« Back To Motion Control & Motors News
Heidenhain Corp. Logo

Member Since 2015

LEARN MORE

DR. JOHANNES HEIDENHAIN GmbH is a leading manufacturer of precision measurement and control equipment. Our superior technology is utilized in high-precision motors, motion control and machining systems worldwide. With over 100 years in the precision measurement industry, HEIDENHAIN is setting the future standards of position feedback accuracy today. Our primary industries are metalworking, machine tool, semiconductor and electronics, motor/drive, general automation, and medical, but can be of service anywhere highly dependable precision measurement and motion control is needed.

Content Filed Under:

Industry:
Mobility Mobility

Application:
N/A

Compact, Sturdy Inductive Multiturn Rotary Encoder with Enhanced Control Quality, Expanded Mounting Tolerances, and Functional Safety up to SIL 3

POSTED 05/13/2015

Users of machine tools, robots and general automation technology are constantly demanding solutions that are high in performance and low in cost. The response to this are dynamic, compact and energy-efficient drives with safety functions. Due to the differing requirements of the individual applications, functionally safe encoders are needed both for the high-end range of machine tool drives as well as for standard applications. In this respect, compatibility with previous versions and products, both for mounting and the interfaces, presents a decisive advantage (Figure 1: ECI 1119 / EQI 1131 FS).

With the series launch of their bearingless ExI 11xx rotary encoders with inductive scanning and a size of 37 mm, HEIDENHAIN now offers an expanded portfolio for the realization of highly dynamic and energy-efficient servo drives for safety-related applications.

Latest-generation inductive scanning
Inductive rotary encoders from HEIDENHAIN have become the ideal solution for standard drives in applications for automation technology. The EnDat 2.2 purely serial interface with up to 16 MHz clock frequency can attain very short control cycle times of as little as 32 µs, which is an indispensable prerequisite for dynamic drive control.

Thanks to the new ASIC, a functionally safe solution that needs only one inductively scanning encoder per axis has also become available for 37 mm encoders (the 56 mm inductive rotary encoders of the ExI 13xx Generation 3 series have been available since 2012). The safety integrity level SIL 2, category 3 PL d is supported by the encoder. Additional measures in the control can even make applications with up to SIL 3 or category 4 PL e possible. The mechanical fault exclusion against loosening of the shaft or stator connection is an additional benefit. Moreover, the system accuracy in comparison with previous versions was improved by more than a factor of 2 to ± 120 angular seconds. Thanks to its contamination-tolerant inductive scanning principle, these encoders are characterized by high durability. In order to make the use of the new ECI 1119 FS (singleturn) and EQI 1131 FS (Multiturn) inductive rotary encoders as simple as possible, a mounting design was developed for compatibility with the optically scanned rotary encoders with integral bearings of the ECN/EQN 11xx FS series (Figure 2: Common mounting ExI / ExN 11xx). The mechanical and electrical compatibility made possible with this design ensures scalability according to the control requirements of the application, and the number of motor variants can be minimized.

Compared with the previous generation, the permissible mechanical tolerance in particular has been greatly extended, for example by a factor of 2 for the permissible axial motion. This dramatically simplifies compliance with the installation dimensions for the motor manufacturer and expands the possible applications for these encoders. For simple verification of the mechanical installation quality, the encoder generates a value for the mating dimension that can be read out from the EnDat interface by the servo inverter. A comparison with the nominal value given in the product specifications can then be used to evaluate the quality of the installation. During operation, functionally safe rotary encoders from HEIDENHAIN can also be monitored through the integrated diagnostic functions.

In addition, the new inductive rotary encoders support the monitoring or temperature by the encoder (through an integrated temperature sensor) and motor (through connection of an external temperature sensor. Both temperature values are evaluated by the encoder and can be read our digitally by the subsequent electronics over the EnDat 2.2 without the need to interrupt the control cycle.

Control quality compared with optically scanned encoders
In order to choose the best-suited rotary encoder for a particular application, it is necessary to compare control characteristics. In the following example, a comparison was conducted with a real motor. To keep possible influences of the motor to a minimum we have chosen a model with extremely low cogging torque for the measurements. A high-precision angle encoder was mounted on the output shaft of the motor (measuring accuracy of better than ± 1 angular second) to assess the drive system with regard to accuracy and rotational speed stability.

The comparison used an EQI 1130 of generation 1.2, an EQI 1131 of generation 3, and an EQN 1135 as high-end solution (see Table 1).

