Revolutionary and robust – the MILE encoder

Small, robust and accurate, maxon motor’s inductive micro-encoder is a milestone in encoder technology. Impervious to the influences of EMC, dust and oil, it has a programmable index pulse as well as integrated commutation signals - and is just 6mm in diameter!
Rotary and position encoders are useful precision instruments, but as such, they are very delicate. Today, optical rotary encoders can achieve nanometre precision, but are very sensitive to anything that disturbs the light beam, such as dust, water and oil (including vapours), smog and frost. Such challenges can be overcome, but require expensive sealing measures.
Magnetic encoders, on the other hand, are produced for use in industrial applications in harsh ambient conditions. Depending on their structure, they can easily cope with dirt and water, but integration into a motor exposes their greatest weakness which is sensitivity to magnetic fields. For magnetic-resistant (MR)and Hall sensor based encoders, there are now compensation measures for correcting homogeneous constant fields. However, these measures are not effective enough in strongly inhomogeneous alternating fields which tend to arise in the direct environment of motor connections. In small motors, customers still require motor and encoder connections to be in a single cable, or at least at the same end of the motor. In most cases there is no shielding as this would be too expensive. All in all, this means that we have to live with these interference fields.
To get round the dilemma of optical/magnetic systems, other principles must be taken into consideration, such as capacitance or inductance. Encoders of this nature have already been featured in large motors, but have not been produced in micromotors until now. For this reason, maxon motor has invested in the development of highly miniaturised inductive encoders. The result is the MILE Encoder, maxon’s Inductive Little Encoder, which is the smallest of its kind in the world. Typical areas of application are medical technology, robotics and industrial applications in harsh environments.
How “standard” inductive encoders work
Inductive encoders exploit the inductance changes of one or more coils due to changes in the magnetic configuration, induced by a moving target disk. . For example, a semi-circular iron core (representing the target disk) could be directed past a coil which then changes its inductance. This could, for instance, be measured as the resonance frequency of an LC oscillator. Improved implementations provide differential structures or measure the compensated mutual inductance of a multi-coil configuration rather than the self-inductance (which carries a strongly temperature dependent offset).
Despite advanced signal conditioning measures, there are many drawbacks to simple (iron or ferrite core based) inductive encoders. The temperature dependence of the soft iron/ferrite requires compensation measures, and external magnetic fields can change the permeability of the material used, well below saturation point. Ferrous or ferrite inductive encoders are therefore more robust against magnetic fields compared to magnetic ones, but in strong fields, even these inductive encoders may change their characteristics significantly. Consequently, they are not suitable for fine interpolation unless additional measures are taken. This is why high-precision inductive encoders are ironless, as the contrast is generated with eddy currents.
How an eddy current-based inductive encoder works, as illustrated by the MILE encoder
The MILE is an inductive encoder where the contrast is generated with eddy currents: the target comprises a structure of non-magnetic, good conductive metal, such as copper or aluminium. Upon exposure to a high-frequency field, said field will not be able to penetrate very far. The eddy currents induced by the field create an opposing field, with the effect that the resulting field remains on the surface (skin effect); the field lines must find a way around the target instead of penetrating it. This is how a contrast is generated between air and metal. In order to achieve this effect with a reasonable metal thickness and to ensure that there is an adequate signal in a highly miniaturised system, an excitation frequency in the Megahertz range has been chosen. At this frequency, thin copper layers can be used (100 •m suffice). The chosen frequency lies well-above standard PWM frequencies, which means that there is little interference, even from very strong PWM signals.

Figure 1: How the MILE sensor works
A doubly differential design has been chosen to minimise the dependencies of the unwelcome effects of system temperature and the target disk’s pitch. The excitation coil (marked in red) receives a symmetrical signal, creating a transformer with the internal coils. Both internal coil pairs are connected anti-serially. As the structure is symmetrical, the expected voltages of the individual coils cancel each other out: no target disk, no signal. If a target disk (shown semi-transparently in Figure 1) is placed on the structure but only partially covers the coils, a measurable voltage is established at the differential amplifier’s inputs. The amplitude and phase of this voltage varies depending on the target disk. The system therefore acts as a resolver.
The signal is demodulated, filtered and processed. Ideally the angle of the target disk would be determined using the arc tangent of the demodulated“sine” and “cosine” signals. However, this would lead to interpolation errors from assembly tolerances which result in misalignment and eccentricity of the target disk. A factory programmable look-up-table (LUT) is therefore stored in every sensor to compensate the resulting errors. The result is an almost perfectly linear sensor. As this resolver principle is based on a single-turn absolute encoder, further benefit can be generated using the information gained. This includes a programmable index, a single-step selectable pulse rate and adjustable commutation signals for electronically commutated (brushless dc) motors. Absolute values can also be accessed through a synchronous serial interface (SSI) rather than incremental signals.
The whole unit is contained in a small chip measuring 3.2 x 2.7 x 0.4 mm which is applied onto a flexprint using flip-chip technology. The encoder fits into a 6 mm motor; other combinations with large motors are to follow.

Figure 2: MILE Sensor, flip-chip on flex, chip dimensions 3.2 x 2.7 x 0.4 mm

Figure 3: maxon EC 6 motor with MILE encoder. 64 pulse. 3 channel with line driver. Max. speed 120,000 rpm. Integrated commutation outputs.

Application fields
As the MILE encoder is compact, the sensor is suitable for all applications where space is limited and an accurately controlled drive is required. The high power density of maxon’s motors can be fully exploited and will not be compromised by a bulky encoder that protrudes over the edge of the motor diameter. For example, there is limited space for circuit boards in pick-and-place machines where electronic components measuring less than 1 mm are picked up, conveyed and put in place. The feeders are situated very close to each other. The strap drivers must be millimetre-thin in diameter and have no protruding parts. Utmost precision is required during positioning as the components are minuscule. This also applies to the mounting head which positions and correctly mounts the electronic components with the utmost precision (tolerances of micrometres and less). Increasing miniaturisation requires 3-D drive units in particular to be as compact as possible to prevent restriction of movements. The same applies to camera adjustments in optical systems as well as mini-robots and medical equipment. The MILE encoder is ideal whenever a given diameter has to be strictly observed, such as in highly dynamic hexapod tables where drives are fitted into each of their six feet. The high resolution of the movement achieved by the MILE encoder.
In battery-driven equipment, the MILE encoder’s low power consumption compared to an optical solution is a distinct advantage. If only limited electrical output is available, such as solar cells, batteries etc., MILE encoders are an ideal combination with ironless maxon motors which are highly efficient.

 

Author: MMAG/23.08.2010