Plasma Diagnostics
Plasma Diagnostics
Plasma Diagnostics
4.1. INTRODUCTION
The diagnosis of hot, dense and short-lived plasmas requires special and
because the plasma conditions achieved vary over a wide range; their time scale and
the spatial dimensions are such that temporal and spatial resolution of sometimes
One of the major tools to investigate the phenomena occurring inside the
pinched plasma without disturbing them is to try to obtain physical information from
the radiation (photons and particles) emitted naturally from the plasma. Since for
different phenomena (e.g. hard X-rays and electrons) the emission mechanisms are
related, different information coming from different detectors can offer a better
For the investigation of the radiation and the electrons emitted from our
plasma focus, as well as for monitoring the electromagnetic parameters of the device,
75
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
configurations. One was for the measurement of the discharge current and the other
one for the voltage measurements. These two diagnostics are discussed in the
following sections.
The Rogowski coil (or Rogowski belt) [135] is a multi turn solenoid bent into
the shape of a torus. It is used to measure the current flowing through the inner
There is no restriction in the position, direction and sense of the current, and it
is not necessary to have a solid carrying medium (e.g. a wire) for this current. Fig. 4.1
r Scope
76
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
In our experiments, two Rogowski coils were used: one of them monitored the
main discharge current, while the other was employed to measure the electron beam
current.
The coil is terminated with a low inductance resistor, r. The equivalent circuit
is given in Fig. 4.2, where Lc and rc are the inductance and the resistance of the coil,
respectively, I is the discharge current and i the circuit current through the coil. The
+ (r + rc ) i = k
di dI
Lc (4.1)
dt dt
di
a) if Lc >> (r + rc )i , then from Eq. 4.1 one can obtain i=(k/Lc)⋅ I, so the output
dt
voltage is proportional to I:
rk
V0 = ri = I (4.2)
Lc
The simplest way to achieve this condition is to make Lc /(r+rc) much bigger
di
b) (r + rc )i >> Lc , then from Eq. 4.1 the output will be
dt
⎛ rk ⎞
V0 = ri = ⎜⎜ ⎟⎟ dI dt (4.3)
⎝ rc + r ⎠
77
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
Due to their open design, the Rogowski coils are strongly sensitive to
electromagnetic noises. The simplest way to avoid such disturbances in the signal is to
use a noise shield, but this will add distributed capacitances which will filter off the
higher frequencies. The coil used for electron beam measurement was shielded using
copper tin. Only a small slit (0.8 mm) along the inner surface of the torus allowed the
Rogowski belt for a certain application is the magnetic coil characteristic time,
Lc /R = Lc /(r+rc). For a good derivative coil, this characteristic time has to be as small
as possible, so the coil will be able to follow all the signal peaks. Frequencies of the
same order of magnitude or lower than the characteristic frequency will be integrated.
For our Rogowski coils, this characteristic time was chosen to be very small for the
electron current detector (0.75 ns) and higher (85 ns) for the main discharge current.
a simple RC circuit [136]. Fig. 4.3 shows the Fig. 4.3: RC passive integrator.
but since
dQ dVC dEout
i= =C =C (4.5)
dt dt dt
one gets
78
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
dEout 1 1
+ Eout = Ein (4.6)
dt τ τ
the signal, the second term can be neglected and after integration one obtains
1
Eout ≅
τ ∫E
in ⋅ dt (4.7)
The time constant, τ, has to be carefully chosen, since a very high value will
give a low-integrated output. The high frequencies will not be integrated due to the
parasitic effects (e.g. the capacitance of the resistor and the inductance of the
capacitor). A low value for the time constant will distort the signal at low frequencies.
