JPH0533356B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Application The present invention relates to a semiconductor radiation detection device, and this semiconductor radiation detection device is used in the fields of science and engineering, such as the field of nuclear medicine.

(b) Prior art In room-temperature radiation detectors using compound semiconductor (including semi-insulating) elements such as CdTe and HgI ₂ , the mobility of holes, μh, is generally significantly higher than the mobility of electrons, μe. Since the charge collection time, ie, the rise time of the output signal of the preamplifier, shows large variations depending on the radiation absorption position. In particular, when measuring gamma rays of approximately 60 KeV or more by entering them from either the anode or cathode side, the thickness of the crystal (distance between electrodes) or the thickness of the depletion layer is 0.5 mm from the viewpoint of detection efficiency. Since there is a limit to the applied voltage, the above-mentioned tendency becomes noticeable. When such a signal is waveform-shaped and amplified by the main amplifier and then analyzed by a multi-channel analyzer (hereinafter abbreviated as MCA), the following two results are obtained.
Due to two reasons, the wave height distribution of the total energy absorption peak becomes asymmetrical with a tail toward the lower energy side (7th
(see Figure a), it is difficult to obtain good energy resolution.

The μh・τh product (τh; average lifetime of holes) is small,
In other words, imperfection in charge collection due to holes being easily captured by the hole trapping center.

Dependency of the output wave height of a waveform shaping circuit on the input signal rise time.

Among these, τh is a property unique to the crystal element, and τh varies greatly depending on the quality of the crystal, that is, the concentration of hole capturing centers. Regarding this, please refer to (1) Eiji Sakai; "Current status of HgI ₂ radiation detectors" Applied Physics, 46(10) 1034 (1977).

is purely a problem of the response characteristics of the electronic circuit,
It depends on the type of waveform shaping circuit (filter) selected and its shaping time constant.

Next, a more detailed explanation will be given using FIGS. 5 to 9. As shown in FIG. 5, the room temperature radiation detector 1 is constructed using a compound semiconductor element 11 such as CdTe or HgI ₂ as a radiation detection element, and is constructed with a cathode 12 and an anode 13. Normally, in the case of high-resistance type CdTe or HgI ₂ , the element 11 is the crystal itself due to its semi-insulating properties, and both electrodes 12 and 13 are coated with red or metal evaporated (in the case of HgI ₂ , the element 11 is the crystal itself).
low-resistance CdTe
In this case, the element 11 is a surface disorder type or a pn junction type. Therefore, in the former case, the entire crystal is a sensitive layer, whereas in the latter case, the sensitive layer is a depletion layer. A negative bias voltage is applied to the cathode 12, and the anode 13 is connected to a preamplifier 2 to obtain a voltage signal Vp. Furthermore, the energy signal Ve is transmitted through the main amplifier (waveform shaping amplifier) 3.
This enables wave height analysis using, for example, MCA.

For convenience of explanation, a case is shown in which the detector 1 is sandwiched between parallel plate electrodes and a signal is extracted from the anode 13 by direct current coupling, but the same applies to other configurations.

Assuming that the thickness of the sensitive layer is 1 and the distance from the cathode 12 to the position where radiation is absorbed is x, for radiation events in a region where x/1 is sufficiently small, the distance traveled by holes is short and the distance traveled by electrons is The signal mainly comes from movement, the charge collection time is short, and the charge collection is complete. However, for events where x/1 is larger,
Since the distance traveled by the holes becomes longer, the charge collection time becomes longer, and the aforementioned cause, that is, the incompleteness of charge collection, becomes more noticeable. This tendency can be expressed as the x/1 dependence of the peak value Vpmax of the output signal of the preamplifier 2, as shown by the solid line in FIG. In addition, as the horizontal axis of FIG. 6, instead of the absorption position information x/1,
It may also be expressed in terms of the charge collection time, that is, the rise time tr of the output signal of the preamplifier 2.

Furthermore, the waveform shaping amplifier 3 has the above-mentioned cause, that is, dependence on the rise time of the input signal.
Normally, if the input signal rise time is increased under the condition that the input signal wave height is constant, the output signal wave height Vemax tends to decrease. In other words, the larger the rise time, the smaller VemaxVpmax becomes.
The x/1 (or tr) dependence of Vemax is shown, for example, by the dashed-dotted line in FIG.

