellerer and H.
Roos, Radiat. Environ. Biophys.,1994, 33, 353.
9.) A. M. Kellerer, Radiat. Environ.
Biophys., 1996, 35, 117.