Power supply build quality pictorial. part 2

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

Why?

Because single forward design.
That SG Chip on the secondary side is a good giveaway...

Besides:
If the PFC is faked, it's very improbable that the rest is any better. So I'd dump that PSU...

[370forlife]

Re: Power supply build quality pictorial. part 2

Because single forward design.
That SG Chip on the secondary side is a good giveaway...

Besides:
If the PFC is faked, it's very improbable that the rest is any better. So I'd dump that PSU...

Its not even that lucky. Its a half-bridge unit. If designed well, a half bridge unit can deliver the advertised wattage no problem. 13009's are not sufficient in this case, and this unit is not a stellar design.

[ReeceyBurger123]

Re: Power supply build quality pictorial. part 2

Its not even that lucky. Its a half-bridge unit. If designed well, a half bridge unit can deliver the advertised wattage no problem. 13009's are not sufficient in this case, and this unit is not a stellar design.

As I suspected 400w max if that, I could upgrade the 13009,s but its not really worth doing as it has no pfc.

[Behemot]

Re: Power supply build quality pictorial. part 2

Because single forward design.
That SG Chip on the secondary side is a good giveaway...

Besides:
If the PFC is faked, it's very improbable that the rest is any better. So I'd dump that PSU...

Single forward is used mostly in old >250W units, so no.

I vote for try'n'load. Because the reality is you would condamn 80 % not-so-stellar PSUs on the market which can actually deliver the power…

[Pentium4]

Re: Power supply build quality pictorial. part 2

I vote for try'n'load. Because the reality is you would condamn 80 % not-so-stellar PSUs on the market which can actually deliver the power…

Of course he does, he's from JG forums. They're total elitists about power supplies, they think most PSU's are garbage because they're used to seeing such good stats from the top of the line PSU's all the time.

[kaboom]

Re: Power supply build quality pictorial. part 2

Of course he does, he's from JG forums. They're total elitists about power supplies, they think most PSU's are garbage because they're used to seeing such good stats from the top of the line PSU's all the time.

- don't even get me started....

[Pentium4]

Re: Power supply build quality pictorial. part 2

I'm glad at least one person knows what I'm talking about!

[Behemot]

Re: Power supply build quality pictorial. part 2

Don't think so. It is a relief I finally meet some people who also care about quality when talkign about picking up new PSU, but nonetheless there is still many people advertising crap in there.

Not to mention if I was editor in there, the score of most power supplies would be half it is now and many won't pass at all. It has terrible voltage regulation, OK than, one point down. It does not even comply with ATX? Oh that's OK, we just take half a point down…under me such power supply does not pass, period. They are too soft on them…

[kaboom]

Re: Power supply build quality pictorial. part 2

I'm glad at least one person knows what I'm talking about!

Let's not forget that they always whore out- those "reviews" are so positive and "sparkly" (hehe). But it's like newegg "reviews"- give five stars for "experience" (or lack thereof), and suddenly we know what we're talking about, either on NE or JG.

JG power supply "reviews" are an insult to anyone who's worked on, built, or otherwise has experience with them.

They should stick to reviewing cases. Preferably the nasty ones with razor-sharp edges, deep inside waiting to "surprise" someone.

Don't think so. It is a relief I finally meet some people who also care about quality when talkign about picking up new PSU, but nonetheless there is still many people advertising crap in there.

Not to mention if I was editor in there, the score of most power supplies would be half it is now and many won't pass at all. It has terrible voltage regulation, OK than, one point down. It does not even comply with ATX? Oh that's OK, we just take half a point down…under me such power supply does not pass, period. They are too soft on them…

They've sold out, plain and simple, like so many other useless "tech" sites.

Tells me all I need to know when they keep referring to the RFI/EMI filter as a "transient filter" (it is NOT) and proudly showing pix of questionable caps. Then upholding a 9.5/10 beacuse "the ripple rejection" is "excellent."

Sorry JG- good ripple rejection is part of any decent PS- don't pull something outta your ass and use it to whitewash those junk caps, which you and/or your buddies took pix of.

Just because "antecrep" may have fooled some people, remember there's a certain group that will never be fooled.

I refer to that bunch as "JonnyGOOFUP." Why not- same initials!

When L&C comes out with one of those large mostly 12V supplies, with internal converters to derive the other voltages, there will be an inside joke:
Deer and Buck (converter) season!

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

because the reality is you would condamn 80 % not-so-stellar PSUs on the market which can actually deliver the power…

Well, yeah, kinda.
Because too much PR/Marketing bullshit and not a really balanced design.

Something like a cheap sleeve bearing fan, secondarys like Su'scon, CapXon, SamXon, next to no heatsinks but a big 'japanese cap(s)' sticker on the box, because they used on the primary side...
And the usual 8pin protection IC like a Weltrend WT7502 (witch is fine, when combined with say an LM339 or something like that as used in some old enhance or HEC designs)...

So in the end you have to say:
The best PSU out there comes with a computer. They don't have some fancy shit like sleeves or cable managment or other stuff...

Especially those double PCB PSUs like those NMBs I have here...

[Behemot]

Re: Power supply build quality pictorial. part 2

Samxon caps are among the best being produced. Does not apply to GF and GK series, though.

