FPGA-Based Data Acquisition System for Muon Scattering Tomography

This paper presents the development and validation of a scalable, multi-channel FPGA-based data acquisition system featuring NINO ASIC front-end electronics and a 500MHz sampling rate, which successfully demonstrates accurate two-dimensional muon position tracking for non-destructive muon scattering tomography applications.

The Gist

Imagine you have a giant, sealed treasure chest, but you can't open it without ruining the contents. You want to know what's inside: Is it gold? Is it lead? Is it just empty space?

This is the problem scientists face when trying to inspect large objects like nuclear waste containers, ancient pyramids, or even the structural integrity of a volcano. They can't cut them open. Instead, they use Muons.

The Invisible Rain

Think of Muons as tiny, ghostly raindrops that are constantly falling from space. Unlike normal rain, these "muon drops" are super-penetrating. They can pass right through mountains and thick steel walls. However, when they hit something heavy or dense (like lead or gold), they get bumped off course, scattering slightly. When they hit something light (like wood or air), they barely change direction.

Muon Scattering Tomography (MST) is the art of tracking these ghostly drops. By placing detectors above and below an object, scientists can see how much the muons "bounced" as they passed through. If they bounced a lot, there's something heavy inside. If they went straight through, it's likely empty or light.

The Problem: Too Many Eyes

To see this clearly, you need a lot of "eyes" (detectors) watching the muons. The more eyes you have, the sharper the picture. But here's the catch:

• If you want a high-resolution image, you need thousands of tiny sensors.
• Each sensor sends a signal when a muon hits it.
• Connecting thousands of sensors to a computer is like trying to plug 10,000 USB drives into a single laptop at once. It gets messy, expensive, and slow.

The Solution: The "Smart Switchboard"

This paper introduces a new, clever way to handle this data traffic. The authors built a system that acts like a super-fast, smart switchboard for these muon signals.

Here is how their system works, broken down into simple parts:

1. The Front-End: The "Translator" (NINO Chip)

Imagine the muon hitting a detector and shouting a message in a language only the detector understands (an analog electrical signal).

• The NINO chip is like a super-fast translator. It listens to the detector's shout, converts it into a digital "Yes/No" signal (called LVDS), and measures exactly how long the shout lasted.
• This chip is tiny, uses very little power, and is incredibly fast. It's the first step in turning a chaotic electrical spark into clean data.

2. The Back-End: The "Traffic Cop" (FPGA Board)

Once the translator (NINO) speaks, the FPGA board (a reprogrammable computer chip) acts as the Traffic Cop.

• The Speed: This cop is incredibly fast. It checks for muon hits 500 million times every second (500 MHz). That's like checking a stoplight 500 million times a second to ensure no car gets through without being counted.
• The Timing: It has a "trigger window." Think of it like a camera shutter. When a muon is detected, the cop opens the shutter for a tiny fraction of a second (256 nanoseconds) to snap a picture of the data, then closes it immediately. This ensures it doesn't get overwhelmed by noise.
• The Delivery: Once the data is snapped, the cop sends it to a computer via a standard USB cable (using a protocol called UART), just like sending an email.

3. The Magic Trick: Scalability

The coolest part of this system is that it's modular.

• Imagine you have a small room with 8 sensors. You use one "Traffic Cop" board.
• Now, imagine you need to monitor a whole warehouse with 64 sensors. Instead of building a giant, custom, expensive machine, you just plug in more of these same boards.
• The authors showed that you can link these boards together in a "Master-Slave" team. One board tells the others, "Hey, a muon just hit! Everyone, send your data!" It's like a conductor leading an orchestra of identical musicians. You can expand the system from 8 sensors to 64, or even more, just by adding more boards without rewriting the whole software.

Why Does This Matter?

Before this, building a system to track muons for large objects was expensive and required custom-made electronics that took years to design.

This team proved that you can use off-the-shelf development boards (like the ones hobbyists use for robotics) to build a professional-grade scientific instrument.

• It's cheaper: No need for custom circuit boards.
• It's faster: They can process data at lightning speed.
• It's flexible: If you need more sensors tomorrow, you just add another board.

The Bottom Line

The authors built a "camera" for invisible particles. By using a smart translator chip and a super-fast traffic cop board, they created a system that can take detailed 3D pictures of the inside of objects without ever touching them. And the best part? It's a system that can grow as big as you need it to, making it a perfect tool for inspecting everything from ancient ruins to nuclear waste.

Technical Summary

1. Problem Statement

Muon Scattering Tomography (MST) is a non-destructive imaging technique used to inspect the internal structure of large objects (e.g., civil structures, geological formations) by tracking the scattering of cosmic ray muons.

