The use of external power supplies (EPS) has been growing significantly over the past three decades. California was the first to have its government organization put minimum legal requirements on the energy efficiency of the EPSs, which historically had poor efficiency. The U.S. Federal Government and other international governments followed California’s decision by initially introducing voluntary requirements, which turned into mandatory requirements a few years later.
These government entities have been gradually raising the bar since nearly 15 years ago, and the latest energy efficiency standards are not the last. We will briefly overview the history of power supply efficiency requirements and their impacts. We will outline the major aspects of the latest Department of Energy (DOE) efficiency standard, which is DOE level VI, and compare it with the latest European efficiency standards COC Tier-1 and Tier-2. We conclude with possible upcoming requirement updates and changes in the next version of the DOE standard, which is due as early as 2 years.
In the 1990s, the use of electronic devices started growing exponentially. These electronic devices mostly used different voltage magnitudes and types than the AC grid voltage, and therefore, required power supplies to convert the grid voltage to a suitable conditioned voltage for their use.
To step back, there are two types of power supplies, internal power supplies (IPS), which are used inside kitchen appliances, TVs, and desktop computers, and External Power Supplies (EPS), also known as power adapters, which are used to deliver power from the grid to laptop computers, mobile phones, game consoles, and other electronic devices.
At that time, due to the proliferation of electronic devices, National Resources Defense Council (NRDC) estimated that there were more than 1 billion EPSs operating in the United States alone, and of those, most were using an inefficient design called linear regulator technology with the average efficiency around 50 percent. Moreover, the power consumption of the EPSs was substantial when the end device was turned off or detached from the EPS (which is referred to as “no-load” condition). This poor efficiency of EPSs was not principally due to the lack of technology, since IPS could gain considerably higher efficiencies by employing switched-mode circuits, but because of lower cost and lack of market incentives. Studies showed that if the efficiency and no-load power consumption of the EPS remained at the same level, EPSs would be accountable for around 30 percent of the total energy consumption over the next two decades.
NRDC had also estimated that by boosting the EPS efficiency by 15-20 percent, it could save 32 billion KWhr/year cutting the annual national energy bill by $2 billion and reducing the carbon dioxide emission by 24 million tons per year. The U.S. Environmental Protection Agency (EPA) was the first organization to start a voluntary program to encourage higher energy efficiency and reduce pollution, which eventually became the Energy Star program. At that time, the main focus of the Energy Star program was to promote the sleep mode or no-load power consumption and lacked a regulatory program addressing the active mode efficiency of EPSs.
The state of California soon realized that with the accelerated rate of energy consumption, its demand would outpace its energy generation. Thus, to decrease the energy waste in its grid, the California Energy Commission (CEC) introduced the first mandatory standard for energy efficiency of external power supplies in 2004, which became effective in July 2006. This decision was then followed by other organizations around the globe who tightened the efficiency and no-load power consumption limits, many of which are still evolving. The CEC’s standard limit was set to be 49 percent at 1 W with a sliding scale of 49 to 70 percent between 1 W and 10 W; sliding scale of 70 to 80 percent between 10 and 30 W; and sliding scale of 80 to 84 percent between 30 and 49 W. The efficiency limit of EPSs greater than 49 W and less than 250 W was set to be at the fixed amount of 84 percent, and no limit was introduced for EPS above 250 W. To meet this standard, manufacturers had to dramatically increase the efficiency of their existing EPSs by nearly 20 percent for units below 50 W.
Having the CEC standard as a reference, the U.S. Congress passed the Energy Independent Security Act (EISA) in 2007, in order to harmonize the energy efficiency standards across the U.S. Known as Level IV, these standards boosted the efficiency limits of the CEC standard by 1 percent across the 1 to 250 W range. In 2011, Level IV was upgraded to Level V, dividing EPSs into two categories based on their output voltage; standard voltage for output voltages above 6 V and low voltage (LV) for output voltages below 6 V, and output current of greater or equal to 550 mA. Compared to Level IV, Level V raises the efficiency limits for standard and LV EPSs at 10 W by 6 percent and 3 percent respectively; by 1 percent at 30 W; and by 2 percent and 1 percent respectively for above 50 W. A similar scenario happened for the mandatory limits on the no-load power consumption of EPSs, where the CEC set the first mandatory requirement at 0.5 W for EPSs below 10 W. Level IV then extended the output power range to 250 W and set 0.75 W as the limit for the no-load power consumption. These limits were substantially below the industry norm for the no-load power consumption of around 1 to 2 W. In Level V standard, the no-load power consumption limit was further lowered to 0.3 and 0.5 W, respectively, for 0 to 49 W and 50 to 250 W power range for both standard and LV EPS. The European Union enacted its ErP phase 2 directive with similar efficiency and no-load power requirements in 2011, to harmonize with Level V.
