Jump to content
Corsair Community

Corsair AX760


fifo

Recommended Posts

I am trying to compare the new corsair AX760 , to the seasonic SS-760XP2 760 W platinum psu. Am I right in saying that the corsair AX760 is the SS-760XP2 inside or is it the older version

The x-750 km3 with slight modifications.

 

Can ayone tell me if any what is the difference between the SS-760XP2 760 W platinum, the x760/750 km3 and the corsair AX760.

Does the SS-760XP2 760 W suffer from the coil whine problem as the km3? Has something been done about it?

Link to comment
Share on other sites

  • Corsair Employee

You're seriously taking the opinion of people that don't work for Corsair over mine? ;):

 

KM3 is essentially a cost down version of the XP2 so they look very similar internally, but trust me... the AX is not KM3. It's XP2.

Link to comment
Share on other sites

Thanks for the reply & confirmation, i didn't know you work for corsair,

 

the thing is there are many references that say the AX760 is based on the old km3 series and the old xp. Here is one of them on the corsair forum itself.

 

http://forum.corsair.com/v3/showthread.php?t=121024

 

I have not seen any references that say the AX760 is based on the new Seasonic XP2 internals. The things is i want to purchase the latest seasonic unit, I've heard they put extra glue on the XP2 to eliminate the coil whine problem, (Seasonic or Corsair) but i prefer it to buy it in Corsair form because you get more cables.

 

There is also another thing according to the fan speed graph on the corsair website, the fan comes on at 70% load on the Ax760, and on the Seasonic 760 XP2 the fan comes on at 30 % load. Why so much difference if they are same internally?

Link to comment
Share on other sites

  • Corsair Employee

One end user made the assumption that the AX is based on the KM3. Don't take it as gospel. ;):

 

The XP2 and KM3 are VERY similar. And the old AX (750W, 850W) were based on the KM2.

 

But all that aside, the potential noise issues ("extra glue", which is NOT the only solution) is not resolved on one platform vs. the other. It was a revision change across ALL Seasonic platforms that included using different X caps, improving QC on the windings of the coils and transformers, etc.

Link to comment
Share on other sites

Here is a review done by HardwareSecrets that also confirms that the AX760 is based on the KM3 platform.

 

http://www.hardwaresecrets.com/article/Corsair-AX760-Power-Supply-Review/1687/1

 

 

The Seasonic SS-760W-XP2 seems very recent its been reviewed on the 3/Dec/2013

 

Heres the Seasonic SS-760W-XP2 Review,

It also compares Efficiency/noise/ripple with the corsair AX760,

 

http://uk.hardware.info/reviews/5011/11/seasonic-platinum-series-v2-660w760w860w-review-extremely-efficient-conclusion

 

From this review you can see that the Seasonic XP2 has outstanding efficiency at low loads (right down to 20W) where as the Corsair AX760/860 is no where near. Above 150W the efficiencies are quite similar. The ripple on the XP2 is also outstanding compared to the AX760.

 

Please don't take this wrong, but it seems like the AX760 is quite different from the XP2.

Link to comment
Share on other sites

  • Corsair Employee

Gabriel at Hardware secrets doesn't "confirm" it. He assumes it. He neither works for Seasonic or Corsair and I doubt he contacted either about who makes what.

 

KM3 uses a pair of IPP50R250CP FET for the PFC circuit as opposed to the pair of IPP60R199CP which have a lower RDS(on) (which is better).

 

The KM3 uses four IPP50R399CP FETs for +12V. The AX uses four IPP50R250CP FETs, which are higher capacity.

Link to comment
Share on other sites

  • Corsair Employee

Oh... and I misspoke about the old AX. There were two versions. The original AX Gold was KM2. The newer AX Gold (750 and 850 with the fan speed switch on the modular interface) was KM3. But the new AX Platinum is XP2.

 

Of course, I always doubt myself when someone is so determined to say I'm wrong, so I dug through my Outlook and I'm looking at emails in my inbox that come from Seasonic themselves that say the same thing. AX evolution: KM2, KM3, XP2. That's it.

Link to comment
Share on other sites

Thanks for the reply and i appreciate you looking into this further, With all due respect my intention is not to prove you wrong in any way i am just trying to clarify the AX760 Internals. Now there weren't two versions of the AX750, I have owned the latest AX750 purchased almost a year and half ago and it did not have a fan switch at the back, then it was discontinued and superseded with the AX760 (which has a fan switch at the back) Unless your referring to the RM 750 Gold.