When the wave height of such an energy signal Ve is analyzed by irradiating single-energy γ-rays, the total energy absorption peak shows an asymmetric distribution with a tail toward the lower energy side, as illustrated in FIG. 3a. Figures 3 b to d are
It shows that the wave height distribution differs depending on the radiation absorption position, and the larger x/1 means that the pre-energy absorption peak shifts to the lower energy side.

To explain the waveform shaping amplifier 3, its functions are mainly "amplification", "reduction of circuit noise", "formation of short pulse width to improve count rate characteristics", and "formation of pulses with easy-to-process shapes". There are four ways in which the pulse can be formed.Here, "shape that is easy to process" means that, for example, when performing pulse height analysis, the peak part of the pulse is relatively flat with respect to the time axis. means. There are various types of waveform shaping circuits (filters) used in the amplifier 3, and each has advantages and disadvantages regarding three functions other than "amplification". Circuit noise mainly originates from the semiconductor detector and the input stage of the preamplifier, and includes series noise, parallel noise,
Classified as fritska noise and others. The definition of the shaping time constant (hereinafter abbreviated as time constant) of a waveform shaping circuit differs depending on the type of filter, but in general, the smaller the time constant of any filter, the greater the series noise, and the larger the time constant, the greater the parallel noise. The total circuit noise is minimized when a time constant is selected such that the noise tends to increase and the two are about the same level (flicker noise does not depend on the time constant). On the other hand, the larger the time constant, the longer the dead time and the lower the count rate characteristics, so the time constant is usually selected as a compromise between circuit noise reduction and count rate characteristics.

However, if the variation in charge collection time is large and the above-mentioned cause, i.e., the tr dependence of the fixed wave height of the waveform shaping circuit, cannot be ignored, the time constant (for example, peak arrival time) can be adjusted to minimize the effect of the variation in charge collection time. It is necessary to set a sufficiently long value compared to , which causes the problem that a time constant that satisfies the minimum noise condition cannot necessarily be selected and that the count rate characteristics deteriorate.

Normally, a quasi-Gaussian filter (generally composed of a combination of one stage of differentiating circuit and multiple stages of integrating circuit) is used in the waveform shaping circuit, but this has a relatively simple circuit configuration and provides good circuit noise reduction. This is because it is possible to obtain In other words, as shown in Figure 9,
Differentiator circuit and amplifier 3 as waveform shaping amplifier
1, an integrating circuit and an amplifier 32, and a differentiating circuit and an amplifier 32.
1 usually uses one stage of CR differentiator circuit, and in particular, pole-zero cancellation is commonly used. In the integrating circuit and amplifier 32, an RC integrating circuit or a low-pass active filter is generally used as the integrating circuit, and the integrating circuit may have one stage of integration, or a filter that has several or more stages of integration. Note that the integrating circuit and the amplifier 32 often include a baseline restorer, and the amplifier 30 also often includes a pile up extraction circuit or the like. However, this case has the disadvantage that the dependence of the causes is relatively large. for example,
Vemax/Vpmax for quasi-Gaussian filter
An example of tr dependence is shown by the dashed-dotted lines a, a', a'' in FIG. 8, and the time constant (peak arrival time) increases in the order of a, a', a''.

Further, a single delay line clipping filter has a characteristic that the dependence on the above-mentioned causes is very small when compared under the same time constant conditions, but it has the disadvantage that it inherently has large circuit noise.

On the other hand, for systems where the charge collection time varies, such as a large-capacity coaxial Ge detector (a good case where imperfection in charge collection like the cause described above is hardly recognized), a type of time-varying filter is used. By using a method in which the signal shaped by a quasi-trapezoidal filter, that is, a quasi-Gaussian filter, is further integrated for a certain period of time by a gate integrator, it is possible to almost eliminate the dependence of the above cause with a short time constant, and the result is good. It has been reported that excellent energy resolution and count rate characteristics were obtained. Regarding this, for example, (2) V.Radeka; “Trapezoidal Filtering of
Signal from Large Germanium Detectors
at High Rates.” IEEE Trans.Nucl.Sci., NS−19(1)412
(1982) (3) FSGoulding, et al.; “Signal Processing
for Semiconductor Detectors.” IEEE Trans.Nucl.Sci., NS−29(3)1125
(1982) etc.

