A machine cutting a metal rod, producing sparks.

What's our Position on Position Sensing?

In the previous blog in this series, we explored the current control loop and the current and voltage sensing circuits that provide the feedback variables. This blog post explores the other important sensed variable in many drives – motor position sensing and feedback.

 Servo Drive Architecture Diagram

Figure 1: Servo Drive Architecture Diagram

The blocks under consideration are highlighted once again in the architecture diagram shown above. The majority of these blocks do not actually reside within the motor drive itself, but rather inside a position feedback device known as an encoder, that is typically integrated with the motor itself, or added on as a separate device that is closely coupled with the motor. The purpose of the encoder is to measure in real time the changing position of the motor rotor (the part that rotates), and in the case of permanent magnet or synchronous motors, the position of the rotor magnetic field as well as its purely mechanical position.

Another blog series from ADI, commencing with this blog post, covers encoders in detail, and examines the various technologies and electronic signal chains that can be used inside encoders to detect position, from magnetic field-based measurement to optical-based sensing.  The difference between incremental and absolute position feedback is outlined, as well as some recent trends in encoder technology. Instead of going over these details again here, this post addresses some specific matters related to position feedback as pertains to the drive itself.

Is position feedback always needed?

It must be pointed out that many lower performance motor drives can operate without position feedback when connected to induction motors. This is due to the fact that the magnetic fields within an induction motor are induced (hence its name) by the applied stator voltage and so the motor rotation frequency will automatically follow the applied three-phase voltage frequency (with a small differential known as slip). In these open loop drives, no encoder is attached to the motor and no position feedback is used in the controller.

What function does position feedback fulfill in the drive?

In permanent magnet/BLDC motor servo drives, position feedback is required for two reasons. Firstly, position feedback is needed in order to synchronize the motor magnetic field position with the applied motor voltages and currents from the servo drive. Motor torque is developed by the interaction of the rotor magnetic field and the magnetic field generated by the application of three-phase currents in the stator windings. These fields must be in synchronism with each other or useful torque will not be generated.  In order for maximum torque per Ampere of motor current, in fact, these magnetic fields must be orthogonal to each other in space, as well as being synchronized in time (this is known as field oriented control, FOC).

The second function fulfilled by position sensing is as the feedback variable in a position control or velocity control loop, so that the motor speed or position can be accurately controlled. Motor position control is important in applications such as robotics, pick-and-place, and other dynamic applications. Motor velocity control is also important in applications such as machining, conveyors, rollers, winding etc. For velocity control, the velocity feedback value is generally derived as the derivative of the position feedback.  These uses of the position feedback signal in a motor drive controller are shown in Figure 2, where the drive has an outer position control loop and inner velocity and current control loops.

 How Motor Position Feedback is used in a Drive

Figure 2: How Motor Position Feedback is used in a Drive

How is position feedback brought into the motor drive from the encoder?

The majority of encoders used in industrial applications utilize either square-wave incremental encoder feedback (known as ABZ or ABI) or digital protocol-based feedback. So most servo drives need to be capable of interfacing to either of these signal types. Some encoders also provide 3 square wave signals from magnetic Hall sensors embedded in the motor windings that indicate the absolute position of the rotor magnet fields at low resolution. Digital protocol-based feedback utilizes specific protocols such as EnDat [1], Hiperface [2], BiSS [3]. Most of these are vendor-specific and are based on RS-422/RS-485 as the physical layer. In all cases, robustness to Electrostatic Discharge (ESD)/ Electrical Fast Transients (EFT) and noise-induced common mode voltages, and the need to potentially interface to long cables is a requirement. Some part suggestions for these different interfaces are shown in the Table below.

Feedback Type Drive Interface Requirement Part Suggestions
ABZ Standard digital I/O – may need to be differential, e.g. RS-422 MAX14890
Hall Standard digital I/O Any standard buffer IC e.g.MC74AC125
Protocol Robust RS-485, up to ~30Mbps, generally not isolated, but sometimes can be

ADM3065E 

MAX22500E (features for long cables)

The next blog post will look at how both of the control loops discussed in this and the previous blog - position and current control - are brought together and implemented inside the overall drive controller electronics.

References

[1] https://www.heidenhain.us/addl-materials/enews/stories_1012/EnDat.pdf

[2] https://cdn.sick.com/media/docs/5/65/865/operating_instructions_specification_hiperface%C2%AE_motor_feedback_protocol_en_im0064865.pdf

[3] https://biss-interface.com/c/downloads/biss-interface/