Regarding the AX760 being SP2, The results of the ripple and low load efficiency (below 100 W) does not seem to match up between the Seasonic SP2 and the AX760. I expect there to be minor variations due to manufacturing processes, but difference seem to significant. At 50W load the seasonic has an efficiency if 85.2 % and the corsair AX860 i (the flagship series - suppose to be better than the AX760 non i) has an efficiency of 79%. The ripple on the SP2 is 20.2 mVtt and the Corsair AX760 is 43.8 mVtt. More than double the ripple why so much difference. In fact the SP2 seems to be more comparable to the AX760 (i) rather than the non i. So unless your referring to the i version being SP2 but that is made by Flextronics.

Confused!

Link to comment
Share on other sites

  • Corsair Employee

If someone is measuring 79% efficiency at 50W and 43.8mV p-p ripple on the +12V then they either have a defective unit or their testing methodology is flawed or their test equipment is failing.

 

Not giving you lame excuses. Just telling you that's probably the case. Those numbers are even out of spec for what Corsair tells Seasonic a unit is going to have to perform at for it to even pass the QC stage. Ok?

 

Let's use a non-Seasonic/non-Corsair example so we can consider this more open mindedly. Let's take for example when OC3D had a look at the Thortech Thunderbolt 1200W "the worst PSU ever" The results were horrendous. Certainly worthy of the "worst PSU ever" label:

 

[ame]

[/ame]

 

But if you have a look at my forums where we discuss the matter further:

 

http://www.jonnyguru.com/forums/showthread.php?t=9559

 

As well as the review I did of the 1200W myself:

 

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=222

 

(My feelings aren't hurt if you didn't recognize my name as someone that's been reviewing power supplies for the last decade)

 

You can see that EVERYONE ELSE'S results are completely different. YET!!!!.......

 

Overclock3D has reviewed other make/model PSUs with results that are similar to other what other review sites have found... so is the problem the equipment or testing methodology? And other review sites have had success with their very first Thortech sample while OC3D had two that acted buggers, so is the problem Thortech's quality control?

 

We can only speculate. (And for the record, I'm not knocking Thortech. I actually have an 800W I use at home from before I started working at Corsair because I thought the display feature was "neat".)

 

As for the KM2 vs. KM3... that was my mistake and I'll explain where that came from because it makes for a good story. Since working for Corsair, I've been in a bit of a bubble of Corsair products and that start happened right when the AX Platinum came out. :biggrin: When we discontinued AX Gold for AX Platinum, we created a price delta between the two to help the Gold liquidate. But instead of simply liquidating, sales shot through the roof. So Corsair decided to keep the Gold on for a while if we could get a better price from Seasonic. Seasonic didn't propose a better price on KM2, but did offer the KM3, which had the switch on the back. Unfortunately, the latter version of the AX (the KM3 based one with the switch) didn't meet the same performance criteria as the older KM2 based model, which makes sense because KM3 is supposed to be a cost down to KM2, not just a next generation. So the project was axed before a production unit fell into customers hands.

Link to comment
Share on other sites

Thanks for the info, I found it quite surprising how the thunderbolt psu was rated worse psu of the year and yet another review gave it 6 or 7 stars out of 10. Goes to show you cannot rely on one psu review.

 

I accept that two different reviews testing the same psu using different measuring equipment is going to yield different results as i have an Electronics background myself.

 

But what i don't get is the review that Hardware secrets did on the

seasonic XP2.

 

http://uk.hardware.info/reviews/5011...ent-conclusion

 

In this review the ripple of the Seasonic XP2 and the corsair AX760 is measured with the SAME measuring equipment and it shows the ripple of the AX760 being twice as much as the seasonic XP2. Now wouldn't you agree with me that both units are XP2, just different labeling (unless you've made slight alterations in the AX) the ripple of which being measured by the Same Oscilloscope, then besides the accuracy, the results should be fairly close to each other. What do you think?

 

Even when they did the fan/Coil Whine noise measurement, the

Seasonic was 20 - db No Fan running, Just coil Whine

AX760 was 18 dB - db No Fan running, Just coil Whine

 

Hence the Ax being 2 dB quieter, Same PSU same Sound meter.

OK 2db's is neglible, another AX could give 2 db more.

 

I read the review of the AX760 on the Guru3d site which was good.

 

http://www.guru3d.com/articles_pages/corsair_ax760_psu_review,8.html

 

On this review there is a link to the Chroma test report shown below.

 

The chroma test report - voltage noise test at 100% load .

