🔬 applied physics

Real-Time 3D Magnetic Field Camera for a Spherical Volume

This paper presents the design, calibration, and performance evaluation of a real-time 3D magnetic field camera utilizing a spherical $t$ -design array of Hall magnetometers to achieve efficient, 10 Hz volumetric measurements with approximately 1% uncertainty.

Original authors: Fynn Foerger, Marija Boberg, Niklas Hackelberg, Philip Heinisch, Katharina Ostaszewski, Jonas Faltinath, Florian Thieben, Fabian Mohn, Paul Jürß, Martin Möddel, Tobias Knopp

Published 2026-02-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Fynn Foerger, Marija Boberg, Niklas Hackelberg, Philip Heinisch, Katharina Ostaszewski, Jonas Faltinath, Florian Thieben, Fabian Mohn, Paul Jürß, Martin Möddel, Tobias Knopp

Original paper licensed under CC BY 4.0 (http://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 Idea: A "Magnetic Camera" for 3D Space

Imagine you want to understand the shape of a wind current inside a room.

The Old Way (Sequential): You hire a single person with a wind meter. They stand in one corner, take a reading, walk to the next spot, take another, and so on. To map the whole room, they have to walk thousands of steps. It takes hours, and by the time they finish, the wind might have changed.
The New Way (This Paper): Instead of one person, you hang 86 wind meters all over the walls, ceiling, and floor at the exact same time. You flip a switch, and instantly, you have a complete 3D map of the wind inside the room.

This paper describes the invention of that "instant" system, but for magnetic fields instead of wind. The team built a Real-Time 3D Magnetic Field Camera.

How It Works: The "Smart Sphere"

1. The Hardware: A Ball of Eyes

The researchers built a hollow sphere (about the size of a grapefruit) made of plastic. On the surface of this ball, they stuck 86 tiny magnetic sensors (called Hall magnetometers).

The Arrangement: They didn't just stick them randomly. They placed them in a very specific, mathematically perfect pattern (called a "spherical t-design"). Think of it like arranging seeds on a sunflower so they cover the surface perfectly without gaps.
The Speed: Because all 86 sensors read the magnetic field at the exact same moment, the system can take a "snapshot" of the magnetic field 10 times every second (10 Hz).

2. The Magic Trick: Math as a Shortcut

You might ask, "If I only measure the surface, how do I know what's happening in the middle?"

The Analogy: Imagine a balloon. If you know exactly how the rubber is stretched on the surface, you can mathematically predict exactly how the air is pushing inside, assuming the air is behaving normally.
The Science: Magnetic fields follow strict rules (Laplace's equation). The team uses a mathematical tool called Spherical Harmonics (a fancy way of saying "3D polynomials"). By measuring the surface with their 86 sensors, they can calculate a single mathematical formula that describes the entire magnetic field inside the ball.
The Result: They don't need to measure every single point inside the ball. They measure the surface, do the math, and instantly know the field strength and direction at any point inside.

Why Was This Hard? (The Calibration)

Building the sphere was easy; making it accurate was hard.

The Problem: When you 3D print a ball and glue 86 sensors onto it, they aren't perfectly placed. One might be tilted 1 degree to the left; another might be slightly higher than the others. If you don't fix this, your "wind map" will be wrong.
The Solution: They created a special "training" process. They put the sphere in a known, perfect magnetic field (like a giant, invisible magnet) and rotated it in different directions. By comparing what the sensors said they saw versus what they should have seen, they calculated a "correction map" for every single sensor. This is like teaching a group of people with slightly different eyesight how to see the world correctly by giving them glasses.

The Results: Speed vs. Accuracy

The team tested their new camera against the old "one-person-walking-around" method.

The Old Method: Took 3 minutes to map the magnetic field. It was accurate but painfully slow.
The New Camera: Took 0.1 seconds (100 milliseconds) to do the exact same job.
The Trade-off: The new camera is about 1,800 times faster.
Accuracy: The new camera was about 99% accurate compared to the old method. The tiny errors (about 1%) were due to things like temperature changes or the fact that the 3D printed ball wasn't a perfect sphere.

Why Does This Matter? (Real World Applications)

Why do we need to see magnetic fields this fast?

Medical Imaging (MPI & MRI): In advanced medical scanners, magnetic fields are used to "paint" images of the body. These fields change rapidly. If the scanner doesn't know exactly what the field looks like right now, the image gets blurry. This camera can watch the field change in real-time, allowing for sharper, faster medical images.
Robotics & Manipulation: Imagine using magnets to move tiny robots inside the human body (like a blood clot remover). You need to know exactly where the magnetic force is pushing at every millisecond. This camera acts as the "eyes" for that control system.
Future Scaling: Because the system is so simple (just a ball of sensors), it can be made bigger or smaller easily. It's a "plug-and-play" solution for any size room.

The Bottom Line

The researchers turned a slow, tedious process of mapping magnetic fields into a high-speed, real-time video feed. By combining a clever arrangement of 86 sensors with some smart math, they created a device that can "see" invisible magnetic forces instantly, opening the door to faster medical scans and more precise magnetic control systems.

