Digital Control versus Management

Digital control and digital management are two-separate and distinct functions that can be designed onto a power supply. However, confusion continues in the user environment as to what each of these actually means for the user. Digital control refers the digitisation of the error associated with the main control loop and it use by the control loop of the power supply to adjust the main power stage, in order to converge on the required set point. Stability of the loop is also set by software parameters, where traditional poles and zeros can be set by software parameters.
Digital monitoring refers to the design taking a series of analogue inputs, converting them to digital, and making these available for interrogation by a system host. With this approach, analogue circuits maintain the main loop functionality along with all protection circuits of the power supply. While the user may have access to information on the performance of the power supply, this feature set does not typically provide the option to vary any of these. The details of how the user interfaces with these can vary greatly, and the industry is attempting to normalise some such methods.

Closing the loop
Closing the loop digitally has many significant advantages over traditional analogue loops. The obvious one being that poles and zeros can move on the fly to provide the optimum stabilised loops over all line and load conditions. This is of particular advantage when dealing with transient demands, but also to meet the increasing demands on power supplies to meet performance criteria over all line and load conditions. With an analog loop, resistors and capacitors set these and in general are designed for a particular operating condition; the user must be satisfied with subsequent performance if they use it outside of these boundaries. There are also advantages to setting the profile of a power supply as it goes through various stages, prior to steady state operation, such as power-up and -down stages.
It also allows for the customisation of such performance by recoding the main controller, which can be done on production-finished units if necessary. However, care must be taken in order to meet new challenges. More digital aspects on the design mean more square waves, and higher EMI content in their harmonics. If not dealt with at the design stage their implementation in a system could mean compliance with conduced and radiated noise norms will be more difficult to achieve.
With digital control the design ‘automatically’ gets digital power management. All of the relevant information is pulled into the central controller and most of the associated ICs will have a digital power management capability added to it in the form of an on-board microcontroller. However, it is also possible to implement a digital power management feature set with a traditional analogue control loop. A sub $1 microcontroller from any of the leading suppliers will typically have three Analogue to Digital convertors and, say, a Digital to Analogue controller onboard.
It will also have up to 20 general I/O ports. With this feature set it is possible to read back voltage, current and temperature, convert these to digital and read back to the onboard I2C peripheral. The designer can also implement addressing, logic control etc. via the remaining I/O ports. So in closing the loop digitally, digital management is inherent, but it is also possible to operate digital management in isolation from the control features.

Digital monitoring
Of course, the range of digital power monitoring is not limited to just a small number of key points such as voltage, temperature and current. There are literally hundreds of items that can be pulled from a power supply. Some industry standards such as the PMBus define several hundred commands that a power supply can supply information on. These include setting bits in one or two byte commands that determine how the power supply should react, for example, in the event of a non-conformance.
Closing the loop digitally is not a new concept. In fact it has technically been feasible for well over 10 years now, and in the last 5 to 7 years it has been commercially viable also. So, the obvious question is why haven’t we seen wider adoption in the market. The answer lies in the end customer usage In some risk adverse cultures, there is a concern about relying on a piece of software to control the output of a power supply. The main concerns being that an error in algorithm or coding can result in an uncontrolled voltage appearing on the output. In the industrial world this can result in catastrophic damage to expensive equipment, in the medical world in can result in patient fatalities.
Some companies have been successful in bridging the gap between customer needs and requirements with respect to digital control. However, most of these have circumvented the issue by proving a digital control loop using either clock edge driven state machines, or FPGAs. As these parts do not depend on a clock counter, they can be considered non-software driven. But the reality is that if the code is properly constructed and compliant with the relevant IEC standard, it should be a strong enough argument to negate any such concerns. However, the perception of the end customer remains the leading perception of the industry, and will thus determine the acceptance rate and product offering.
Digital has opened up a range of new ideas and is drawing in new core competencies into the power supply industry. Gone are the days when a full working knowledge of analog circuit design will suffice for a power supply design engineer. The traditional skill set of power engineers is now opening up to a wider range of experts. Subsequently we will experience a much higher influx rate of graduates being attracted to power design. With them will come new ideas and we can expect to see more functionality integrated onto power supply designs. This will include power supplies being added to a the growing interest in the Internet of Things. In order for this to manifest itself, there must first be an adoption by the industry and by end customers, that adding digital power monitoring to the feature set of a power supply can only be a positive thing. From there we will see adoption and with correct designs we will see an acceptance of digital power control onboard even the most risk averse industries.