Enermax Galaxy DXX Series PSU Review!!

[sladesurfer]

Design for the hard-core gamers that demand nothing less then perfect. The Galaxy DXX is equipped to handle dual 300w video cards thanks to the new PCI E 2.0 standard. Enough juice to handle the meanest dual quad core CPU system and quad graphics card on the market today.

:) http://enthusiast.hardocp.com/article.html?art=MTMyMCwxLCxoZW50aHVzaWFzdA==

http://img396.imageshack.us/img396/4015/11755692354fcth5mvej25gv0.gif (http://imageshack.us)

http://img396.imageshack.us/img396/881/11755692354fcth5mvej26kz1.gif (http://imageshack.us)

http://img396.imageshack.us/img396/4159/11755692354fcth5mvej27ol5.gif (http://imageshack.us)

http://img101.imageshack.us/img101/6974/11755692354fcth5mvej31lrq2.jpg (http://imageshack.us)

http://img296.imageshack.us/img296/6198/11755692354fcth5mvej32lov1.jpg (http://imageshack.us)

http://img403.imageshack.us/img403/1475/11755692354fcth5mvej33lkb7.jpg (http://imageshack.us)

http://img403.imageshack.us/img403/7629/11755692354fcth5mvej37lfj1.jpg (http://imageshack.us)

[EternityZX9]

I think that will be my next psu unless something better than that, which is modular, comes out by the end of the year.

[yaddam205]

Nice to see the 8-pin PCI express connectors.

[Kasparz]

That drop on 12v3 doesn't look promising. Good thing that this one have rubycon's.

[NapalmV5]

lol @ specs, 5th grader can do better...

there should be law against multi-rail psu!

[revenant]

contrary to popular lore, this is really a 2 rail psu..

http://www.silentpcreview.com/article726-page1.html


Five +12V Rails
Now here's an interesting change of pace. Really! There are two sets of electrically isolated 12V transformers instead of the typical single 12V transformer with common ground. Benefits of the split transformers may include things like improved cross loading.

also, each of the two 12v rails are theoretically capable of making 80 amps each.. looking at the components inside.. crazy amounts of power with this psu.. it's prolly more like a 1300W psu.. very conservative 1000W rating here..

[deathman20]

Intresting PSU. Definatly overkill for my needs but intresting never the less.

How big is this beast?

[fornowagain]

also, each of the two 12v rails are theoretically capable of making 80 amps each.

Where are you getting that from, do you have further information? The specs give about half that per rail, 41/34A.

[Nanometer]

Where are you getting that from, do you have further information? The specs give about half that per rail, 41/34A.

It's fact. It has two physical sources of 12 volts. The information is there, you can find it. I'm sure johny has talked about it in his review about the galaxy, so that's the first place to look.

[fornowagain]

It's fact. It has two physical sources of 12 volts. The information is there, you can find it. I'm sure johny has talked about it in his review about the galaxy, so that's the first place to look.
What is? 80A on each? I can see two primaries sure. I'd just like to know where Revenant got the 80A per side from? The specs don't support that is all.

[Nanometer]

No. It's about 40 amps each. Just read the reviews and you will learn more about it.

[fornowagain]

No. It's about 40 amps each. Just read the reviews and you will learn more about it.

Which is what I said 3 posts up. I think were at cross purposes, read Revenant's post again.

[Nanometer]

The specs don't say this because that is not what the PSU is designed to deliver. He is saying that the components, transformer and caps, are so "beefy" that they rated for well above what they are being used for. I don't know if that's true, but I wouldn't be surprised. The Galaxy is among the heaviest of desktop PSUs, and it takes big components to make it so. Running a component 50% below specifications isn't unusual specifically for power delivery equipment. I can promise you that the transformer isn't rated at a max of 40 amps at 12 volts. It's probably rated like the guy above said, 80 amps or much more.

[turbox997]

If you think about it, every few years the "standard" PSU goes up in watts. One day we'll all be having a 1000watt PSU..if the trend continues. That'll be one hell of an electricity bill when that day comes.

Mind as well start today~ haha, jk..

Seriously though, all the people getting SLi and CF for this current gen will probably need one.

[serialk11r]

If you think about it, every few years the "standard" PSU goes up in watts. One day we'll all be having a 1000watt PSU..if the trend continues. That'll be one hell of an electricity bill when that day comes.

Mind as well start today~ haha, jk..

Seriously though, all the people getting SLi and CF for this current gen will probably need one.

Uh, no. Then all computers will be crazy energy wasters, and the average guy with a crap computer will start complaining. Right now, 1kw isn't needed for anything pretty much.

[Helmore]

What is? 80A on each? I can see two primaries sure. I'd just like to know where Revenant got the 80A per side from? The specs don't support that is all.

You can find more in this little review (http://www.hardwaresecrets.com/article/365) (no performance testing though, just a dissection), it is a review of the older unit but I'm pretty sure that the internals haven't changed that much (maybe only the way that the limiters are set). In that little dissection they state that every single component in that PSU is made to handle much more than what they are making it handle. So it indeed is a pretty beefy PSU.

