## HARDWARE SECRETS

Uncomplicating the complicated

[nextpage title=”Introduction”]

Efficiency is traditionally an overlooked power supply specification. It says how much power is being wasted while you are using your PC. Problem is that you are paying for this wasted power. The 80 Plus certification was created to allow consumers to know which power supplies are the most efficient ones and, as the name implies, guarantees that the power supply is able to present efficiency of 80%. With more and more manufacturers trying to design power supplies with efficiency well above 80%, they decided to create three new certification categories: Gold, Silver and Bronze. Learn what they really mean.

First let’s give you some background on efficiency. Efficiency, which is also represented by the Greek letter Eta (η), is given by dividing the power being delivered to the equipment (output power, i.e., DC power) by the power being consumed by the power supply (input power, i.e., AC power).

If you have a PC that consumes 250 W and a power supply with 75% efficiency, this means that you are pulling (and paying for) 333 W from the wall. If the same computer had a power supply with 85% efficiency, it would be pulling 294 W from the wall and you would be saving 39 W. Thus a power supply with higher efficiency will allow you to save money on your electricity bill.

Power supplies do not present a constant efficiency. The graph for efficiency is a bell curve, where the power supply presents its best efficiency when delivering 50% from its labeled load. We have an example in Figure 1.

Figure 1: example of an efficiency curve

Because of this effect it is recommended that you buy a power supply with double the power you are actually going to pull. This explains the offer of high-wattage power supplies above 700 W. Manufacturers don’t expect you to pull the full power from their units, but that you operate them around 50% load for a higher efficiency (during our reviews, however, we need to see if the power supply can really deliver its labeled power, because if a power supply is labeled as, let’s say, 600 W unit, we want to be capable of pulling 600 W from it, if we want to). The only disadvantage to this approach is the price of a higher wattage unit. But at the long run it is a good idea, as you will save money on your electricity bill, you computer will run cooler, you have enough headroom for a future upgrade and you won’t face any stability problems when playing games at their maximum quality for hours.

The second thing you need to know about efficiency is that power supplies present a higher efficiency when connected to a 230 V (“220 V”) power grid, and the efficiency numbers announced by manufacturers are measured at this voltage. Thus if you live in a country or region where the power grid is 115 V (“110 V”) – like the United States – your power supply is likely to present an efficiency lower than the number announced by the manufacturer. Our power grid is 115 V and in our reviews we test power supplies under this voltage.

[nextpage title=”The 80 Plus Certification Methodology”]

The 80 Plus certification is conducted by an independent laboratory. Manufacturers willing to get their products certified need to send their samples and pay for them to test the units. This fee also allows them to use the 80 Plus logo if the units pass the certification process.

Like us, they only test one sample from each product. Manufacturers are advised to send at least two samples to their lab, but the extra samples are used only if the first sample fails.

It is important to take a closer look at how they test power supplies. It is very similar to the tests we do while reviewing power supplies – i.e., plug the power supply to a given load, measure how much power the power supply is pulling from the wall and voilá, we have the efficiency number – but with some important differences.

The good:

• 1. Although their methodology paper says that can test power supplies under 115 V or 230 V, they usually test power supplies at 115 V, so numbers won’t be inflated by the use of a 230 V power grid. The exception is power supplies targeted to datacenters, that can only work at 230 V and thus this is the voltage they use to evalute this power supply class.
• 2. To be certified power supplies must present a power factor (PF) of at least 0.90 (0.95 for the Platinum and Titanium certifications). Click here to understand what this is. We measure this in our tests.

• 1. Their tests are conducted inside a thermal chamber with a constant temperature of 23° C (73.4° F) ±5%. This is ridiculous as no computer in the world works internally at such low temperature. The problem is that as temperature increases power supplies start consuming more from the power grid in order to deliver the same amount of power on their outputs, so efficiency typically decreases with temperature. Our tests here on Hardware Secrets are conducted with a temperature between 45° C and 50° C (113° F and 122° F) inside our thermal chamber, as we want to measure power supplies under real-world conditions.
• 2. Power supplies are tested only under three loads: 20%, 50% and 100% (called “light,” “typical” and “full,” respectively). At one hand the use of these three loads is enough for having an overall idea of the power supply efficiency. On the other hand, for a more precise measurement it is our opinion that they needed to do tests under several different loads, especially when they are charging for doing so. In our tests we test power supplies under five different loads: 20%, 40%, 60%, 80%, and 100%. On the other hand, the new Titanium certfication provides minimum requirements for 10% load, which is excellent.
• 3. They don’t disclose the exact equipments (e.g., manufacturers and models) they use on their testing.

Let’s see now the differences between the existing 80 Plus certifications.

[nextpage title=”The Available Certifications”]

The difference between the six 80 Plus certifications available is summarized in the tables below. The numbers presented are the minimum efficiency a power supply must present under each load in order to be granted a given certification. As explained in the previous page, their numbers may be different from the numbers we see during our tests due to differences on the measuring equipment and especially on the room temperature (23° C on their tests vs. 45° – 50° C in our tests).

The requirements for each certification level depends on the market the power supply is targeted to. Non-redundant power supplies (i.e., the kind everybody uses) are tested at 115 V and use the table below. The Titanium certification is the only one that has a minimum requirement for 10% load.

 115 V Non-Redundant 10% Load (Ultra Light) 20% Load (Light) 50% Load (Typical) 100% Load (Full) Power Factor 80 Plus Standard NA 80% 80% 80% 0.90 at 50% 80 Plus Bronze NA 82% 85% 82% 0.90 at 50% 80 Plus Silver NA 85% 88% 85% 0.90 at 50% 80 Plus Gold NA 87% 90% 87% 0.90 at 50% 80 Plus Platinum NA 90% 92% 89% 0.95 at 50% 80 Plus Titanium 90% 92% 94% 90% 0.95 at 20%

Requirements for redundant power supplies (used on servers), however, are slightly different, as you can see in the table below. Also, they are tested at 230 V, because this is the voltage used in datacenters (datacenters use a 230 V power grid basically because power consumption is lower at 230 V and since they have hundreds of servers running, they can save money in their electricity bill).

 230 V Redundant 10% Load (Ultra Light) 20% Load (Light) 50% Load (Typical) 100% Load (Full) Power Factor 80 Plus Standard NA NA NA NA 0.90 at 50% 80 Plus Bronze NA 81% 85% 81% 0.90 at 50% 80 Plus Silver NA 85% 89% 85% 0.90 at 50% 80 Plus Gold NA 88% 92% 88% 0.90 at 50% 80 Plus Platinum NA 90% 94% 91% 0.95 at 50% 80 Plus Titanium 90% 94% 96% 91% 0.95 at 20%