how long will this power supply last? superflower leadex gold

[ChaosLegionnaire]

according to this disassembly of a superflower leadex gold by johnny guru, this psu seems to use what looks like ncc kzj? on the psu mainboard which is a known badcap. ncc could have fixed their kzj problems but kinda skeptical on the choice of caps here.

next, if you also look at the first page of the review, one of the pictures on the back of the retail box of the psu says "100% japanese capacitor (mainboard) 105°C" but it says in smaller print "modular backpanel: A grade 105°C taiwan capacitor". now that's just lies. if you look at the last picture of the disassembly, it uses capxons for the modular backpanel which are just plain horrible caps according to many forumers. 'A' grade caps my behind...

however, oklahoma wolf remarked that "the modular board isnt critical for filtering with the good stuff". i also note that the voltage output in the review was stable despite the use of crap caps.

so any explanation for this peculiar choice of caps aside from cost-cutting measures? wont the capxons destabilise the output voltages since they are the last line of caps to output the voltages? if one day the capxons go bad, will it be "adios" for the psu? or will the good caps on the mainboard still keep the output voltage within spec?

thanks for your views.

[Pentium4]

Re: how long will this power supply last? superflower leadex gold

Those aren't KZJ on the output filtering, they're KZE which is a much more reliable cap. Chemi-Con seized production on KZJ.

I agree, I do NOT like seeing CapXon in any PSU. However, it's a little bit easier on them being at the end of the filtering. The KZE caps will take the brunt of the noise, and the CapXon would do a little more smoothing before being sent through the modular cables. They look like CapXon GL, which I have seen fail in low stress areas before. I guess I wouldn't immediately worry, but I would replace them as soon as the warranty period is up. It helps that the PSU is so efficient, so it won't be creating as much heat.

[Stefan Payne]

Re: how long will this power supply last? superflower leadex gold

1st:
Those caps are probably not filter caps but rather some decouppling caps.

2nd:
I don't mind Capxon there at all...

Don't understand why everyone hates them so...
They do their job according to the spec sheet...

You can not expect a 2000h part to last 10 years or so.
The problem is that we do not know how much stress those caps have to endure.
And you may want to think that those CapXon caps are build more spot on to the datasheet...

Anyway, those ones on the modular panel are either Polymers or they could be GL Series. They should be pretty OK, if treated properly. (the last thing is the biggest problem!)

And we can only conclude how caps were a couple of years ago as whe have to see them fail over the years...

How long it may last?!
I don't know. And no one really can tell...
If you use the same rig until it dies, it may be OK for a while.

But you never know what will change, how the power consumption will be in future devices, how much power they may need...
It isn't total impossible that there may be something that succeeds ATX and has, for example, a 24V Rail...

[c_hegge]

Re: how long will this power supply last? superflower leadex gold

Don't understand why everyone hates them so...
They do their job according to the spec sheet...

Only for about 3-5 years before they fail, which is completely unacceptable.

You can not expect a 2000h part to last 10 years or so.

Yes, you absolutely can. The Japanese parts all manage it with no problems. 2000 hours is simply how long they stay in spec when run at 105*C and at their maximum rated ripple current. The difference is in how much longer they last in more realistic scenarios. For junk caps, it's only marginally longer. For Jap caps, it's many times longer.

As far as this PSU is concerned, though, I kind of agree with you. I recon that, even if, erm, sorry. When they do fail, the ripple would still be in spec and it wouldn't kill hardware. It just won't have those same single digit figures that Wolf saw in the review.

[TELVM]

Re: how long will this power supply last? superflower leadex gold

Shit happens. The EVGA Supernovae G2 also boast of 100% jap caps, then you peel off the wires and find this:



http://www.overclock.net/t/1494167/t...stealth-lacing

I'd prefer Chemicons, but anyway it's better to have Crapxons than no caps at all there.

