By Dr. Chiming Zhao, Senior Product and Marketing Manager, iSentek
The advent of 5G communication is driving the need for base station antenna azimuth perception. On the one hand, this is to meet the directional angle accuracy requirements of Massive MIMO antennas; on the other hand, as 5G base station deployments increase massively, relying on real-time remote monitoring or online optimization to reduce operation and maintenance costs becomes essential. Antenna interface standards are catering to the latest AISG 3.0 specification, officially incorporating orientation perception requirements as a new antenna specification.
Common technologies for antenna azimuth perception include e-compasses, dual GPS, and sum-and-difference beams. Among them, the e-compass is the mainstream solution chosen by antenna manufacturers due to its small footprint (making integrated antenna design easier), relatively low cost, and real-time measurement capabilities. However, because geomagnetic signals are susceptible to interference from the antenna system itself, fixed structural mechanics, and environmental magnetic materials, extra attention is required during implementation to ensure directional accuracy. The following details iSentek's experience in supporting antenna manufacturers with e-compass applications.
Optimization of E-Compass Pointing Capability for Antenna Applications
For antenna engineers designing antenna applications, the optimization of the e-compass can be approached from four aspects: sensors, hard iron interference, soft iron interference, and calibration methods:
1. Choose an Accurate Geomagnetic Sensor Less Susceptible to Environmental Interference: Besides basic sensitivity and noise, the parameters that should be focused on are hysteresis and temperature drift.
Sensitivity: This is the ability to measure the amount of magnetic field change; generally, the higher the sensitivity, the better. iSentek's IST8308 e-compass for antennas has a sensitivity of 1,320 LSB/Gauss, while the earth's magnetic field strength is about 0.25~0.6 Gauss, meaning the sensor's resolution can divide the earth's magnetic field into 330~792 units. Sensitivity must be looked at alongside the noise value. If the sensitivity is very high but the noise is even higher, then when encountering minute output changes, the system cannot determine whether it is an azimuth change transmitted by the sensor or a noise output.
Hysteresis: The residual magnetism phenomenon that occurs after a magnetic material approaches a magnetic field and then leaves is called hysteresis. iSentek's e-compasses use Anisotropic Magnetoresistive (AMR) technology and feature built-in degaussing circuits. Even when subjected to strong magnetic interference, the output will revert to near its original state after the interference leaves. Figure 1 shows that the IST8308's hysteresis error is close to 0 Gauss after being subjected to a strong magnetic field of nearly 20 Gauss, and the maximum residual magnetic field after leaving is under 38mG, whereas a comparative product's maximum hysteresis is 342mG. The general earth magnetic field strength is around 350mG. If there is a 38mG residual magnetic field in the system that hasn't been calibrated, the maximum error in the angle measurement output will be about 6.5 degrees. Comparing this to the comparative product's 342mG residual magnetic field, it will cause an error of about 78 degrees.
Temperature Drift: This refers to the measurement error generated by changes in the sensor's operating temperature. Base station antennas operate outdoors, so sensors need high temperature stability. Temperature drift characteristics can be divided into zero-point drift and sensitivity drift. Figure 2 shows that the IST8308's zero-point drift is controlled within -40 to 80 degrees, with a maximum of under 10mG, and the sensitivity drift is under 2%. Choosing an e-compass with appropriate parameters can guarantee stable and accurate pointing capabilities across different operating environments and temperatures, and normal pointing capabilities can be restored after a sudden magnetic shock. For base station antenna applications that cannot be calibrated frequently, magnetic hysteresis performance is extremely important.
2. Avoid Hard Iron Interference: This can be categorized as originating from the antenna system and the external environment. Hard iron interference that may occur within the antenna system includes permanent magnetism retained by the inductor or wireway after being magnetized, motors, or magnets within solenoids. DC traces can use a twisted-pair layout to mutually offset magnetic fields and reduce the impact. Pay special attention when designing specific timing operation traces, such as lines with large current variations, and keep them far away from the sensor. To evaluate the specific magnetic field size, simulation software such as COMSOL, ANSYS, or MAXWELL can be used. External environmental sources include brackets and the environment, which can be evaluated using high-precision Gauss meters or high-precision magnetic sensors. If the sensor's magnetic hysteresis performance is poor, the antenna will suffer a strong magnetic field shock when moving close to strong magnetic sources like magnets or motors, causing permanent offset of the magnetic components and resulting in hard iron interference. Therefore, antennas must not be close to strong magnetic sources during transport. iSentek provides simulation assessments for antenna clients, performing evaluations early in the design phase to assist clients in altering hardware design and enhancing pointing capability.
3. Avoid Soft Iron Interference: This can also be divided into the antenna system and the external environment. Soft iron interference that may occur within the antenna system includes large filtering capacitors and inductors near the main magnetic sensor, and iron-cobalt-nickel materials inside the antenna or outer casing, such as galvanized iron sheets. Antenna manufacturers commonly using fiberglass or plastic ABS casings can significantly reduce soft iron interference. Minor components within the system can be avoided via layout, followed by calibration. Evaluation measurements for the external environment must be strictly conducted. It is recommended that antenna manufacturers add a specific measurement process during installation to determine the level of external environmental influence and assess whether the current environment is viable before proceeding with the calibration process.
Figure 1: IST8308 Hysteresis Performance Comparison Figure 2: IST8308 Temperature Drift Performance: Zero-point Drift and Sensitivity Drift Figure 3: Magnetic Field Simulation of Hard Iron Effect (Area with lower than 1 Gauss interference)
System Calibration Process
4. Plan an Appropriate Calibration Process: Depending on the antenna system's soft iron and hard iron interference risks, the following three procedures can be selected:
Circuit Board Level Calibration: If the antenna casing uses non-magnetic materials, the mechanics and installation do not use magnetic materials, the circuit board has undergone proper evaluation, and potential hard/soft iron interference mainly comes from non-circuit board components, then circuit board level calibration can be performed. This involves designing a fixture to rotate the entire circuit board and collect data in a uniform magnetic field in 3D space, and then using a calibration algorithm to correct it to a sphere.
System Calibration: If the antenna casing uses magnetic materials, such as galvanized iron sheets, and magnetic materials are used during mechanical design and installation, system-level calibration must be conducted due to the impact of these magnetic materials. At this time, circuit board level calibration only minimizes the impact on the circuit board itself, and special attention must be paid during transport to avoid proximity to strong magnetic sources to prevent hysteresis effects.
On-site Installation Calibration: The main thing to note is whether there are iron-cobalt-nickel materials in the fixed bracket components. If there are, it will cause a soft iron effect. In this case, the farther the geomagnetic sensor is from the fixed bracket, the better.
In addition, since the MCU computing capability and storage space on the antenna system are both limited, it may not be suitable to run the calibration algorithm locally. In this scenario, coordinate data points need to be collected and sent to an external computer for calculation, and then the calibration results are sent back and stored in the antenna system. iSentek also has clients who plan to read the values and transmit them to a cloud server to run calculations before sending the calibration parameters back to enable real-time calibration.
