Unit 8
Detector Electronics
1 Nuclear Instrumentation and Measurement
MANE-4350
Chapter 16: Pulse Processing and Shaping
Thus far, we have generally treated the elements of a radiation detection system, or “signal
chain” as separate items. However, we clearly know that the elements must be tied together in
order for the system to work.
In any connection, the voltage is transferred from the source to the load based upon the
characteristic impedance, such that
ZL
VL = Vs
Z0 + Z L
If the output impedance is low compared to the input impedance, the maximum signal is
transferred to the next step in the chain (the output impedance of one step is the input impedance
of the next step).
In most cases, a signal is transferred from one device to another with the aid of a coaxial cable.
When using cables, we must consider:
• The speed of the signal through the cable:
o If the dielectric in the cable is a gas, the signal travels near the speed of light, but
if it is a solid, the speed is about 0.66c (19.8 cm/ns, or 5.1 ns/m)
o Real cables have some attenuation of the signal
o Mostly a concern for high-frequency or fast rise-time signals
• Proper grounding
o “Ground loops” can lead to current flow and noise pickup
• Characteristic impedance:
Different behavior for “slow” or “fast” pulses
Rise time >> transit time => slow pulse
Rise time fast pulse
2 Nuclear Instrumentation and Measurement
MANE-4350
For slow pulses, capacitive loading is most significant, and increases linearly with cable length.
Usually only an issue between detector and preamplifier.
For fast pulses, the characteristic impedance is the chief concern. It depends upon the resistance
at the terminating end of the cable:
• Terminating resistance = characteristic impedance – cable appears to the voltage source
to be infinitely long
• If this is not the case, the current varies at the termination, and a signal is reflected back
along the length of the cable – for fast pulses, this can cause distortion
If the impedance is not matched between the cable and the device, we must use one of a set of
add-ons to properly match the impedance:
• Terminator
• Attenuator
• Splitter
• Inverting transformer
Pulse Shaping
The main purpose of pulse shaping is to avoid pulse pile-up and improve signal quality. These
can be done in an “analog” form, as in the good-old-days, using RC circuit construction, or can
be done digitally (but in a similar fashion) in more modern equipment (discussed later)
3 Nuclear Instrumentation and Measurement
MANE-4350
CR Differentiator
dEout dE
Eout + τ = τ in
dt dt
if RC is sufficiently small
dEin
Eout ≅ τ
dt
High frequency passes through
Fast rise time signal passes through, but the long tail is “differentiated” away
4 Nuclear Instrumentation and Measurement
MANE-4350
RC Integrator
dEout 1 1
+ Eout = Ein
dt τ τ
if RC is sufficiently large (compared to signal time),
dEout 1
≅ Ein
dt τ
1
Eout ≅ ∫ Ein dt
τ
Low frequency signal passes through
CR-RC Shaping
A combination of a CR differentiator and RC integrator results in a more attractive pulse shape
for pulse height analysis, as well as improved signal-to-noise characteristics. When the same
time constant for each stage is used, the output is given by
5 Nuclear Instrumentation and Measurement
MANE-4350
t
Eout = E e − t /τ
τ
Other pulse shapes are possible through combinations of CR and RC stages, such as Gaussian,
triangular, or trapezoidal shapes.
CR-RC networks cannot conduct direct current signals – the current must be alternating. In
practice, this means that the “zero” level of a signal, must appear below the true zero voltage –
this is the baseline shift, and it will vary with the pulse spacing. To avoid the problem use:
• Baseline restorer (better noise characteristics)
• Bi-polar signal (CR-RC-CR network or other circuitry)
Chapter 17: Linear and Logic Pulse Functions
There are two types of pulses that are used in nuclear instrumentation systems:
• Linear Pulses – the magnitude of the signal carries information
o Fast linear pulses – usually come from the output of a detector with a small time
constant
o Tail pulses – pulse generated when the output of a detector is collected across a
circuit with a long time constant. The time of the leading edge corresponds to the
charge collection time.
o Shaped pulses – tail pulses that have been shortened through the shaping methods
discussed previously
• Logic Pulses – the pulses generally correspond to a yes/no or on/off state
6 Nuclear Instrumentation and Measurement
MANE-4350
7 Nuclear Instrumentation and Measurement
MANE-4350
Preamplifier:
• Most detector output is too small to be practically useful without initial amplification
• Preamp is located as close to the detector as possible to minimize capacitive loading
• Conventionally outputs a linear tail pulse with short rise time and long tail
• Pulse pile-up is inevitable, but if there is too much pileup, the signal level can saturate
and distort the signal
• Bias voltage is often supplied through the preamp – preamps must be able to handle up to
several thousand volts
• The low voltage power required to operate the preamp is usually supplied through a
standard multi-pin connector, built into a NIM bin holder
8 Nuclear Instrumentation and Measurement
MANE-4350
Pulse discriminators:
• Integral discriminators – counts pulses above a defined level
• Differential discriminator (Single Channel Analyzer) – counts pulses between two
discriminator set levels
Other counting system elements:
• Counter
• Timer
• Count rate meter (response time varies with changes in storage capacitor – shorter or
longer response times are possible)
Dead time:
• Dead time is usually electronics dependent
• A uniform recovery time is desired (even for paralyzable detectors)
• Electronics dead time may vary with pulse amplitude
• In practice, the dead time is often artificially induced to maintain regularity
Pulse Height Analysis
For relatively low counting rates, pulse analysis is fairly straightforward. When the pulse rate
becomes more rapid, consideration of pulse pile-up becomes very important.
9 Nuclear Instrumentation and Measurement
MANE-4350
• The ballistic deficit refers to the reduction of amplitude between a shaped pulse and the
pulse height from an infinite time constant.
• Signal to noise ratio – varies with pulse shape
• Pulse pile-up
o Pulse that does not return to zero may result in next pulse being artificially high
o Pulse undershoot that does not return to zero may result in next pulse height being
artificially low
Everything that can be done with analog systems as we have discussed, can also be done with
digital pulse processing systems. Analog-to-digital converters are used to transform the pulse
signal into a digital signal. Digital signals can be treated much more precisely, with more
complex mathematical operations, that allows things like deconvolution of piled pulses or timing
methods that are not possible in an analog system
10 Nuclear Instrumentation and Measurement
MANE-4350
Chapter 18: Multichannel Pulse Analysis
Resolution:
FWHM FWHM
R= or H =
H R
FWHM and H can be represented in units of channels. To adequately show the peak shape, the
FWHM must cover a minimum of 5 channels. We can therefore determine how many channels
are required to meet that condition at all levels for a given value of R (= 5/R).
Counting statistics in an MCA apply on a channel by channel basis. When more channels are
used, there are fewer counts per channel, and thus more statistical variation. Changing the
number of channels trades off resolution with statistical noise.
Energy calibration is required for all MCAs – the system only knows about channels, and cannot
link an energy to the channel without user calibration. Calibration parameters include
• Gain
• Zero offset
An ideal MCA behaves linearly with all parameter changes.
11 Nuclear Instrumentation and Measurement
MANE-4350