Time of Flight Measurement System using Time to Digital Converter (TDC7200) Mehul J. Gosavi 1, Rushikesh L. Paropkari 1, Namrata S. Gaikwad 1, S. R Dugad 2, C. S. Garde 1, P.G. Gawande 1, R. A. Shukla 2 1 Department of Electronics and Telecommunication Engineering, Vishwakarma Institute of Information and Technology, PUNE-411048 2 Tata Institute of Fundamental Research, Mumbai - 400005 Abstract:- Silicon Photo-Multiplier (SiPM) is a sensitive photon detector which is being aimed as replacement to conventionally used vacuum Photo Multiplier Tubes (PMTs), because of its numerous advantages. SiPM is used in many high energy physics and nuclear physics experiments as a secondary detector coupled to a scintillator which is primary detector for many high energy radiations. The SiPM signal strength and signal arrival time measurement with high accuracy and resolution are extremely important for correct physical interpretation of data. We have developed a data acquisition system for arrival time measurement of the SiPM signal (pulse) using Time to Digital Converter (TDC). The system is designed around commercially available TDC7200 integrated circuit (IC) and PIC18F87J50 microcontroller. The microcontroller was programmed to configure and readout the TDC IC over high speed SPI line and log the read-out data to PC over UART. TDC data acquisition system was tested with calibrated delay inputs to characterize different modes of operation thoroughly. The collected data was analysed statistically using ROOT framework and the measurement accuracy of about 150 ps has been demonstrated. Here we discuss the methodology of measuring arrival time of the SiPM signal to sub 150 ps accuracy, development of data acquisition system and its performance validation in this article. Keywords:- TDC, SiPM, SPI, UART. I. INTRODUCTION :- a) Introduction to SiPM SiPM is an array of Avalanche Photo Diodes (APDs), all connected together. The device is operated in reverse bias to bias the APDs in Geiger mode. The incident photon causes avalanche breakdown to produce high gain. Thus SiPM is highly sensitive photon detector. Silicon Photo- Multiplier has many advantages over conventionally used photo multiplier tube (PMT) like high gain(~10 6 ), low bias voltage(30-100v), higher photon detection efficiency [1]. SiPM can even detect extremely low light flux [2]. At cosmic ray experiment GRAPES-3 PMTs are used to convert scintillation light to electric signals [1]. But due to advantages offered by SiPM, PMTs are being planned to be replaced by SiPM [1]. Electric signals from SiPM contain valuable information about the primary high energy particle detected in scintillator. Timing information of 252
arrival of the signal is very important parameter for analysis of the cosmic rays. The arrival time is measured with respect to a system trigger. To measure time difference with such high precision TDC7200 can be used. Before STOP pulse reaches TDC it is preconditioned using amplifier, splitter, and discriminator. Typical system block diagram is shown in figure 1. Fig 1 [6] : Block diagram of system to collect data of high energy particles. b) Introduction to TDC [4] TDC performs the function of a stopwatch and measures the elapsed time between a start pulse and stop pulse. The analog output of SiPM is preconditioned using high speed amplifier and digitized using discriminator then given to TDC7200. TDC7200 has range of 12ns to 8ms and resolution of 55ps. In TDC7200 two counters are used to measure time: Course counter and Clock counter. The Coarse Counter counts the number of times the ring oscillator (core time measurement mechanism of TDC7200) wraps. The Clock Counter counts the number of integer clock cycles between START and STOP events. There are two modes of measurements, mode 1 and mode 2. The time difference between the START and STOP pulse is regarded as Time of Flight (ToF). Mode 1:- Mode 1 is used to measure ToF from 12 ns to 500 ns. To measure time in mode1 internal ring oscillator and coarse counter is used. That is why this mode is recommended for time measurements of <500ns. Mode 2:- Mode 2 is used to measure ToF from 250 ns to 8ms. In mode 2 clock counter is used whereas coarse counter is used to measure fractional part of measurement. Due to clock counter this is recommended for measurements with high ToF. To test the system we have measured ToF from 5 ns to 750 ns using mode 1 and 250 ns to 5 ms using mode 2. 253
