Motor control isolation scheme using iCoupler technology
Robotic applications require precise control of the motors that drive many machine joints. The control system needs to know the positioning positions of the various robot arms and actuators to ensure safe and reliable operation. To be efficient, you need to know more about the rotor motion in the motor housing in more depth.
Without information on the rotor angle (easy to slide under high loads), the electronics controller may provide too much current, which is simply wasted by heat. In order to sense the position and rotor state, an important variable of the control algorithm is the current level of the motor windings. Conceptually, this is a low-cost variable that is easy to monitor because it only involves providing a link from the motor to the control circuit. However, there are many factors that need to be considered to ensure that the signal is as accurate as possible. Errors can result in inaccurate detection of locations and increased unnecessary energy consumption.
The most commonly used current sensors in motor control are shunt resistors, Hall effect sensors and current transformers. The latter two devices provide isolation, which, while increasing overall cost, is important when dealing with high power. Shunt resistor circuits are typically limited to measuring currents of 50A or less, but have the advantage of having the highest linearity of response in sensor-type devices and lower cost. These devices are also suitable for AC and DC measurements.
Accurate and sensitive results can be achieved by coupling the shunt resistor to the delta-sigma modulator. Triangular integral sampling and filtering techniques help to suppress transient noise effects and support well above 12-bit resolution. Texas Instruments' ADS1203 is a delta-sigma modulator designed for instrumentation applications including motor control. This device is a single-channel, second-order delta-sigma modulator designed for high-resolution analog-to-digital conversion from DC to 39kHz. The output of this converter is a series of numbers 1 and 0 whose time average is proportional to the analog input voltage. A key advantage of using a filtered delta-sigma modulator signal is that the quantization noise source and the transient noise source can be converted to high frequencies, making it easier to filter out through the low-pass filter.
By using a modulator instead of a complete analog-to-digital converter, designers can adjust the digital filtering performance to best meet motor control requirements. This includes strict synchronization with transistor switching events in the H-bridge circuit that supplies power to the motor itself. The filter itself can be implemented using a digital signal processor (DSP), microcontroller, or field programmable gate array (FPGA), depending on cost and performance goals. By using a custom filter, it is better to choose between the transient response and the final sampling resolution. A higher oversampling rate results in higher accuracy, but results in a lower value update rate - reducing oversampling reduces resolution but provides a higher refresh rate.
In terms of data processing, there is a comparison with a traditional successive approximation (SAR) analog-to-digital converter. By using a SAR converter, sampling can be performed with the aid of a sample-and-hold circuit, which allows the system designer to tightly control the timing of the sampling instant. On the other hand, the triangular integral conversion uses a continuous sampling process, so the sampled value has no defined trigger time. Conversely, the sampled value at this point in time is a weighted average of a series of 1-bit sample values that may span the value of this point in time represented by this sampled value.
Filtering a 1-bit bitstream and extracting it to a lower rate multi-bitstream sample value can be done in two different phases. A very common approach is to use a SINC filter that performs both tasks in one phase. The third order, commonly referred to as sinc3, is currently the most common choice for these applications.
The filter is largely a weighted sum of a sampled value window that gives more weight to the sampled values at the center of the sequence, while giving less weight to the sampled values at the beginning and end of the sequence. In view of the influence of the switching component of the power transistor in the measurement current, this effect needs to be considered, otherwise the feedback algorithm will be affected by aliasing and the like.
The impulse response of the sinc3 filter is symmetric with the contribution of the sample value prior to the center sample value, and the center sample value is the same as the sample value following it. The switching component of the current is also symmetrical along the average current point: so that the sum of the switching components is zero. If the center of the sampling window is aligned with the PWM sync pulse used to drive the H-bridge, the phase current is allowed to be measured without aliasing, but care must be taken to ensure that the sample values are properly aligned when reading data from the filter. Filtering imposes a delay so that the sampled value output of the filter will be from multiple previous time periods when the PWM sync pulse is used. This has a significant impact on software program scheduling compared to SAR-based current measurements.
In the case of SAR, the PWM sync pulse can trigger the analog to digital converter to perform a series of conversions. When data is prepared for the control loop, the system generates an interrupt and begins executing the control loop. These sample values are continuously generated by using a delta-sigma modulator and filter, but important sample values for phase current measurements are ready after a fixed delay. Timers or counters should be used to generate an interrupt when a PWM sync signal is present. The delay in counting sample values is actually half of the sinc3 impulse response.
In a typical control system, the zero-order hold effect of the PWM timer is much more than half of the impulse response, so the SINC filter does not significantly affect the loop timing. By utilizing a delta-sigma modulator and a custom filter, the user can freely switch the SINC filter delay to obtain sampled value resolution. This flexibility is a big advantage when designing motor control algorithms. Usually some parts of the algorithm are sensitive to delay but less sensitive to the accuracy of the feedback. The rest of the algorithm is used in conjunction with lower dynamics and benefits from accuracy but is less sensitive to delays.
Consider a proportional integral controller (PI) algorithm. The P part and the I component can use the same feedback signal. However, the P path and the I path can be separated and the feedback signal can be combined with different types of filtering functions. In the PI controller, the P component is mainly used to suppress the rapid change effect of load and speed. Therefore, it needs to be able to respond to rapid changes in signal levels. The I component focuses on steady state performance and is more focused on measurement accuracy. Therefore, the P component can benefit from a low resolution, fast update rate current feedback signal, meaning that the sinc3 filter has a low oversampling and decimation rate. The I component will benefit from a higher oversampling rate and can withstand the resulting increase in update rate.
It is important to note that when using a delta-sigma modulator in a system that handles large loads, another factor to consider is isolation. One option is to use only the isolation amplifier and use a non-isolated modulator for analog-to-digital conversion, or place an optocoupler between the output of the modulator and the input of the device for digital filtering. Alternatively, an isolated delta-sigma modulator can be selected. By using an isolated modulator, the analog overcurrent protection circuit can be eliminated because the digital filter can also be configured to eliminate overcurrent effects.
The AD7403 is provided by AnalogDevices, an example of this. By implementing a second-order modulator, this device allows for flexible selection of shunt specifications and provides more than 14 bits of significant bits and an output stream rate of 20 MHz. By utilizing an appropriate digital filter, the device achieves a signal-to-noise ratio of 88dB at 78,100 samples/second. This isolation scheme uses the company's iCoupler technology, and the company claims that it exceeds the performance of a typical optocoupler arrangement.
With the addition of features such as isolation and the increasing filtering performance of microcontrollers and programmable logic devices, designers can continue to optimize motor control for robotic applications.
If you want to buy a medical device motor, please pay attention to Precision Medical Motors.