I am not saying they are any good, but at least most of them nowaddays can deliver the power. They improved at least so far.

[ReeceyBurger123]

Re: Power supply build quality pictorial. part 2

Well another typical antec came to me full of blown fuhjyyus but apart from that a decent little cwt psu, rated for 480w semi modular need the cables though to be able to use this.
Had a nasty burnt 1800 ohm resistor on the primary side.
Specifications listed:
Ball bearing dynatron fan.
Double banks of output rectifiers
I couldn't see what the primary transistors were
Output capacitors listed:
5v 3* 3300uf 16v samxon rs
3.3 2* 3300uf 16v samxon rs
12v 2* 3300uf 16v samxon rs
-12v 1* 1000uf 16v Panasonic fr
-5v 1* 1000uf 16v Panasonic fr
5vstb ic design 1* 470uf 25v ucc ky 1 *220uf 25v jamicon (they are on the 5v stb weird values)
Smaller capacitors jamicon
330uf 400v rubycon usc

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

What's that big white thing screwed to the secondary heatsink for?

[mariushm]

Re: Power supply build quality pictorial. part 2

It's a 5-10w rated ceramic/cement resistor.. probably to act as a minimum load for one of the rails and keep the output regulated in some cases.

[ReeceyBurger123]

Re: Power supply build quality pictorial. part 2

What's that big white thing screwed to the secondary heatsink for?

It is a dummy resistor to reduce high pitch whine on the 5vstb.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

It is a dummy resistor to reduce high pitch whine on the 5vstb.

SRYSLY?!
Hard to belive that is...

Samxon caps are among the best being produced. Does not apply to GF and GK series, though.

The rule of the thumb: "You get what you paid for" applies to everything. Capacitors are no exeption...

GF are 1000-4000h types.
GK are 2000-5000h types.

The life depends on the diametre of the cap.


GT and GY Types are 4000h-10kh types, so it's safe to assume that they will last longer...
But they will also be more expansive than the GF and GK types...

Another point:
Japanese capacitors sounds better than long life (minimum 5000h types) chinese/taiwanese capacitor...

For exaple:
A NCC KZE cap is also just a 1000-5000h type...

But most 'PSU testers' will rate a NCC KZE Cap higher than for example a Teapo SY, although the datasheet of the Teapo SY states that it lasts from 2000-6000h, so it's a bit better than the NCC KZE.
The same applies to the SamXon lineup. The ones that probably will last longer are more expansive. And you can't really put them into end user PSUs because of marketing/PR reasons. You'd rather use 'japanese caps', even though they are worse because of that...

Buttom line:
The really good stuff is not available in a box on a shelf in a store!!
Most PSUs we can buy aren't really worth it...

The really good stuff (like the Delta 500QB you can buy for about 15€ in Germany + shipping) is put into PCs from big OEMs like DELL, HP. And some of the stuff you can buy with a server case...

[c_hegge]

Re: Power supply build quality pictorial. part 2

^
The lifetime ratings on the datasheets are endurance lifetime ratings. ie. how long the cap can run with its maximum rated ripple current at its maximum rated temperature before the specs go out of tolerance. Any cap will last longer than that in real world conditions. The difference between good caps and cheap caps is how much longer (and possibly how much the manufacturer decides to overstate the endurance lifetime).

[mariushm]

Re: Power supply build quality pictorial. part 2

The lifetime rating isn't in relation with technical parameters.

The lifetime depends on chemical formula of electrolyte, materials used (plastic/rubber/etc bottom, metal of leads) , dimensions of capacitor and a few other things.

It's an estimation of how many hours the capacitor will stay within specs (around some technical parameters specified) if you keep the capacitor at those high temperatures.

With lower temperatures, the lifetime rating increases... a non official rule says rating doubles with every 10c decrease in average temperature.

ps.

It's possible for a company to have a whole catalogue of very good capacitors but also have a few series that are simply bad. In particular with very low esr capacitors, this can happen, as the electrolyte inside them is more... let's say unstable. Some manufacturers get the process better, others don't. Samxon GF may be a victim, just as UCC KZG and a few other UCC series were, and like Nichicon HM and HN had problems around 2004.

[ReeceyBurger123]

Re: Power supply build quality pictorial. part 2

"Hard to belive that is..."


Why is that hard to belive ? That resistor is on the 5vstb and does help reduce that high pitch whine when in standby !

[Behemot]

Re: Power supply build quality pictorial. part 2

^
The lifetime ratings on the datasheets are endurance lifetime ratings. ie. how long the cap can run with its maximum rated ripple current at its maximum rated temperature before the specs go out of tolerance. Any cap will last longer than that in real world conditions. The difference between good caps and cheap caps is how much longer (and possibly how much the manufacturer decides to overstate the endurance lifetime).

Well, not exactly. The difference between good caps and bad caps is goodcaps can actually run that long, and possibly (much) more. Bad caps usually are not capable of even delivering this. They can run adequately long while being cool, but try heating crapxon, teapo, OST to 105 °C and run it for several weeks (1000 hours is aprox. 6 weeks) or months with some ripple on it!