Seasonic Platinum Series 860W - Platinum without the Fuss
by Martin Kaffei on February 24, 2012 12:00 AM ESTVoltage Regulation
| +3.3V Regulation/Ripple and Noise | |
| Load | Voltage |
| 5% | +1.28% (5mV) |
| 10% | +1.24% (6mV) |
| 20% | +1.12% (8mV) |
| 50% | +0.24% (10mV) |
| 80% | -1.23% (11mV) |
| 100% | -1.67% (14mV) |
| 110% | -1.79% (14mV) |
| Crossload +12V max. | +0.61% |
| Crossload +3.3V/+5V max. | +0.00% |
| +5V Regulation/Ripple and Noise | |
| Load | Voltage |
| 5% | +1.20% (7mV) |
| 10% | +0.82% (7mV) |
| 20% | +0.40% (9mV) |
| 50% | -0.02% (11mV) |
| 80% | -0.94% (12mV) |
| 100% | -0.92% (12mV) |
| 110% | -0.86% (13mV) |
| Crossload +12V max. | +0.28% |
| Crossload +3.3V/+5V max. | -0.04% |
| +12V Regulation (Worst Rail)/Ripple and Noise (Worst Rail) | |
| Load | Voltage |
| 5% | +1.13% (4mV) |
| 10% | +1.04% (5mV) |
| 20% | +0.98% (22mV) |
| 50% | +0.48% (16mV) |
| 80% | +0.01% (20mV) |
| 100% | -0.31% (24mV) |
| 110% | -0.67% (26mV) |
| Crossload +12V max. | -0.17% |
| Crossload +3.3V/+5V max. | +0.13% |
Noise Levels
| Loudness | |
| Load | Opinion |
| 5% | low fan noise and small electrical side noise |
| 10% | low fan noise and small electrical side noise |
| 20% | low fan noise and small electrical side noise |
| 50% | fan noise and small electrical side noise |
| 80% | strong fan noise and small electrical side noise |
| 100% | strong fan noise and small electrical side noise |
| 110% | strong fan noise and small electrical side noise |
Efficiency and PFC
| Efficiency (AC input minus DC output) and Power Factor | ||
| Load | Efficiency | PFC |
| 5% | 79.83% | 0.712 |
| 10% | 86.53% | 0.898 |
| 20% | 90.77% | 0.942 |
| 50% | 93.03% | 0.979 |
| 80% | 91.99% | 0.988 |
| 100% | 91.40% | 0.989 |
| 110% | 91.27% | 0.990 |
The ball bearing is (almost) inaudible under high loads. During low loads you might hear the PFC choke if you are close enough, but this shouldn't be a problem when the PSU is mounted in a closed case. If desired the fan can be turned off through up to 40% load by a small switch on the front of the casing, which would make the PSU completely silent at lower loads (other than the electrical noise we noted).
Efficiency is much higher than expected. Seasonic is well above the requirements for the 80 Plus Platinum certificate (90%/92%/89% efficiency at 20%/50%/100% load). Even at lower loads Seasonic is able to surpass most results from competing products. In addition the Platinum Series 860W performs much better than what we saw from the Enermax Platimax 750W. It's only at 5% load that this PSU is below 80% efficiency, and even then it's basically a rounding error. So far this is an exceptionally good result; PFC could be a little better, but we don't want to overdo things.
During all loads voltage regulation is simply perfect. Even under crossload conditions all voltages are close to their optimal values. -1.67% is the "worst" result on 3.3V while 12V is much better (-0.31% as long as we stay at <100% load). Another highlight is the low ripple voltage. We would like to see such results more often, since 26mV on 12V is only 0.22% ripple and noise—according to the ATX specification 1% is allowed. Enermax had no problems with this level as well and makes stable outputs too, but ripple was noticeably higher. Given these results, Seasonic is clearly the winner in our head-to-head comparison.
