Currently, there are several efficiency certifications, based on very basic methodologies that leave much room for questionable results. The significant downsides that we spot in the current efficiency certifications are the following:
A limited number of sample measurements: All current efficiency certification programs only take a limited number of efficiency readings to classify the under test PSU into one of their categories. This simply means that anyone can easily tune a PSU to perform best under those specific load levels, sacrificing efficiency on all the rest. Also, those plain methodologies can be easily fooled by Golden samples, which perform much better than the normal retail ones with 10%, 20%, 50% and 100% of their maximum rated capacity loads. This is something not to be taken lightly because for example a supposedly higher efficiency level PSU can have lower overall efficiency than a lower certified PSU, while normally it should be the other way around. Our approach allows the calculation of an overall efficiency rating that covers the entire operating range of the PSU. This means that the final result besides highly accurate, also takes into account all possible usage scenarios. A single efficiency rating number takes away the complexity of our test procedures, making it much easier for all consumers the decision on their PSU purchase, according to their efficiency needs.
Very low ambient temperature: The current efficiency certification programs conduct all measurements under a very low ambient, which in the majority of cases is restricted at 23 °C. Such a low ambient is highly unrealistic for the conditions that usually apply inside a PC chassis. As the operating temperature increases, the performance of a PSU deteriorates, and this affects efficiency. In some units, this performance deviation can be small, but some PSUs are greatly affected by higher operating temperatures.
The current efficiency certification programs don’t take into account the power consumption at standby: Every PSU that it is connected to the mains network consumes energy, even when it is shut down. You see, the +5VSB converter inside a PSU is continuously working and the only way to shut it down is to either remove the power source from the PSU or switch it off through its AC power switch (if it has one of course). The problem with this converter is that even with no load at all on its outputs it still consumes energy (like all converters). The amount of this energy is called phantom or vampire power since it goes wasted. According to the ErP Lot 6 2013 directive, every power supply that is sold in the EU market should have less than 0.5 W power consumption in standby mode.
No 5VSB measurements: There are no efficiency measurements on the 5VSB rail in the most popular efficiency certification program, so we can find PSUs with high certifications to have lower efficiency 5VSB rails compared to PSUs with lower efficiency certifications. This is something unacceptable since when a user invests in a high-efficiency PSU he/she expects it to be highly efficient on all rails, including of course the 5VSB one.
Fake efficiency badges: Currently, the most recognizable efficiency program doesn’t have an effective method to deal with some manufacturers that use fake efficiency badges. This is something of immense importance since it can mislead consumers into believing that they buy an efficient PSU. There are methods, thankfully, to track any of the performance rating badges that an organization provides, which we are going to exploit in our methodology.
Given the deficiencies mentioned above that the current efficiency certification programs have, we believed that it was high time to propose something better and more modern, addressing the increasing demands of our time. The older approaches clearly show their age, and their inventors aren’t able to make dramatic changes now since this would be the end of their entire databases and all PSUs should be re-evaluated. A new methodology is required for measuring accurately a PSU’s efficiency, which can be easily modified in the future without sacrificing the entire database of test results. According to our experience, the more data a test methodology offers, the easier it is to modify/adjust it when this is needed.
