⚡ electrical engineering

Architectural Engineering and Mathematical Modeling of the Hybrid Fusion Metal Detector (H-FMD) for Deep-Subsurface Extraction and High-Resolution Target Discrimination

This paper introduces the Hybrid Fusion Metal Detector (H-FMD), a novel system combining high-voltage transient magnetic generation with sub-VLF synchronous pilot-wave injection and advanced DSP processing to achieve high-resolution target discrimination and deep-penetration capabilities (3.0–4.0 m) in highly mineralized soils where traditional VLF and Pulse Induction technologies fail.

Original authors: Salar Farzaneh

Published 2026-06-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Salar Farzaneh

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Problem: The "Depth vs. Detail" Dilemma

Imagine you are trying to find a lost coin buried deep in a muddy field. You have two tools, but both have a major flaw:

The "Flashlight" (Pulse Induction): This tool sends out a super-bright, powerful flash of light. It can see very deep into the mud (up to 4 meters). However, the light is so bright and blurry that it can't tell the difference between a shiny gold coin and a rusty iron nail. It just sees "something metal."
The "Microscope" (Very Low Frequency): This tool uses a gentle, steady beam of light. It is incredibly sharp and can easily tell a gold coin from an iron nail. But, the light is too weak to penetrate deep mud. It stops working after about 1.5 meters, especially if the soil is heavy with minerals (like volcanic rock).

For decades, engineers had to choose: either find deep things without knowing what they are, or know exactly what a shallow thing is without seeing deep.

The Solution: The "Hybrid Fusion" Detective

The author, Salar Farzaneh, has built a new machine called the H-FMD (Hybrid Fusion Metal Detector). Think of this machine as a detective that carries both a super-bright flashlight and a high-powered microscope, but it uses them in a clever, synchronized dance so they don't get in each other's way.

Here is how it works, step-by-step:

1. The "Big Bang" Pulse (The Flashlight)

The machine sends a massive, high-voltage electrical pulse (25 Amps!) through a giant 1-meter square coil. This creates a powerful magnetic "boom" that punches deep into the earth (up to 4 meters).

The Trick: Usually, after a "boom," the machine has to wait a long time for the noise to settle before it can listen again. This H-FMD uses a special "active brake" (like a super-fast airbag) to stop the noise in less than 12 microseconds. This lets it listen almost immediately.

2. The "Whisper" Signal (The Microscope)

While the machine is listening to the "echo" of that big boom, it quietly whispers a steady, low-frequency tone (1.5 kHz) into the ground.

Why this frequency? Most metal detectors use high-pitched whistles that get swallowed by mineral-rich soil. This specific low pitch is like a submarine signal; it can travel through heavy, rocky soil without getting lost.

3. The "Brain" (The Computer)

The machine uses a very fast computer chip (an ARM Cortex-M7) to listen to two things at the same time:

How fast the echo fades: Heavy metals (like gold or copper) keep the echo ringing for a long time. Rusty iron stops ringing very quickly. The computer measures this "ring time" to guess the material.
The angle of the whisper: The computer checks how the low-frequency whisper bounces back. Different metals twist the wave in different directions.

By combining the "ring time" (from the big boom) and the "wave twist" (from the whisper), the machine creates a 2D map. On this map, iron junk clusters in one corner, and precious metals cluster in another. It's like sorting laundry by color and fabric type simultaneously.

The Results: What Did They Find?

The author tested this machine in very difficult, rocky, volcanic soil (the kind that usually confuses detectors). They buried different metal objects at depths between 3 and 4 meters.

Gold and Copper: At 3.5 meters deep, the machine correctly identified these as "precious metals" 94.2% of the time.
Iron: At 3.8 meters deep, it correctly ignored (rejected) iron junk 98.5% of the time.
Bronze: Even at the extreme depth of 4 meters, it correctly identified a bronze vessel 89.1% of the time.

The Bottom Line

This paper claims to have solved a problem that has existed for a long time: You no longer have to choose between depth and accuracy.

The H-FMD is a single device that can dig as deep as the most powerful industrial machines (4 meters) but can still tell the difference between a gold nugget and a piece of scrap iron, something standard deep-search machines cannot do. It does this by using a "hybrid" approach: a powerful pulse to reach deep, and a steady whisper to identify the target, all processed by a fast digital brain.

