Current Sensor Selection

Automation and Control Sensors Transducers
1

How to Select the Right Current Sensor Architecture and Manage Power Density

This document is part of a comprehensive series that provides guidance about current sensor selection. Here are the main points from the article. For more information and to view the article in it’s entirety please use the link below.

Key Takeaways

  1. Current Dictates Architecture: The choice between an integrated, coreless, or core-based architecture is driven primarily by the continuous current and the resulting thermal limits of the system.
  2. Integrated is for Simplicity: For applications that require less than 200 A, an integrated-conductor sensor (like the ACS37220 or ACS71010) provides a factory calibrated, “plug-and-play” solution with the smallest footprint.
  3. Coreless for Maximum Power Density: For measurement of 100 A to 4000+ A within a tight space, a coreless differential sensor offers a lossless, contactless method that eliminates bulky magnetic cores. The tradeoff is mechanical tolerance. A shift in busbar placement can result in a sensing error.
  4. Core-Based for Harsh Environments: If the design requires high accuracy and immunity to external magnetic stray fields at very high current, a core-based solution with a ferromagnetic concentrator is a good solution.
  5. Layout is a Critical Component: Regardless of the architecture, the PCB or busbar design acts as both a magnetic interface and a thermal heat sink. For long term reliability, adherence to the best practices for copper volume and thermal vias is essential.
  6. TMR is the High-Speed Frontier: For applications that require bandwidth up to 5 MHz, an Allegro TMR based integrated sensor offers a significant performance improvement compared to traditional Hall-effect technology. TMR also offers better sensitivity, which can allow for even smaller integrated packages or better performance at lower current levels within a high-current architecture.
Feature Integrated Core-Based Coreless
Typical Current Range 2 A to 400 A 20 A to 1000+ A 100 A to 4000+ A
Sensing Method Current passes through the sensor; all-in-one solution, but has the highest thermal bottleneck Current passes near the part, but the magnetic field is “captured” and “shaped” by a physical core; accurate but bulky Sensor is placed above or near the conductor; physical contact does not occur; bulky core is not present; greatest-power-density and lightest-weight solution for high current
Primary Advantage Smallest footprint; factory calibrated for ease of use High accuracy; inherently shielded from external fields by core Lossless; lowest cost for very high currents; magnetic/core saturation does not occur
Primary Disadvantage Thermally limited by I2R losses in the package Bulky, heavy, costly, and susceptible to magnetic-core saturation/ hysteresis Sensitive to mechanical alignment, PCB/busbar design, and external stray fields
Isolation Type Inherent galvanic isolation; level depends on package High galvanic isolation; determined by core and conductor spacing Inherent (noncontact); depends on air gap/insulation of conductor
Bandwidth High: 100 kHz to 1 MHz (TMR versions up to 5 MHz) Low-medium: limited by eddy-current losses in the core High: limited by sensor
Accuracy/Drift High (factory programmed) High (must account for core thermal drift and hysteresis) System-dependant (system-level calibration for best results)
Power Dissipation Highest: conducts current through internal leadframe Negligible: sensor IC is isolated; core losses are minimal Negligible: IC consumes very little power; current remains in busbar or PCB

From Amps to Kiloamps: How to Select the Right Current Sensor Architecture and Manage Power Density | Allegro MicroSystems