A look inside
When I first popped the cover, I thought this unit must have been used, looking at the PFC coil, but it’s not actually dust. It’s glue. With my initial panic attack out of the way, I can now start taking this unit apart and looking at the components. Starting at the input filtering, this unit has two X capacitors, two common-mode chokes, five Y capacitors (including the one after the rectifier) and an MOV. It’s all the recommended components plus an extra X-cap and two Y-caps. The unit has two 8A bridge rectifiers which are mounted to a heat sink. In theory, they would allow the unit to deliver 1508.8W from a 115V power grid assuming the 82% efficiency required for 80 Plus Bronze. Mounted to the main primary heat sink are two Infineon IPA60R190C6 PFC transistors rated at 20.2A and two Toshiba 2SK3934 Switching transistors (in two-transistor forward topology) rated at 15A. They are clearly more than good enough to allow this unit to deliver 600W. They are driven by a Champion Micro CM6802 PFC/PWM Controller. The PFC capacitor is a 390μF part made by OST.
The unit has four 12V rails, in the configuration that I assumed based on the colour coding of the wires. There are four groups of 12V wires, each with its own OCP shunt. Three of them are visible in the above left picture, while the fourth one is hiding under the wires. The solder side is fairly clean, although there are a few component legs which could have been cut a bit shorter. I’ve seen much, much worse, though.
The selection of silicon components is somewhat unusual on the 12V rail. Rather than using identical rectifiers like most manufacturers, In Win chose to use different parts. There are two PFC Devices PFR30L60CT 30A Schottky rectifiers for the direct rectification path and two PFR40L60CT 40A rectifiers for freewheeling. This gives us a theoretical maximum output of 140A for the 12V rails (with a 43% duty cycle; using a higher or lower duty cycle will reduce the maximum output), but of course, it has to be shared with the 3.3v rail. The 5V rail uses two PFR30L45CT 30A schottky rectifiers, which translates to a theoretical maximum of 60A, which is overkill for a modern PC, since all the power hungry devices eat from the 12V rail. The rectifier on the left end of the heat sink is an STPS2045CT. It’s used by the 5Vsb rail. The 3.3V rail has a separate DC-DC VRM module. It contains two STD85N3LH5 MOSFETs rated at 80A each and two 330μF Teapo polymer caps. The capacitors elsewhere are all made by the Taiwanese company OST, which is dissapointing. Not only are they not very reliable, but when they fail, they don’t often show any visible signs of it. The secondary monitoring chip used is a Weltrend WT7525. It supports Over-Voltage, Under-Voltage and Over-Current protections.
As I thought, the fan is indeed an ADDA *sigh*. At least the heat sinks are nice and chunky with plenty of surface area, so it won’t have to work too hard to keep the unit cool. It wasn’t very loud during the load testing. It was audible above 400W load, but still not what I would consider loud.