![](bg2.png)
channel MOSFETs (pMOS) (Sarrabayrouse and
Siskos, 1998; Ma and Dressendorfer, 1989; Hughes
et al., 1988).
During readout, the V
T
determination can be
carried out by extracting the complete current-
voltage characteristic curve of the device (Asensio et
al., 2006) or by using simpler methods based on
constant current measurements. Basically, the
measurement methods consist in recording the drain-
source voltage while the transistor is being polarized
with a constant drain current, before and after
irradiation. Under this configuration, the source-
drain voltage shifts equals, approximately, the
increase in V
T
(Sarrabayrouse and Siskos, 1998;
Asensio et al., 2006; Carvajal et al., 2010).
Most commercial dosimetric systems based on
MOSFETs, measure increases in drain-source
voltage at constant drain current
(Thomson&Nielsen, 1991; Halvorsen, 2005).
Usually, in order to minimize thermal drift, the drain
current selected is the Zero Temperature Coefficient,
I
ZTC
, because the drain-source voltage is invariant
when temperature increases. The dosimetric
parameter used by one of the commercial dosimetric
systems is the difference of the increase in drain-
source voltages of two transistors under different
polarizations. With this arrangement, a wide linear
range and thermal compensation are achieved in the
biased mode (Thomson&Nielsen, 1991). There is
another commercial system based on transistors in
the unbiased mode, but the linear range is up to 5 Gy
and its use is limited to only one irradiation session
(Halvorsen 2005; Sicel Technologies, 2005, Best et
al., 2005).
In previous works (Asensio et al. 2006, Carvajal
et al., 2010; Carvajal et al., 2010b), we showed the
feasibility of using a general-purpose low-cost
pMOS as dosimeter, irradiated in unbiased mode,
with a significant increase in linear range compared
with similar systems and reduced thermal drift.
These improvements are based on gate-source
voltage measurements at three different drain
currents during readout instead of the usual method
with a single polarization current.
Although this work has focussed on the pMOS
dosimeter, it is important to point out that the
proposed methodology could be applied to other
sensors based on MOSFET, such as CHEMFET or
ISFET, where the magnitude under study is
extracted from the threshold voltage. In these cases
(Fraden, 1996), the threshold voltage is also
measured at constant drain current.
In the following section, a complete
measurement system description is carried out,
including the measurement algoritmh and an
explanation of the electronic system. After this, the
experimental setup used in our experiments is
described. We report experimental results of the
dosimetric characterization of the dosimeter
presented here and its technical specifications are
shown. Finally, the main conclusions are drawn.
2 DOSIMETER DESCRIPTION
In this section, a detailed description of the
measurement algorithm and the electronic design of
the dosimeter will be carried out.
2.1 Measurement Algorithm
First, the theoretical background of the measurement
process, developed in previous works, will be
resumed (Carvajal et al., 2010; Carvajal et al.,
2010b). This is the basis on which the electronic
design relies on. The operation method of the dose
measuring of the presented dosimeter is the same of
most previous devices based on MOSFET sensors.
Before the sensor irradiation, an electronic reader
unit measures the electrical characteristics of the
sensor; this is usually named as sensor zeroing (first
readout period). Then, sensor is irradiated
(irradiation period) and, finally, another readout
process is carried out to measure the MOSFET
electrical characteristics shifts caused by the
irradiation. Here during the irradiation period, we
have configured the sensor in the unbiased mode, i.e.
short-circuiting all its terminals and without
polarization, obtaining a sensor without wires in this
period.
During the dose readout under constant drain
current bias, the gate and the drain terminals of the
MOSFET are usually short-circuited and grounded,
and the bulk and the source are also inter-connected
(Sarrabayrouse and Siskos, 1998; Best et al., 2005).
In this configuration, the transistor operates in the
saturation region, where
SD
VI − can be modelled
for pMOS (Sze, 1981):
()
2
2
TSD
VV
β
I −=
(1)
where the drain current,
D
I , depends on the source
voltage,
S
V , the threshold voltage,
T
V , and the
parameter,
W
C
oxp
μβ
=
, where
p
is the carrier
mobility in the channel,
ox
C is the oxide
BIODEVICES 2011 - International Conference on Biomedical Electronics and Devices
24