• The Challenge: Accurate MST requires detectors with high position resolution (hundreds of micrometers) and large area coverage. This necessitates a high number of readout channels with fine granularity (strips of a few millimeters or less).
• The Bottleneck: Traditional readout systems struggle to handle the massive channel counts required for large-scale experiments cost-effectively. There is a critical need for a scalable, cost-efficient, and high-speed Data Acquisition (DAQ) system capable of processing signals from gaseous detectors like Resistive Plate Chambers (RPCs) without compromising timing or spatial resolution.

2. Methodology

The authors developed a multi-channel DAQ system comprising two main stages: Front-End Electronics (FEE) and Back-End Electronics (BEE).

A. System Architecture

• Front-End Electronics (FEE):

  • Utilizes the NINO ASIC (Application-Specific Integrated Circuit), a low-power, ultra-fast amplifier-discriminator originally developed for the ALICE experiment.
  • Function: Converts analog signals from RPC readout strips into Low Voltage Differential Signaling (LVDS) pulses.
  • Key Feature: The pulse width of the output is proportional to the input charge, providing Time-Over-Threshold (TOT) information, which aids in precise timing and charge measurement.
  • Configuration: Uses a board developed by the INO Collaboration operating at ±4V.

• Back-End Electronics (BEE):

  • Built around an Intel/Altera MAX-10 FPGA development board.
  • Signal Processing:
    • Directly acquires LVDS signals from the FEE via dedicated I/O pins.
    • Uses a custom connector board with 100Ω termination resistors to maintain signal integrity.
    • Implements a custom VHDL IP core containing a digital delay module, controller, FIFO memory, and UART transmitter.
  • Timing & Sampling:
    • Operates with a 500 MHz sampling clock (generated by an onboard PLL), achieving a 2 ns timing resolution.
    • Applies a 128 ns digital delay to align incoming signals with a 256 ns trigger window.
  • Data Transmission: Transmits processed data to a PC via a UART interface (using a CH340G USB-UART chip) for offline analysis using Python.

B. Experimental Setup

• Detector: A single-gap glass RPC prototype (2mm × 30cm × 30cm) with a gas mixture of 95% Freon and 5% Isobutane.
• Readout: Two orthogonal readout panels (8 copper strips each) to capture 2D (X, Y) position data.
• Trigger System: Validated using plastic scintillators (SCN1, SCN2, SCN3) in various coincidence configurations (e.g., SCN1 & SCN2 & SCN3) to ensure valid muon events are captured.

3. Key Contributions

1. Direct LVDS Acquisition: The system successfully implements direct acquisition and processing of LVDS signals from the NINO ASIC using an off-the-shelf FPGA board, eliminating the need for complex intermediate analog-to-digital conversion stages.
2. High-Speed FPGA Implementation: Utilization of the MAX-10 FPGA allows for a 500 MHz sampling frequency, providing a 2 ns timing resolution, which is sufficient for measuring TOT pulses and reconstructing muon trajectories.
3. Scalability: The architecture is designed for modularity. The authors demonstrated a Master-Slave configuration where multiple FPGA boards can be linked. This allows the system to scale from an 8×8 channel configuration to 16×16 or larger with minimal software changes.
4. Cost-Effectiveness: By using commercially available development boards (MAX-10) rather than custom-designed PCBs, the authors significantly reduced development time and costs while maintaining performance comparable to custom solutions.

4. Results

• Signal Acquisition: The system accurately acquired digital outputs corresponding to analog RPC signals. Comparison with oscilloscope data confirmed the fidelity of the DAQ system.
• Timing Resolution: The system achieved a 2 ns timing resolution, validating the 500 MHz sampling capability.
• Trigger Validation: Different trigger logic conditions were tested and successfully validated. The system correctly identified muon events and generated 2D histograms of muon hit distributions.
• Scalability Test: A Master-Slave configuration was tested with two FEE boards. The system successfully aggregated data from the slave board and transmitted it to the PC, proving the feasibility of expanding the channel count for larger tomography setups.
• Efficiency: The RPC strip detection efficiency was analyzed, showing performance consistent with standard electronics, with minor variations attributed to voltage and threshold settings.

5. Significance

This work presents a robust, scalable, and economical solution for the data acquisition challenges inherent in Muon Scattering Tomography.

• Application Impact: The system enables the construction of large-scale MST detectors necessary for inspecting civil structures, geological surveys, and archaeological sites.
• Technical Advancement: It demonstrates that high-performance, high-channel-count DAQ systems can be built using standard FPGA development boards, lowering the barrier to entry for MST research.
• Future Outlook: The modular nature of the design allows for easy expansion to finer readout granularities (e.g., 1 cm strips), making it a viable platform for next-generation non-destructive evaluation techniques.