Facebook
Twitter
RSS
69 Comments
View All Comments
kmmatney - Friday, February 24, 2012 - link
Yup. I'd really like one of these, but find it hard to justify the big price jump to get to platinum efficiency. You can get all the PSU you need (Corsair GS800) for $115. It is nice, though. I'm still running a Corsair TX650 which I bought 5 years ago for $90 or so.Hrel - Friday, February 24, 2012 - link
Most seasonics have NO noise under about half load. Why is this one noisy? Also what's with the "electrical noise". I don't wanna hear a damn thing. "Best power supply ever", psh, not if I can hear it.palindrome - Friday, February 24, 2012 - link
Martin,It is hard to take much stock from your review. Especially when you write a garbage conclusion that claims Super Flower products a bad product because of their name. The last Super Flower manufactured PSU reviewed on Anandtech was 10/16/2008 by Christoph Katzer. Also, your testing equipment and methodology is not listed in this article. For all we, as readers, know, you could be using a kill-a-watt and a RadioShack multimeter. You should also list your error tolerance for your readings, no matter what equipment you are using. When I read this article, I thought I might have mistakenly been redirected to Tom's for a moment.
I'm sure reviewers such as yourself all have opinions and incentives to write nice reviews for the generous companies who give you free, expensive toys to play with. However, you should let your test results do the talking and provide some of the basic information, such as methodology, just like we all learned in 5th grade science class, so that you can be much more credible. This isn't the worst PSU review I've seen from a major tech site (Tom's has that locked down tight), but I certainly expect more from Anandtech.
Time to get flamed by the herd.
smilingcrow - Friday, February 24, 2012 - link
We’re not worthy.I have the 400W Gold fanless version and we’ve set a date for the autumn.
Crypticone - Friday, February 24, 2012 - link
I would really like to see some real world usage testing. Clearly every uses computers in a different manner and I am not sure the best way to go about it.Much like you have a very standardized testing bed for computer cases. It would be nice to have some measure of power usage over a certain period of time.
Using the same computer setup run the system for two hours at idle speed measuring power usage. Then two hours looping some gaming type benchmark. Then two hours of scripted web browsing. Then an hour fully loading the CPU with prime. Then maybe two hours of streaming media to a media player of some sort.
Then it would be nice to see how much power is saved between the units. It would be interesting to see where the mock breakeven point is power savings vs the added cost of these new pricey power supplies vs a gold or silver rated PSU.
This might be more of a feature article then something done for every PSU you review.
John Doe - Friday, February 24, 2012 - link
That SuperFlower you just threw off in a whim makes a unit that outdoes the Platinum 1000; the LZP-1000.http://enthusiast.hardocp.com/article/2012/01/04/k...
It gets about 30mV ripple against the 60's of the SeaSonic, has quicker transient turn-on response and sells for $70 cheaper.
One of the main goals of a review is to be subjective, which is what this review is not. This is a load of garbage.
smilingcrow - Saturday, February 25, 2012 - link
From the review you linked to:“If you want the absolute best power supply we have seen then you still want the Seasonic Platinum-1000.”
And you want a review to be objective not subjective!
Martin Kaffei - Saturday, February 25, 2012 - link
John,I got an XFX PSU which is exactly like the 1KW version from Seasonic and I never saw more than 30 mV as well.
Does HardOCP measure with capacitors conformable to ATX specification? You need them to simulate system load, otherwise the results do not apply to a real PC.
Hereunder Seasonic might be worse than SuperFlower, but your PC doesn't get the pure ripple. With ATX test environments Seasonic is always as good as or even better than SuperFlower.
John Doe - Saturday, February 25, 2012 - link
What're you on about? Yes, Paul puts all the units in an incubutor for at least 8 hours. And you? Have you even explained your test methodology? You aren't making any sense and aren't fooling anybody with little SMPS knowledge.If you look at Oklahoma Wolf's results, you can find the same. The Golden King 1000 (which you labeled as a joke just because of it's name) usually has lower ripple, better transient response, better cooling and sells for cheaper. The only thing the SeaSonic has over it is better regulation.
It's clear that either you're unknowledgeable or you have a personal agenda, with the latter making more sense. Excuse my bluntness, but you're diviving into pure nonsense.
DanNeely - Saturday, February 25, 2012 - link
You've got 28 pins on the PSU for a 24pin cable.