For comparison, the information regarding the pseudo trapezoidal filter is shown below.
An example of the tr dependence of Vemax/Vpmax is shown by the solid line in FIG. However, this shows the case where the output of the quasi-Gaussian filter shown in FIG. The dead time at this time was longer than that shown in a of the same figure, and was about the same as that of a'.

There are basically two possible methods for improving the deterioration of energy resolution due to the above-mentioned causes in a room-temperature radiation detector using a compound semiconductor element as described above. That is, to explain using Figure 7, [A] Measurement is performed only for events where x/1 is sufficiently small as shown in Figure 7 b. [B] In Figure 7, b, c, d and x/1 are As the energy increases, the total energy absorption peak shifts to the lower energy side.
Vemax with larger amplification for events with larger x/1
By amplifying (or adding or subtracting from Vemax a correction amount that is a function of both x/1 and Vemax), the total energy absorption peak does not depend much on x/1 with respect to the corrected energy signal wave height. I can think of an improvement method of the order of magnitude of .

Regarding improvement method [A], for example (4) LTJones; “The Use of Cadmium
Telluride γ Spectrometer is Monitoring
Activity Deposited in Nuclear Power
As introduced in "Statios" Rev.Phys.Apl., 12, 379 (1977), the rise time tr of the preamplifier is measured for each event, using the depth information x/1 of the radiation incidence position. , is a method of removing signals with slow rise times tr. However, such method [A] has the drawback of significantly reducing detection efficiency, although energy resolution is significantly improved.

On the other hand, improvement method [B] has the effect of improving energy resolution without substantially reducing detection efficiency. Regarding this, (5) R.Kurz; “A Novel Pulse Processing
System for Hgl ₂ Detectors.” Nucl.Instr.and Meth., 150, 91 (1978) (6) M.Finger, et al.; “Energy Resolution
Enhancement of Mercuric Iodide
Detectors.” IEEE Trans. Nucl. Sci., NS−31(1)348
(1984).

The method of document (5) is similar to document (4), and the rise time tr of the preamplifier is measured for each event as depth information x/1 of the radiation incident position. but,
Since a single delay line clipping filter is used, the circuit noise is large and the circuit is complicated (two delay line clipping filters are used).
Constant Fraction Discriminator and Time−to
- Amplitude Converter is required).

In the method of reference (6), the output of the preamplifier is input to two types of waveform shaping amplifiers: a quasi-Gaussian filter (with a large time constant) and a fast pulse processing circuit, and the output signal S [slow] of the former is Depth information x/1 of the radiation incident position is obtained using the ratio of the latter output signal S[fast], and S[slow] is corrected. In a fast pulse processing circuit, in order to obtain almost only the signal due to the movement of electrons, the pulse processing circuit starts from the point where about 50 nsec has elapsed from the beginning of the rise of the output signal of the preamplifier.
It is configured to integrate over 100nsec.
Therefore, the larger x/1 is, S[fast]/S
The [slow] ratio tends to become smaller. However, it has the disadvantage that the pulse processing circuit has a complicated configuration, and errors tend to become large when measuring relatively low-energy γ-rays (for example, 60 KeV to 140 KeV).

Note that in the methods of literature (5) and literature (6), the time constant of each filter is set to a sufficiently large value so that the influence of the above-mentioned causes is practically negligible.

In the above explanation, for convenience, the separation of holes from the hole capture center has been omitted, but in reality, the problem is a little more complicated due to the effects of this separation. However, in that case, the improvement method [A]
The improvement in [B] is the same, but the strength of the correction and the nonlinearity of the correction with respect to x/1 are slightly different in the case of [B].

(c) Purpose This invention solves the problem of incomplete charge collection (described above) and rise time dependence of the response of a waveform shaping circuit (described above) that occur when a compound semiconductor element is used as a radiation detector for room temperature. It is an object of the present invention to provide a semiconductor radiation detection device that can improve the deterioration of energy resolution caused by the irradiation without reducing the detection efficiency.