1. Problem Statement

Accurate volumetric magnetic field measurements are critical for applications ranging from medical diagnostics (MRI, Magnetic Particle Imaging - MPI) to industrial processes. However, conventional measurement methods face significant limitations:

Sequential Scanning: Traditional approaches use a single 3D sensor moved robotically to grid points within a volume. This is time-consuming (often taking minutes for a 3D volume) and mechanically complex.
Scaling Issues: As the measurement volume increases, the time required for sequential scanning grows prohibitively.
Hardware Complexity: Existing high-speed solutions (e.g., NMR-based field cameras) often require complex electronics, offset fields, or fail to measure the full field vector, limiting their applicability and speed.

The core challenge is to achieve real-time, volumetric magnetic field mapping with high accuracy and low hardware complexity.

2. Methodology

The authors propose a Spherical Magnetometer Array that combines physical sensor arrays with mathematical field reconstruction.

A. Hardware Design

Sensor Array: The device consists of 86 independent 3D Hall magnetometers (Texas Instruments TMAG5273) mounted on the surface of a 3D-printed polyamide sphere (radius $R = 45$ mm).
Geometry: Sensors are positioned according to a spherical $t$ -design ( $t=12$ ). This specific geometric arrangement allows for efficient numerical integration (quadrature) of the field over the sphere's surface.
Electronics: Sensors are connected via flexible printed circuit boards to a central microcontroller. The system uses a shielded USB interface for data transfer and includes temperature sensors for compensation.
Performance: The system samples all sensors simultaneously, achieving an effective data rate of 10 Hz.

B. Theoretical Framework

The method relies on the fact that in regions with zero current density, the magnetic field components satisfy Laplace's equation ( $\nabla \cdot B = 0, \nabla \times B = 0$ ).

Spherical Harmonic Expansion: The magnetic field inside the sphere can be expressed as a series of solid spherical harmonics.
Quadrature: By sampling the field at the $t$ -design positions on the surface, the expansion coefficients can be calculated exactly using a discrete quadrature sum.
Reconstruction: For a $t=12$ design, the system can reconstruct the magnetic field inside the sphere using a 6th-degree polynomial. This allows the field to be evaluated at any point within the volume, not just at the sensor locations.

C. Calibration Strategy

To ensure accuracy, a two-step calibration process is employed:

Local Sensor Calibration: A linear model ( $B_n = R_n b_n + O_n$ ) is used to correct for individual sensor scaling, skew, rotation, and offset. This is achieved by exposing the array to known homogeneous magnetic fields in multiple orientations.
Global Orientation Correction: Since the physical mounting of sensors on the 3D-printed sphere may have slight deviations from the ideal CAD model, a global rotation matrix is calculated. This aligns the theoretical $t$ -design positions with the actual sensor positions derived from the calibration data, minimizing spherical distance errors.

3. Key Contributions

Real-Time Volumetric Mapping: The development of a camera-like device that captures the full 3D magnetic field vector within a 9 cm diameter sphere at 10 Hz, a significant improvement over sequential methods.
Efficient Reconstruction: Demonstrating that a minimal set of 86 sensors, arranged via $t$ -design, can mathematically reconstruct the entire volume field using a 6th-degree polynomial.
Scalable Hardware: A design that avoids the complexity of NMR probes or massive sensor grids, utilizing cost-effective Hall sensors while maintaining high accuracy.
Comprehensive Calibration: A robust calibration pipeline that accounts for both individual sensor errors and the geometric misalignment of the array itself.

4. Results

The system was validated by comparing its output against a sequential single-sensor measurement (using a robot and a Lake Shore 460 sensor) in a magnetic gradient field.

Speed: The magnetometer array captured the data in 100 ms, whereas the sequential robot-based method took approximately 3 minutes. This represents a speedup factor of ~1800.
Accuracy:
- The average deviation between the array and the reference sensor was ~1% (approx. 131 $\mu$ T).
- The maximum deviation was ~4% (403 $\mu$ T).
- Discrepancies were attributed to minor inhomogeneities in the calibration field, temperature drift, Earth's magnetic field (not yet compensated), and slight non-sphericity of the 3D-printed holder.
Dynamic Tracking: The system successfully tracked the trajectory of a Field-Free Point (FFP) moving in an oval path within the sphere at 10 Hz, demonstrating its capability for dynamic field monitoring.

5. Significance and Applications

This work introduces a paradigm shift from point-by-point scanning to simultaneous volumetric capture, enabling new applications:

Magnetic Particle Imaging (MPI): Crucial for real-time planning and characterization of complex multi-patch field sequences and tracking the FFP, which directly impacts image resolution and reconstruction accuracy.
MRI Shimming: Facilitates rapid, iterative magnetic field mapping to correct inhomogeneities, allowing for real-time feedback during the shimming process.
Magnetic Manipulation: Enables precise control and monitoring of magnetic fields used for manipulating particles or medical devices.
Scalability: The approach is inherently scalable to larger volumes or higher sensor counts. Future work aims to extend the technique to ellipsoidal geometries (better suited for cylindrical imaging systems) and increase the frame rate up to 100 Hz.

In summary, the paper presents a highly effective, low-complexity solution for real-time 3D magnetic field mapping, bridging the gap between high-precision measurement and the speed required for dynamic medical and industrial applications.