This article reviews ring encoders and frameless resolvers, compares their merits, and presents an alternate approach. It’s especially suitable for engineers designing gimbals and electrical actuation systems for high-performance motion control.
Ring encoders and frameless resolvers are generally the engineer’s first choice for accurate angle measurement of large diameter shafts. As soon as a shaft or through bore diameter is larger than 2 or 3 inches, the choice of angle sensor narrows and cost increases sharply. This can lead a few engineers to ‘indirect’ arrangements using a small shaft encoder, driven off a large shaft between pulleys or gears. Such arrangements can work well if requirements for dynamic or precision motion are modest but high-performance motion control will require direct measurements from a hollow shaft or large bore angle sensor.
A resolver, whose axial height is small compared to its diameter, can be referred to as either a slab resolver, a frameless resolver, or a pancake resolver. Strictly speaking, ‘frameless’ typically means that the resolver’s housing has been eliminated but many engineers will resort to using the term frameless when referring to a resolver with big diameter and low height.
Figure 1. A frameless resolver with low axial height and large diameter.
In recent times, most resolvers are brushless instead of brushed, but all resolvers are based on transformer principles – that is, they are inductive angle sensors. As the position of a resolver’s rotor differs relative to its stator, the electromagnetic coupling between the stator and rotor differs. This can be seen as the resolver’s output signals differ with respect to the input signal or excitation.
Some resolvers are labeled ‘single speed’, ‘two speed’, ‘four speed’, etc. This denotes the number of times that the resolver’s output uniquely varies over 1 revolution. A single speed resolver’s output is unique over 1 rev; a two speed resolver’s output is unique over any 180 degrees within 1 rev; a four speed resolver’s output is unique over any 90 degrees within 1 rev and so on.
Resolvers have an outstanding track record in safety-related applications, particularly in civil aerospace. There are numerous reasons including non-contact operation, ‘friendly’ failure modes, and insensitivity to environmental conditions.
In numerous safety-critical or safety-related applications, the most dangerous type of failure is not one which results in no output signal but rather one which creates a credible but wrong output signal. For example, consider an aircraft aileron control – a wrong but credible angle measurement could have disastrous ramifications. Given the operation and construction of a resolver, the probability of a failure which generates a credible but wrong signal is significantly small.
Resolvers will be the first choice for most of the older generation engineers in the defense, aerospace, and oil & gas sectors because they are aware of the resolvers’ track record as well as the corresponding AC analog electronics. Some knowledge of analog electronics is needed to specify and select the electronics needed to power a resolver and decode its signals into a digital format. Younger generation engineers are more acquainted with digital electronics so they will struggle to decide definitely a suitable resolver.
The perceived complexity of analogue electronics has been an important factor in the demise of resolvers since the 1990s. This has led to a consistent decrease in the number of companies producing resolvers and an increase in the number of companies producing optical encoders. Perhaps surprisingly, this reduction in the availability of resolvers has led to an unequal increase in their cost and of large diameter resolvers in particular. Lead times of >6 months are common for high specification resolvers in the defense and aerospace sectors.
Advantages: Robust, reliability, accuracy (in multi-speed arrangements), and wide environmental operating envelope.
Disadvantages: High cost, bulky, heavy, complicated electrical interface, tight installation tolerances, and limited availability.
Ring encoders are also known as large through shaft encoders or large hollow bore encoders. Similar to frameless resolvers – all such terms denote an encoder whose axial height is small compared to its diameter.
Figure 2. A ring encoder with low axial height and large diameter.
An angle ‘encoder’ is strictly described as a device which translates (i.e. encodes) angle or change in angle into an electrical signal. Although this may be true, the term ‘encoder’ is thought by many engineers to denote an optical angle sensor that generates a digital signal, most usually in the form of pulses. While this is incorrect, nonetheless, most large bore encoders do use an opto-sensor as the basis of their measurement. Generally, a read head shines light on to an optical grating arranged around the periphery of a ring. The grating alters the light reflected back into the read head’s opto-sensor and the pattern of light indicates angle or change in angle of the ring relative to the read head.
Modern printing or etching technology can generate microscopic features repeatedly on the optical grating. In turn, this allows optical encoders to offer theoretically high angular measurement performance. This sensitivity acts as both a benefit and disadvantage. The benefit comes from the optical sensor’s ability to measure small changes in angle—in other words, high-resolution measurement. The disadvantage comes from the potential for dirt or foreign matter (notably water and condensation in particular) interfering with the optical signal and causing a misread. The sensitive nature of high-resolution opto-sensors generally also restricts their environmental envelope. They are usually not adapted to applications with extreme temperatures or high shock and vibration.
More practically, the high measurement performance mentioned in the datasheet needs such tight installation tolerances that the claimed accuracy figures are unattainable unless two read heads are employed (and paid for). This is frequently not realized by many engineers until the angle encoder is compared against a higher accuracy calibration system.
Advantages: Widely available, accuracy (in arrangements with multiple read heads), easy (digital) electrical interface.
Disadvantages: Tight installation tolerances, sensitivity to dirt and moisture, restricted limits for temperature, shock, and vibration.
Inductive encoders—or more generally ‘incoders’—use the same basic physics as resolvers but present the same digital electrical outputs as an optical encoder. This means they present the same reliability and robustness as a resolver but with an easy-to-use, electrical interface.
Figure 3. A large bore, low height inductive encoder.
Unlike resolvers, all the electronics needed for operation are inside the incoder’s stator. This means that the electrical interface is usually a low voltage DC supply which generates a digital data output representing absolute angle or change in angle.
Unlike a ring encoder, the incoder’s measurement is not just made at one point but rather across the full planar faces of stator and rotor. This means that incoders are much less prone to inaccuracies from non-concentric rotation, thus making their installation fairly easy.
Incoders are especially appropriate in applications such as turntables and gimbals where low axial height is needed as well as a large through bore.
Incoders are available in incremental and absolute formats with high-resolution digital outputs (up to 4 million counts per rev) such as SSI, SPI, BiSS-C and A/B pulses. Some devices are also available as replacements for potentiometers with 0–5 V and 0–10 V outputs.
Advantages: Easy to specify and use, high accuracy without tight tolerance installation, proven track record, insensitive to dirt and moisture, less expensive than optical encoders or resolvers of equivalent performance.
Disadvantages: Higher cost than the least expensive optical encoders.