 

 

Chroma test report of the PSU

STEP.16(UUT Test seq.16) : Noise Test(80+ 100% Load 90Vac) ---------------- PASS

Vin (V)= 90.00

Fin (Hz)= 47.00

Delay Time (ms)= 1000.00

 

Load Noise Range Noise B.W

Name (V) (MHz)

------------- ----------- ---------

5V 0.400 20.00

3.3V 0.400 20.00

12V1 0.400 20.00

12V2 0.400 20.00

12V3 0.400 20.00

-12V 0.400 20.00

5VSB 0.400 20.00

12V4 0.400 20.00

12V5 0.400 20.00

 

Load Loading Vpp Noise Vpp Noise Vpp Noise

Name (A/Ohm/V) Max Min (V)

------------- --------- --------- --------- --------

5V 12.690 0.035 ******* 0.0160

3.3V 12.690 0.030 ******* 0.0154

12V1 10.620 0.060 ******* 0.0340

12V2 10.620 0.060 ******* 0.0272

12V3 10.620 0.060 ******* 0.0319

-12V 0.420 0.060 ******* 0.0353

5VSB 2.530 0.040 ******* 0.0093

12V4 10.620 0.060 ******* 0.0290

12V5 10.620 0.060 ******* 0.0316

 

Am i right in saying that the Vpp noise is the ripple,

If so the ripple is about 32mV and max ripple is 60mV.

 

Would you agree with me that the efficiency at low loads, ripple and

noise of the AX760 should be the same as the seasonic XP2 as they both are the same unit. Has there been any slight changes to the AX, such as different FETs etc.? If not then buying the AX760 or the Seasonic XP2 shouldn't make any difference as they are both essentially the same.

Link to comment
Share on other sites

  • Corsair Employee

The 60mV is what they set the Chroma to trigger a failure at. It's not a reading from the PSU. The 32mV is the actual ripple measurement. This is better, but still a bit high.

 

Still, if anything, this should prove to you that you should read more than one review and that no two units, or two testing methodology/results are the same.

 

Like I said, the AX Platinum and XP2 are the same. The KM3 is different.

 

As for a review... jonnyguru.com was passed over for an AX Platinum review for whatever reason. But I happen to be in Corsair labs this week, so I can see about grabbing a unit and throwing on the Chroma real quick to see some results.

Link to comment
Share on other sites

  • Corsair Employee

Here you go....

 

Load Name Loading/other(A) Loading/Noise(A) Min Meas Max

5V 12.640 12.640 0.000 0.014 0.035

3.3V 12.640 12.640 0.000 0.013 0.030

12V1 10.407 10.407 0.000 0.018 0.060

12V2 10.407 10.407 0.000 0.016 0.060

12V3 10.407 10.407 0.000 0.018 0.060

-12V 0.500 0.500 0.000 0.029 0.060

5VSB 2.518 2.518 0.000 0.010 0.040

12V4 10.407 10.407 0.000 0.017 0.060

12V5 10.407 10.407 0.000 0.022 0.060

Link to comment
Share on other sites

Thanks for the quick test JonnyGuru,

 

it would be nice if there were spaces between the columns so i can see which reading is which column.

Thus from these readings 12V1 for example am i right in saying that the

instrument is set to trigger at a max of 60mV and a min of 0mV and the 18mV

is the measured ripple. If so that's really good. What load were these readings taken at? How you tried worse case load (low or High)?

 

12V1 10.407 10.407 0.000 0.018 0.060

12V2 10.407 10.407 0.000 0.016 0.060

12V3 10.407 10.407 0.000 0.018 0.060

 

Have you got any efficiency measurement at low loads such as

20W

50W

100W

 

I am interested in the low load efficiency.

 

When my pc is idle

My EVGA 460GTX consumes 19W

Carviar Green hDD 4W

Crucial M4 128GB SSD ?

Core 2 Duo E8400 - Not much

Gigabyte GA-EP45-UD3LR ?

Corsair Denominator DDR2 4GB Memory

 

Thus it probably consumes about 50W most of the components are energy saving.

 

Aside from that,

Is the Seasonic X series 760 Gold the km3 version (Cost down version)

and Seasonic X series 750 Gold, the km2

 

if so what performance issues did the km3 suffer?, it is still being sold by seasonic as the Gold Version.

Link to comment
Share on other sites

  • Corsair Employee
Thanks for the quick test JonnyGuru,

 

it would be nice if there were spaces between the columns so i can see which reading is which column.

 

Yes it would. But formatting was lost in the copy and paste and I'm a busy guy. Feel free to do it yourself.

 

Thus from these readings 12V1 for example am i right in saying that the

instrument is set to trigger at a max of 60mV and a min of 0mV and the 18mV

is the measured ripple. If so that's really good. What load were these readings taken at? How you tried worse case load (low or High)?

 

12V1 10.407 10.407 0.000 0.018 0.060

12V2 10.407 10.407 0.000 0.016 0.060

12V3 10.407 10.407 0.000 0.018 0.060

 

The first set of numbers after the rail identification is the load. 10.407 = 10.407A. This was a full load test (750W).

 

Have you got any efficiency measurement at low loads such as

20W

50W

100W

 

I am interested in the low load efficiency.

 

Aren't we all? I only ran the one test for you and don't have time right now to do the others. Sorry.

 

 

Is the Seasonic X series 760 Gold the km3 version (Cost down version)

and Seasonic X series 750 Gold, the km2

 

Correct.

