Articles: Cases/PSU
 

Bookmark and Share

(3) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 ]
Cross-Load Voltage Stability

We’ve been disappointed with the +12V stability of the above-discussed OCZ ZX 1000W, but the Seasonic SS-850KM is even worse in this respect. The +12V voltage gets 4% off the required level. About half of all permissible loads (including our three reference configurations) are within the 3% zone.

The +3.3V voltage is more stable, deflecting by more than 2% only when there’s very low load on all of the power rails.

The +5V rail is in fact blameless. At load distributions typical of real-life computers, this voltage is only 1% off the required level when there is near-maximum load on the +12V rail.

But if we consider all the voltages, the Seasonic SS-850KM turns out to produce the worst result in this test among all the PSUs in this review. The reason is the poor stability of the most demanded +12V voltage. The better stability of the other rails can’t save the day for this PSU.

By the way, we do not think that this performance is an accident because we’ve seen a similar picture with the earlier-tested 660W and 760W products from the same series. This is a shame especially as the user manual claims that these PSUs are going to keep their output voltages within 3% of the required levels on every power rail.

Output Voltage Ripple

Seasonic’s top-end PSUs are blameless in terms of the output voltage ripple. The high-frequency ripple is very weak on each power rail.

The graphs are almost flat lines at the double frequency of the mains, too.

Temperature and Noise

Like the other products of this series, the Seasonic SS-850KM is cooled by a 7-blade 120mm fan Sanyo Denki San Ace 120 (the 9S1212F404 model with a rated speed of 2200 RPM). Such fans are very high-quality things, up to individual impeller calibration, and produce no unwanted sounds at work. All you can hear at any speed is pure aerodynamic noise.

The fan mostly remains idle at low loads, turning on at minimum speed for brief periods of time. It switches to constant operation at a load of 150 watts but works at a very low speed (from 300 to 550 RPM), so there is no discomfort for the user. The speed remains low until a load of 600 watts but then rises up rapidly. The fan becomes audible at a load of 700 watts. Reaching a speed of 1700 RPM at full load, the fan is still comfortable enough, thanks to its high quality.

However, you can see from the diagram that there is no need for such a rapid increase in speed at high loads. The fan accelerates so fast that the difference in temperature between the incoming and outgoing air shrinks. If the speed increased at a slower rate, just enough to keep the temperature at the same level, the Seasonic SS-850KM would probably be a record-breaking PSU in terms of acoustic comfort.

As it is now, the SS-850KM is similar to the Cougar PSUs in terms of maximum noise. The latter have a lower fan speed at high loads, but the Seasonic’s fan is somewhat smaller and doesn’t rattle.

Efficiency and Power Factor

The power factor of the Seasonic SS-850KM is almost as high as 99% at high loads.

The PSU was 90%, 92.6% and 89.5% efficient at the three reference loads of 20%, 50% and 100%. Its peak efficiency was achieved at 50% load. We can also add that it is 90% or more efficient in the load range of 160 to 780 watts.

Standby Source

The standby source does its job without any problems.

Summary

There are three problems we can find about the Seasonic SS-850KM. It is expensive, its fan accelerates too rapidly at high loads, and its +12V voltage is not very stable. If it were not for the last downside, the first one might be ignored because the competitors are obviously inferior across most parameters. As for the fan, PSUs do not often work at their full capacity in real-life computers, especially as most users prefer to buy PSUs with some reserve of wattage. It means that you are unlikely to hear the full sound of this PSU’s cooling fan in real usage scenarios.

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 ]

Discussion

Comments currently: 3
Discussion started: 03/19/12 07:59:00 AM
Latest comment: 04/08/12 03:07:54 PM

View comments

Add your Comment