[ChaosLegionnaire]

Re: how long will this power supply last? superflower leadex gold

Only for about 3-5 years before they fail, which is completely unacceptable.

interesting, cuz this psu has a 5 year warranty.

so it seems like we could say that the manufacturer planned for the psu to fail just as the warranty expires aka planned obsolescence.

looks like pentium4 gave excellent advice. recap as soon as the warranty expires cuz thats when the manufacturer planned for the psu to fail lol!

[Wester547]

Re: how long will this power supply last? superflower leadex gold

Only for about 3-5 years before they fail, which is completely unacceptable.

Well, it depends on how hot the PSU runs. If it's a well cooled and ventilated PSU then yeah, 3-5 years is a premature failure. That said, all wet electrolytics, especially those sealed with rubber, do eventually dry up no matter the brand.

Yes, you absolutely can. The Japanese parts all manage it with no problems.

Well, excepting the ultra low ESR parts from Chemi-con (most of which are out of production) like P4 said, along with early HMs/HNs and maybe a couple other series. Honestly, KZGs and KZJs are some of the worst series ever produced by any brand. I only consider Sacon FZ, LTEC LMB, TK ATWY/ATWB, OST RLA, and Teapo SZ to be worse.

On the other hand, I see what you are saying. Even 2,000 hours @ 85C capacitors from the Japanese brands will last much longer than the parts from the crap brands (though those GP parts lucidly aren't suited for low ESR applications anyway).

I'd prefer Chemicons, but anyway it's better to have Crapxons than no caps at all there.

When the CapXons fail, they can wreak major havoc upon the PSU as well as the system to which it is connected. It would be better to have CapXon polymers than lytics at least, like Stefan said... (at least that's the impression I'm of)

[Stefan Payne]

Re: how long will this power supply last? superflower leadex gold

so it seems like we could say that the manufacturer planned for the psu to fail just as the warranty expires aka planned obsolescence.

No...
And I can't hear thes 'planned obsolesence' bullshit no more...

To be blunt:
They could make a PSU with guaranteed 10 Years of working.
But you do not want to pay for that...
As this PSU gets pretty damn expansive...

Everytime someone uses this bullshit phrase some people like to use, they do not have a clue about what really happens.
What's missing is the price tag of what they want...

Like washing mashines. Do you really think that the cheap ones do last more than the warranty?!
Well of course not! But that has nothing to do with 'planned obsolescence' but with the price you pay...

Here in Germany there is a brand that's famous for Quality: Miele.
And their cheapes washing machine is about 700€.
And a cheapish Samsung one for about 300€...

What do you think you'll get for that extra ~400€?! t

The same goes for everything else - it all comes down to a price...

And real good caps from Nippon Chemicon are something like KZM or KYB series.
What do you normally find in Power Supplys? KZE or KY...

[Stefan Payne]

Re: how long will this power supply last? superflower leadex gold

On the other hand, I see what you are saying. Even 2,000 hours @ 85C capacitors from the Japanese brands will last much longer than the parts from the crap brands (though those GP parts lucidly aren't suited for low ESR applications anyway).

That may be because of a different mentality.
The japanese may be closer to the germans.
So they may kinda overdeliver.

While the chinese are more spot on. If you spec something with +/- 10% you will get most of it barely passing on the lower end.
For example a fan with 2000rpm +/- 10%, you will get fans with a round 1850rpm.

I think it's the same with everything the chinese do...

Well, that's kinda the thing with capitalism...
You try to maximize your profit...

When the CapXons fail, they can wreak major havoc upon the PSU as well as the system to which it is connected. It would be better to have CapXon polymers than lytics at least, like Stefan said... (at least that's the impression I'm of)

Isn't it the same with every cap failing, not just CapXon? ;-)

On the other hand:
I prefer to have more caps than less (in parallel)
Because when a cap fails it's not that bad. And they are way less stressed...

It's the question what's better:
1 4700uF japanese brand one
4 1200uF Taiwanese brand (like Teapo, CapXon)

[stj]

Re: how long will this power supply last? superflower leadex gold

if your sneaky you put a ceramic across the leads under the board.
old design trick you dont see anymore!