Gary Bocock, Technical Director, XP Power
Digital control in power supplies and power systems broadly fits into two implementations. The more common approach is a digital interface between the traditional analogue control system and the outside world providing signals & alarms and various levels of control via a communication bus.
Simple, low-cost microcontrollers have also been implemented in power applications for many years for functions such as fan speed control, protection functions & alarm detection.
Increasingly manufacturers are using digital signal processing (DSP) via a micro-controller for power system control bringing more sophisticated functions and greatly enhanced flexibility, allowing user programmable features and characteristics.
Click image to enlarge
Figure 2. DSP controlled digital power supply
Without doubt, DSP is higher cost when compared to an off the shelf analogue controller but the cost of a microcontroller capable of implementing full DSP control has decreased over time, making this an increasingly attractive and desirable solution given the significant benefits it provides, especially as the power rating increases. The mixed domain architecture required, combining power analogue design principles with efficient code and stabilisation of the control loop in the discrete time or z-domain, rather than the frequency or s-domain, is well proven and understood by the product design & development teams within the major power manufacturers.
While the development, documentation, verification and approval of efficient, robust firmware takes significant time and resources to ensure a robust and reliable power supply, once the initial investment has been made the significant benefits of digital power and the ability to reuse the firmware across a broad range of products and platforms with relatively minor changes can be realised.
Digital control loops have the advantage of being insensitive to changes in environment, temperature, ageing and tolerances of components. They can be calibrated at the point of manufacture to further improve accuracy and they enable monitoring of the performance of the power system in real time and adjust parameters to tune for optimal performance at the operating point, increasing efficiency and reducing power losses.
Features & Benefits of Digital Signal Processing (DSP)
Fully digital power supplies are able to offer unparalleled flexibility and adjustability to suit a wide range of applications without the hardware changes and adaptations which traditional analogue control systems have historically demanded.
DSP control loops bring the capability of output voltage and current adjustment over ranges as wide as 0 to 105-110% by tailoring the converter operational mode to the demand. They ease the implementation of constant current overload characteristics, which can be complex and costly in modern resonant switching topologies, without the need to compromise efficiency. This entails the employment of multiple switching schemes and control algorithms in the same power conversion stage to achieve optimal performance at the required operational point, an extremely complex, if not impossible task, in a traditional analogue control scheme with fixed hardware drive and compensation schemes. This wide range control can be implemented as a continuously variable power supply to maximise system flexibility and efficiency or, exploited during the system development phase, to optimise the supply characteristics to the application without the need for hardware updates.
DSP also enables the user to determine start up ramp times, soft start characteristics and slew rates in software, another feature that would result in hardware changes in traditional control systems.
Warning levels and fault conditions such as input over/under voltage, output over/under voltage, output under/over current, temperature warnings & fault conditions can be set by the user to suit the application via software. The use of DSP further allows the response type & delay times applied to individual warning or fault conditions to be user specified. Options under warning or fault conditions may be as varied as continuing operation for a short delay & then disable, continuing operation indefinitely, disable & retry (including how many times to retry & time between retries before shutdown) and disable & resume when OK or disable & latch, all user selected.
Digital control systems also allow users to set the polarity of signals, alarms and controls to suit the system demands. A good example is the ability to set the remote on/off control to operate as inhibit or enable simply by toggling a digital switch.
Information from the power system is readily available through the communication interface enabling reporting and status such as model, revision, serial number, run time, operating temperature & fault/event logs.
This level of flexibility and user control is possible as the latest microcontrollers for digital power applications contain DSP functionality that allows the digital control loop to execute within a fraction of one switching period, every switching period. In the simplified example below, the output voltage is sampled once per switching cycle. An ADC conversion time of a few hundred nanoseconds is typical.
Click image to enlarge
Figure 3. Simplified example of control loop & spare bandwidth
The time that the MCU does not spend executing the controller is spare bandwidth and this spare bandwidth can be used to perform other tasks or functions. Low priority tasks are run in a slow loop and are interrupted whenever a high priority task occurs, such as the ADC interrupt to run the control loop code.
The provision for analogue control of digital power is usually also provided for systems that use traditional 0-5V or 0-10V control signals by the implementation of an Analogue to Digital Converter (ADC) within the power supply and all alarms & controls can usually be accessed through conventional connections as well as digitally through the communications bus.
Communication and control are increasingly important with the rise of connected, smart factory and IoT applications benefiting from real time status information from power systems as well as adjustment and control inputs, allowing real time adjustments to maximise process efficiency where it benefits from accurate voltage and/or current supplies and the ability to tune these to suit the environment and application.
Digital power products are able to report warnings, fault conditions, power delivery information, run time, thermal data and event logs in addition to enabling the real time adjustments to output voltage, current and power delivery to maximise system efficiency in sensitive processes or test applications.
A range of digital interfaces and protocols ranging from the commonly used I2C/PMBus & RS232/RS485 serial buses to DeviceNet & EtherCAT enabled interface solutions are available to suit a wide range of environments, applications and requirements.
While not all end applications require communication to the outside world, the ability to communicate with and adjust the parameters of the power system within an end equipment can enhance features & operating characteristics and has the potential to save cost by replacing the external hardware controls that are required for traditional fixed output supplies. DSP enabled power supplies can support dynamic requirements for output voltage, current and power delivery that are normally associated with far higher cost laboratory supplies, where tolerances are acceptable, and enable complex test, burn-in & process routines directly from a cost-effective power source.
In end equipment where there is no requirement for communication, either external or internal, there are still benefits to be gained in tailoring the power supply to suit the application, easing integration and removing the need for an application specific solution requiring a modified standard or custom power solution.
Output voltage, output current, power delivery, warnings, alarms, protection & controls can be adjusted, evaluated, amended and finalised during the development stage, creating a set of unique characteristics in firmware which are then implemented by the power supply provider at the point of manufacture for the end equipment production phase. These iterations of characteristics can be implemented on the same standard product saving considerable time and cost compared to the hardware changes required in traditional power products.
End applications employing analogue controls for voltage or current adjustment still benefit from the ability to determine the warning and fault condition settings & responses and polarity settings of alarm and control signals again, without resorting to application specific or fully customised power solutions with the time delays, risks and inevitable costs that are involved in development, EMC and safety agency approvals.
Manufacturers of digital power supplies commonly offer a Graphical User Interface (GUI) to enable users to define the requirements for just this purpose as well as enabling speedy evaluation of capabilities for connected applications.
In summary, there are clear and realisable advantages to implementing DSP control in power supplies and power systems in many, usually higher power (1kW+), applications which benefit from the flexibility and the time and cost savings it brings. For simple, lower power applications it is likely to be overly complex and cost prohibitive, with standard off the shelf analogue controllers enabling fast time to market combined with low acquisition cost for the commonly used topologies.