Position Sensors Enter the
New low-cost sensors are moving magnetostrictive technology from the factory floor to your vehicle.
Jesse L.J. Russell, MTS Systems Corp.
Photo 1. Automated manufacturing techniques have enabled the development of a line of low-cost magnetostrictive sensors that are proving attractive to the automotive market. From right to left are the fully integrated sensor package, the remote electronics package, and the basic sensor for customer-produced electronics.
The automotive industry, and, specifically, shock absorber manufacturers, have long been attracted by magnetostrictive sensors’ noncontact design, ability to be mounted internally, high performance, absolute measurement, and adaptable output. Until recently, however, these devices were assigned to factory automation applications because they are built to order in relatively small batches, holding the cost structure high.
Thanks to changes in its manufacturing process, MTS Systems has developed a lower cost line of magnetostrictive sensors (see Photo 1) to meet the automotive market’s needs (see the sidebar “Lowering Manufacturing Costs”).
The term magnetostriction describes the tendency of some materials to change shape, constrict, or expand in the presence of a magnetic field. Normally, a material’s magnetic domains are randomly oriented. If a magnetic field is applied, those domains will align, causing a change in shape or lengthening (see Figure 1).
Figure 1. Magnetic fields cause magnetostrictive materials to change shape (H = magnetic field intensity).
A magnetostrictive position sensor takes advantage of this effect to induce a mechanical wave or strain pulse in a specially designed magnetostrictive wire called a waveguide. The time-of-flight of this pulse is measured and can be equated to distance because the speed of traverse is very constant and repeatable. The pulse is created by momentarily causing an interaction between two magnetic fields (see Figure 2).
Figure 2. The interaction of magnetic fields causes the waveguide to twist.
One field originates from a permanent magnet that passes along the outside of the sensor tube. The other field, encompassing the entire waveguide, is created when a current interrogation pulse is applied to the waveguide.
At the interaction point between these two magnetic fields, which is the current magnet position, a strain pulse is produced. The strain pulse, or wave, travels at the speed of sound in the waveguide alloy (~3.54 µs/cm of travel) along the waveguide until the pulse is detected at the head or coil end of the sensor (see Figure 3).
Figure 3. By measuring the amount of time between the electronic pulse's launch and the strain pulse's arrival, a magnetostrictive sensor can determine the precise position of the moving magnet.
The position of the moving magnet is precisely determined by measuring the elapsed time between the launching of the electronic pulse and the arrival of the strain pulse. Noncontact position sensing is thus achieved with absolutely no wear to any of the sensing elements.
Figure 4. Because each magnet along the sensor generates a pulse, a single sensor can monitor a shaft's linear and rotational motion. Either two absolute position measurements or a differential measurement can be made.
The strain pulse is small, on the order of 20–30 µ, resulting in virtually no fatigue of the waveguide. Interrogation intervals are determined by the length of the sensor, but typical rates are 1–3 kHz.
Because a return pulse will be generated for each magnet located along the sensor, multiple marker-magnet sensors can be designed to allow two positions to be monitored with one sensor. This way, differential measurements can be taken or two absolute positions can be monitored. Sensors from MTS use this feature to monitor both the linear motion of a shaft and its rotational motion with a single sensor (see Figure 4). One application for this is in automotive manual transmissions where, as it goes through the H pattern, a shift shaft’s linear and rotary action are automated to emulate an automatic transmission.
Current and Future Applications
Custom magnetostriction position sensors are currently used in several automotive applications, including the Mercedes S class Automatic Body Control system (ABC system) and Carrera racing shocks (see the sidebar “CSE Sensors Hit the Road”). Other markets include medical equipment and appliances. These new custom and standard magnetostrictive sensors reduce implementation costs dramatically in high volume. And they offer design engineers an alternative sensor with noncontact durability, versatile packaging, and output implementation choices.
Lowering Manufacturing Costs
Photo 2. Switching from a monolithic to a modular structure enabled faster design and production of new models. The basic sensor element became the foundation for a line of lower cost products.
In the industrial magnetostrictive sensor market there is a rising demand for more variety in sensor models and higher performance in the form of smart sensors. Despite cost reduction efforts on the part of magnetostrictive sensor manufacturers, however, industrial model prices produced in quantities of 10’s still range from several hundred dollars on up. This precludes their use in high-volume applications.
