Power supply build quality pictorial. part 2

[Stefan Payne]

Re: Power supply build quality pictorial. part 2



Something new, pretty good unit, especially those connectors. You just have to put them in at least once in your life. No, really you do

[Stefan Payne]

Re: Power supply build quality pictorial. part 2


And doesn't cost more than competitiors, while more efficient and more capacitance on +12V...

[Quaddro]

Re: Power supply build quality pictorial. part 2

Delta DPS-450DB..

Pretty sexy..

[Wester547]

Re: Power supply build quality pictorial. part 2

That looks like a nice PSU, although it's a bit dusty. I see 2003 datecodes on the unit, confirmed by the PVC-green-sleeved KZEs. Looks like that PSU uses independent regulation on the output, also a plus.

[Quaddro]

Re: Power supply build quality pictorial. part 2

This is why enermax naxn 500 is called graphic card killer..

[Behemot]

Re: Power supply build quality pictorial. part 2

And why exactly is that? Half of the cheaper mainstream units is made by CWT on this or similar cheap platforms. This is some modified GPA.

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

Concur with Behe. The 500W rating is bogus, but it's probably good for 250W-300W. The crap-caps probably limit its life without repairs to 4-6 years. OTOH, the +5VSB regulator is a Power Integrations TinySwitch, which is better than some notorious MB killer Stby circuits (I see PI products in some server PSU Stby circuits; Delta Products used PI products in their desktop computer and workstation PSUs back around the turn of the century).

Please do explain further.

[Quaddro]

Re: Power supply build quality pictorial. part 2

And why exactly is that? Half of the cheaper mainstream units is made by CWT on this or similar cheap platforms. This is some modified GPA.

Easy..
Coz many people think this power supply is a "real" and a "good" 500w psu..
Just because it's "enermax" and it has 6 pin & 8 pin graphic card power connector, they put power hungry graphic card in their system..
And voila, artifact is here and there..

This power supply is very popular in my country (indonesia)..

Well, i think 1000uF junfu class capacitor in 12v rail is not enough to supress the ripple even at the medium load..

[Behemot]

Re: Power supply build quality pictorial. part 2

All the CWT units I had always delivered rated power. Ripple was in spec most of the time, though hold-up time and crossload regulation often not.

[dmill89]

Re: Power supply build quality pictorial. part 2

All the CWT units I had always delivered rated power. Ripple was in spec most of the time, though hold-up time and crossload regulation often not.

X2, Any reviews I've seen of the Enermax Naxn units indicate that they can do their rated power. Yes they are cheap CWT units with sub-par caps that you probably shouldn't expect more than 3-5 years out of (without a re-cap anyhow), but you can definitely do worse, and a good chunk of other units in that price range are the same CWT platform with different stickers.

[momaka]

Re: Power supply build quality pictorial. part 2

Coz many people think this power supply is a "real" and a "good" 500w psu..
Just because it's "enermax" and it has 6 pin & 8 pin graphic card power connector, they put power hungry graphic card in their system..
And voila, artifact is here and there..

Video artifacts are caused by GPU going bad, not PSU.

I really wish the myth about PSU "not delivering enough power to the GPU" would stop, because there is NO such thing. If the PSU doesn't deliver enough power, either system will crash and restart or the PSU will shut off (and sometimes explode, if it's a cheapie).

[Behemot]

Re: Power supply build quality pictorial. part 2

There are however strange phenomenons happening like quality high-end PSUs making the VGA run 10 °C cooler and so on. We have eye witnesses and while that may only be wishfull thinking, based upon my initial tests, there is some merit to it.

Using as simple thing as higher-capacity lower-ESR capacitor(s) I have been able to increase PSU efficiency by as much as 3 %, just check my CX550M experiments http://www.hardwareinsights.com/wp/c...esign/7/#Bonus Voltage stability and ripple can have great impact on VRM stress. And considering there is VRM on practically every device in a modern computer (often many of them), the quality of the power you provide greatly impacts their stress thus heat/efficiency thus stability and lifespan. It is not only about raw power really.