(d) Configuration In the semiconductor radiation detection device of the present invention, a compound semiconductor is used as a radiation detector for room temperature, and the output of the preamplifier is shaped into a waveform including a quasi-Gaussian filter consisting of a combination of differentiating means and integrating means. The charge collection time information for each event, that is, the depth information of the radiation injection device is obtained by inputting the signal into the amplifier and using the signal wave height related to the output of the differentiating means of the quasi-Gaussian filter, and according to this, the charge collection time information for each event is obtained. By correcting the output signal wave height of the waveform shaping amplification means, deterioration in energy resolution caused by incomplete charge collection is improved.

(E) Embodiment In FIG. 1, the room temperature radiation detector 1 and preamplifier 2 are the same as those in FIG. 5.
Further, the differentiating circuit and amplifier 131 and the integrating circuit and amplifier 132 in FIG. 1 are similar to the differentiating circuit and amplifier 31 and the integrating circuit and amplifier 32 in FIG. 9, respectively, and include a quasi-Gaussian filter as a whole. An amplifier is configured. The output Vd of the differentiating circuit and amplifier 131 is input to the delay circuit and amplifier 100 and the trigger generation circuit 101, respectively. The trigger generation circuit 101 includes, for example, a short time constant differentiator, a baseline restorer,
A discriminator using a comparator etc.,
It is composed of a pulse generation circuit, a threshold voltage supply circuit, etc., and outputs a trigger pulse T only when the signal Vd or its differential signal exceeds the threshold level. Note that although the leading edge trigger method or zero crossing method of the above configuration example may be used for the trigger generation circuit 101, a constant fraction timing method may also be used using a delay line or the like. The trigger pulse signal T is input to the timing control circuit 102, and timing signals a to d, which will be described later, are output at appropriate timings. It is also assumed that there is a function to detect and remove a pileup event, and a function to not accept processing when a signal related to a certain event is generated while a signal related to the next event occurs.

On the other hand, the signal Vd is transmitted through the delay circuit and amplifier 100.
After being delayed for a certain period of time, the signal is input to an integrating circuit and amplifiers 132 and 133.
Vg and Vd' are output. In particular, if a differentiating circuit with a time constant shorter than that of the differentiating circuit and amplifier 131 (which normally matches the time constant of the integrating circuit and amplifier 132) is included in the amplifier 133, the signal Vd' When the pulse width becomes short and the charge collection time tr is relatively small, the wave height of the signal Vd' becomes highly sensitive to fluctuations in tr (however, noise increases relatively). The signals Vd' and Vg are input to peak detection/hold circuits, that is, pulse stretcher circuits 140 and 141, respectively, which are controlled by the timing signals a and b (for example, reset signals, gate signals, etc.), and the signal wave height is
Vd′max and Vgmax are each output. signal
Vd'max and Vgmax are each input to a divider 110, and a signal proportional to Vd'max/Vgmax is output, which is taken into a sample and hold circuit 151 according to a timing signal c and held. The output of the sample and hold circuit 151 is subjected to nonlinear conversion via the nonlinear amplifier 111, and then input to the multiplier 112 together with the output of the sample and hold circuit 150, and a signal proportional to the product of both is output. Furthermore, the output signal of the multiplier 112 is inputted to the adder 113 together with the output signal of the sample and hold circuit, and the obtained added signal is sent to the sample and hold circuit and the driver 113 according to the timing signal d.
2 and held, the energy signal
CORR.Vg is obtained.

Note that in applications where the signal CORR.Vg is analyzed by MCA, the timing control circuit 102
Configure so that the MCA GATE signal is output.

The operating principle is as follows. Energy signal after correction with respect to energy signal wave height Vgmax before correction
CORR.Vg is expressed by the formula.