II. SYTEM DESIGN As shown in figure 1 we have three major parts of this system TDC, microcontroller PIC18F87J50 and PC.. Fig 2 : Block diagram of data acquisition system Communication between PC and TDC happens through PIC. Here PIC acts as Communication Bridge between PC and TDC7200. PIC configures TDC and reads data from TDC using SPI protocol. Open ttyusb Configure the Port for UART Send command and register data to PIC Wait to read data from PIC and store the data in file After receiving required data from PIC calculate ToF and store it in file. Fig 3. Algorithm for PC code Algorithm of PC software is shown in Figure 3. As shown in figure first we configure a USB port of PC for UART communication (Operating system used in PC is LINUX). PC is sends the command to PIC through FTDI232 using UART protocol. Command contains information about selection of mode and stop pulse selection and address of register in TDC. After sending command PC waits to receive data from PIC. When PIC sends measurement data to PC it calculates ToF and stores data in a file so it can be used for further analysis. PIC reads registers of TDC which store the time measurements. ToF is calculated from these register values using formulae given in datasheet of TDC7200. Formula varies according to the mode used for measurement. 254
Initialize the PIC Configure SPI (SSPCON and SSPSTAT register) [5] Configure UART (set baud rate and select asynchronous mode Wait for Command form PC Configure the TDC registers as per specifications and received address (PIC is writing data in registers of TDC through SPI) Wait for the TDC to complete measurement and then read data from TDC registers (PIC is reading data from registers of TDC through SPI) Receive data for TDC through SPI and send it to PC through UART Fig 4. Algorithm for PIC code 255
Fig 4 shows the algorithm and instruction flow of the micro-controller firmware written to communicate with PC and TDC7200. First PIC receives the address of TDC7200 registers and their configurations from user (PC software) as shown in Figure 3. Then PIC configures registers of TDC which are addressed by PC using specification received. After proper configuration of the TDC7200, the measurement is initiated with arrival of START pulse and measurement is completed with arrival of STOP pulse. PIC waits till the measurement is completed, which is signalled over an interrupt line by the TDC7200. Once the measurement is completed PIC reads the TDC registers in which the measured time information is store. After reading PIC sends these register values to PC through UART protocol. These register values are used by PC to calculate ToF. As can be seen in algorithm PIC uses both UART (to communicate with PC) and SPI (to communicate with SPI) protocol for communication while PC use SPI to communicate with SPI. Using this type of communication system we were able to take 400 measurements per second. Collected event data is stored into an ASCII file and then analysed with analysis program developed using ROOT framework. Testing methodology has been described in following section. III. TESTING METHODOLOGY:- Block diagram of the test setup is shown in figure 5. START and STOP pulses are generated from an Arbitrary Function Generator (AFG3202) which provides user defined calibrated delay between two output channels. These outputs are fed to respective inputs of TDC. The STOP pulse has been generated in such a way that it has specific delay (set delay) as compared to START pulse. Fig 5: Block diagram of testing setup We are measuring this delay using our system. This set delay is varied step by step to test the system. Minimum delay measured is 5 ns while maximum delay measured is 4ms. Then configure the TDC7200 using PC, TDC7200 measures the delay after completion of measurement PIC sends register values to PC. PC calculates time of flight using these register values. To analyse the measured values of time intervals we have used open source software ROOT developed by CERN in LINUX. A program is written in C to extract the data from file in which measurements of TDC7200 are stored and then they are plotted in ROOT. To test the system extensively we took samples over large range of time starting from 5 ns to 5 ms. For each delay setting at least 1000 events were collected and histogrammed. The plotted histogram was fitted with Gaussian function to get mean and standard deviation (sigma). Mean and sigma for each delay was further plotted for final analysis. One such histogram for mode 1 and mode 2 measurements have been shown in figure 6 and figure 7 respectively. 256
