Fundamentals of position feedback

In the world of automation, it is commonplace to find some sort of position feedback on pretty much every design. Often, the feedback is static in the form of a limit or proximity switch, and this suits many applications. However, sometimes we want to know about the places in between the goal posts.

The points along the way can be achieved in several ways. If it is just a few points, we might add additional sensors at appropriate positions but, if those points might change depending on operation, then a more dynamic way to determine position would be appropriate. This might come up when the speed of the movement might require anticipating the position rather than a definite position that is the same, regardless of speed.

It might also come up if the controlled device only has two options—fully open or fully closed, extended or retracted. While the controller might issue a command to open and one might use sensors to indicate when it is closed, half-open and open, for example, ultimately that device is going to go to full open and stop when the commanded travel reaches the physical limit. Converting motion into an electronic signal is achieved by using an encoder.

There are two main types of encoders, absolute and incremental. The output of an absolute encoder represents the current angular position of the shaft. The position is retained during a power cycle. An incremental encoder, on the other hand, provides output based on the relative position of the shaft. With no particular starting point, the incremental encoder presents the motion of the shaft and that value can be used to represent position, speed and distance. It is important to understand that incremental encoders do not retain their value through a power cycle.

Get your subscription to Control Design’s daily newsletter.

One way to convert linear position or displacement is to convert motion into a proportional electrical signal using a linear variable displacement transformer (LVDT). Highly accurate, this device uses a contactless, electromagnetic transducer to convert position into an ac or dc value. This lends itself to applications like a linear cylinder or to measure properties, such as stress, tension and compressibility, in flexible materials but may be limited by length of motion.

A more common way to represent is to monitor the driving device—for example, a motor-driven linear actuator, where we use an encoder mounted to the motor to represent the relationship between motor turns and linear travel of the actuator. In a similar manner, an encoder mounted on a conveyor can translate rotations into linear travel.

On a larger scale, a motor-mounted encoder where the motor has a gearbox can represent the travel of the driven device, like a robot arm or similar device.

Encoders have evolved greatly over the years. I remember early encoders that provided individual wires for each bit in the resolution of the encoder. Code in the PLC would convert the incoming signals into a Gray code integer that could be used for position indication. Patented by Frank Gray, the method ensures that only a single bit can change between consecutive numbers to eliminate spurious results. For example, if the binary signals from an encoder represented 00000010, one would expect 00000011 to be the next number so we would not expect to see 00100011.