It is not simple to choose the right time constant for our purposes, since, for example,
the main current has a period of 11.9 µs, but the fast dip corresponding to the focus
When the Rogowski coil is used to measure the integrated current (as a current
circuit, the following formula [137] may be used for the calibration:
πC0V0 (1 + f )
I1 = (4.8)
T
C0 = 31.70 µF
V0 = charging voltage
f = reversal ratio V2 /V1= 1/4 (V5 /V4+V4 /V3+V3 /V2+V2 /V1), where Vi (with
i = 1 to 5) are successive absolute values of the recorded signal peaks
T = periodic time (average value of several cycles)
I1 = the first peak value of the discharge current which has to be calculated
79
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
The coils used in our experiments were both derivative-type. The output from
the coil used for the discharge current was first recorded directly (via a 10:1 resistive
passive RC integrator) to give the discharge current. The calibration factor for the
whole detector chain in our experiments is k1 = 18×103 A/V. The coil used for the
electron beam signal was coupled to a 2 µs passive integrator. The calibration factor
high resistance voltage dividers of conventional design are usually employed), and
probes are employed. The basic designs are either resistive or capacitive dividers, or
mixed. The use of resistive dividers is limited by the quality of the resistors; the
characteristics of the experiments usually require very high resistances, in high power
configurations, which normally display high inductances; therefore, the time response
of the divider limits the use of such dividers in the case of very high frequency
transients. The capacitive probes are limited at the other end of the spectrum; a
80
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
normal capacitive divider does not work properly at low frequencies, typically
below 107 Hz [126]. Another limitation for the capacitive dividers is the quality of the
insulation of the capacitors, mainly the low capacity one, which has to be able to
withstand the peak value of the discharge voltage; this constraint makes the whole
ad-hoc, resulting in a high error in the capacitance and therefore in the divider ratio.
On the other hand, an important advantage for the capacitive probes is the better
In this case, the main limitation arises again from the quality of the resistors (the stray
inductances), as well as from stray capacitances caused by the shielding and the
relative positions of different components. Both the inductances and the capacitances
can cause the whole circuit to start oscillating, and a strongly non-linear frequency
The resistive dividers are commonly used, because of their easy design and
fabrication. In our experiments, the voltage along the plasma column was measured
using resistive probes connected at the lower end of the focus tube across the anode
One of the problems in measuring the pinch voltage in a plasma focus device
is related to the position of the probe. For data interpretation, the important parameter
is the value of the voltage along the plasma column, i.e. at the inner end of the
therefore, the probes are usually positioned at the other end of the electrodes,
connected to the voltage plates. In this case, in Eq. 3.2 we have to separate in the
81
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
second term the fixed inductance of our electrode system (34 nH) from the plasma
V = R plasma ⋅ I plasma +
d
(Lplasma ⋅ I plasma ) + Lelectrodes ⋅ dI plasma
dt dt
The first two terms represent the voltage along the pinch, while the third one is
the supplementary term introduced by our point of measurement. Since for the pinch
phase the current is decreasing, this formula predicts that the measured voltage for the
radial phase is lower than the real one. Although the Lee model predictions for the
value of the last term for our device could lead to a pinch high voltage
much smaller value, which is less than 5% of the peak value. This error is smaller
than the other experimental errors, and the high voltage signal is strongly affected by
the electromagnetic noise during the pinch phase, which leads to a higher uncertainty
in the scope readings; therefore, the error introduced by the point of measurement in
The original voltage probe for the UNU/ICTP PFF device is a resistive
divider, which measures the voltage between the high voltage plate and the ground
plate of the dense plasma focus device. Fig. 4.4 shows the structure of the divider
82
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
Fig. 4.5.
purposes. Therefore, a new voltage probe was built and successfully implemented on
The PSpice software was used for preliminary studies on the behaviour of a
the measuring circuit, consisting of probe, cable, and terminator or end-divider. The
simulations showed that the inductance of the resistors used modify the time response
of the probe, and also induces small phase shifts in the signal. On the other hand,
resistance and the ground, influence the output more dramatically: resonant
frequencies appear, and the whole probe starts oscillating, even for very small values
for the capacitance (few pF). The effects are more intense as the capacitances are
positioned towards the grounded end of the probe. Another result showed that the
resistors used in the terminators and end-dividers should have a low, but non-zero
83
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
inductance, because for resistors with zero inductance the terminator will start
oscillating.
Following the PSpice simulations, the following conclusions have been drawn:
- the number of resistors should be as small as possible (two) to avoid the increase
- the shielding should be avoided, and the resistors should be “seen” as little as
- the resistors should see also as little as possible each other, to minimise their
- a proper resistor should be found for the terminator and, if necessary, the end
divider.
