Hall-effect sensor technology has advanced and now offers accuracy, consistency, reliability and new feature sets at reasonable cost. The sensors are used in high-volume automotive applications where low cost, quality, reliability and the ability to withstand harsh environments are key factors. Hall-effect devices have become the preferred technology for many critical safety and performance applications that involve the sensing of motion, position, speed, direction, proximity and electrical current.
The basic element in a Hall-effect device comprises a small sheet of semiconductor material. When a constant voltage source is applied to the element, it forces a constant bias current to flow in it. The output takes the form of a voltage, which can be measured across the width of the sheet.
On its own, this voltage has negligible strength, but if the biased Hall element is placed in a magnetic field with flux lines at right angles to the Hall current, the voltage output is amplified, becoming directly proportional to the strength of the magnetic field. This principle was discovered by E.F. Hall in 1879 and remains the basis of all linear Hall-effect IC devices today.
Using modern semiconductor manufacturing techniques and circuits, the basic Hall element can be augmented by adding a voltage regulator to provide a stable power source over a range of input voltages and an amplifier to increase the signal. The combination of these elements is the building block in most practical Hall applications, including linear Hall devices.
This building block can be combined with a Schmitt trigger threshold detector with built-in hysteresis (a mechanism that turns on and off depending on a predefined level of voltage or magnetic field strength), and either an open-collector N-P-N or an open-drain MOSFET output transistor to create a digital switch.
The transistor used is normally a saturated switch that shorts the output terminal to ground whenever the applied flux density is higher than a set upper limit, known as the operate trip point. When the magnetic field falls below the trip point by a certain margin, referred to as the release point, the trigger provides a clean transition from on to off, without contact bounce or chatter.
The defined, built-in hysteresis eliminates oscillation (spurious switching of the output) by introducing a magnetic dead zone in which switching is disabled after the threshold value is passed.
This type of switch is typically compatible with all digital logic families, when used with a pull-up resistor. The output transistor can typically sink enough current to drive many loads directly, including relays, triacs, SCRs, LEDs and lamps. Such a circuit is usually limited to 24V and 25mA. For inductive loads, such as relays, an external flyback diode is usually required.
Switching higher voltages or currents usually demands an additional relay, or a discrete power device such as a bipolar or MOSFET transistor, an SCR, or a triac with biasing resistors.
Hall-effect switches are classified according to their mode of operation in various magnetic fields. These classifications are based on the magnetic operate and release characteristics.
For example, unipolar switches operate and release with respect to the south pole while omnipolar switches operate and release with respect to either the south or north pole. Then there are bipolar latches that operate with respect to the south pole and release with respect to the north pole, whereas bipolar switches operate in three alternative modes, all typically attempting to switch as close to zero gauss as possible. These modes are: unipolar mode which operates and releases with respect to the south pole; latch mode which operates with respect to the south pole and releases with respect to the north pole; and negative unipolar mode which operates and releases with respect to the north pole.
Each of these magnetic switching characteristics is important, depending on the application requirements. For example, in appliance applications, the unipolar or omnipolar modes of operation should meet the vast majority of user needs.
Early Hall-effect sensor designs used a single Hall element or plate that is susceptible to both thermal and mechanical stresses, causing output voltage to be inconsistent due to changes in temperature, pressure and mechanical stress. To address this dependence, recent designs use a four-plate Hall-element array that can be considered as a resistor array similar to a Wheatstone bridge. The quadratic array places four Hall plates in parallel, providing a 'mechanically averaged' Hall voltage.
Offset errors and mechanical stresses normally cancel each other out, although not entirely. A tenfold performance improvement can be realised in both stability and stress immunity by using this quadratic element scheme.
Most Hall effect sensors are now designed using a 'chopped' Hall plate. Terms such as 'chopper stabilised' or 'dynamic offset cancellation' are used to describe this function, which again uses a single Hall plate. In this scenario, a four-terminal element is 'chopped' (that is, electrically rotated) at a high frequency (typically 100kHz to 500kHz), depending upon the sensor function and the manufacturer.
This technique has resulted in a superior and stable device that minimises the effects of thermal and mechanical stress and effectively eliminates offset and mechanical stress error.
MINIMISING CURRENT CONSUMPTION
A typical Hall-effect switch requires from 3mA to 8mA of supply current to operate properly. For some applications, this current draw is too high. However, a low average power scheme can be employed, in which internal timing circuitry activates the sensor for a very short time (60µs) and deactivates it for the remainder of the period (240µs or 60ms, depending on the device). The result, at 3.0V to 5.5V, is a typical average current draw in the range from 5µA to 11µA when a 60ms period is employed, and from 295µA to 460µA when a 240µs period is employed.
The short 60µs 'wake' time allows for stabilisation prior to sensor sampling and data latching, which occurs on the falling edge of the timing pulse. During the 'sleep' time, the output is latched in the last sampled state, and the device supply current is not affected by the output state.
The ability to operate on a north pole or a south pole (omnipolar) is advantageous in several applications because it obviates problems with orientation of the magnet. Omnipolarity is accomplished by using dual comparators after the amplifier, but before the Schmitt trigger.
Various Hall-effect devices now come with short-circuit protection, to prevent overcurrent conditions from damaging the device and its outputs. The device monitors the current at the output and protects itself by turning off, or folding back, the voltage that is allowed to the load. Thermal protection features turn the device off and on to limit current, until the resulting junction temperature returns to a safe operating level. Short-circuit protection provides greater robustness in comparison with earlier products, while improving reliability, attained at little to no additional cost.
Reverse voltage or 'biasing' of a semiconductor can cause failures and catastrophic damage. The latest Hall devices include an internal diode to block voltage from being applied in the wrong direction. This reverse voltage protection also results in greater robustness and reliability than was previously possible at little to no additional cost.
It is advantageous to trim critical parameters on the Hall-effect device to improve the accuracy or the performance of the system for many applications. Previously, accurate switch points, quiescent voltages, sensitivity and other parameter improvements have been achieved by a combination of extreme attention to detail regarding tolerances, as well as exacting specification of the mechanical aspects of the system, the magnet structures and the Hall devices themselves. This approach led to increased system and development cost and greater complexity.
The development of programmable Hall-effect devices that enable users to set the critical parameters means a far more relaxed approach can be taken to mechanical, magnetic and semiconductor device parameters, all of which can be optimised for the system and application during final assembly. The programmable device is usually more expensive than a non-programmable device, but as the cost of semiconductor technology keeps falling (despite increasing circuit complexity), the price differential is becoming less of a concern.
BENEFITS AND APPLICATIONS
Hall-effect devices will never completely replace mechanical switches, but they do offer significant benefits in many applications. Their major advantage over other switch technologies is that they offer contactless, bounce-free switching which virtually eliminates failures induced by physical wear and tear. Nor are they affected by dirt, dust or other environmental factors normally associated with the harsh conditions encountered in the automotive and industrial sectors.
The latest developments outlined in this article point the way forward to more versatile and higher-performance Hall devices that will find use in yet more diverse application sectors.