Today’s sensor technologies for robotic and medical applications include many devices that have evolved from industrial applications. Because of this general migration, they are not ideally suited to the requirements in new robotic and medical products. Specifically, conventional contacting and non-contacting sensing solutions do not offer the customization, small package size(s), and environmental and durability requirements that are so necessary in these custom applications.
Shaft angle sensors have matured in the industrial marketplace over the past 10 years, especially with the increasing penetration of absolute position sensors, which broadcast the true position (shaft angle) at power on. These sensors do not need supplemental electronics or initial motion to trigger an index pulse to determine true position. Their suitability in medical and robotic applications has met with mixed success. When these devices are selected for use in medical equipment or robots, the inherent weaknesses of each technology becomes more pronounced. Some examples include:
- Existing sensor package is too large to fit in the medical device or robotic subassembly
- Durability of the industrial device is not matched to the medical or robotic specification
- The industrial specified device offers only incremental position information
- The industrial device is ‘over spec’d’ and cannot meet cost targets in these new designs
Potentiometers are the most mature technology under consideration and have two key weaknesses. When used in a shaft angle sensing (SAS) applications, the termination points of the resistive track do not allow 360° of position feedback. The practical limit is approximately 340°. This is due to the need for terminating points at the ends of the resistive tracks, creating a ‘dead zone’ (no feedback) in this region. This can be overcome with gear reduction (achieving 360° of mechanical travel and feedback). However, this adds substantially to the design in terms of cost, part count, and mechanical hysteresis.
Second, the electrical contact has to maintain very tight dimensional repeatability over its entire mechanical travel. Large mechanical tolerances between these two create unstable conditions for the electrical contact, resulting in noise, excessive wear, or open circuit conditions.
Incremental and Absolute Encoders
Incremental devices are the simplest form of an optical encoder — a device that “increments” from one pulse to another. This technology requires that the counting be done upstream in the external electronics. Incremental devices cannot keep their true position when the power is turned off. When the power is turned back on, the device must be rotated to an index pulse to determine its true position.
Both types of optical encoders have three key design advantages: 1) the ability to resolve to very high resolutions within one rotation; 2) very fast response times that make them ideal in high-speed feedback applications; and 3) they are relatively immune to axial play on the shaft. The code disc can float up or down, and the available light hitting the photodiode remains in position. However, these optical devices have very little tolerance for radial play as the available light moves off of the target (photo-receptive) area of the diode. For low-speed SAS applications, the available resolutions are far greater than are typically necessary for shaft angle position feedback. Lastly, optical devices cannot tolerate dust or dirt on either the light source or the photodiode. This attribute is a strong consideration over the lifetime of a device when plastic bearing features are employed.
GMR Technology Applied in Shaft Angle Sensors
Current Giant Magnetic Resistance (GMR) technology mandates that the magnet rotate centered above the multi-layered sensing element (typically an Integrated Circuit) to detect the orientation of the magnetic field. To enable this technology in through-hole rotary applications, gears and a satellite magnet(s) must be employed. One gear is placed around the shaft that drives a separate gear holding the sensor’s magnet. This results in an indirect measurement. Moreover, when the GMR sensor (a true non-contacting technology) is used with these gears, it becomes a ‘contacting’ solution with the inherent disadvantages and complexities that gear trains imply: such as mechanical play, audible gear noise during mechanical life, irregular torque, hysteresis; many of these characteristics contribute to reductions in absolute linearity.
Hall Effect Sensors
The current 3-axis rotary Hall effect sensors, like GMR technology, require that the magnet rotate directly above the Hall application-specific integrated circuit (ASIC) to accurately detect the changes in the magnetic field. These are typically end-of-shaft applications. To enable this technology in through-hole rotary applications, gears and a satellite magnet must be employed. One gear is placed on the shaft driving the second gear that houses the sensor’s magnet. This results in an indirect measurement of the true position of the steering shaft. Like the GMR sensor, Hall devices are a true non-contacting technology that lose their inherent benefit of contact-free position sensing when gears are employed. This gear train burdens the device with mechanical complexity. It now becomes a contacting solution with the mechanical play, audible gear noise, irregular torque, and hysteresis all contributing to reductions in absolute linearity.
PIHER's PST Technology
PIHER's new proprietary Hall-effect-based technology is called PST: Position Sensor Through-hole. PST provides simplified design characteristics in applications that require through-shaft, 360° absolute feedback. This patented non-contacting technology is designed specifically for through-hole sensing applications.
The key feature in this technology is the ability to sense true absolute 360° position of a shaft using only a ring magnet and one ASIC. This is desirable not only in shaft angle applications, but other rotary/pivot point sensing applications common in robotic arm or medical devices, where direct shaft sensing has been difficult to package.
Using a single bipolar ring and an ASIC, PIHER's PST technology requires that the ring magnet be permanently fastened to the rotating shaft. This ring magnet is magnetized in a simple bipolar North/South manner divided at 0° and 180°. The technology requires that the Hall effect sensing chip be placed anywhere around the 360° orbit of the ring magnet. The technology can be configured as a sector device using an arc magnet, which can be magnetized to provide full-scale output over the sector.
The Hall chip can be placed as close as 0.5 mm (air gap) from ring magnet in a typical 25-mm diameter application. The Hall sensing device is capable of all normal signal processing and error correction providing a stable analog, PWM, or Serial output (up to 14 bit resolution). A key feature of this technology is its tolerance for misalignment, both in terms of axial and radial play.
The proprietary feature of PST is its use of the application's existing bearing assemblies for the rotating shaft. The sensing element is in a fixed position, typically common to the bearing mountings. As the bearings approach their end of life tolerances and create play on the rotating shaft, the PST fixed sensing element accommodates this unstable shaft state, maintaining its original linearity specification. The inherent technology is immune to these unstable shaft conditions up to 0.5 mm of run out.
The density of these two magnetic fields allows for a consistent and repeatable sensing of the fields even when there are changes in radial and axial position. Phrased differently, even with excessive bearing play, the technology can still achieve highly accurate position feedback, typically in the ±1% range that is common in SAS applications.
This characteristic of radial and axial tolerance is a breakthrough in rotary angle sensing design in terms of simplicity and robustness. With only two key components, designers are liberated from the constraints of designing a robust and exotic bearing assembly that will outlast the rotational life of the part. A simple rotor that holds the ring magnet in a loosely concentric position now covers half the design task. Fixing the Hall chip to remain in the target sensing region of the ring magnet is the second task. Given total design control, the engineer can now design a device without a conventional rotor and housing bearing. This creates a larger tolerance in this area as an intentionally designed-in air gap or bearingless sensor, depending on the sealing requirements in the final application.
This technology was done by Piher Sensors and Controls S.A., Tudela, Spain. For more information, visit http://info.hotims.com/40434-185.
- P. Cain et al. Piher Electronic Products Pot vs. Sensor: Choosing the right term assists with the right component selection, November 2010
- I. Tokunaga, H. Okumura, Y. Shonai and T. Ogawa et al. ALPS Electric Co. SAW No. 2007-01-0395
- J. Gruber et al. Robert Bosch GmbH SAE No. 970382
- W. Granig, S. Hartmann et al. Infineon technologies Austria AG SAE No. 2007-01-0397
- C. Mahn et al. Dynapar, PD&D News: Choosing the Right Feedback Option November 2009
- Melexis Microelectronic Integrated Systems MLX91204 Data Sheet
- Paul Rako et al EDN News: Automotive Hall-effect rotary-position sensor uses novel flux-concentrator technology October 2007