CORR.Vg=Vgmax{1+f(tr)} =Vgmax+Vgmax・f(tr)...Here, f(tr) is a correction coefficient, which is a function of the depth information x/1 of the radiation incident position mentioned above, that is, the charge collection time tr. It is a function. For example, if the tr dependence of Vgmax is expressed by the tr dependence curve of Vemax shown in Figure 6, if the function f(tr) shown in Figure 2 a is corrected in the equation, the charge due to the above cause can be collected. It is possible to improve the deterioration of energy resolution due to imperfections in the waveform shaping circuit and the tr dependence of the response of the waveform shaping circuit. By the way, the peak value Vdmax of the output signal of the differentiating circuit and the amplifier 131 is as shown in FIG.
In this case, the vertical axis in Figure 8 is tr as shown by the dotted line c, which can be read as Vdmax/Vpmax (relative value).
(however, it differs depending on the differential time constant, as in the case of a, a', a'' in the same figure, but this curve c shows the case where the time constant is about the same as curve a) ), showing greater tr dependence than Vgmax.
Furthermore, if the amplifier 133 includes a differentiating circuit with a short time constant, the peak value of the output signal
Vd'max has tr dependence, for example, as shown by the dotted line d in FIG. 8 (in this case, the vertical axis in FIG. 8 can be read as Vd'max/Vpmax (relative value)).
In such a case, R≡k・Vd'max/Vgmax... exhibits a dependence on tr as shown in FIG. 2b, for example, and represents the depth information x/1 of the radiation incident position or the charge collection time tr. can be used as a parameter (however, the coefficient k is k for the event tr0)
Vd′max = Vgmax). That is, f(tr) can be expressed as a function P(R) of R in the equation, and for example, from FIGS. 2a and 2b, a nonlinear curve as shown in FIG. 2c is obtained.

Referring to FIG. 1, the output of divider 110 is proportional to R in the equation, and nonlinear amplifier 11
1 to a signal proportional to P(R)=f(tr), and multiplier 112 converts it into a signal proportional to Vgmax・f of the equation.
In other words, a signal corresponding to (tr) is obtained.

The above shows one embodiment, and various changes can be made in the structure without departing from the spirit of the invention.

For example, since the integrator circuit and amplifier 132 generally have a longer propagation time than the amplifier 133, the delay circuit and amplifier 100 may not be common, and the integrator circuit and amplifier 132 may be configured to receive input earlier than the amplifier 133. Good too.

In addition, peak detection/hold circuits 140, 14
It may be configured such that each of the outputs Vgmax and Vd'max of 1 is once captured and held in another sample-and-hold circuit and then inputted to the divider 110 and the sample-and-hold circuit 150.

Furthermore, the curve of FIG. 2c can be expressed as shown in FIG. 2d. That is, by setting ΔR≡1−R=(Vgmax−k·Vd′max)/Vgmax…, the correction coefficient f(tr) can be expressed as a function Q(ΔR) of ΔR. An example of this configuration is to provide a subtracter before the divider 110 in FIG. 1 so that the output of the divider 110 is proportional to ΔR in the equation instead of R in the equation,
The response can be made into a monotonically increasing function with respect to the input as shown in Fig. 2d.

As a first approximation of this modification, Q(ΔR) as shown in Fig. 2d is set in a relatively small range of ΔR,
Assume that the following equation holds.

Q(ΔR)≡f(tr)≒α・ΔR... Here, α is a proportional coefficient. By substituting the expressions into the expressions, CORR.Vg ≒Vgmax+α(Vgmax−k·Vd′max) ≡CORR.Vg′ . . . is obtained. An example of this configuration is shown in FIG. The peak detection/hold circuits 140 and 141 are the same as those shown in FIG. Each is output at the appropriate timing.
Both the Vgmax and Vd'max signals are input to the subtracter 214, and a signal proportional to (Vgmax-k·Vd'max) is obtained. This signal is input to the adder 213 together with Vgmax to perform the calculation corresponding to the equation, and the output is taken into the sample hold circuit and driver 252 by the timing signal d' and held, and the energy signal CORR.Vg' is obtained. Even if the accuracy of the correction is increased as shown in the solid curve in Figure 2 d, the error in the correction is essentially large in the range where ΔR is large and the reliability is poor. As shown in FIG. 3, this approximation method is very effective because the circuit configuration is extremely simple since multipliers, nonlinear amplifiers, etc. are not required.