Technical Summary: Architectural Engineering and Mathematical Modeling of the Hybrid Fusion Metal Detector (H-FMD)

Problem Statement
The paper addresses a fundamental physical limitation in geophysical prospecting: the inability of conventional single-topology metal detectors to simultaneously achieve deep subsurface penetration (3.0–4.0 meters) and high-resolution target discrimination. The authors identify that signal attenuation in deep soil follows an inverse sixth-power law ( $V_r \propto 1/r^6$ ), causing returning voltages to plummet into the thermal noise floor.

Pulse Induction (PI) systems offer deep penetration and soil immunity but lack the ability to distinguish between ferrous and non-ferrous metals due to reliance on time-decay constants alone.
Very Low Frequency (VLF) systems provide precise phase-based discrimination but fail at depth because continuous waves are absorbed by soil mineralization (eddy current losses) and dielectric attenuation, typically capping operational depth at 1.0–1.5 meters in magnetite-rich soils.

Methodology and System Architecture
The proposed solution, the Hybrid Fusion Metal Detector (H-FMD), integrates PI and VLF technologies into a unified dual-domain processing framework. The system utilizes a 1.0×1.0 m square dual-coil search array and an ARM Cortex-M7 (STM32H743VIT6) DSP core.

Hybrid Signal Generation:
- PI Domain: A high-voltage (48V rail), high-current (25 A) transient magnetic pulse is generated via an H-bridge MOSFET stage. An active 500V damping clamp forces the primary coil current to zero in under 12 µs, minimizing the "blind time" and allowing immediate sampling of the decay tail.
- VLF Domain: During the inter-pulse relaxation windows, the system injects a continuous 1.5 kHz sub-VLF pilot wave. The authors mathematically demonstrate that at 1.5 kHz, the skin depth ( $\delta$ ) in heavy magnetite soil ( $\sigma \approx 0.1$ S/m) is approximately 4.1 m, enabling penetration to the 4.0 m target depth where higher frequencies would fail.
Dual-Domain Signal Processing:
- Time-Domain Analysis: The DSP captures voltage decay across eight sampling gates (G1–G8). It calculates a size-independent material signature coefficient, $R_{decay} = V(G_8) / V(G_2)$ . Noble metals sustain eddy currents (high $R_{decay}$ ), while ferrous materials decay rapidly (low $R_{decay}$ ).
- Frequency-Domain Analysis: The system performs synchronous quadrature demodulation on the 1.5 kHz pilot wave to extract the in-phase ( $I$ ) and quadrature ( $Q$ ) components, deriving the true phase shift ( $\Delta\phi$ ).
- Classification: The system fuses $R_{decay}$ and $\Delta\phi$ into a 2D coordinate space to generate a Visual Discrimination Indicator (VDI) ranging from −90 to +90, separating ferrous and non-ferrous clusters more effectively than single-domain methods.

Key Results
Empirical validation was conducted in volcanic magnetite soil (Fe3O4 content >15%) with targets buried between 1.0 and 4.0 meters. The H-FMD achieved the following performance metrics:

Depth and Discrimination: 94.2% accuracy in discriminating non-ferrous structures (e.g., gold, copper) at 3.5 m depth.
Iron Rejection: 98.5% accuracy in rejecting iron targets at 3.8 m depth.
Deep Target Confirmation: 89.1% classification accuracy for a bronze vessel at the 4.0 m boundary.
Signal Integrity: The system maintained detectable signal magnitudes significantly deeper than both VLF-only and PI-only single-topology systems, as illustrated in comparative signal magnitude vs. depth graphs.

Significance and Claims
The paper claims that the H-FMD successfully shatters the historical trade-off between depth and discrimination in electromagnetic prospecting. By synthesizing high-voltage pulse induction with synchronized sub-VLF phase extraction, the H-FMD creates a "dual-domain" classification space that neither PI nor VLF systems can achieve individually.

Theoretical Contribution: The authors provide a rigorous mathematical model (Eqs. 2–7) defining the interaction between time-domain decay constants and frequency-domain phase shifts in deep, mineralized soil.
Practical Impact: The system offers a unified hardware solution that eliminates the need for operators to switch between different detector types for deep versus shallow targets.
Future Outlook: The authors outline a roadmap for integrating machine learning classifiers into the firmware, developing 3D spatial mapping arrays, and miniaturizing the architecture for commercial production, aiming to establish a new standard for deep-subsurface extraction and geophysical instrumentation.