In 2014, the US Department of Energy (DOE) published the more stringent Level VI standard, which became effective 2 years later. Adapting with this new standard was somehow challenging for manufacturers, as it reduced the no-load power requirements significantly and increased the efficiency limits of the active operating mode. The no-load power dissipation limits were aggressively cut to almost one third compared to the previous standard, which sets the limit to 0.1 W for EPS below 49 W and 0.21 W for EPS above 49 W and below 250 W. The efficiency limits were increased, especially for lower powers, where an approximate 5.3 percent boost was introduced at 10 W compared to Level V, setting the limits at the 81.9 percent and 78.7 percent, respectively, for standard and LV EPS. This efficiency boost was around 3.5 percent at 30 W which sets the limit at the 86.9 percent and 85.1 percent, respectively, for standard and LV EPS. The efficiency limits then increase nonlinearly and reach to 88 and 87 percent at 50 W for standard and LV EPS, respectively, and stays flat up to 250 W with a 1 percent jump comparing to the Level V.
Moreover, the efficiency limits for EPSs above 250 W, multiple output EPS and AC/AC EPS were established in Level VI for the first time. Level VI does not apply to indirect operation EPS, where the EPS cannot operate the end product without the assistance of a battery. Additionally, Level VI exempts direct operation EPSs where the Federal Food and Drug Administration (FDA) approval is required, or where the EPS has output voltage less than 3 V and output current of greater or equal to 1 A that charges the battery of a product that is primarily or fully motor-operated.
In order to meet the Level VI requirements, power supply manufacturers had to eliminate some features, such as shortening the length of the cable and removing the LEDs. However, after advancements in LED technology, which decreased the power consumption per lumen, LEDs came back again. The tight Level VI limits also pushed IC manufacturers to design new ICs with modern control schemes to improve the efficiency especially at light load conditions (below 25 percent) where EPSs were historically struggling with low efficiency. Upgrading the ICs imposed layout change and recertification costs to original equipment manufacturers (OEM), where the EPS is a part of a more extensive system because the whole system must pass the safety and quality tests such as EMI.
In 2014, the European Union (EU) introduced the voluntary Code of Conduct (COC) Tier-1 to harmonize with the U.S. Level VI standard. Some differences with the U.S. Level VI are less stricter no-load power requirements, up to 3 percent less stringent efficiency requirement below 40 W, and 1 percent stricter requirement for EPS above 49 W and below 250 W. There are no efficiency requirements for EPS above 250 W, and Tier-1 does not have any requirements for AC-AC EPS, and unlike Level VI, Tier-1 does not differentiate between direct and indirect operation EPS. However, the main difference is adding a new 10 percent load efficiency measure where most EPSs have been dramatically inefficient. The EU further tightened the limits for the no-load power consumption and active mode efficiency by COC Tier-2 standard which was introduced in 2016 as a voluntary guideline like Tier-1. The EU COC Tier-2 has a slightly higher limit for active mode efficiency, especially for EPS above 25 W, and cuts the no-load power consumption by 25 to 30 percent compared to the DOE Level VI. The EU COC Tier-1 and Tier-2 had been initially scheduled to become effective in 2017 and 2018, respectively. As of the time this article was written, they have not yet been enforced, but it may happen soon. Figure 2 compares DOE Level VI and EU COC standards.
The EPS efficiency requirements have been evolving and the next DOE standard update is due in a few years. We should expect the limits to become even tighter and potentially some categories to be added in the upcoming standard. One possible update would be increasing the efficiency limits and reducing the no-load power loss to harmonize with the EU COC Tier-2 standard. Power supply manufacturers have already been investigating new technologies such as using wide-band-gap semiconductor devices, such as Silicon carbide (SiC) and Gallium nitride (GaN) semiconductors, with superior switching and thermal capability compared to the conventional Si-based semiconductor devices, advanced soft-switching techniques, and better packaging and thermal management systems. The next option could be introducing efficiency limits for very light load operating conditions like COC Tier-1 and Tier-2 standards. The light load efficiency limit in the EU COC standards are set at 10 percent load, but DOE might set the measurement at a different load condition between 5 to 20 percent. Although some applications usually run at light load (i.e., laptop computers), light load operation might not be a typical operating condition for other applications.
