X-bit labs

Three ASUS Power Supply Units Roundup (page 4)

Category: Other

by Oleg Artamonov

[ 07/07/2004 | 01:09 PM ]

Fan Rotation Speed

All the three units can automatically control the rotational speed of the fan (or fans, like with the A-30H) using the sensor on the heatsink with the diode assemblages. This control works properly as our measurements show – see the table below (we performed the test under 21°C room temperature; after setting the load in, we “warmed up” the PSUs for 15-20 minutes).

The dual-channel A-30H turns to be the quietest unit, while the A-30G couldn’t make a worthy competitor – the blades of its 120mm fan produced a distinctly perceptible buzz combined with the noise of the air stream. Of course, the cheaper A-30F couldn’t compete with the A-30H, either, since its fan sped up to about 3000rpm.

Well, the powerful fan in the A-30G is both a shortcoming and an advantage – depends on how you view it. Its Adda AD1212MS-A71GL fan creates a 80CFM air stream at the maximum rotational speed, and that’s more than twice higher above the capabilities of the fans in the A-30F (38CFM at the maximum speed) and A-30H (31CFM for the rear fan and 22CFM for the top fan). Thus, the A-30G will not only cool down itself, but will also contribute to the cooling of the entire system case.

Voltage Pulsations

We examined voltage pulsations in the three units at two frequencies: the frequency of the pulse-width regulator (several tens of kilohertz) and the double frequency of the mains supply (100Hz).

+5V bus, 10uS/div

+12V bus, 10uS/div

The amplitude of oscillation at the frequency of the pulse-width regulator was small, just over 15mV. It is negligible considering the acceptable levels on the +5V and +12V power rails (50mV and 120mV, respectively).

+5V bus, 4uS/div

+12V bus, 4uS/div

The oscillations at 100Hz frequency are somewhat worse – their swing was 40-50mV on the +12V rail and 20-25mV on the +5V rail. Anyway, these numbers are much below the acceptable ceiling, so there’s no need to worry. As for the reason for that, some flaws in the design of the PCB or of the power transformer may bear blame (the third possible cause – the insufficient capacitance of the high-voltage rectifier’s capacitors – evidently is not applicable here).