A new compensation method for dead time loss and its implementation

    Abstract: Using the counting method to solve the peak missing problem of the pulse amplitude analyzer system, and then making compensation according to the number of missing pulses, is an effective method to solve the dead time problem in the amplitude analysis system. This article introduces in detail the principle of this dead time loss compensation method and its practical application. Keywords: pulse amplitude analysis system dead time

   

Introduction

    The pulse amplitude analysis system is an important part of the nuclear radiation detection instrument. Based on the analog-to-digital conversion principle, it converts the analog signal output by the pre-circuit (sensor + preamplifier + main amplifier) ​​into digital information proportional to its peak value . The main device for signal conversion is the analog-to-digital converter. During the signal conversion and memory storage process of the analog-to-digital converter, the signal subsequently output by the pre-circuit will be lost. Usually, the minimum time interval that can distinguish between two pulses is called the dead time of the counting system. For the pulse amplitude analysis system, the dead time mainly depends on the sum of the conversion time of the analog-to-digital converter and the storage time of the memory. One approach is to accumulate dead time for each count and then extend the measurement time to compensate for the loss of counts due to dead time. This dead time loss compensation method will cause a large error due to inaccurate accumulation of dead time and other reasons. At the same time, when the counting rate is not large and the conversion speed of the analog-to-digital converter is fast enough, and the counting is basically not lost, the dead time It still exists that compensation for dead time loss at this time is obviously unreasonable. To this end, we have adopted a new dead time loss compensation method, which can effectively overcome the above disadvantages.

Method Principle

    There are two commonly used counting system dead time characteristic models: extended response and non-expanded response. These models represent idealized dead time characteristics, and the response of an actual counting system will often closely resemble one of these two models. It is assumed that each true event that occurs within the "live time" of the counting system follows a fixed dead time τ. Although true events within the dead time are not counted, they follow the missing events and extend another dead time with a period of τ. , called extended response dead time. It is assumed that each true event occurring within the "live time" of the counting system follows a fixed dead time τ. True events occurring within this dead time will be lost without any impact on the characteristics of the counting system, which is called a non-extended response. Dead time.

    The dead time of the pulse amplitude analysis system is non-extended dead time, because another pulse entering the system within the resolution time after a pulse causes counting will not cause counting, but it will not further cause an increase in dead time. The correction formula for dead time is:

    N 0 =N/(1-Nτ)

    where N0 is the total number of pulses entering the system per unit time; N is the recorded count rate; τ is the system dead time; the physical meaning of Nτ is unit The sum of dead times within time. In fact, the dead time of the pulse amplitude analysis system is certainly an important technical indicator of the system, but the ultimate purpose of calculating the dead time is to compensate for the loss of dead time. If the method of extending the measurement time is used to compensate for the dead time, in The extended measurement time will generate new dead time, which is obviously detrimental to the correction of dead time. To this end, we adopt the pulse number compensation method, that is, to compensate for the number of pulses lost during the accumulated dead time during the measurement process. Suppose n0 is the total number of pulses entering the system during the measurement time; n is the count recorded during the measurement time. The new dead time loss compensation method is to compensate for the number of lost pulses (i.e. n0-n). The specific method is: use the T1 timer of the microcontroller to cumulatively count the number of pulses output by the pre-circuit, and get n0; at the same time, in the microcontroller program (the counter of the microcontroller can also be used), count the number of pulses converted by the analog-to-digital converter. Cumulative counting, n is obtained; when the set measurement time arrives, the difference between the two is the number of pulses lost by the system (i.e. n0-n). At this time, the system continues to work. When the analog-to-digital converter converts n0- After n pulses, the system stops working and displays that the measurement time is up, thereby realizing compensation for dead time loss.

    Obviously, this pulse number compensation method is also suitable for the compensation and correction of extended dead time.

Hardware design

    Figure 1 shows the control circuit of the system, which can also be regarded as a buffer circuit for the input signal. It is used to identify the rising edge of the signal and provide peak information of the input signal. Among them: capacitor C1 is the holding capacitor of the signal peak value; diode D1 ensures that C1 collects the input signal peak value to prevent it from changing with the decline of the input signal; U1:A is the signal start judgment comparator, and its reverse end is connected to a The reference voltage is 0.5V, the positive phase terminal is connected to the input signal; U1:B is the signal peak amplitude judgment comparator, the reverse terminal is connected to the capacitor C1, and the positive phase terminal is connected to the input signal. During the rising edge stage of the signal, the voltage at the negative terminal is It is 0.3V lower than the positive-phase terminal voltage; the waveforms of the output terminals A, B and C of the two comparators are shown in Figure 2. Point C provides signal rising edge information, which is used to control the subsequent peak hold circuit; U1:A is actually a threshold voltage discrimination comparator. Signals less than 0.5V are considered noise. Point C does not provide control information, and the ADC does not pre-conversion, so the output terminal A of U1:A can be used as the input terminal of the total signal count and connected to the T1 timer of the microcontroller (ie, P3.5 pin).

Software design

    The design of microcontroller software related to dead time loss compensation mainly involves the following modules:

    (1) Timing module. The T0 timer of the microcontroller is used to implement system measurement timing. This module is actually the overflow interrupt service program of the timer T0.

    (2) Signal total count accumulation module. The T1 timer of the microcontroller is used to realize the total counting of system signals. This module is actually the overflow interrupt service program of the counter T1.

    (3) ADC conversion pulse number accumulation module. Interrupt query mode is used for A/D conversion. The status port STATUS of ADC is used as the external interrupt source of INT0 of the microcontroller. Every time an A/D conversion is completed, the external interrupt 0 service program reads the conversion result of ADC and writes it into the memory. , at the same time, the number of pulses converted by the ADC is increased by 1. It can be seen that the pulse number accumulation module for ADC conversion is actually part of the external interrupt 0 service program.

    The software implementation flow chart related to the calculation of dead time is shown in Figure 3. The console sends the measurement time Ts to the pulse amplitude analysis system and starts the measurement. The pulses from the pre-circuit enter the analysis system, the timer T0 starts timing (assuming the timing time is Tn), and the counter T1 starts recording the total number of pulses entering the system. (n0), every time the ADC converts a pulse, n increases by 1. When the measurement time reaches Tn=Ts, the counter T1 stops counting. n0 at this time is the total number of pulses entering the analysis system within the set measurement time Ts. If n0 at this time is equal to the number of pulses converted by the ADC If n is equal, it means that there is no peak loss in the system and the dead time is 0; otherwise, the ADC needs to convert n0-n pulses to compensate for the lost pulses.

Conclusion

    The peak amplitude analysis system in the CD-10 portable X-fluorescence full spectrum measuring instrument developed by Chengdu Institute of Technology initially used the dead time calculation method in the traditional sense. Since the timing reference pulse period (0.25μs) is different from each ADC conversion The time (8μs) is in the same order of magnitude and other reasons, making the dead time calculation inaccurate; after adopting the new dead time calculation method, the peak amplitude analysis system can accurately reflect the dead time of the system, improving the accuracy of the instrument. This method is also suitable for other signal peak amplitude analysis occasions and has great promotion and application value.