Another possible new category could be power factor (PF) limit in different load conditions where EPSs usually have poor performance over much of their operating range. PF is a measure of current quality, and for instance, an EPS with a PF of 0.5 draws two times the current compared to an EPS with a PF of 1. Improving the PF has energy-saving benefits both on the customer side and the utility side. To comply with this potential PF limits, EPS manufacturers might have to use PF correction circuits and/or EMI filters which imposes additional cost to the system. However, the limits mentioned above are toward reducing energy waste and air pollution, which benefits the whole global community, and we should adapt our products with them and welcome them despite the potential cost impacts.
Table 1: DOE Level VI standard performance thresholds. |
||
Single-Voltage AC-DC EPS1, Standard Voltage |
||
Nameplate output power (Pout) |
Min Average Efficiency in Active Mode (decimal) |
Max Power in No-Load Mode(W) |
Pout ≤ 1 W |
≥0.5× Pout +0.16 |
≤ 0.100 |
1 W<Pout ≤ 49 W |
≥0.071 × ln(Pout ) – 0.0014× Pout + 0.67 |
≤ 0.100 |
49 W<Pout ≤ 250 W |
≥0.880 |
≤ 0.210 |
Pout >250 W |
≥0.875 |
≤ 0.500 |
Single-Voltage AC-DC EPS, Low Voltage2 |
||
Nameplate output power (Pout) |
Min Average Efficiency in Active Mode (decimal) |
Max Power in No-Load Mode(W) |
Pout ≤ 1 W |
≥0.517× Pout +0.087 |
≤ 0.100 |
1 W<Pout ≤ 49 W |
≥0.0834 × ln(Pout ) – 0.0014× Pout + 0.609 |
≤ 0.100 |
49 W<Pout ≤ 250 W |
≥0.870 |
≤ 0.210 |
Pout >250 W |
≥0.875 |
≤ 0.500 |
Single-Voltage AC-AC EPS3, Standard Voltage |
||
Nameplate output power (Pout) |
Min Average Efficiency in Active Mode (decimal) |
Max Power in No-Load Mode(W) |
Pout ≤ 1 W |
≥0.5× Pout +0.16 |
≤ 0.210 |
1 W<Pout ≤ 49 W |
≥0.071 × ln(Pout ) – 0.0014× Pout + 0.67 |
≤ 0.210 |
49 W<Pout ≤ 250 W |
≥0.880 |
≤ 0.210 |
Pout >250 W |
≥0.875 |
≤ 0.500 |
Single-Voltage AC-AC EPS, Low Voltage |
||
Nameplate output power (Pout) |
Min Average Efficiency in Active Mode (decimal) |
Max Power in No-Load Mode(W) |
Pout ≤ 1 W |
≥0.517× Pout +0.087 |
≤ 0.210 |
1 W<Pout ≤ 49 W |
≥0.0834 × ln(Pout ) – 0.0014× Pout + 0.609 |
≤ 0.210 |
49 W<Pout ≤ 250 W |
≥0.870 |
≤ 0.210 |
Pout >250 W |
≥0.875 |
≤ 0.500 |
Multiple-Voltage EPS4 |
||
Nameplate output power (Pout) |
Min Average Efficiency in Active Mode (decimal) |
Max Power in No-Load Mode(W) |
Pout ≤ 1 W |
≥0.497× Pout +0.067 |
≤ 0.300 |
1 W<Pout ≤ 49 W |
≥0.075 × ln(Pout ) + 0.561 |
≤ 0.300 |
Pout >49 W |
≥0.860 |
≤ 0.300 |
1-Single-Voltage AC-DC EPS: An EPS that is designed to convert AC grid voltage into lower-voltage DC output and can only convert to one DC output voltage at a time.
2-Low-Voltage EPS: An EPS with a nameplate output voltage less than 6 volts and nameplate output current greater than or equal to 550 milliamps. Standard voltage external power supply means an EPS that is not a low-voltage EPS.
3-Single-Voltage AC-AC EPS: An EPS that is designed to convert AC grid voltage into lower-voltage AC output and can only convert to one DC output voltage at a time.
4-Multiple-Voltage EPS: An EPS that is designed to convert AC grid voltage into more than one simultaneous lower-voltage output