The final choice for the divider resistors was the ceramic-based, non-inductive
with ratios 1000:1 (50 kΩ / 50 Ω) and 100:1 (5 kΩ / 50 Ω) respectively. For the first
the signal on the scope was too big. A Fig. 4.6: Schematic diagram of the new
voltage probe.
probes is shown in Fig. 4.6. The whole construction was carefully insulated and in the
end encapsulated in melted plastic, to avoid high voltage breakdowns which could
84
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
cause health hazards or damage the oscilloscope. The probe itself was not shielded,
but the cable which connects the probe to the scope had a supplementary external
The new probes were tested on a pulse circuit built in the lab. The time
response was compared to a standard Tektronix P6114B 10:1 voltage probe. The time
response for our probe was slightly faster than the Tektronix probe, and it was found
to be better than 2.5 ns, which was the limit for our pulse generator. The attenuation
coefficient was found to be constant up to 600 V (the voltage limit of our generator).
For real discharge testing, the probes were fitted in identical positions on the
UNU/ICTP PFF device. The cable lengths were identical, and the signals were
recorded on the same scope. For all the test discharges, the t = 0 moment is arbitrary.
A comparison between the old and the new 1000:1 probes is presented in fig. 4.7.
100
80 New probe
Old probe
60
Voltage (kV)
40
20
-20
Fig. 4.7: Comparison between the old and the new 1000:1 probes.
The first thing to discuss is the oscillation appearing on both the traces. This
signal does not originate in the probes. It can be seen in all the electrical
85
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
system. Because of its origin, any attempts to modify the measurement circuit
parameters in order to avoid it or to minimise its effects (change in the cable length,
resistors, etc.) will give no results. For the UNU/ICTP PFF device, this frequency is
about 30 MHz.
The oscillation was found to start only after the voltage signal reaches its first
peak. Therefore, on plasma focus devices, only the first peak appearing on the voltage
trace can be considered accurate, and can be used in any kind of correlations with
By comparing the two oscillograms, it is obvious that the new voltage probe
design is much faster than the old one, and is therefore able to follow in a better way
the high voltage plate potential. The high inductance of both the probe itself and the
10:1 attenuator used in the old design gave rise to the phase shift which is
100
40
20
-20
Fig. 4.8: New design probe – 1000:1, versus new design probe –
100:1 coupled to the old design 10:1 end-attenuator.
86
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
In order to further test the new design, a comparison between the 1000:1 probe
and another new probe, 100:1, was also performed. Due to the higher amplitude of the
signal, this second probe used a 10:1 attenuator at the oscilloscope input, identical
with the one employed for the old probe. The results are shown in Fig. 4.8.
By comparing Fig. 4.7 and Fig. 4.8, it is clear that the end-attenuator is
responsible for part of the loss in the time response (which in turn limits the amplitude
of the measured voltage), as well as for the small phase shift in the signals.
After intensive testing, we concluded that the new high voltage probe design
offers a much better solution for measuring the high voltage on our plasma focus
device. Although only the first peak can be considered as offering reliable
information, and the design is more prone to be affected by electromagnetic noise, the
results are in much better agreement with the plasma focus models and, as will be
shown in chapter 6, with other plasma diagnostics, mainly the electron and X-ray
spectra.
During our experiments, we did not reach the time response limit, as measured
with the pulse generator. Instead, the fastest parts on the voltage traces seem to be
divider (e.g. change in the resistance values, and therefore in the divider ratio, caused
by the high power passing through the resistor during the pinch phase – more than
50 kW) or by the signal processing in the scope, which was working at its limits in
The slew-rate-type behaviour seems to be the main cause for the limitation in
amplitude for our high voltage signals; the maximum measured voltage exceeded
90 kV only for a few discharges, for which the higher values seem to be caused by a
87
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
positive addition of the electromagnetic noise. This aspect will be discussed also in
section 6.4.7.1. Anyway, the measured voltages are among the highest ever reported
The studies of the X-rays emitted from the plasma focus provide a powerful
measurements correlated with the other diagnostics improve the overall picture of the
temporal evolution of the transient plasma. These measurements are helpful in finding
the mechanism of the emission of the X-ray photons and other particles.
