FAQ - Deadtime, pulse width- pile up

If you have a question about this topic, the answer is probably in here!
Post Reply
User avatar
Richard Hull
Site Admin
Posts: 10911
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

FAQ - Deadtime, pulse width- pile up

Post by Richard Hull » Tue Nov 21, 2017 6:03 am

This will be a long and informative post aimed at the person who will build or assemble a counting system. It will also inform those merely interested in how counter detectors work within the framework of the front end electronics.

The physical detector –

I am not going to get deeply into these, but suffice it to say that any gas based detector working outside of the region of “limited proportionality” has a physical construction such that from the first instant of the gas reacting to a suitable radiation, will have a physically dictated “dead time”. This is normally the case in all forms of high voltage operated tubes that rely of an ionizing event which may or may not trigger a cascade of other ionizations. Some such detectors are Geiger tubes, Corona tubes, fission tubes, neutron detectors (3He tubes, BF3 tubes and boron lined tubes), etc. Dead time in these detectors can be as short as a few microseconds to as long as 500 microseconds, in modern tubes, manufactured post 1945.

Dead time –

This is the time the tube is active during a detection event. On detection, some form of ionization takes place and may spread through the detector over time. The event ends only when the tube returns to its pre-detection event condition. In short, the tube cannot issue a new count during its dead time.

For the counter designer it is nice to know the manufacture’s stated dead time as this will help him design the “front end” of the electronics designed to amplify the signal and output it as a single detection event. Detector dead time can, ultimately, determine the upper counting rate limit for a given detector and, thereby, the counter, itself.

The front end electronics –

The pulse out of a detector tube is shaped, most often, by a passive integrator consisting of a capacitor and a resistor. The time constant, (width of this passive pulse shaper’s output), is the time of a recorded pulse seen by the actual counter, active electronics.

As normally configured, this integrator consists of a bias resistor connected to the tube’s central wire from the high voltage, (usually, between 1 megohm and 100 megohms). A capacitor couples this pulse to the counter electronics, being placed in series from the tube's central wire to the counter. Note: the return path is via the grounded detector metal shell which is grounded to the counter electronics.

The high ohm, bias/current limiting/quench, resistor form an integrator with the combination of the detector's internal capacitance and the coupling capacitance. Often the bias resistor value is recommended by the manufacturer. This relates to the “quenching” and needed current limiting of the detection discharge activity within their detector tube.

As the resistor and detector tube's capacitance is often fixed, the coupling capacitor in the integrator can determine the input pulse width.

NOTE: it is important to know that the active electronics at the input can further shape and change the incoming pulse. However, in commercial circuitry, this is rarely done as shaping the pulse before it reaches the active circuitry is cheaper and requires fewer components. This is often important in manufacturing the more inexpensive counters to arrive at a price point.

The battle…dead time to pulse width versus counting rate -

Here is the crux of the issue. we would have no problem choosing a capacitor for the integrator such that the pulse width was a small fraction of the tube’s dead time. This is bad, for the counter would be ready for another pulse while the tube is still dead to receive another pulse. Noise in the decaying tube’s detection signal could trigger more false pulses. By selecting a capacitor such that the integrator pulse is just about 20% longer than the tube's dead time, assures us that no extra pulses will be detected. We now have one detection event and one, and only one, pulse counted with the detector ready to count again.

The ideal is that we try and get the shortest counting pulse to the electronics so that a really radioactive item, (many counts per unit time), can be accurately counted.

Pile-up –

Radiation from any radioactive item is a probabilistic thing. Given: An item with a half-life of a million years has a counted average activity of only 300 cpm. The counts recorded are never 300 evenly spaced particle detections; it is, statistically, virtually impossible. Any given atom in the radioactive substance can decay in the next microsecond or in three million years from now! Thus, there is some small possibility that two of those 300 counts per minute can occur 5 microseconds apart and the second particle will not be counted due to our pulse detection time window. This explains part of the variability of the counts per minute averaging.

When extremely high count rates occur, a condition called “pile up” can create a problem. As the time between detection events gets closer and closer to the input pulse width, the probability of losing counts increases, dramatically. Thus, events will not be counted due to those events arriving during the dead time of the detector.

It is incumbent upon the designer of counting systems to take all of the above into account. This demands good data on the detector tube to be in hand and the use of a digital storage oscilloscope to verify the design of the input circuitry.

Example: A Corona tube has a dead time of 500us. We need to add 100us (20%) to this in the integrator to not go afoul of the dead time. Thus, we need to configure our integrator for a 600us time constant. How fast might we count? 1/600us - 1666 counts per second or almost 100,000 CPM. However, this assumes a new count evenly spaced arrives immediately after each count. This will never occur!! I, personally set a 1/10th limit on this count before pile-up needs to be accounted for. Thus, in reality the count limit might be more like 10,000 CPM before pile-up begins to become an issue.

Knoll's book on radiation detection is one of the best on the above subject. For those who seek out or have a specific need for the manufacture of multiple constructions of detection gear, please consult this text.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

Post Reply