[fornowagain]

The specs don't say this because that is not what the PSU is designed to deliver. He is saying that the components, transformer and caps, are so "beefy" that they rated for well above what they are being used for. I don't know if that's true, but I wouldn't be surprised. The Galaxy is among the heaviest of desktop PSUs, and it takes big components to make it so. Running a component 50% below specifications isn't unusual specifically for power delivery equipment. I can promise you that the transformer isn't rated at a max of 40 amps at 12 volts. It's probably rated like the guy above said, 80 amps or much more.
As you say running a line 50% below rating is extremely unlikely, purely from a cost point of view. Maybe the driver for the primary, forward mosfets yeah I can see that. But the entire rail? i.e. full rec, PFC & PWM? nah, if the have it would very unusual, maybe. So I'll leave it there, just wanted to know where the 80A came from.

[Nanometer]

The components, not the actual circuitry. I am sure that the capacitors are rated well above double the supply voltage, and that the transformers are working at least half of what they are capable. Can the whole PSU handle 2000 Watts? Likely, but there's no proof to show that.

[fornowagain]

nvm

[revenant]

Where are you getting that from, do you have further information? The specs give about half that per rail, 41/34A.

Hi - sorry for the late reply.. I'll have to dig up the URL I read which went through this psu component by component.. it was a really good read.. I was trying to find it when I posted my comment but was out of time.. at work, etc.. hopefully I can find it now.. brb

edit: ok, found it..


The +12 V CPU output is produced by two 40CPQ060 Schottky rectifiers connected in parallel, which can also deliver up to 40 A each, thus the maximum theoretical current the +12 V CPU line can deliver is of 80 A, which equals to 960 W. As mentioned above, the maximum current this line can really deliver will depend on other components, especially the transformer, the coil, the capacitor and the wire gauge used.

Another two 40CPQ060 Schottky rectifiers produce the second +12 V output, thus the maximum theoretical power this output can deliver is of 960 W.

from here, on page 5: http://www.hardwaresecrets.com/article/365/1 a very good read.. prolly the best psu review I have ever read.. very informative. I got a bit over zealous with saying 1300W, because in the review is says


Enermax could simply inflate the labeled power rating of this power supply as the majority of manufacturers do – they could easily label this unit as 1,200 W and people would probably believe. anyways.. the theoretical max output depends on many components, so I doubt each 12v rail would be able to yield 80 real amps each, maybe 50 peak? but none the less, a pretty nice power supply from enermax here.. I'm happy with mine! :)

[fornowagain]

anyways.. the theoretical max output depends on many components, so I doubt each 12v rail would be able to yield 80 real amps each, maybe 50 peak? but none the less, a pretty nice power supply from enermax here.. I'm happy with mine! :)

Thanks for the reply. I was going to comment earlier but didn't want to drag the thread OT. Simply though, over-rated components is a long way from double the output. i.e thermal dissipation, arc gaps, insulation, wire gauge..et al. But I see where the 80A came from, they've quoted package current. A 3 pin package, so usually connects to a center tap. I'd say that's the forward current per diode, or two in parallel is 40A. And the other thing is components derate a lot depending on the case temp, so individual package ratings are not accurate indicators of design load. Although it does looks a high spec PSU, just not 2KW good.

[Nanometer]

The Galaxies are all rated at advertised wattage and stability at 50C. So case temp has been already factored in. And I don't believe that the PSU is based on a center tap transformer, though I could be wrong.

[fornowagain]

The Galaxies are all rated at advertised wattage and stability at 50C. So case temp has been already factored in. And I don't believe that the PSU is based on a center tap transformer, though I could be wrong.
Eh? In this application the case and junction temperatures are for the component package, Tc & Tj. Just looking but Tc=160C. But yes of course temps are factored in, its just that a casual inspection can't tell the thermal resistance of each HS mount. Taking individual package ratings and saying that is the lines theoretical maximum current is very misleading. The component when selected can run anywhere on a curve and maybe running at a less than optimal c/w with the corresponding derating of the maximum current. The rec specified is a three pin, so two diodes, normally used in full wave center tap rectification. Maybe some clever use in PWM circuit for each leg, without a schematic dunno? A tap would only be on the secondary, you could be right they are a lot bigger. Anyway regardless its not important, the maximum instantaneous forward current is "per leg" for each part of the duty cycle. The total package current quoted is not the same thing, I think in that instance they've used the wrong one in the article.

[sladesurfer]

Eh? Of course temps are factored in, just that a casual inspection can't tell the thermal resistance of each HS mount. Taking a package rating and saying that is the lines theoretical maximum current is very misleading. The component when selected can run anywhere on a curve and maybe running at a less than optimal c/w with the corresponding derating of the maximum current. The rec specified is a three pin, so two diodes, normally used in full wave center tap rectification. I suppose it could be some clever use in PWM circuit for each leg, without a schematic? Regardless, the maximum instantaneous forward current is per leg for each part of the duty cycle. The total package current quoted is not the same thing.

In English please :p:

[STEvil]

In english: its 1KW RMS good, but not 2000w RMS good.

[fornowagain]

In english: its 1KW RMS good, but not 2000w RMS good.

Lol, thats what I said!! :soap: