Next-Generation Power Firmware Simplifies Energy Star-Compliant Designs

The U.S. Environmental Protection Agency (EPA) and Department of Energy established the Energy Star program in 1992 to encourage energy-efficient products and practices. Since then, the program has grown to encompass more than 35 product categories for the home and workplace. Yet even as late as 2007, different countries have had different views on how to drive manufacturers to design more power-efficient products.

Related Articles

To promote trade, many countries have created a common set of specifications, procedures, and labeling so consumers can readily identify the efficiency of various products. In addition to the United States, Energy Star serves Australia, Canada, the European Union, the EU Free Trade Association, New Zealand, Switzerland, and Taiwan. In 2007, negotiations with China began on implementing the program there, though there are issues with its harmonization.¹

Evaluation Criteria And Test Methods

The criteria for passing the Energy Star tests are called out in each of the respective 60 product categories, such as computers, workstations, set-top boxes, and dishwashers. The results are based on use cases that describe the amount of energy (energy = power * time) within a year of use that the product will see. The weighted power consumption (P_TEC) for a workstation according to the Energy Star program requirements for computers as calculated per Equation 1 must be less than or equal to the maximum weighted power consumption requirement (_{PTEC_MAX}) as calculated per Equation 2:²

P_TEC = (P_OFF * T_OFF) + (P_SLEEP* T_SLEEP) + (P_IDLE * T_IDLE) (1)

where:

P_OFF = measured power consumption in off mode (W)

P_SLEEP = measured power consumption in sleep mode (W)

P_IDLE = measured power consumption in idle mode (W)

T_OFF, T_SLEEP, and T_IDLE are mode weightings as specified in Table 1.

P_{TEC_MAX} ≤ 0.28 * {P_MAX + (N_HDD * 5)} (2)

where:

P_MAX = measured maximum power consumption (W)

N_HDD= number of installed hard disk drives (HDDs) or solid state drives (SSDs)

It is clear from Table 1 and Equation 2 that Energy Star requires a workstation product to consume no more than a fixed number of kilowatt hours (kWh) per year based on the number of HDDs, maximum measured power, and measured power in the off, sleep, and idle modes.

Optimizing Power Consumption During Idle Mode

In the above example, optimizing the power consumption during the idle mode will contribute more to reduce the P_TEC since it is weighted at 55%. There are several ways to reduce the power during the idle mode in system software, hardware design, and power supply design. One method is to change the operating mode of the regulators from continuous conduction mode (CCM) to pulse frequency modulation (PFM) to increase low power efficiency.

A synchronous buck regulator with low-R_DS(ON) FETs is used for high efficiency in normal operation where the output current is high and there is continuous positive ripple current in the inductor. This is known as continuous conduction mode (CCM). When the load current is reduced in idle mode, the inductor ripple current will remain the same as in the high load current state, and power will still be dissipated in the FETs R_DS(ON) and inductor even though the output power is very low. This results in very low power efficiency for a synchronous buck regulator.

1. In a typical analog power management system, the power manager is simply used to configure the individual sync buck regulators. It does not provide the means for controlling the switching mode.

In a conventional analog approach to this design philosophy, there is no way to control the transition point between CCM and PFM mode, as it is predetermined by the design of the controller IC (Figures 1 and 2). Also, all the regulators require external components for loop compensation and voltage setting, increasing cost and space requirements. Code has to be written for the microcontroller for sequencing, under-voltage lockout (UVLO), delays, and more. In general, analog controllers have a fixed point that they will transition from CCM to PFM mode (if they have a PFM mode). Additional components and signal paths are required to monitor current and change of the PFM to CCM transition.

2. Analog synchronous buck regulators of the type shown in this simplified schematic require numerous external components for loop compensation and voltage setting, adding to cost and increasing space requirements.

There are several design techniques to overcome poor low power efficiency such as turning off the bottom side FET and allowing discontinuous conduction mode (DCM) or pulse frequency modulation (PFM) where the regulator skips pulses when the output current drops below a threshold current set by the controller. It would be even more advantageous to directly control the threshold point that the power system transitions from PFM to CCM. If this were controlled by software and could be changed simply by loading a register in a power manager, it could be quite simple to implement.

Programmable power enables better control and easier Energy Star compliance by allowing the designer to fine tune the transition point between CCM and PFM to minimize the idle mode power. Also, programmable power allows rapid system upgrades as future specification revisions to the Energy Star standard occur or to contend with standards that other countries may implement.

One Solution

Exar offers a simple solution to optimizing the power consumption in Energy Star systems such as these. Its Power family of products, including silicon, software, and development tools, allows a much improved design methodology compared to conventional analog power converters with external digital-to-analog converters (DACs), analog-to-digital converters (ADCs), and power management controllers to monitor and control system power. This family of products incorporates all of the above features and adds features such as current and device temperature monitoring and the ability to drive external N or P channel MOSFETs as load switches.

No external components are required for compensation, current sense, UVLO, or voltage setting. All parameters including sequencing times, delays, and power-good parameters are all set in a user-friendly graphical user interface (GUI) via the interface I²C bus. Development tools include a USB to I²C interface that works under Windows, interface code examples, and programming services.

3. The complexity of a four-channel analog power management design (a) contrasts starkly with the simplicity of Exar’s fully integrated XRP7714 quad-channel programmable power management system (b).

To appreciate these advantages and particularly the reduction in solution complexity, Figure 3 shows a real design comparison from an actual system. Table 2 summarizes the considerable saving in component count achieved by Exar’s programmable power management system, reducing the number of active components from seven to four and the number of discrete components from 148 to just 33.

Summary

Exar’s Power programmable power management system provides a cost-effective and easy to change solution for Energy Star-rated products both in the United States and worldwide. The ability to configurethe power system via an easy to use GUI allows Energy Star compliance to be quickly achieved for idle and standby modes.

References

1. “Communication from the Commission on the implementation of the Energy Star programme in the European Union in the period 2006-2010,” http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2011:0337:FIN:EN:PDF

2. www.eu-energystar.org/downloads/specifications/20081118/final/Computer_Spec_Version%205%200_%20Final%20Nov08.pdf

Ron Vinsant joined Exar Corp. as a principal field applications engineer in July 2010. With more than 25 years of experience in analog and digital power systems design, he has demonstrated success in engineering and application engineering roles at technology companies both large and small, including PowerVation, Fairchild, Zilker Labs, SOMA Networks, Linear Tech, and Teledyne. Prior to joining the power industry, he studied physics at U.C. Berkeley, California.