In the mid 90s, MTS switched from a monolithic structure to a modular structure in its products (see Photo 2). Though the purpose was to speed the design and production of new models, the basic sensor element became a foundation strategy for lower cost products.
Photo 3. Production of the CSE magnetostrictive position sensor element is automated, reducing manufacturing costs even for short-run orders.
Enter the CSE
In 1998 MTS brought on line a new basic sensor element module (CSE) and the automation necessary to produce it at low cost. The CSE was designed for automatic production on a transfer line (see Photo 3). Components are placed in bins at appropriate stations, each of which selects one and either performs a dedicated assembly function or does gluing or welding, and testing. The result is a process that can produce a basic sensor element module every 17 s.
Orders are typically in 2500- to 10,000-unit quantities, allowing one machine setup cost to be spread over the cost of many sensors. Because the quantities are high, parts are cheaper to produce. Product quality is ensured because each station does the same job the same way every time, and each station has an automated QA check.
These low-cost CSE sensor element modules are combined with high-volume electronics modules and packaging to form the final assembly at targeted costs for high-volume custom applications and new standard products. MTS frequently works with automotive packaging suppliers to complete custom assemblies, or the customer can take on that role.
Photo 4. The new line of commercial CSP sensors (lower right) are much smaller overall than their industrial cousins (top left), as well less expensive in higher volumes (2500-50,000 units/yr.).
Automotive and appliance applications normally range from 50,000 to hundreds of thousands of units per year. Because the technology is very malleable, the individual requirements of each application can be accommodated in a custom assembly and still keep the price down.
Low Cost for Smaller Commercial Production Runs
Customized packaging and electronics can be frequently justified by the customer requiring volume ranges of 50,000 units/yr. and above because any R&D costs can be amortized over a large quantity. But that’s not so true in quantities of 2500–30,000/yr. To reach this portion of the commercial market, MTS combined its CSE basic sensor element module with standard electronics and wrapped it all with a low-cost package to create a line of standard complete sensors (see Photo 4). By standardizing the packaging and electronics, R&D funds for unique assemblies are not necessary and R&D lead times are reduced.
CSE Sensors Hit the Road
The first implementation of the new CSE-based line of sensors was the Mercedes S class Automatic Body Control system (ABC system). Each corner has a strut equipped with an integrated hydraulic actuator and a CSE-based custom sensor (see Photo 5).
Photo 5. The Mercedes ABC strut is a hydraulic cylinder sitting on top of the damper. A CSE-based sensor measures ride height.
In this case, the electronics were placed along the axis of the sensor inside the rod. The actuator controls preload and thus the ride height at that corner.
Microcomputers calculate the amount of pressure and the duration applied to each spring depending on the information received by the magnetostrictive sensors. The driver can set the suspension for a sporty ride or a softer setting for greater comfort. Beyond this, the main advantage is the safety advantage that the system provides: the ABC system stabilizes the body of the car within milliseconds. The roll angle, for example, is reduced by 68% when this system is used.
Each cylinder can move at frequencies up to 5 Hz in reaction to vibrations. The system also includes an automatic self-leveling system based on the load of the car. In addition, the driver can change the height of the car by 25 or 50 mm at low speeds. At speeds of 140 kph or greater, the car body automatically sinks an additional
10 mm to reduce the air resistance.
Photo 6. Carrera's new MagneShock uses a CSE sensor to measure stroke. Position information is used by a controller to change the damping characteristics.
Carrera Racing Shocks give more control and adjustment to the drivers and suspension tuners, which translates into better track performance over changing conditions. To achieve this, Carrera uses two innovations in its new double-adjustable MagneShock models. The first is a magnetorheological fluid that changes viscosity when exposed to a magnetic field. Carrera uses this property to dynamically change the damping rates of the shock. The second is the Temposonics CSE that measures shock and motion and feeds the data back to the controller.
A coil within the shock’s piston provides the magnetic field for the magnetorheological fluid. By modulating the coil, the rate of flow can be controlled. This effectively replaces the valves found in conventional shocks. Gun-drilling the piston rod for the coil wires to enter provides a convenient place for the CSE sensor to be located. The head is in the specially designed upper mounting eye, along with some buffer electronics. The body extends down the inside of the rod. A magnet positioned at the end of the tube where the rod exits the shock activates the sensor. All the wires for both the coils and the sensor exit at the upper eye (see Photo 6).