Seasonic X-Series 750W: Potent for the Price
by Martin Kaffei on April 13, 2010 11:13 PM EST- Posted in
- Cases/Cooling/PSUs
- Seasonic
- PSUs
X750 - Voltage Regulation and Quality
Looking at the graphs, the +5V rail could be regulated much better for an expensive power supply. In fact, it drops very close to the tolerance limit at full load. However, ripple is much more important and a 5% deviation is still ok with a high load—something you will never reach with a normal system. All the other voltages perform fine and are within specification. The all-important +12V rail is still right at 12.00V even at 100% load.
When Seasonic switches to the resonant topology on higher loads, ripple and noise get much lower and overall they can keep up with Enermax. +12V shows a maximum 24.4mV ripple and the smaller rails are always below 20mV. Clean voltage helps all PC components last longer, so these results are very impressive.
Facebook
Twitter
RSS
38 Comments
View All Comments
MadMan007 - Wednesday, April 14, 2010 - link
I hope they come out with some milder wattage units for those of use running 'normal' systems with a single GPU and maybe some non-ragged edge CPU and GPU overclocks. 83% efficiency at 10% of load is great.7Enigma - Wednesday, April 14, 2010 - link
This was going to be my comment as well. I build average systems with the highest performance/$ so currently have a C2D with 4870 and will be looking in the next year to make the jump to quad cpu /5XXX gpu. With the fantastic idle consumption of both cpu and gpu the amount of time at or below that 20% level for this PSU is considerable.Give me the 500w version that is 85% @ 100w for <$125 and I'm sold.
jonup - Wednesday, April 14, 2010 - link
Genuin question (not being surcastic):What is the big obsession with efficiency? 85% vs 87+%? Does it really matter?
I use a OCZ's modular power supply and at full load (OCCT PSU) voltages remain stable, it's quite and I do not justify paying extra $100 for 3-5% efficiency. It's not like I am using $30 PSU. What am I missing?
Alexstarfire - Wednesday, April 14, 2010 - link
Lower power bills. IF you maxed out a 750w then a 3% difference equates to at least 22.5w. Sounds like nothing, but if you leave your computer on 24/7 it adds up pretty fast. That's 540w per day so let's say 1Kw every 2 days. $.10, obviously this will vary the most since cost per Kw of electricity varies depending on location, every 2 days, that's roughly $18 per year.I doubt most would get anywhere near this, but expecting about $10 per year could be realistic depending on your computer specs and usage habits. So if you can get an extra 3-5% for less than $30-$40 then it's probably worth it. Though that's assuming the PSU lasts that long. If you can get more than 5% then it's probably almost always worth it. :P
Lots of variables that really just depend on the individual. Very useful for someone like me that does a ton of gaming and video conversions.
LordanSS - Wednesday, April 14, 2010 - link
Since we're on the topic of wattage and efficiency, I was wondering if I could get input from people about what "good", wattage would be for a PSU running a quad-core CPU (125W TDP, thereabouts), a 5970 (or wattage-equivalent video card combo) and 3 to 4 7200rpm mechanical HDDs?I was initially thinking about going with an 850w PSU, but if I can get "lower" with a big enough headroom for future component upgrades (like GPUs, that might consume more power), that'd be good.
Thanks in advance.
sviola - Wednesday, December 29, 2010 - link
Try using this tool:http://extreme.outervision.com/psucalculatorlite.j...
ekerazha - Wednesday, April 14, 2010 - link
On the contrary, the JonnyGURU review says the Seasonic X-Series performance are better than the Modu87+ series performance... who is wrong?C'DaleRider - Wednesday, April 14, 2010 - link
Given that AT always has spurious results in their load testing when compared to other reputable power supply testers, like at JonnyGuru, HardOCP, and Hardware Secrets, I'd tend to think AT still has yet to set up their Chroma properly.The aforementioned sites all showed a 1% voltage regulation on all rails yet AT has the 5V rail at 5%......inferior testing from AT. And when will AT learn those graphs are almost worthless? I'd really suggest going to JonnyGuru and look at charts with numbers, so you can see what the actual load being put on the power supply during testing and what stability the various rails demonstrated, along with actual captures from the oscilloscope of the ripple/noise generation and captures of the overshoot transient tracings.
AT, while great at motherboard, cpu, and video card testing, is waaaay behind in power supply testing and while better than some sites, is almost becoming worthless in their results, esp. when compared to more reliable testers on the web.
ElBurroMaron - Wednesday, April 14, 2010 - link
*lol* and what exactly makes you the expert? Just because you don't understand the graphs you conclude they're bad? And you think testing with a shoddy 3k Sunmoon is so much better than doing it with a 30k Chroma?I'd also love to see o-scope shots and a little more on loading scales but hey, maybe it'll still come one day. Maybe asking nicely and suggestion what could be added is better than posting such a stupid comment as yours?
MrSpadge - Wednesday, April 14, 2010 - link
I got my Modu 87+ 500W for 110€ and it's just perfect for me.. I love it! Kind of.