Table 1: Specifications of the rotary encoders
EQI 1130 Generation 1.2 EQI 1131 Generation 3 EQN 1135 Generation 2
  • EnDat 2.1 (18-bit singleturn, 12-bit multiturn), ≤ 2 MHz clock
  • No analog signals (purely serial)
  • Without integral bearing
  • Without support for temperature evaluation
  • ± 280” system accuracy
  • Permissible axial motion of measured shaft: ± 0.2 mm
  • Vibration resistance as per EN 60068-2-6: stator: ≤ 30 g, rotor: ≤ 30 g
  • For applications with moderate requirements on control quality and accuracy
  • EnDat 2.2 (19-bit singleturn, 12-bit multiturn), ≤ 16 MHz clock
  • No analog signals (purely serial)
  • Without integral bearing
  • Evaluation of an integral and an external temperature sensor
  • ± 120” system accuracy
  • Permissible axial motion of measured shaft: ± 0.4 mm
  • Vibration resistance as per EN 60068-2-6: stator: ≤ 40 g, rotor: ≤ 60 g
  • Optimal for modern production machines with functional safety
  • EnDat 2.2 (23-bit singleturn, 12-bit multiturn), ≤ 8 MHz clock
  • No analog signals (purely serial)
  • With integral bearing
  • Evaluation of an integral and an external temperature sensor
  • Permissible axial motion of measured shaft: ± 0.5 mm
  • ± 60” system accuracy
  • Vibration resistance as per EN 60068-2-6: ≤ 30 g
  • For high-end drives, for example in machine tools with functional safety

System accuracy
The system accuracy was compared and evaluated as an important criterion for the evaluation of suitability for a specific application.

Exact positioning in the application plays a decisive role for the encoder accuracy. Figure 3 shows the recorded accuracy curves of the three encoders. With the new EQI 1131 inductive rotary encoder, in particular the error within one signal period was greatly reduced in comparison with the EQI 1130. As expected, the EQN 1135 optically scanned encoder produced the highest accuracy results in the comparison. While the mounting tolerances of the EQN 1135 with integral bearing were compensated by the stator coupling, in the EQI 11xx without integral bearing they went directly into the system accuracy. For encoders without integral bearing, it is important to note that the device is to be mechanically mounted under the best possible conditions. Thanks to the even inductive scanning around the circumference and its greater tolerance for mounting error, the attainable accuracy is better than for rotary encoders with similar resolution but without bearings and with only one (e.g. optical) scanning point.

Dynamic behavior in the control loop
The investigations of the dynamic behavior in the control loop were made with a cascading control structure. With this structure, a speed controller is superposed on the motor current controller. This speed controller used a differentiated position signal from the rotary encoder as actual value. For position control, a further control loop is superposed on this arrangement. This control loop uses the encoder position as actual value. To achieve a good level of comparability, the control parameters were set identically for all the encoders investigated.

To achieve a high dynamic level, a control cycle time of 100 µs was chosen as well as a high proportional gain of the speed controller (KP = 1400 1/s). Additional filtering of the velocity formed from the differential quotient of the position values was done without, in order to avoid a loss of dynamics due to the additional filter time. Because of the generally good group run time (also known as “Data Age”) of the encoders’ integrated signal processing (EQI: ≤ 15 µs, EQN: ≤ 2 µs), very good dynamic performance was achieved.


Market Intelligence News & Insights: 

How Will Apple’s $500 Billion Investment for US Manufacturing, Education, and AI Play Out?

On February 24th, Apple announced a “commitment… to spend and invest more than $500 billion in the U.S. over the next four years”, with plans for new manufacturing plants, enhanced support for manufacturing partners via their Advanced Manufacturing Fund, and a myriad of other initiatives. As part of this plan, Apple projects 20,000 new hires, with roles focused on R&D, silicon engineering, software development, and AI/ML engineering.

Read More


 

Figure 4 shows the amplitude frequency response of the closed speed control loop for the three encoders (Bode plot). Because all tested units show a negligible dead time in comparison with the time constant of the speed control loop, the measured curves are nearly identical. The drive system achieved a control bandwidth of approx. 600 Hz with all three encoders. No points of resonance were detected. The achievable dynamics depends only on the controller parameters set.

The large bandwidth attainable with HEIDENHAIN rotary encoders therefore makes the realization of powerful systems possible in which the attainable dynamic performance of the total system is not limited by the encoder, but usually rather by the complex mechanical controlled systems.