In addition, as another modification, a waveform shaping amplifier,
A pseudo-trapezoidal filter, which is a combination of a quasi-Gaussian filter and a gated integrator, may be used. In that case, as shown by the solid line b in Figure 8, the above cause, that is, the dependence on tr of the waveform shaping circuit response, is very small, so the cause of the deterioration of energy resolution is mainly due to the above cause, that is, the imperfection of charge collection. be. This modification is shown in FIG. Integrator circuit and amplifier 13
2. The amplifier 133 and the peak detection/hold circuit 141 are the same as those shown in FIG. 1, and by inputting the trigger pulse signal T to the timing control circuit 302, the timing signals a, bg, br,
Each of c, d (or d'), MCA GATE signal, etc. is output at appropriate timing.
The output signal Vg of the integrating circuit and amplifier 132 is input to a gate-controlled integrator 160 and integrated for a certain period of time to obtain a signal Vgi. Here, the gate controlled integrator 160 receives a timing signal as shown in the figure.
If the gate switch is controlled by bg and the reset switch is controlled by timing signal br, a hold function can also be provided. Also, providing a voltage-to-current converter before the gated integrator 160 improves the linearity of the integration over a wide range of input signals. In addition, the signal
Regarding Vgi processing, please refer to the signals shown in Figures 1 and 3.
It is similar to Vgmax, i.e. signal Vgi and signal
The energy signal is obtained by correcting the signal Vgi from the relationship of Vd'max.

Furthermore, in the configuration as shown in FIG. When the time constant τ of the filter is made variable, the signal Vd (and hence the signal Vd') changes in accordance with the time constant τ, which may cause problems in use. Therefore, the time constant τ ₀ of the differentiating circuit stage of the differentiating circuit and the amplifier 131 is fixed, and the time constant τ of the integrating circuit and amplifier 132 including the differentiating circuit stage and the integrating circuit stage whose time constant τ is variable is fixed.
may be configured to be a quasi-Gaussian filter that is variable within the range τ≦τ ₀ .

(f) Effects According to the present invention, in a room-temperature radiation detector using a compound semiconductor element, problems caused by incomplete charge collection (the cause described above) and the dependence of the waveform shaping circuit response on the input signal rise time. Deterioration of energy resolution (the cause mentioned above) can be improved with almost no reduction in detection efficiency. Moreover, this is possible with a relatively simple circuit configuration.
It is particularly effective in applications where elements are arranged one-dimensionally or two-dimensionally to perform position detection. Further, the circuit noise is smaller than when a single delay line clipping filter is used. Furthermore, since there is no need to make the time constant extra long to compensate for the above-mentioned causes, high count rate characteristics can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 a, b, c, and d are graphs for explaining the embodiment, and FIGS. 3 and 4 are block diagrams showing modified examples, respectively. 5 and 9 are block diagrams of the conventional example, and FIG. 6, FIG. 7 a, b, c, d, and FIG. 8 are graphs for explaining the conventional example. 1... Radiation detector for room temperature, 11... Compound semiconductor element, 12... Cathode, 13... Anode, 2... Preamplifier, 3... Main amplifier (waveform shaping amplifier), 30... Waveform shaping amplifier using a quasi-Gaussian filter. ,3
1,131...Differential circuit and amplifier, 32,13
2... Integrating circuit and amplifier, 100... Delay circuit and amplifier, 101... Trigger generation circuit, 102,
202, 302...timing control circuit, 110...
Divider, 111... Nonlinear amplifier, 112... Multiplier, 113, 213... Adder, 133... Amplifier or differentiation circuit and amplifier, 140, 141... Peak detection/hold circuit, 150, 151... Sample hold circuit, 152, 252 ...Sample hold circuit and driver, 214...Subtractor, 1
60...Gate controlled integrator.

Claims (1)

[Scope of Claims] 1. A radiation detection means for room temperature using a compound semiconductor as a detection element, a preamplification means, a waveform shaping amplification means including a quasi-Gaussian filter consisting of a combination of a differentiating means and an integrating means, means for obtaining charge collection time information for each event by comparing the signal wave height regarding the output of the differentiating means with the output signal wave height of the waveform shaping and amplifying means; 1. A semiconductor radiation detection device comprising: correction means for correcting the output signal wave height of the shaping amplification means to improve deterioration in energy resolution caused by incomplete charge collection. 2. The semiconductor radiation detection device according to claim 1, wherein the waveform shaping amplification means includes a pseudo-trapezoidal filter configured to integrate a signal related to the output of the quasi-Gaussian filter for a certain period of time using a gate integrator. Device. 3. The means for obtaining the charge collection time information comprises inputting the output of the differentiating means into a differentiating circuit with a shorter time constant and comparing the wave height of the pulse signal obtained by inputting the output of the differentiating means with the output signal wave height of the waveform shaping and amplifying means. A semiconductor radiation detection device according to claim 1 characterized by: