Digital power will replace and surpass analog power

This article introduces digital power supply technology that replaces traditional analog control: Digital power supply has great advantages over analog solutions, not only in terms of performance (efficiency, transient response, stability, etc.), but also in terms of time to market and total cost of ownership. Digital power supply is completely changing the design method and implementation of power supply.

Basic Information

Intersil's ZL2008 second-generation adaptive digital power controller for DC/DC power conversion is an industry-leading digital power controller. It is suitable for non-isolated buck, boost, buck-boost and isolated single-tube forward or flyback converters. Inside the 6mm×6mm QFN package (Figure 1) is an advanced power controller that integrates power conversion control, power management, fault management and telemetry functions. In addition, it also contains an integrated microcontroller that can run complex algorithms and can adapt to operations that exceed the performance of analog solutions. It represents the latest technology in cost-effective digital power. Basic specifications are shown in Table 1.

Digital Power Architecture vs. Analog Architecture

Figure 2 shows the progression of power conversion control architectures from analog (a) to modern digital control (c). The analog PWM controller generates a proportional duty cycle by using a ramp error signal. The error signal is compensated using a resistor and capacitor network to modify the signal to stabilize the control loop.

The earliest attempts at digital power (Figure 2b) generated a proportional duty cycle through a digital counter (DPWM) whose count was determined by a digital signal processor. While this approach was very powerful in digital implementation, it proved too expensive and required too much quiescent current for most practical applications.

In modern digital power control (Figure 2c), the duty cycle is still generated by a digital counter, but now the counter is controlled by a digital state machine. This state machine is designed specifically for power controllers (rather than general-purpose DSPs), so this solution is more cost-effective and requires less quiescent current.

The architecture of Figure 2c uses a proportional, integral, and derivative (PID) compensator to stabilize the power supply without the need for a full DSP to compensate the power supply. The three elements of the error voltage, the proportional error, the integral error, and the differential error, combine relative weights to achieve stable operation.

Note some of the advantages that digital power supplies have over analog power supplies in terms of architecture: Digital control requires no external components for compensation. This not only reduces component count, but also allows compensation to be easily changed, either on the fly or even adaptively as the load changes.

Typically there is no external voltage divider for a digital controller. The internal reference can be scaled so that no external voltage divider is needed. This obviously reduces component count, but it also helps to accurately calibrate the controller at the factory so that the user can benefit from high accuracy without having to use expensive precision resistors for voltage division.

The digital architecture allows easy use of digital communications so that operation can be configured, controlled, and monitored with virtually no external components.

A digital power supply controller

Figure 3 shows the basic architecture of a modern digital power controller. In this architecture, the output voltage is sensed using a differential amplifier. This analog signal is compared to a reference to generate an error signal. This error signal is digitized (ADC) and the result is processed through a digital compensation network, which will be described later in this article. The output of the digital compensation is a duty cycle command that sets the duration of the digital PWM. The digital PWM control can then control the FET drivers and switch the power supply.

The output voltage, input voltage, output current, and temperature can all be detected using an auxiliary analog-to-digital converter (ADC), which can be multiplexed to various detection points.

Configuration can be accomplished by pin jumpers, resistor configuration, or by means of commands through the I2C interface. The power supply can be controlled by pins or the I2C interface. Configuration, operation, and environmental condition monitoring are accomplished through the I2C interface.

Advantages

1. A higher level of integration

Figure 4 shows a typical application schematic for analog PWM and digital PWM. Although both controllers share the same number of power train components (power FETs, inductors, input and output capacitors), the analog controller still requires more external components. This is because the digital controller integrates many functions and features that are not integrated into the analog controller. As shown in the figure, the digital controller reduces more than a dozen components. In actual implementations, digital controllers have been proven to reduce external components by up to 60% in medium to high complexity designs.

2. Stability

Figure 5 shows a typical power conversion circuit. The power converter consists of a PWM controller with a fixed modulation gain Gfix, high-side and low-side switches, an output stage containing an inductor and one or more capacitors, a load, and a feedback or control loop. In this case, the feedback control is shown as a Type 3 (or III) amplifier, but it can be any feedback controller. The purpose of the control loop is to compare the output with a known reference, VR, and adjust the PWM signal to correct the difference between the output and the reference.

In addition to the advantage of reducing the number of components, digitalization provides the further advantage that the values ​​of the integrated components can be represented as values ​​stored in digital registers. This makes it easy to change these values ​​according to different designs, even on the fly, or to adapt to changing conditions.

Any change made to a control system introduces a disturbance into the system. To be a robust and useful system, the system must remain stable in the presence of this disturbance. In fact, it must remain stable in the presence of a whole host of disturbances, including input voltage changes, load changes, and even temperature changes.

We can describe the stability of a system by how close the feedback path gain is to -1. That is, how close is the feedback to a gain of -1. Since the feedback has a magnitude (gain) and phase relative to the output, we can express stability in terms of gain margin, where the gain margin is measured relative to unity gain when the phase is 180 degrees, and phase margin, where the gain margin is measured relative to unity gain when the phase is 180 degrees, and how close is the phase margin to a phase of 180 degrees when the gain is unity.

Phase margin and gain margin can be determined from either a Nyquist plot or a Bode plot. Since the Bode plot has an easily readable frequency range, it is a convenient tool and this is what we will use in this article.

If there is no feedback, the simplified transfer function of the system shown in Figure 5 can be expressed as:

in:

ωesr is the zero point produced by the output capacitor esr, ωn is the natural frequency of the output stage, and Q is the quality factor of the output stage.

For the purposes of this article, we will ignore the contribution of the capacitor esr zero and focus on the remaining poles of the transfer function. In other words, let's focus on the transfer function:

This equation has two poles. For Q < 0.5 (damped case), both poles are real. For Q > 0.5 (underdamped case), both poles are complex conjugate.