Fig 6: Histogram for mode 1, set delay of 200 ns We measured ToF for 1001 times for constant delay. Here we used function generator to set the delay. Then histogram was plotted for these 1001 readings. As shown in histogram of Figure 6 the measurement is taken for 1001 times for same set delay of 200 ns in mode 1. Mean of the histogram is 199.9 which is very close to set delay. As can be seen in histogram the measurements of TDC varies from 199.8 ns to 200.2 ns which means that variation of readings is in the range of ±0.2 ns that is very low. Fig 7: Histogram for mode 2, set delay of 1000 ns Figure 7 shows histogram for mode 2. For this histogram the set delay is 1000 ns, measurement is carried out 1001 times. Mean for 1001 readings came out to be 999.9 as shown in figure 7. Also in mode 2 variations inmeasurement is ±0.2 ns. But for mode 2 sharp bar occurs at 999.9 ns which implies that mode 2 has very high consistency. IV. RESULTS AND DISCUSSIONS:- We have tested this system for both modes mode 1 and mode 2. Below given are graphs of measured ToF vs Set delay in ns and graphs of Error vs Set delay. Error is the difference between measured ToF vs Set delay in ns. 257
Fig 8: Measured delay vs Set delay(ns) Figure 8 is the plot of time interval between start and stop pulse given by function generator (Set delay) vs the time interval measured by TDC7200 (measured delay mean) for mode 1. The measurement was fitted with straight line with p1 being slope and p0 is y intercept. Linear behaviour with very small deviation was observed. As y intercept is very close to zero offset error in measurements is negligible. The delay for modes are decided according to the ranges of modes as given in datasheet of TDC7200. Fig 9. Error vs set delay in ns Figure 9 is plot of error (ns) vs set delay. Error is difference between measured value of ToF and expected value (with straight line fit) at each set delay. As can be seen in graph the error varies within the range of ±150 ps in the measurement range of 5 ns to 750 ns (our range of interest). Though preferred measurement range of mode 1 is 12 ns to 500 ns we have tested mode 1 of TDC7200 up to 750 ns. 258
Fig 10: Set delay vs measured delay. Figure 10 is plot between time interval given between two pulses using function generator and time interval measured by TDC7200 in mode 2. Also for mode 2 like mode 1 the measurements were fitted in straight line with slope 1 (p1 being slope and p0 is y intercept). In mode 2 also linear behaviour with very small deviation was observed. In mode 2 the offset error is negligible considering wide range of measurements.. Fig 11: error vs set delay in ns. Figure 11 is plot of error (ns) vs set delay. Error is difference between measured value of ToF and expected value (with straight line fit) at each set delay. As can be seen in graph the error varies within the range of -0.08 ns to 0.06 ns in the measurement range of 250 ns to 5 ms. By observing both graphs it can be concluded that error is not more than 150 ps in any mode. Maximum speed to take measurements achieved was 450 Hz that is 450 measurements per seconds. V. CONCLUSION:- SiPM is becoming a versatile photo-readout element for experiments in high energy physics like GRAPES-3. Arrival time of photon on SiPM is of very much importance and for high resolution timing measurement different types of TDC are widely used. We have demonstrated capabilities of commercially available TDC7200 TDC IC with extensive characterization. The developed data acquisition system using TDC7200 has provided timing measurement accuracy within 150 ps, which is required for timing measurements using SiPM. The data rate capability up to 450 Hz was achieved. Furthermore building upon developed expertise, the system will be updated to support more readout channels. ACKNOWLEDGEMENT We express our gratitude towards VIIT and TIFR for giving us opportunity to work on this project. We take this opportunity to show our gratitude to our seniors, Mr. Ravi Kesharwani and Mr. Akshay Manjare for their and 259
kind suggestions and constant encouragement. Their dynamism has always inspired us. Without them this would have been tougher journey. We thank Mr. A. S. Naik and Mr. N. M. Bhoj for their support and timely guidance. We are happy to express our gratitude to Prof. Dr. B.S. Karkare, Principal, VIIT for her appreciation and Prof. Dr. Y. H. Dandawate, HOD of Department of Electronics and Telecommunication, VIIT for his constant motivation. VI. REFERENCES:- [1] Multi-channel programmable power supply with temperature compensation for silicon sensors, R.A. Shukla, V. G. Achanta, S.R. Dugad, J. Freeman, C.S. Garde, S.K. Gupta, P.D. Khandekar, A.M. Kurup, S.S. Lokhandwala, S. Los, S.S. Prabhu. [2] Development of Micron-Resolution Optical Scanner for Silicon Detectors,, S.R. Dugad, S.P. Duttagupta, C.S. Garde, A.V. Gopal, S.K. Gupta, A.M. Kurup, S.S. Lokhandwala, S.S. Prabhu, R.A. Shukla. [3] J. Christensen, HPTDC High Performance Time to Digital Converter. [4] Datasheet of TDC7200. ( http://www.ti.com/product/tdc7200/datasheet) [5] Datasheet of PIC18F87J50. (http://www.microchip.com/wwwproducts/en/en027172) [6] K. Patil, Q. Jawadwala and F. C. Shu, "Design and Construction of Electronic Aid for Visually Impaired People," in IEEE Transactions on Human-Machine Systems, vol. PP, no. 99, pp. 1-11. doi: 10.1109/THMS.2018.2799588 260