High energy photons falling upon a suitable scintillation material produce light
produces a greater light output for a given energy deposited in the crystal than a
plastic scintillator, and resolves lower energy photons from the photomultiplier noise.
plastic scintillator is commonly bonded by an optical joint (with a silicon grease such
as high viscosity silicon grease Dow Corning DC 200) to the photomultiplier window.
For high light collection efficiency, the scintillator and the light pipe are generally
coated with white tygon paint. As far as the linearity of plastic scintillator is
88
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
concerned, Bailey [138] reported that the light output for NE102A versus energy
The tubes vary widely as to the spectral response, quantum efficiency, gain, noise
level, maximum current output, physical size and overall structure. When a suitable
tube has been chosen, there remains a selection of an appropriate tube base circuit
suitable for the signal duty cycle of the experiment. For example, the provision of a
linear anode output to high current pulses over a long duty cycle will require
i ⋅ dt
capacitors on each stage large enough to ensure that ∫ C
<< V for each stage. The
capacitors of several µF, or a separate power supply may be necessary for the last
termination time constant RC and a scintillator time constant τ much smaller than the
∞ dN (E , t )
V (t ) ~ ∫ f (E ) dE (4.10)
0 dt
dN (E , t )
where is the number of photons reaching in one unit of time the unit of
dt
surface of the scintillator, while the function f(E) includes, among other parameters,
photomultiplier tube (e.g. the overall gain of the dynode chain and the quantum
efficiency of the photocathode), and geometrical factors (related to the light collection
efficiency).
fast multiplier and a selected scintillation material. The detection efficiency for
89
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
photons in a plastic scintillator depends upon the photon energy, the geometry and the
scintillation material.
For the time-resolved soft X-ray measurements from the plasma focus, the
choice of the detector depends on the information required. The purpose of our
experiment was to get information on the temporal evolution of the soft X-ray
emission.
diodes, etc. The semiconductor detectors are also sensitive to charged particles,
provided the particle deposits part of its energy in the sensitive layer of the detector.
needed to create an electron-hole pair. The collected charge is also proportional to hν.
For silicon at room temperature, ω = 3.62 eV [139]. The quantum detection efficiency
is given by [139]
hν
Q= e − µ d t d (1 − e − µ s t s ) (4.11)
ω
where the first exponential term describes the absorption in the dead layer (on the
front surface of the detection area) of thickness td , and the second one gives the
90
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
efficiency for stopping X-rays in the depletion (sensitive) region of thickness ts . It can
be seen that there is an energy range over which semiconductor detectors can be used.
The lower limit is determined by the absorption in the dead layer and the upper limit
The pair creation energy, ω, is, in general, a function of the photon energy
[140]. It is usually assumed that ω is a constant for photon energies well above the
band gap of Si (~10×Eg [141]). However, recently, Monte Carlo simulations of the
pair creation energy [142] show significant photon energy dependence up to a few
hundred eV.
Since both the electrons and the holes have large mobilities, and since the
collection distances are short, it is possible to achieve relatively short collection times.
0.20
sources. 0.15
91
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
A five-channel X-ray spectrometer with filtered BPX-65 PIN diodes was used
in our experiment. The details about this spectrometer are provided in chapter 5.
