Lorentz-Force Magnetometers

The market for low-cost MEMS sensors is growing rapidly, driven partially by the exponential growth of smart phones and driver assistance systems. A compass, or magnetometer, is an essential component for providing navigation and location-based services in these devices. While magneto-resistive (MR) and Hall-effect magnetometers are dominant technologies in existing electronic compasses, Lorentz-force magnetometers have some notable advantages including:

  • No requirement for any specialized magnetic material
  • No need for magnetic concentrators to sense fields parallel to the device, and
  • Easy integration with MEMS gyroscopes and accelerometers currently used in consumer electronics

Due to these advantages, there are intensifying R&D efforts to develop practical MEMS-based Lorenz-force magnetometers.  Work in this area is focused on reducing power consumption and noise [1], as well as integration with 9 degree-of-freedom monolithic inertial MEMS [2].

Design Challenges

Sensitivity, noise, hysteresis, linearity, dynamic range, reliability, and yield are all critical parameters for MEMS-based magnetometers. Design engineers need to optimize device performance with respect to these criteria, while also reducing cost, form factor, and time to market.

Lorentz-force magnetometers are multiphysics in nature.  Design of these devices must include an analysis of resonant frequencies, Lorentz force, and capacitive output. The accompanying circuit design typically requires an accurate magnetometer model in order to design the AC current drive control, capacitance readout amplifier, ADC, and output signal processing. System design might also include noise analysis for the sensor and circuitry, and an investigation of temperature stability.

SEM image (left), with the white arrows indicating drive current in the y+ direction and sensing circuitry (right) of a Lorentz-force magnetometer from [1]. Reused with permission of the author.

Fig. 1: SEM image (left), with white arrows indicating drive current in the y+ direction and sensing circuitry (right) of a Lorentz-force magnetometer from [1]. Reused with permission of the author.

Multiphysics Design Methodology

Sensing frequency and Lorentz force application
The magnetometer shown in Figure 1 (referred to as the “UC Davis magnetometer”) uses the micro structure displayed in Figure 2 to sense a two-axis B field. An AC drive current is sent through the structure in the y+ direction.

<em>MEMS</em>+ model of the UC Davis magnetometer (top), and Lorentz force due to interaction of the drive current and Bz (bottom left), and Bx (bottom right) fields.

Fig. 2: MEMS+ model of the UC Davis magnetometer (top), and Lorentz force due to interaction of the drive current and Bz (bottom left), and Bx (bottom right) fields.

To maximize sensitivity, the frequencies of the AC drive current are matched to the mechanical mode frequencies corresponding to x-direction, in-plane motion and z-direction, out-of-plane motion. Results of a modal analysis in MEMS+ are shown in Figure 3.

<em>MEMS</em>+ modal analysis for the Bz sensing in-plane mode at 49.7kHz (left), <br /> and Bx sensing out-of-plane mode at 112kHz (right).

Figure 3. MEMS+ modal analysis for the Bz sensing in-plane mode at 49.7kHz (left),  
and Bx sensing out-of-plane mode at 112kHz (right).

Capacitive sensing of the Bz and Bx fields
Bz sensing is provided by differential capacitance between the x+ and x- electrostatic comb fingers, and Bx sensing is provided by capacitance to a single electrode beneath the micro structure, as shown in Figure 4.

 <em>MEM</em>S+ model showing differential capacitance sensing for Bz (left, only x+ capacitor is highlighted), and single capacitance sensing for Bx via the electrode beneath the micro structure (right).

Figure 4: MEMS+ model showing differential capacitance sensing for Bz (left, only x+ capacitor is highlighted), and single capacitance sensing for Bx via the electrode beneath the micro structure (right).

Sensing results for the Bz and Bx fields are shown in Figure 5. Note that the frequency of the AC current for each sensing field (left) matches the predicted frequency for each sensing mode in the MEMS+ modal analysis (right).

Capacitive sensing output for Bz (left) and Bx (right) predicted by <em>MEMS</em>+

Figure 5. Capacitive sensing output for Bz (left) and Bx (right) predicted by MEMS+

References

[1] Area-Efficient Three Axis MEMS Lorentz Force Magnetometer, V. Rouf et al, IEEE Sensors Journal, Vol 13, 2013.

[2] A Monolithic 9 Degree of Freedom (DOF) Capacitive Inertial MEMS Platform, I. Ocak, et al, IEEE Electron Devices Meeting (IEDM), 2014.