Voltage Regulation and Quality

+3.3V regulation
Load Voltage
10% +1.72%
20% +1.64%
50% +1.09%
80% +0.81%
100% +0.15%
110% +0.04%

 

+5V regulation
Load Voltage
10% +2.43%
20% +2.22%
50% +1.82%
80% +1.65%
100% +1.16%
110% +0.90%

 

+12V regulation
Load Voltage best/worst
10% +1.73% / +1.91%
20% +1.66% / +1.82%
50% +1.46% /+1.61%
80% +1.24% / +1.48%
100% +1.09% / +1.30%
110% +0.84% / +1.03%

Results are measured under room temperature and 115 VAC. HuntKey has one of the best voltage regulation results out of all the high-end PSUs we've tested so far. The DC-to-DC VRMs are more than enough for the power levels listed on the label, so the voltage has low drops on +3.3V and +5V. Even +12V (V1/V2) is absolutely stable. Of course, the need to only handle 300W helps here. Some higher wattage PSUs with DC-to-DC have problems with ATX spec loads, since the VRM PCBs use almost the same dimensioning for their components.

Ripple and Noise

+3.3V ripple quality
Load Ripple and noise
10% 2mV
20% 3mV
50% 4mV
80% 6mV
100% 9mV
110% 10mV

 

+5V ripple quality
Load Ripple and noise
10% 8mV
20% 10mV
50% 13mV
80% 17mV
100% 21mV
110% 23mV

 

+12V ripple quality (worst rail)
Load Ripple and noise
10% 19mV
20% 26mV
50% 30mV
80% 22mV
100% 26mV
110% 28mV

+12V has the same problem as the first X-Series with relatively high ripple and noise results at low loads and lower results at high loads. Somewhere between these loads they change the switching type and frequency. Even so, the results are still very good and well within specification. Moreover, most manufacturers don't care about the "unimportant" 3.3V rail, but HuntKey has no more than 10 mV ripple there. Even if it's not the most important rail these days, we appreciate the attention to detail.

Internals and Components Noise, Efficiency, and Power Factor
Comments Locked

46 Comments

View All Comments

  • Sabresiberian - Tuesday, March 8, 2011 - link

    Why can't Anandtech ever run PSUs that normal people use? No one cares about these low-power pieces of junk!

    /end sarcasm

    (To those who don't get it, this is a jab at all those posters who whine and cry when Anandtech runs articles on high-powered PSUs and make the same kind of comment. Anyone who has read more than one PSU article on Anandtech knows they do high quality reviews of PSUs that are across the board in power rating and usefulness to people who buy them, or just want to know more about them. Hats off and /bow to Martin for once again giving us something interesting and informative.)

    ;)
  • MrSpadge - Tuesday, March 8, 2011 - link

    Would be nice to see efficiency measured for us European / Australian / whatever people as well. I know at ~220 V efficiency should be 1 - 2% higher, but it's still nice to see the actual measurement.

    MrS
  • Tinhead - Wednesday, March 9, 2011 - link

    Last 9months i have heard more and more friends with PSU problems.
    Most of their PSU's are about 2-5years old and they either die on some "rails" or completely.
    Most of them has been used for servers, on 24/7 etc..

    One thing im wondering about is how long should we expect to get a PSU to run?
    Can we expect 24/7 for 5years or 2years?
    The parameter is not that popular since it is hard to test, the users that are driving the enhancements are considered mostly to be gamers, and they like power and efficiency more then MBTF.

    Another factor is how does the PSU preform with under or over voltage?
    And ofc how it handles spikes, and other "dirt" in the wires. Which is more common then most users know.
    (one reason why i use a UPS to my equipment.)

    Also it would be interesting if you can test with 220V AC since that is a common standard and it does affect the performance and operation of the PSU.
  • Martin Kaffei - Wednesday, March 9, 2011 - link

    I know what you mean. AC stress is a problem as well as a good way to test power supplies. In fact it is usually easier for a PSU to handle DC loads than transients. Nevertheless it is indispensable to have an AC source with a clean output for better comparability in all tests. Later that year, I hope the AC side gets more attention after adding some components.

    24/7 ist basically no problem for a PSU but the lifetime depends on many factors. capacitor aging is a problem, EMI immission is, temperatures are. In my opinion servers should get other PSUs than usual PCs. The letter will usually be changed after 2 or 3 years, so voltage/ripple/EMI during a relative short time is more important than endurance (which includes "long life" capacitors). Some older PSUs died in my PCs after a year or two, but I'm sure the quality is better today if you buy known brands. Specially brands who have to make sure, that the PSU works 5 years because of their warranty program. But that's no guarantee for all products. We aren't able to make long time tests over a year, even manufacturers can't do that. Their "long time" tests take a week and the rest is just math.
  • HangFire - Wednesday, March 16, 2011 - link

    >One thing im wondering about is how long should we expect to get a PSU to run?

    The capacitors define the lifetime of a quality power supply. They are rated for a certain number of years at a given temp. Typical numbers are 3 and 5 years as you go up the price ladder. Unless there is a bad batch (it happens), quality Japanese capacitors can be counted on to last their rated lifetime.

    For very low end PSU's (and there are a lot of them out there), besides the capacitors it is components being run over-spec that will define the lifetime (burn out first). For example a 400W rated power supply made up of 250W level components will have one life expectancy at 150W and another at 250W. You can read some recent reviews of such power supplies over on Hardware Secrets.

    The warranty is supposed to give you some idea of what to expect, but not really. Seasonic's lowest end series have a 1 year warranty but you can expect them to be useful much longer. Diablotek offers 3 years on some of theirs, but recent published reviews cast doubt on their lifetime if used at their rated capacity.
  • Onus - Friday, March 11, 2011 - link

    After seeing Gabriel blow up many overrated Huntkey pieces of dung, I'm leery of the brand, but this is the sort of PSU I'd like to see more often. I've built a couple of project PCs to test various ideas, and a PSU like this would be perfect for a mini-ITX gamer, running a HD5770 or GTX460. I'd be worried about ventilation with a fanless PSU in that application, but this would be good. Hopefully they'll do a modular version.

    The reason GPU manufacturers list outrageously high minimum wattages is they are trying to account for all the Chokemax PSUs out there that can barely manage 60% of what's on their labels.

    With the 80+ Bronze EA-380D at $45, this can't cost any more than $60 if Huntkey expects it to sell. In this wattage range, an efficiency difference of even 10% doesn't amount to all that much.

Log in

Don't have an account? Sign up now