 

if so what performance issues did the km3 suffer?, it is still being sold by seasonic as the Gold Version.

 

More than anything, just voltage regulation isn't as tight. But Seasonic plans to address that with the next generation.

Link to comment
Share on other sites

  • 3 weeks later...

I've heard it has the Yate loon fan which has a reputation to be noisy rather than the sans ace fan which the non i has. I wonder why they chose a different fan for the i series.

 

Is the fan turned on at start in the i series and then turned off according to load?

rather than being off at start.

Does the small improvement in voltage reg and ripple help you pc components?

 

I have heard of cases where components have been damaged due to a psu malfunctioning on this forum with the AX series. Is there some sort of protection where if the voltage mainly 12V deviatetes by a certain amount the psu cutts off to save your components? I'm mainly worried about ny Raid Hard drive which has important data on them.

Link to comment
Share on other sites

  • Corsair Employee

The Yate Loon fan isn't louder. But San Ace is a generally better quality fan. The fan is off at start. But if you want to turn it on, you can either set it to turn on in the Link software if you want it to run all of the time. If you just want to make sure the fan works before installing the PSU, there is a test button.

 

The better the voltage regulation and ripple suppression, the less stress on other components that have to work these imperfections out:

 

http://www.corsair.com/en-us/blog/2013/march/why-does-a-better-power-supply-mean-a-better-computer-experience

 

Any components can be damaged by any PSU malfunctioning. This is not limited to Corsair PSU's. There is protection in all (quality) PSUs that should shut down the PSU if voltages are out of spec. The DSP just latches off quicker and has a tighter tolerance for measuring out of spec voltages because of the DSP.

Link to comment
Share on other sites

Thanks for the reply jonny,

 

Basically i am trying to decide whether to go for the AX760 (Seasonic XP2)

or the AX760i (Flextronics).

 

Now i have read quite a lot of customer reviews on the AX760 i on newegg.com

 

and customers are generally complaining about similar issues. Firstly you cannot get one of the screws in as there is a circuit board in the way (manufacturing fault - could cause a potential short),

The red self test light comes on sometimes,

Fan is on 100% till windows is started.

The link software not accurate, very dull UI, psu defaults to fanless mode

if link software does not run on your os then you cannot set the fan to cooling mode as it can only be done by the link software.

You get this big dongle that you have lto ocate somewhere in your case. (why isn't the dongle part inside the psu?)

One of the problems with fanless mode customers complain about is that when you draw 50% load or more the fan does not turn on (OK) but you get a lot of heat from the psu being dissipated in your case, as a result the case fans start ramping up trying to remove the heat. Thus where is the benefit of fanless if you have to compromise with heat.

 

Does the non i seasonic version suffer from this problem?

 

The difference in voltage regulation and ripple suppression between the AX760 and the AX760i, is it large enough to affect components as the AX760 has good ripply suppression anyway.

 

I have also read that in the i version flextronics use Chinese caps where as in the Seasonic version there are Japanese caps - better quality . Why?

Link to comment
Share on other sites

  • Corsair Employee

The dongle does have to be put somewhere, but it's not that big of a part. There are several reasons the components of the dongle are not within the PSU itself.

 

Other than that, every single one of those problems (except the Link accuracy, which is being addressed) are isolated incidents.

 

As for the capacitors: the only Chinese capacitors in the AXi version are on the modular interface. They're very small, low stress capacitors that's used as a final filtering stage. Sometimes they're Japanese, sometimes they're Chinese. It didn't seem to matter because of the duty and position of the capacitors.

 

Now, if you don't mind my asking, and please take no offense... Are you trolling me?

Link to comment
Share on other sites

What makes you think i am trolling you. I am only asking questions about corsair psu's, not creating interpersonal conflict.

 

I am just trying to decide whether to go for i version or non i,

The reason why i have been asking you these questions is from what i read from customer reviews, basically these psu's are quite pricey and i want make sure i get the right one that suits me.

 

I have read 5 consecutive reviews abut the Fan being on 100% on the i version till windows is started. Thus i just wanted to confirm that this is the case thus the fan must be turned off by the Link Driver or something which windows boots.

 

Another thing i want to clarify is when you draw 45% load the fan does not turn on (OK) as it is only suppose to turn on the about 50 % load but as a result of this do you get a lot of heat from the psu being dissipated in your case,?

Or is the psu still cool at this load? This goes for the non i as well. Is the psu still cool <40 C before the fan turns on?

 

Is the difference in voltage regulation and ripple suppression between the AX760 and the AX760i quite small, such the AX760 will be good enough anyway?

 

Basically i am not going to fiddle with the link software so i'm not to bothered about that, but if the regulation and ripple is close to the non i Ax760 that its not going to make much difference to components then i might as well get the non i seasonic build.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...