Great review about PSU from X-bit labs.

[burebista]

Here (http://www.xbitlabs.com/articles/other/display/atx-psu5.html).
Six solutions from Antec, CoolerMaster, Corsair and High Power.

[bignose]

http://www.xbitlabs.com/articles/other/display/atx-psu5.html

Hello everybody,:D

i found this on x-bits labs site. really good staff, so i would share it .

Thank you .:eek:

[clsA]

Thanks I'll check it out later tonight

[Oklahoma Wolf]

The sky must be falling - that Phantom has Panasonic caps on the secondary :eek:

Still doesn't excuse the Fuhjyyus I still see in there, but it's a start.

[jonnyGURU]

Well, it seems that CWT recognizes the problems with Fuhjyyu's the same way Andyson does: They don't like heat and the Phantom is bound to get hot.

The problem is, they don't recognize it to the degree that Andyson recognizes as the Fuhjyyu's shouldn't be used in anything that could be considered a "quiet" PSU.

[MrWicked1968]

nice review, I suppose. My only complaint has to do with choosing 3 Antec PSU's when one should have sufficed.

I was interested to see the High Power PSU though. There are a lot of those up for grabs on ebay.

[jonnyGURU]

I don't know. They're three Antecs, but they're very much different models. I think it's good to show how much the brand can vary in performance and construction.

[burebista]

OK, now part VI (http://www.xbitlabs.com/articles/other/display/atx-psu6.html).
10 power supply models from Hiper, Mushkin, OCZ, Seasonic, Silverstone, Zippy and Zalman. :cool:

[Bbq]

Hmm, the ripple on the OCZ GXS seems to be... less. I wonder why? Re-test anyone?

[jonnyGURU]

No. Oleg has tested low ripple on the Epsilon before. He scopes ripple at the PWM and not at the load.

[Oklahoma Wolf]

Yet another revised Epsilon based unit - the big coil on the secondary side is different than the 600W I have here. I wonder if the caps are the same... it certainly doesn't look to have the revised PCB the Zalman does. The smart load is a new development too.

[jonnyGURU]

Yet another revised Epsilon based unit - the big coil on the secondary side is different than the 600W I have here. I wonder if the caps are the same... it certainly doesn't look to have the revised PCB the Zalman does. The smart load is a new development too.

What I was pointing out is that he's tested an Epsilon in the past and didn't show much ripple on it either. I don't think the insides of his GameXstream are too different than mine:

http://www.jonnyguru.com/review_details.php?id=35&page_num=3

I might be wrong, but I personally think his ripple readings are low because of where he's probing.

I mean, FSP rates the max ripple/noise on these things at 200mV instead of 120mV for a reason. It wouldn't make sense for it to be the same as everyone else's and then specify it higher just so we can be "pleasently surprised" when we find out it's not as bad as they said it is. ;)

[Oklahoma Wolf]

Yeah... his does look identical to yours... interesting to see how often FSP is revising the design though from the really early ones like this 600W.

I commented on the issue over at SPCR where he'd posted about it yesterday, but he hasn't responded yet :(

[Super Nade]

Looks like the Munchkin PSU came out with flying colours, regulation and ripple even. A Topower first? I like X-bit's reviews, but I just hope thay put more pics of the internals. Maybe they were pressed for time, but internal shots like Jon's or at hardware secrects really make my day. :)

With the Silverstone ZF's, remember Jon we discussed the possibility of insufficient RFI shielding a while ago? Well take a look at this:

Here, each card is a complete and independent PWM regulator that receives a direct voltage on its input. The regulator’s operating frequency is 400kHz, which helps greatly reduce the size of the choke and the capacitance of the smoothing capacitors on the output. Kind of high isn't it? PWM's are notorious RFI sources, so maybe this is the culprit? The sinusoids in the scope outputs would very well confirm this conjecture.

Finally, I would have liked a Yes/No verdict tilted more towards the build quality than fan noise. :)

[Oklahoma Wolf]

Looks like the Munchkin PSU came out with flying colours, regulation and ripple even. A Topower first?

Naw - the Powerstreams usually did well too except the early 520W, and OCZ still dropped them. Could still be unreliable for all we know.

The switching frequency in the Etasis models could very well be a factor in the meter acting up.

[jonnyGURU]

Yeah... I heard the Mushkin 650W was actually pretty decent. That's probably why Mushkin is so mad at me for panning the 550W. Hey.... you want me to review one of your power supplies and give it a good rating, you need to send it to me. ;) Otherwise, I'm not going to go out of my way to obtain one. :D

[Super Nade]

Well, the reustls on the 550 Watter spoke for themselves. Instead of allegding bias and flawed testing methodology and spinning its deficiencies, they could have worked on the unit to improve it. As Wolf mentioned the Powerstream 520W, any idea why the higher multiple rail units kept blowing up, while the 520 W was considered the best in its class?

I'm not sure how X-bit test their units. Do they have a custom built load tester or are they using an ATE? Also, I'm not sure if they perform a hot-ambient test like you do.

[Oklahoma Wolf]

Oleg built himself an ATE using DAC's and mosfets - it interfaces to his computer and custom software. The rail loading is similar to the way the SunMoon works actually.