capacitance in the feedback loop) to produce an output which is proportional
to the charge into the preamplifier input. A large valued resistor in parallel
to the feedback capacitor slowly discharges the capacitor, restoring the
preamplifier output to its original state. Unlike voltage sensitive preamplifiers,
charge sensitive preamplifiers must have low input impedance so that the
preamplifier can easily sink (or supply) charge from the detector.
How are charge sensitive preamplifiers more suitable for use with
particle detectors than voltage sensitive preamplifiers?
Ionizing events within detectors generally produce an amount of ionized
charge that is proportional to the energy of the incoming particle or gamma-
ray. For this reason, the detector preamplifier should be configured in a
way to produce an output that is precisely proportional to this ionized
charge. Voltage sensitive preamplifiers were first used to read out solid
state detectors when they were first developed in the '40s. A problem was
noted, however, in that the signal voltage at the preamplifier input was not
only proportional to the ionized charge, but also inversely proportional to
the input capacitance. Because the detector capacitance is usually a weak
function of the temperature, temperature drifts were causing drifts in the
preamplifier gain and degrading the energy resolution. For this reason the
charge sensitive preamplifier was developed, which has a gain equal to the
reciprocal of the feedback capacitance, and more importantly independent
of the input capacitance. For many decades, charge sensitive preamplifiers
have been the standard design for use in detectors where the energy
measurement of individual ionizing events is of interest.
The decay time of the preamplifier output pulse is quite long. Do I
have to worry that pulses will build on previous pulses and cause a
'pile up' of events?
The point at which to be concerned about the effects of pulse 'pile up' is
after the preamplifier output pulse has been filtered through a shaping
amplifier. The shaping amplifier (also called 'linear amplifier', 'spectroscopy
amplifier', or 'pulse amplifier') dramatically changes the shapes of the pulses,
generally giving them a longer risetime and a much quicker fall time, and
restores the baseline to prevent pile up as much as possible. Events that
appear to pile up before the shaping amplifier often become very clearly
separated after the shaping amplifier.
What is the bandwidth of the CR-110?
The term 'bandwidth' is generally not used when discussing charge sensitive
preamplifiers - instead one describes their rise time due to a delta current
pulse input (which charges the feedback capacitance), and their decay time
due to the discharge of the feedback capacitance through the feedback
resistance. In general, one seeks a fast pulse rise time, but not necessarily
a short decay time. In fact, if the feedback resistor value were substantially
decreased in order to quicken the decay time, the added thermal noise due
to this decreased resistance would be unacceptable.
How can we check to see whether the preamplifiers are operating
within the specified noise level?
The method described here requires the following:
1. A test circuit board (such as the CR-150-X) with an appropriate power
supply.
2. A low noise Gaussian shaping amplifier, having a shaping time of 1 s.
The CR-200-1 s Shaping Amplifier used with the CR-160 evaluation board
would be suitable.
3. A pulse height analyzer.
4. A tail pulse generator or square wave generator.
5. A silicon p-i-n photodiode (Hamamatsu S1223 or equivalent), and a bias
supply (100 volts if using the Hamamatsu S1223).
6. A small
241
Am isotopic source.
To measure the noise of the preamplifier, the gain of the detection system
must first be precisely measured (in keV per channel). To do this, construct
the circuit shown in Figure 2 (the CR-150-AC-X test board could be used
for this). A p-i-n photodiode should be used as the 'detector', and a bias
power supply should reverse bias the detector to its maximum allowed
value. The
241
Am source should be oriented so that its emissions can
irradiate the photodiode. The circuitry and photodiode should be in a well
shielded, light tight box. Route the preamplifier output to the Gaussian
shaping amplifier (1 s), which should have its output routed to a pulse
height analyzer. Acquire a
241
Am pulse height spectrum, in which you
should be able to clearly detect the 60 keV gamma-ray emission (see
Figure 6). Note the channel number at which the 60 keV photopeak is
observed. The gain of the detection system is the ratio: peak channel
number / 60 keV.
Next, disconnect the input lead of the preamplifier (pin 1) from the test
circuit board. This can be done a using a variety of methods, but make sure
that pin 1 is left floating and does not touch the circuit board or other
components. Connect the tail pulse generator (or square wave generator)
to the preamplifier via a small valued capacitor of no more than just a
couple pF. Alternatively, you can use a 'dangling wire' connected to the tail
pulse generator and rely on the small capacitive coupling between the input
and the wire to make this connection (be sure, though, that the wire does
not move during the subsequent measurements). Acquire a pulse height
spectrum of the tail pulse signal, which should appear as a Gaussian
distrubution. Measure the width of this distrubution by measuring, in
channels, the full width at half the maximum value (denoted as FWHM).
The noise can then be calculated by dividing by the previously measured
gain, to yield a figure having units keV FWHM (Si). To convert this figure to
the more generally applicable units of electrons RMS, divide by 0.0036 keV
(the ionization efficiency of silicon) and divide again by 2.355 (converting
FWHM measurements to RMS).
Figure 7 above shows the
power dissipation of the
CR-110 charge sensitive
preamplifier as a function
of power supply voltage.
This assumes the output
is unloaded and the input
unconnected.
0
100
200
300
400
0
200
400
600
800
1000
1200
Figure 6
241Am spectrum
60 keV emission
# count
s
channel #
0
100
200
300
400
6
7
8
9
10
11
12
13
supply voltage
po
we
r di
ssi
pation
(mW
)
Figure 7
Table 4: Ordering Information
Model No.
Description
Order No.
CRXXX
Charge sensitive preamp, SHV, 2kV/10nF, 1.4V/pC
HY100
CR 110
Charge sensitive preamp, SHV, 2kV/10nF, 150mV/pC
HY110
CR 111
Charge sensitive preamp, SHV, 2kV/10nF, 15mV/pC
HY111
CR 112
Charge sensitive preamp, SHV, 2kV/10nF, 1.5mV/pC
HY112
CR 113
Charge sensitive preamp, SHV, 4kV/4.7nF, 1.4V/pC
HY113
Table 1: Sensitivity Versions
preamp
model
gain (mV / pico-
Coulomb)
max. detect.
pulse (e-)
Equiv. noise in sili-
con keV (FWHM)
CR 110
1400
10
7
1.7 keV
CR 111
150
10
8
6.0 keV
CR 112
15
10
9
65 keV
CR 113
1.5
10
10
230 keV
preamp
model
noise (ENC) in
e- RMS*
noise (ENC)
slope e-/pF
rise time
(C
d
=0pF)
rise time
slope
CR 110
200
e-
4
e-/pF
7ns
0.4ns/pF
CR 111
630 e-
3.7
e-/pF
3ns
0.25ns/pF
CR 112
6,800 e-
28
e-/pF
6ns
0.25ns/pF
CR 113
24,000 e-
27
e-/pF
20ns
0.25ns/pF
