Trigger in the context of nuclear physics applications refers to a signal or event that initiates data acquisition and recording processes. It acts as a decision-making mechanism to determine which events or signals are relevant and should be captured for further analysis.
The primary purpose of a trigger is to select specific events or signals of interest from a larger set of data. The ADC of our digitizer produce a continuos stream of data; triggers play a crucial role in reducing data volume and selecting events that meet specific criteria, such as energy thresholds, timing requirements, or specific patterns.
A good trigger should possess certain characteristics to ensure effective event selection and data acquisition. It should have a high efficiency to capture as many relevant events as possible while minimizing false triggers due to noise or unrelated signals. Additionally, a good trigger should have a fast response time, enabling quick event detection and reducing dead time between events. It should also have a low threshold to detect signals of interest, and a high signal-to-noise ratio to distinguish relevant signals from background noise effectively.
The difference between a leading edge and a derivative trigger lies in their operational principles. A leading edge trigger detects an event based on a specific threshold crossing point, typically the rising edge of a signal, to initiate data acquisition. On the other hand, a derivative trigger looks for changes in the signal slope or its rate of change. It detects events based on the magnitude or shape of the derivative of the signal, providing a different criterion for triggering and event selection. Both types of triggers have their own advantages and are suitable for different applications depending on the desired event characteristics to be captured. Since leading edge triggers are based on amplitude thresholds, they have lower sensitivity to derivative noise and small amplitude fluctuations. This can reduce the risk of false triggers caused by noise or unwanted signal variations. On the other hand, derivative triggers can operate at very high rate and have a fast response time, making them suitable for applications that require high speed and low dead time. They indeed can operate even in the presence of pile-up, where multiple events occur in a short time interval.
The trigger should be designed to produce both single clock cycle trigger pulse and ToT signal.
Time over Threshold (ToT) is a technique commonly used in triggering systems. It measures the duration of a signal above a certain threshold, providing additional information about the signal’s characteristics. ToT helps discriminate against noise and determine the signal’s energy deposition, making it useful in particle detection and energy measurement applications.
For this application notes we will use a DT1260, connected to a PMT and to an Oscilloscope to monitor the output signal.
Design a Leading-edge trigger with hysteresis using basic components
This paragraph will cover a basic example of how to design a leading-edge trigger with hysteresis using basic components. The trigger will be based on a comparator that compares the input signal with a threshold. The comparator output will be used to generate the trigger pulse and the ToT signal. The comparator output is also used to drive a multiplexed that select the threshold between two values in order to implement the hysteresis. When the comparator is high the THRS_2 register is used as threshold while when the comparator is low the THRS register is used as threshold. Setting the THRS_2 lower by a few ADC counts than the THRS register will implement the hysteresis.
The LEMO 0 output will generate a ToT (Time Over Threshold) signal that will be used to measure the pulse width of the input signal. The LEMO 1 output will generate a trigger pulse that will be high for one clock cycle when the input signal crosses the threshold.
Setting the threshold to 2400 ADC counts and the THRS_2 to 2300 ADC counts will implement an hysteresis of 100 ADC counts.
Design a Leading-edge trigger with hysteresis using integrated components
Sci-Compiler offers a set of integrated components that can be used to speed-up the design implementing macro-functionalities. A leading-edge trigger with hysteresis is available in the component library. The Leading Edge trigger with hysteresis works in the exact same way of the design implemented with basic components. The only difference is that the hysteresis is calculated automatically by the component based on the THRS and DELTA input, where DELTA is the difference between High Threshold and Low Threshold. The Trigger-LE component internally manage the signal polarity and implement a programmable dead-time to avoid multiple trigger on the same signal.
The Data Out bus is the replica of data in bus temporal synchronized with the trigger pulse. The Data Out bus should be used by downstream processing block as data bus to avoid misalignment between data and trigger.
The derivative trigger calculate the difference between one sample and t-N samples before. If this difference is greater than the programmed threshold the trigger is generated. During the block creation a property window will appear to set the delay used to calculate the derivative. y[n] = x[n] - x[n-delay]
The derivative trigger is insensitive to the baseline level. The trigger is generated when the signal slope is greater than the threshold. The derivative trigger is very useful to detect fast signal with a very short rise time. Due to it’s derivative nature the trigger is very sensitive to noise. The derivative trigger should be used only when the signal is very clean and the noise is very low. The threshold is not referred to the baseline but to the signal slope. A trial and error approach is needed to find the right threshold value. If the rate is high and the Leading Edge trigger can not be used it is recommended to use a fast trapezoidal trigger.
Fast trapezoidal trigger
A trigger based on the output of a fast trapezoidal filter is a mechanism used in signal processing and data acquisition systems to detect specific events or signals of interest. The fast trapezoidal filter is a type of digital filter commonly employed for pulse shaping and noise reduction in applications such as gamma-ray spectroscopy.
In this context, the trigger algorithm processes the filtered waveform obtained from the fast trapezoidal filter. The filter shapes the input signal, enhancing its pulse-like characteristics by applying a combination of integration and differentiation operations. The resulting filtered waveform exhibits a trapezoidal shape, which facilitates the identification of desired signals amidst noise.
The trigger algorithm then examines the output of the fast trapezoidal filter to determine when a trigger condition is met. The fast shaper output is passed in a derivative filter that is used to detect the rising edge of the signal. The derivative filter output is then passed in a leading edge trigger that is used to generate the trigger pulse.
Triggers based on the output of a fast trapezoidal filter are commonly used in nuclear spectroscopy because the trigger path will implement the same filtering strategy of the energy path, but with faster time constant in order to reduce pile-up. Even if a derivative component is present (in order to make the trigger insensitive to the baseline level) the trigger is less sensitive to noise than a pure derivative trigger because the trapezoidal filter will reduce the noise.
The fast trapezoidal trigger can be implemented in several way in SciCompiler:
- Using the trapezoidal shaper followed by a derivative trigger
- Using the MCA-HP or MCA-FAST component that embed both trigger and energy path
- Using the Fast Trapezoidal Trigger component
PK and FT are the peaking time and the flat top of the filter. Typically this parameters are very short for trigger application. The peaking time is the time constant of the filter and it is used to shape the signal. The flat top typically is set to 1 because the trigger do not need to compensate the ballistic deficit.
The M coefficient is the deconvolution coefficient and is equal to 1/(exp(-Ts/tau)-1), where Ts is the ADC sampling time (1/fs) and tau is the exponential decay time.