The debate on digital power is raging and from the analog bastions concerns are being raised about the legitimacy of such digitalization claims. One could counter that while the real world is actually analog; digital is nevertheless flourishing. A power supply is no different than any other real world system: its output is analog (voltage, current, power, etc.) the same way that digital TVs, digital still cameras, and digital cell phones yield analog outputs (video, pictures, sound, etc.).
Are all these instances misnomers? Is this an exercise in semantics? Digital power – beyond the obvious communications aspects which have been around for quite some time – simply refers to the progressive digitalization of some processing elements at the earth of the system, whereas nobody argues that the outer shell (the power train, the voltage reference, etc.) will remain analog.
Digital power designs can help in a number of areas. Applications such as lighting ballast and power factor correction (PFC), which require voltage- and current-profiling, often have features that are better implemented in digital rather than in analog. And when it comes to power conversion, digital power enables digital compensation, as well as the automatic change of a number of phases.
Consider the ballast and PFC applications. Current profiling is required in lighting ballast applications where the intensity and duration of the current can be flexibly set for different lamps, as well as for the three operational phases of pre-heating, ignition and dimming. Current profiling is also required in PFC applications, where the current drawn by a load from the wall outlet must be in phase with––basically have the same shape of––the line voltage. In digital implementations many predictable errors can be corrected via a ROM table. This type of control can be implemented with a device like Fairchild’s FMS7401 microcontroller.
Since conventional digital algorithms are sequential in nature, requiring several clock cycles to execute an instruction, they are also inherently slow and so are not useful in applications requiring fast response. A hard-wired digital power conversion implementation, on the other hand, will inherently produce a faster transient response compared to software code execution. Such digital control architecture for power conversion is one in which the input error signal is converted to digital via an analog-to-digital converter (ADC) and hence the proportional-integral-differential (PID) compensation and digital pulse width modulation (DPWM) is all done in the digital domain.
Hard-wired digital power conversion may be better justified in the case of a notebook or cell phone voltage regulator application where, due to the necessity to save power at light loads, a mode change is required. This typically occurs when there is a change from a PWM algorithm to pulse-frequency modulation (PFM). PFM is a mode in which the frequency adjusts with the load, thereby yielding lower frequencies and hence lowering switching losses at lighter loads.
Such a mode change in an analog system would require an abrupt commutation from one control loop, say PWM, to the other (PFM), typically at the time that the load is changing. This type of algorithm discontinuity would invariably lead to some degree of temporary loss of regulation of the output.
In such systems, digital control may prevent risk of loss of regulation and save additional overhead in bill of materials (BOM) that would be required to mitigate the effect of discontinuities in analog implementations. A product like the FAN5608 two-channel LED driver is an example of such hard wired digital control designed for ultra-portable applications and implemented in piece of silicon rivaling any conventional competitive implementation.
Digital control can provide a more robust solution compared to analog, based on its ability to change, on the fly, parameters like the loop compensation and hence withstand wider load changes and poor layouts. It may also be able to calibrate out-errors in the system due to low-cost component tolerance errors. The advantage is the potential for better yields, lower testing costs and lower BOM costs.
With all of these possible advantages, one may wonder why digital power has caught big share of attention but little market share. The answer is that the market has made it clear that digital will be preferred every time over analog, but at cost parity. The successful digital power products are those that solve a customer’s problem by utilizing lean architectures with minimum overhead, and that do not result in a cost penalty versus their analog counterparts.
Reno Rossetti is director of the Analog Product Group Strategy at Fairchild Semiconductor. Reno has many years of experience in design, applications and marketing in the analog and mixed signal semiconductor industry. Reno’s career has revolved around analog, power ICs and discrete semiconductors. He holds several patents and has authored many articles in these areas. Before coming to Fairchild, Reno worked at ST in spindle motor and voice coil drivers design. Later he headed up National’s Power Management design group. Reno graduated in Electrical Engineering from Politecnico di Torino in his native Italy. He also has an MBA from the Bocconi University of Milan.