I installed the LitePower 450W in an oversized midtower case with bottom-mount power supply. This is probably not a typical application for this particular model, and as a result many of the cables came up a bit short. The worst was the +12v P4 cable, which had to be stretched tight across the PCI slots and under the video card in order to reach the connector, thus precluding the use of any of the other slots. This is a common location on many motherboards so this problem will likely be seen on any case with the power supply located at the bottom. The other cables that were close were the main ATX cable and the accessory Molex cables to reach the top-mount fans and optical drive. If you happen to have a SATA optical drive you will likely be stuck using a Molex-to-SATA power adapter, as the single SATA cable can't possibly reach both the top and bottom of the case simultaneously.
The test system in this setup consists of:
AMD X2 4000+ @ 2.9Ghz 1.45v
Biostar 690G AM2 Motherboard
2x1Gb A-Data PC6400
XFX GTS 250
Measuring voltages with my Fluke multimeter directly from the connectors, and using a Kill-A-Watt meter to observe the wattage consumed, I recorded the three main rails both under idle and load conditions, which were 110W and 260W respectively. All three voltages actually read higher than spec, and fluctuation was limited to the +12v rail, and that only by +/-0.01v. I ran Windows Defrag, OCCT and Furmark simultaneously in order to generate the load conditions.
OCCT records the voltages from the motherboard sensors, which typically read lower than the actual power supply output at the connector. The graphs show the result at one minute into the log when the system is loaded, the vcore drops nearly 0.02v, the +3.3v comes down about the same, while the +12v loses anywhere from 0.1 - 0.5v in a highly fluctuating reading.
Let's wrap up the LitePower 450W review.