Also as suggested, hold-up time is ordinary parameter where even big brands save on in cheaper units. While it seems to be neglected, I wonder how many instability cases are actually at least partially affected by this. Cause what happenes if you have dropouts in the voltage? Yeah, as you said, it crashes. If not the whole system than at least the VGA driver or something.

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

GIGO holds true in the world of power.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

Using as simple thing as higher-capacity lower-ESR capacitor(s) I have been able to increase PSU efficiency by as much as 3 %

You don't have a good AC source, you have a Sunmoon wich isn't known to be especially precise. Quite the contary. So no, I doubt THAT increased the efficiency that much. If it was that easy it would have been done.

So it's rather something like the inaccuracy of your equipment or changed conditions. Like temperature or other factors.

[eccerr0r]

Re: Power supply build quality pictorial. part 2

Powerman microatx HTPC custom PSU. 3845 switch controller (on bottom, not pictured)

Bad caps, bad caps...

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

Looks like a forward converter design, possibly built in late 2007. The output inductor cores look to be -52 material, so the switch frequency could be north of 50KHz. Don't love the 85C input caps; probably a more modern 105C series would have the same capacitance in the same case.

[eccerr0r]

Re: Power supply build quality pictorial. part 2

I don't know exactly how many watts I've been pulling of that ATX PSU, but it does get quite warm on my core2 duo, 1 HDD, 1 DVDRW, onboard GPU, two ATSC tuner cards (or formerly one video capture + 1 low profile GPU). The power estimates when I first got this PSU/case I remember calculating it to be just be on the edge of the power rating of the supply... I seem to recall it was rated at 235W but I don't remember exactly. It has a 60mm fan intended for the PSU itself and a separate 80mm fan for the HTPC case. Except for the times the computer failed (always bad caps, always bad caps...) it's been powered up.

I've had multiple caps fail in this PSU indeed due to the immense heat apparently, both in the motherboard and this PSU. I've really wondered if the heat was "normal" due to the small, low profile case.

The strange thing is that it seems to have thermal sensing, I'm not sure if it's for shutdown or for the fan controller... might be for the latter only...

BTW - what tipped you off for this to be a forward converter... It sort of was confusing because all the UC384x supplies I've seen were invariably flyback. I don't think I've had any dead forward converters yet to do autopsy/resurrection...

[RukyCon]

Re: Power supply build quality pictorial. part 2

Looks like a classic mix of OST and CapXon.
I don't get why they were used together so often in PSU's.

[eccerr0r]

Re: Power supply build quality pictorial. part 2

That capxon might have been my fault, I think that cap blew many years ago, and I didn't have any spares on hand so I dumped a sort-of-okay capxon. Fortunately that capxon is actually still good; many of the remaining OSTs are now over 5 ohms ESR, and another formerly "ok" cap that I dumped in also due to not having spares is now 1 ohm ESR.

This psu is really cramped on the secondary side, and do most people think 200+W is not actually realistic for this PSU?

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

...
BTW - what tipped you off for this to be a forward converter... It sort of was confusing because all the UC384x supplies I've seen were invariably flyback. I don't think I've had any dead forward converters yet to do autopsy/resurrection...

In a flyback, the "transformer" is actually a coupled inductor. The primary gets charged during the switch on time, and in the off time the stored energy is coupled to the secondary. IOW, there's no need for a nice big output inductor like that PSU has.

Also, in a flyback the energy is stored in a gap between the core halves, usually created by paper shims between the two outer legs of the core halves. The transformer core in that PSU is not gapped.

I didn't remember it this morning, but the UC3845 has a maximum duty cycle of 50%, which is especially suited for a forward converter.

FWIW, in a forward converter, half bridge, or full bridge, the powdered iron core of the output inductor(s) has a sort of gap distributed throughout the core, just by the nature of the material mix. The energy stored in the inductor core is stored in that distributed gap.