Fast-reacting optical encoder feedback system for miniature motor-driven applications

Recently, motor controls have been miniaturised, enabling new applications in surgical robotics for healthcare and in drones for aerospace and defence. The challenge for designers is to meet the high accuracy requirement of the position feedback sensor in a high-speed application, while infusing all components into the limited PCB space to be fit inside tiny enclosures, such as a robotic arm.

Motor control

Motor control loops, as seen in Figure 1, are mainly made up of a motor, a controller and a position-feedback interface. The motor turns a rotating shaft that causes the arms of a machine to move accordingly. The motor
controller tells the motor when to apply force, stop, or continue rotating. Meanwhile, the position interface in the loop provides rotational speed and position information to the controller. This data is central to the proper operation of a pick-and-place machine for the assembly of a tiny surface-mount PCB. All these applications require accurate position measurement information about the rotating object.

The position-sensor resolution must be very high – enough to detect accurately the motor shaft position, and pick up and place a tiny component on a board. Also, higher motor rotational speeds lead to higher loop bandwidth and lower latency requirements.

Position-feedback system


In a lower-end application, an incremental sensor along with a comparator may suffice for position sensing, while a higher-end application will require more complex signal chains. These feedback systems comprise the position sensor, followed by analogue front-end signal conditioning, the ADC and its driver – before data gets into the digital domain. One of the most precise position sensors is the optical encoder. An optical encoder is composed of an LED light source, a marked disc attached to the motor shaft, and a photodetector. The disc features a masked pattern of opaque and transparent areas that obscure the light or allow it to
pass through. The photodetectors sense the resulting light and the on/off light signals are converted to electric signals.

As the disc turns, the photodetectors – in conjunction with the patterns of the disc – produce small sine and cosine signals, in the mV or µV level. This system is typical in an absolute position optical encoder. These signals are fed to an analogue signal conditioning circuitry, usually consisting of a discrete amplifier or an analogue PGA to gain the signal up to the 1V p-p range – commonly to fit an ADC input voltage range for maximum dynamic range. Each of the amplified
sine and cosine signals are then acquired by a simultaneous sampling ADC’s driver amplifier.

The ADC must feature simultaneous sampling on its channels, such that the sine and cosine data points are taken at the exact same point in time, as that combination provides the shaft position information. The ADC conversion results are passed to an ASIC or microcontroller. The motor controller queries the encoder position every PWM cycle and uses this data to drive the motor based on the instructions it receives. In the past, system designers would have to trade ADC speed or channel count to fit restrictive board footprints.

Read the full article in the September issue of DPA.