By Gus Pabon, CTO
Since about 1973, power supply controllers have been analog and not much has changed in their basic functionality, until recently. An emerging trend is happening where digital power supply control IC’s are starting to make their way into designs and they show a huge potential for additional benefits that can be brought to power conversion electronics.
In today’s analog technology one approach for regulation and control of the output DC-voltage is achieved using a voltage-feedback method that compares the DC-output voltage that is proportional to a reference voltage that is part of this feedback loop; in this part of the electronics, additional circuitry is used to compensate for amplifier gains over adequate frequency ranges to ensure stability of the output voltage. The difference between these two voltages is then used to set the limit when compared to a voltage ramp generated internally by the control IC; this difference-voltage and the voltage ramp are inputs to a comparator circuit whose output results in a square-wave of voltage with a varying duty-cycle. This duty-cycle-varying square wave drives the power switches in the power supply. By this method, regulation of the constant DC-output voltage is achieved even under conditions where the power supply loading changes.
Now, with the emergence of high-speed digital electronics, this feedback-loop can be implemented in a digital form. The output voltage can be fed to an Analog-to-Digital (A/D) converter whose output is compared in digital (multi-bit) form to a digital-reference. The difference is then fed in its digital form to some type of Digital-Signal-Processor (DSP) as part of the digital-control circuitry. This DSP , among its many functions, performs compensation to ensure stability of the feedback loop and hence the output voltage. The output of this DSP is then fed to Duty-cycle registers that output (as in the old analog-control method) a square-wave of voltage with a varying duty-cycle that is used to drive the power-switches in the power supply.
If the basic functionality described above was all that was required in the operation of a switch-mode power supply, digital-control would be at best equal to the mature analog solutions. It is in the digital circuit’s ability to dynamically change its operating mode where a digital controller shines and outperforms an analog controller. A simple example considers the stability of the feedback loop over a wide operating range of the power supply’s output current; each current load of the power supply determines an operating point. In the analog method, the feedback loop is optimized (compensated) for stability for only a few discrete operating points and outside of these operating points there still exists stability, but the operation of the feedback loop is not at its optimum point. The reason for this is that the stability of the feedback loop in an analog system is achieved by using discrete-components (capacitors, resistors) that set the frequency-dependent gain of an amplifier. As these components are fixed in their values, the compensation and hence the performance of the feedback loop is fixed.
In a digital system, simple monitoring of aspects of the power supply’s operation (such as output current loading) can be fed to the digital-controller and the digital-controller can employ firmware that can digitally change the compensation depending virtually on any operating point of the power supply and ensure close-to-optimum operation of the feedback loop. This is possible because only firmware and not discrete circuit-components determine the compensation of the feedback loop.
Other clear benefits of digital control are that accurate and reliable communication is enabled with a digital-controller. Such communication with the “external” world can be bi-directional. This communication enables the power supply to be configured from an external source by receiving communication or by the power supply sending information, enabling a single power supply to be used for many applications – to achieve universal, interoperable behavior. Communication also enables monitoring of the power delivered by the power supply or the “state-of-health” of the power supply and its components, and is readily available to any host system chartered with overseeing the power system.
Also, every power supply has additional “house-keeping” circuitry whose function is usually for protection and prevention of catastrophic failures; this is additional analog circuitry that adds cost, requires pcb real-estate both of which are not desirable and also as the protection-thresholds are set using discrete circuit components they are not easily modified. In a digital-controller, much of this protection function can be done in firmware thereby eliminating additional circuitry and as describe above the protection-thresholds can be easily and dynamically (on the fly) changed in firmware.
Today digital-controllers are two to three times the cost of their analog counter parts, but as our demands for power-converters increase as is happening due to trends in rechargeable battery systems, solar-arrays, electric cars, etc., so will the complexity in their designs. This increasing complexity in applications will soon be hard-pressed to be met by analog controllers. When this point of inflection is reached, higher volumes of digital-controllers will drastically drive the prices down, resulting in cost-competitive solutions. The trend toward digital control in power conversion is only in its infancy and the potential of the benefits it brings to a power electronics designer are sure to be welcomed as a breath of fresh air.
