When design engineers use the term ‘motion control’, one’s first reaction is to perhaps think of high speed systems with rapid changes in direction and speed. Certainly, high speed, highly dynamic motion control applications have their own technical challenges but this is also true at the other end of the spectrum with very slow speeds. Tight control of slow speed is an important issue for many motion control applications such as industrial test and measurement, specialist automation, weapons systems, telescopes, CCTV cameras, security and surveillance systems to name but a few. This article looks at ultra-slow speed motion control and, in particular, the critical role of position or speed sensors

Let’s consider a long-range camera such as might be used in a gimbal system (sometimes referred to as a pan and tilt system) for border surveillance. Nowadays, it is not uncommon for high zoom optical systems to focus clearly on a target object 20kilometres away. When that distant target object moves, things start to get tricky because the field of view is limited and the target will quickly disappear from view if it is not followed smoothly.


Fig 1 – Pan and tilt system tracking a distant moving target

If the target is moving at say 20km/hr at a range of 20km this equates to a rotational speed of 0.05rpm – in other words, extremely slow. In order to keep the target within the field of view, and preferably central in the field of view, this low speed needs to be controlled accurately and responsively as the target object’s speed and direction changes. To maintain a central position in the field of view, equivalent to say 0.3m at a range of 20km then >200k encoded points are required or 18bits of encoded position. To maintain smooth motion typically 4x this is required, equating to 20bits.

The traditional approach would be to use an encoder on the system’s drive motor and to multiply its counts per rev through a reduction gearbox connected to the motor. The higher the gear reduction, the greater the multiplying effect but also the greater the amount of backlash and the less responsive the overall servo system. Such a system would have limited dynamic range and typically would be unable to track a quickly moving object at modest distances of less than a kilometre as it would lack the required performance at higher rotational speeds.

The alternative, more modern approach is to use a high resolution position encoder on the output shaft of the gearbox – thus avoiding backlash effects and maximising the servo’s dynamic performance. This approach has only recently been more widely adopted because sufficiently high resolution encoders have, until recently, been prohibitively expensive.

Traditional high resolution position encoders

Traditionally, high resolution (>18bits) position sensing has been provided by multi-speed resolvers, precision optical or capacitive encoders. Precision resolvers are notoriously expensive and can present engineers with packaging problems due to their bulk, weight and precision installation tolerances. Similarly, high resolution optical ring or capacitive encoders are expensive and also require precision mechanical installation. Unlike resolvers, which are usually very robust, optical encoders offer limited resilience to shock or vibration (due to their use of glass scales) and their operating temperatures are limited. Optical and capacitive devices both suffer reliability problems from foreign matter such as dust or condensation. High precision optical devices are particularly susceptible to foreign matter because they use fine optical gratings to attain the high number of counts per revolution. Their absolute position output can be thrown by particulates whose dimensions approximate to the features of the optical grating.

New generation inductive encoders

Traditional high resolution position sensing techniques all have their limitations and so design engineers are turning to new type of sensor. These may be thought of as a new generation of inductive encoder which offers especially high resolution measurement of up to 4millions counts per rev (22bits). Inductive encoders (sometimes referred to as ‘incoders’), use the same basic physics as resolvers and in large part this allows them to offer high resolution, non-contact measurement in tough operating environments. Their ability to operate reliably in dirty or wet environments allows the design engineer to eradicate the seals, bushes or O-rings required to protect optical or capacitive encoders.


Fig 2 – Examples of the new generation of inductive encoder

Rather than the traditional wire wound transformer constructions found in resolvers, incoders use printed circuits as their principle components. This provides a further advantage in form factor:- low axial height and a large bore, making it easy for cables, shafts, pipes etc. to pass through the middle of the sensor. Incoders typically offer a simple electrical interface featuring a DC supply and an absolute, digital output.