Several diodes can be manufactured together on the same silicon chip. If the
information (the charge created by the incident photons) can be read separately for
each diode, an imaging detector can be built. Currently there are two main
manufacturing technologies, the result being two different types of light detectors:
Developed in the 1970s and 1980s specifically for imaging applications, CCD
technology and fabrication processes were optimised for the best possible optical
properties and image quality. The technology continues to improve and is still the
holding region, which is shielded from light. The photodiode converts light into
Typically, light is collected over the entire imager simultaneously and then transferred
to the adjacent charge transfer cells within the columns. Next, the charge is read out:
each row of data is moved to a separate horizontal charge transfer register. Charge
packets for each row are read out serially and sensed by a charge-to-voltage
some disadvantages have to be mentioned: the integration of other electronics onto the
silicon chip is impractical; the CCD operation requires several clock signals, clock
92
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
levels, and bias voltages, which complicates the system integration and increases the
A CMOS imager, on the other hand, is made with standard silicon processes in
process. CMOS imagers can also benefit from process and material improvements
and an amplifier section. Overlaying the entire sensor is a grid of metal interconnects
to apply timing and readout signals, and an array of column output signal
interconnects. The column lines connect to a set of decode and readout (multiplexing)
electronics that are arranged by column outside of the pixel array. This architecture
allows the signals from the entire array, from subsections, or even from a single pixel
chip), lower power dissipation (at the chip level), and smaller system size, at the
expense of image flexibility and quality: the built-in electronics reduce the area
available for light capture, and with each pixel doing its own conversion, uniformity
For our experiments, a CCD detector was used as a detector in the X-ray
pinhole camera. As mentioned earlier, the semiconductor detectors are also sensitive
93
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
mainly in the detector area, in order to simplify the maintenance and the usage of such
tight);
camera is presented in Fig. 4.10. The image is inverted. The magnification can be
modified by changing the distances between the pinhole and the object and/or
In the beginning, pinhole cameras used X-ray films [144] as detectors. The
film was moved after each exposure (unless the goal of the experiment was the
reproducibility of the image, because in this case several shots used to be overlapped
on the same position of the film). This feature used to increase the cost and the
complexity of the device, and the later processing of the film was time-consuming and
did not offer very reproducible results. Better detectors are used at present, and the
94
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
If an X-ray band filter is used [104], the image obtained will show the
The pinhole dimension has to be carefully chosen, because the diameter has to
be much smaller than the details to be investigated. Since the resolution depend also
Special designs for pinhole cameras, like multiple pinholes, allow several
images to be captured for the same discharges, for different positions (so resolution on
the third axis can be achieved) and/or with different absorption filters, so the pictures
detectors have been developed. The detectors used for particle detection can be
- with respect to their dynamic properties (dynamic, or active detectors, and static,
or passive detectors).
95
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
the incident particle. Some other detectors (like the Geiger-Muller detector) are not
sensitive to the input energy; they can just count particles without being able to
are used to study the temporal behaviour of the radiation emission. Geiger-Muller
tube, scintillation detectors coupled with photomultiplier tubes etc. are all members of
this class.
The static detectors give a permanent record of the incident radiation, but the
information about temporal behaviour of the radiation flux is normally not available
with these detectors. This family includes, among others, photographic films and solid
(i.e. electrons, protons, deuterons, alpha particles, etc.) transfer their energy to the
stopping medium:
- Coulomb electrostatic force between the particle and the electrons of the target
(Bremsstrahlung);
- direct electrostatic force between the moving charged particle and the target nuclei
can result in the ejection of target atoms from the lattice sites, or out of the
molecular chain.
96
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
4.6.2. Spectrometers
the data recording more convenient even for energy-dependent detectors, a device
which is able to discern between different incoming particles with different energies is
positioned between the particle source and the detector. This device is called
spectrometer.
There is a large palette of methods which can be used as a base for designing a
particle spectrometer. The design is based on one of the many physical processes that
link particles’ energy with a measurable physical parameter. The most convenient
way to discern between particles is to spread them in space and/or in time. By using
electric and/or magnetic fields in a convenient geometry, the charged particles can be
- the spectrometer might be connected to pulsed high voltages, so the whole system,
including power supplies, have to be carefully insulated from the ground, from the
main power source, and from any other devices nearby; and
- the electric fields might create unwanted breakdowns inside and outside the
spectrometer due to inductive effects created by fast varying electric fields created
which the magnetic field is created by a pair of permanent magnets. This enables us to
avoid the above mentioned limitations and to detect the particles without any change
97
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
value of the magnetic induction in the magnetic area (which is not the case in an
electrically-driven configuration);
- the trajectories are more complicated and require a very exact computation to see
The first constraint can be overcome by dealing with more than one pair of
detected. To deal with the second issue, the special computer program, CHAPAT
focus devices similar with the one used in our experiments. Argon ion energies
between 10 and 3,000 keV were measured using a biased ion collector [146], a
method which is not very reliable if only one detector is used (since for energy
calculation it considers that all the ions are emitted at the same moment), or involves
difficult (and not always reliable) mathematical interpretations (if two or more
detectors are employed). On the same device deuteron energies in the 80 – 250 keV
detectors [147]; the automatic processing of the detector is time consuming, and the
energetical calibration might not be very accurate. For the electron energy
98
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
The basic configuration for the analyser was inspired by the one used by
Fig. 4.11: Spectrometer designed at Gunma University; MCP defines the microchannelplate.