The high setting of the speed controller gain ensures that the mechanical interference on the motor shaft is well suppressed. However, this also strongly amplifies the rotary encoder's measurement errors. Thus, in the case of high-performance servo motors, it is the influence of the encoder and not the mechanical influences of the motor (cogging torque, for example) that plays the decisive role for speed stability. Figure 5 shows the speed stability of the controlled motor at different speeds. Deviations from the nominal speed value are primarily the result of the encoder’s short-range position error. The significantly reduced speed ripple can be attributed primarily due to the reduction of such error. Thanks to its excellent signal processing with a high signal period number, the best speed stability can be attained with an optical encoder, whereby the differences from the new inductive encoder is relatively small so that, due to its reliable scanning and higher vibration tolerance (stator: ≤ 40 g, rotor: ≤ 60 g) the inductive encoder is optimally suitable for production machines.

The following text examines the current ripple, which can be used to evaluate the acoustic noise emission and energy efficiency of the drive system. The three encoders compared here all provide a high-resolution digital absolute position value. The singleturn resolution that is decisive for control ranges from 18 bits with the EQI 1130 up to 23 bits with the EQN 1135. Limited resolution manifests itself in drives as quantization noise of the measured speed. The quantization noise plays a role mainly at high frequencies because the derivative of the (quantized) position is required for the speed. This high frequency noise from the speed control loop is coupled into the current control loop which, in addition to the measuring error of current acquisition, leads to ripple of the actual current and is clearly visible in Figure 6. The current noise leads to forces and torques in the motor, which on the one hand can trigger natural mechanical frequencies and also become apparent as acoustical noise emission. Furthermore, this causes higher losses in the motor, which have a negative effect on the energy performance.

This can be remedied by filtering the speed formed from the rotary encoder position. Due to the resulting additional dead time, however, the controller gain of the speed controller has to be reduced, whereby the attainable dynamic performance decreases. Therefore, high position resolution is a must for highly dynamic drives.

In addition, a lower accuracy of the rotary encoder usually leads to an increased current consumption by the drive because its controller attempts to follow the encoder measurement error within one signal period (greater current ripple). This is why the improvement of system accuracy with the inductively scanned newest generation 3 rotary encoders compared with generation 1.2 is so important—because it results in dramatically improved control quality.

Position error in the position control loop
Depending on the task, first the accuracy and/or reproducibility of the rotary encoder is relevant for the positioning accuracy. The reason for the position error around the nominal position lies in the noise sources of the drive system and in the quantization of the quantities to be measured (position, speed and current). If a position is to be approached as fast possible, this requires high loop gains in the speed controller and position controller. However, this means that also the errors of the actual value acquisition are coupled into the control loop. In order to achieve the best possible performance in positioning mode, once again encoders with the highest possible signal quality are required, as provided by HEIDENHAIN.

In order to be able to clearly depict the differences of the three encoders in positioning operation, the position errors (position noise) were measured with constant nominal position.

The noise of the encoder signal plays a role when holding a specific position of the motor shaft. Both the noise that is always present in the encoder already before the position is digitized and the quantization noise from the finite position resolution cause actual value deviations in the control loop. The closed-loop control attempts to compensate these deviations by setting torques that cause small movements in the motor around the desired position. These movements are shown in Figure 7. As expected, the optically scanned encoders score highest thanks to the highest effective resolution.

Summary and outlook
The newly developed ExI 11xx multiturn rotary encoder with inductive scanning of generation 3 enables users to realize functionally safe systems up to SIL 3 with only one encoder. Thanks to their dimensions for mounting and their EnDat 2.2 interface, the new encoders are mechanically and electrically compatible with the optically scanned ExN 11xx rotary encoders, which have been available for several years, and can cover nearly the entire spectrum of possible applications. In particular at low speeds, encoders with optical scanning and singleturn resolutions above 22 bits will continue to represent the high end.

All users, however, can benefit from very good control characteristics with regard to high accuracy, dynamic performance, efficiency, rugged design, compactness and generous mounting tolerances.

Figure 2
ECI/EQI 1100 FS
Generation 3 inductive scanning
ECI/EBI 1100
Generation 2 inductive scanning
ECNEQN 1100 FS
Generation 2 optical scanning
Common mating dimensions

 

Figure 3
Measuring error in angular seconds [“]
Rotation angle in degrees [°]

 

Figure 4
 
Frequency in Hertz [Hz]

 

Figure 5
Speed ripple in revolutions per minute [min–1]
Rotational speed
Nominal rotational speed
Standard deviation
(speed ripple)
Time
Speed in revolutions per minute [min–1]

 

Figure 6
Current ripple in amperes [A]
Current
Nominal current
Standard deviation
(current ripple)
Time
Speed in revolutions per minute

 

Figure 7
Position error in angular seconds [“]
Time in seconds [s]