Thanks, Acoustic Testing and FunniesThe lovely trio of Grace Liang, Jessica Yu and Ceres Wang were always around at FSP to give us a hand with logistics if we ever needed anything, with Jason Wei our technical contact and James Wu, a marketing manager at FSP, around to discuss how the PSU world turns.
Posing with PD on the final day of testing as we shot the PSUs as a group for the article, Jessica, Ceres and James were very camera shy. Standing next to PD does that to a person! Of course PD would be upset if a shot of me didn't get in to this piece somewhere, so here's a picture of the monkey that did all the testing. Note the miniature nuclear reactors in the background (pressure chambers in reality) used to give me the power needed to run 16000W of PSUs at the same time.
Many thanks to everyone at FSP, especially Grace and Yi back in the UK, for letting us cordon off a section of their Taoyuan lab for 5 days so we could test a bunch of PSUs in peace and relative quiet.
Acoustic TestingOn the final day of testing, Jason Wei invited us to test a power supply in their quarter million dollar acoustic testing facility. The custom built anechoic chamber is setup to allow IEEE certified testing of noise levels. FSP use the chamber for obvious reasons, verifying that their PSU designs are within acceptable acoustic limits. Stepping inside the chamber and closing the massive door behind you is a surreal experience. If you've never experienced true silence before, it completely scraps all your previous notions of what's quiet and what's not. When you're inside and you hold your breath, you can't hear a damn thing. It makes you appreciate just how much background noise there is in everyday life that your brain just tunes out.
I can't describe it other than a complete nothingness, devoid of any sound whatsoever. To be as deaf as you think you are in the chamber would be heartbreaking for me, as a lover of music. Speaking to someone in the chamber is similarly surreal. The walls of the chamber are designed to cancel out noise as much as possible. Voices therefore tail off much quicker in the chamber and it almost feels as if the person you're talking to is 10x as far away. It's a very weird place to be!
You can see the internal chamber design (as few reflective surfaces as possible, so sound waves can't bounce around and be amplified) with James and another engineer making sure the PSU is exactly 1 metre away from the recording mic during testing. The complete test system is able to monitor a PSU over a long period of time, under many test conditions. For brevity, we were allowed to test a pair of power supplies in the chamber under full load, to see what noise (or not) they really made.
First up was FSP's FSP300-60GNF, otherwise known as their Zen fanless 300W PSU. Utterly devoid of any moving parts, it made for a highly interesting experiment. The anechoic chamber has a baseline silence level of just 15.5dBA. To put that into perspective, the ambient noise level in an average urban area is ~45dBA and outdoors in a wilderness with nothing around to make a sound is ~30dBA.
Given my description of the chamber above, I hope you can understand how quiet that actually is. The decibel scale is designed around the human hearing system, so 0dBA is the absolute lowest we can perceive, technically (even though it's not the lowest sound that something can make!). The FSP300-60GNF posted a test result, at full 300W load, of 16.0dBA. Here's the proof, click it for the full version.
It literally makes next to no noise, and you'd really never be able to hear it in your PC system. Next up was the AOpen AO700-12ALN engineering sample. A 700W PSU with single 120mm cooling fan, its results were similarly suprising.
Illingworth & Rodkin 2002 states that 30dBA is how loud a soft whisper is in a quiet bedroom. A 700W PSU that's a 'soft whisper' at full load is fine by me. I'm sure it's fine by you too.
FunniesYou don't spend a week in PD's company without hilarity ensuing. As I tested the QTec Triple Fan 650W, PD was sat nearby working on some copy for the site. He heard me laughing as the input power shot up to 900W for just 400W of output power. "Video it, Rys!". So I did. The following download is the last few seconds of that PSU's life, in input power terms, before I turn the camera off and stand back from the test bench to watch it pop. I really wish I'd caught it on camera for you, but I didn't want to stand too close as nearly 1100 watts made that piece of junk go bang.
Click the image below to download the ~600KiB .avi file. XviD is the encoder, so a recent FFDShow will help you play it if you can't already.