After carefully studying the solutions used by the Gunma University group,
we made several important changes in the original design, in order to simplify the
construction and the handling, to improve the performances and the reliability, and, in
the mean time, to keep a low production and maintenance cost for the analyser.
The first important change was in the position of the detector; the Gunma
analyser uses a 180° deflection system, but the dimensions of our detector required a
The second important change was in the magnetic mounting. In the original
design, the magnetic circuit included the vacuum chamber, which was made of soft
iron. We choose to use a separate mounting – an U-shape piece on which the magnets
are fixed. In the whole construction, only this component is made of magnetic
99
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
material (magnetic stainless steel). For all the other components, non-magnetic
materials were used. Several advantages derive from this change: i) the magnetic field
measurements are simplified, since the measurements can be performed outside the
vacuum chamber; ii) different mounting shapes can be used to modify the field
distribution; iii) the position of the magnetic holder can be adjusted, and the result is a
fine tuning of the analyser parameters; and iv) the whole mounting can be changed
The collimator was also simplified. Instead of using two slits, very difficult to
align, we choose to have only one pinhole (200 µm) in front of the collimator, since
the simulations showed that the decrease in the energetical resolution caused by this
Simulations done using the CASINO program for the filter which appeared in
the original design proved that below 100 keV the electrons are strongly scattered
after passing through 3.5 µm mylar, which is a filter thinner than the one used by
Hirano’s group (4.9 µm). At 50 keV, a big fraction of the electrons is scattered at
angles bigger than 45°. Since in the Gunma analyser the filter was placed after the
for low energies. One solution was to place the filter in front of the collimator, but it
turned out that the interaction with the shock waves created by the plasma discharge,
combined with the electron bombardment, destroyed the filters after one or two
The last important change with respect to the original design was to eliminate
the MCP. This costly and extremely sensitive component was introduced not
necessarily to multiply the number of electrons reaching the detector (since the
100
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
amplification factor was kept below 20), but mainly as a buffer, to convert the
incident electrons (especially since those with energies above 45 keV would not lose
all their energy in the sensitive layer of the semiconductor detector) in lower energy
electrons, whose energies are given by an external high voltage biasing circuit.
- there was no vacuum restriction (a MCP requires a pressure below 10-6 mbar,
this pressure is low enough to avoid electron scattering by interaction with the gas
(as CASINO simulation showed), but, in the mean time, high enough to create a
continuous gas flow through the pinhole, in order to protect it from debris coming
- the MCP requires an in-situ calibration, a difficult operation time and resource-
consuming; the Gunma group used nuclear emulsions for proper calibration; and
- a MCP requires high voltage biasing, as well as another high voltage circuit for
accelerating the emitted electrons; the biggest problem is to insulate these power
supplies from the ground and from the mains, since the analyser floats at high
voltage (up to 100 kV); it is the same kind of problem which appears while trying
energies in a spectrometer.
different energy electrons and the detector. Using the CASINO software, based on the
characteristics of the silicon chip (thickness of the dead layer and sensitive layer) we
obtained a correction curve for the detector signal. This aspect will be fully described
in section 5.8.
101
CHAPTER 4 PLASMA DIAGNOSTIC TECHNIQUES
1024Q, together with its driver, Hamamatsu C7615 were used. This linear image
sensor is basically a NMOS detector with only one line of pixels; each pixel
dimensions are 5 mm × 45 µm, and the pixel pitch is 50 µm. The total number of
pixels for our detector is 1024, which leads to a better energetical resolution as
compared to 512 for the detector used in Gunma analyser. The geometrical
optocoupler and an optic fibre. Further details about the detection chain will be given
in section 5.8.
All the analyser electronic components, as well as the other electronic circuits
from the detection chain were powered by lithium batteries or rechargeable batteries
(where this was possible), to avoid the high voltage insulation problems between the
normal power supplies (which would have to float at more than 100 kV) and the main
102