If you're looking for a solid 550W PSU, check this out.

[PCBONEZ]

Re: If you're looking for a solid 550W PSU, check this out.

I'm still waiting for the Aqueous sweat pumps to explain why non-Aqueous FC has a shorter load life than Aqueous FM... ??? ???

[weirdlookinguy]

Re: If you're looking for a solid 550W PSU, check this out.

PCBONEZ, if you replace the 10v caps w/ 6.3v and the 16v w/ 12.5v, could you tell us how that goes? If the PSU still works great afterwards? It sounds like it will work well but I'd like to know how it goes for you.

[acstech]

Re: If you're looking for a solid 550W PSU, check this out.

I'm still waiting for the Aqueous sweat pumps to explain why non-Aqueous FC has a shorter load life than Aqueous FM... ??? ???

I dont know who you're waiting for. I use what fits, and then whatever has the right specs to most closely match what it's replacing. As far as I'm concerned they both have more than adequate lifetime. They're both rated at 105c. The fact that the FM has a longer rated lifetime and higher permissible ripple would seem to indicate that it is the better cap.

Of course none of this makes any difference since neither one of 'em will fit.

[PCBONEZ]

Re: If you're looking for a solid 550W PSU, check this out.

acstech, the wet comments were not referring to anything you said.
I just know one of those guys is gonna pop-in and tell us how important that is even though they won't fit anyway. - Preemptive evasive maneuver.

Where did 12.5 volts come from?
I said something about 12.5mm ~ not volts.

6.3v caps are fine on 3.3v and 5v rails.
16v caps are used on 12v rails.

In PSU make sure caps are on outputs because there may be voltages in there other than output rails use.

~~

Been holding off on projects because I need to order small numbers of maybe 20 odd-ball sizes to be complete on all my current projects. I'm well covered for 6.3v but there are some 16v, 25v, 35v, and 50v caps I need.
- And I can't order until I return from pending trip to Cali.
- Don't want caps sitting on the front porch. U know..

.

[weirdlookinguy]

Re: If you're looking for a solid 550W PSU, check this out.

Oh... sorry, my bad, for some reason I thought you said 12.5v. Well if you do it, let us know how it goes.

[Spacedye69]

Re: If you're looking for a solid 550W PSU, check this out.

[ZL, 10mm]
Aqueous boys are gonna cry tears but if it won't F'king then it won't F'king fit..
WHAAAAA.

.

I'm still waiting for the Aqueous sweat pumps to explain why non-Aqueous FC has a shorter load life than Aqueous FM... ??? ???

Were you abused as a child?

The original reply was to suggest that if you were to buy a NEW SUPPLY and go through the trouble of a preemptive recap to use the supply for SEVERAL YEARS of reliablility, a LONG LIFE, non aqueous cap would be a better choice but are hard to find in the size required.

If you can't find the size required in a long life cap, of course you have to use what you can find.

[PCBONEZ]

Re: If you're looking for a solid 550W PSU, check this out.

Abused as a child?
As a matter of fact yes, thanks for bringing it up.

Lets see,
Go buy a new supply in order to put non-aqueous caps in.
[We won't call them long life at this point.]
-
How?
-
I don't recall ever seeing an ad for a PSU that listed the diameter or uF of the caps.
Nor have I even seen a dealer that allowed breaking warranty seals for a look inside.
Don't recall any reviewers that consistently (or at all) call out both uF and diameter.
-
Just how is one to choose this PSU for a preemptive recap?

~~~~~~~~

>>>>>
SEVERAL YEARS of reliablility, a LONG LIFE, non aqueous cap would be a better choice
<<<<<

May be true 'in theory' but I've seen nothing to indicate this is true in actuality.
There certainly are counter examples:

------------------ Endurance 8mm/10mm/12.5mm

FC (-55c) -- Non-Aqueous --- 2000hr/3000hr/5000hr

FM (-40c) -- Aqueous ------- 4000hr/5000hr/5000hr
HV (-40c) -- Aqueous ------- 6000hr/6000hr/6000hr
ZLH (-40c) - Aqueous ------- 8000hr/10000hr/10000hr


Certainly aqueous'ness' is a factor in cap longevity but it is only one of many factors and can't be called out as the be-all end-all indicator as to a caps reliability or expected lifetime. That approach completely ignores the importance and effects of other additives to the electrolyte and so is inept.
-
.

[acstech]

Re: If you're looking for a solid 550W PSU, check this out.

You seem a little touchy. Just let it go.

I suppose non-aqueous caps would be better... at the south pole!

[kc8adu]

Re: If you're looking for a solid 550W PSU, check this out.

i agree.
drop the childish shit and move on.
if your psu runs at 105 c the rest of the system will die long before the 2000 hr rated caps in the psu does.
i have been putting mbz/mcz in many psu just because i have a cubic shitload of them.
i doubt i will see any issues in the usable life of any psu.

[PCBONEZ]

Re: If you're looking for a solid 550W PSU, check this out.

Yeah, sorry, I was wired the other night. Dunno why.....

It's just that I keep seeing this aqueous vs non aqueous thing come up and while it might matter a tiny bit I think that bit is too tiny to be a -practical- concern.

It's like: A 2.66 GHz CPU is faster than a 2.60 GHz CPU..
.... OMG! You'd better get the 2.66... It's faster.
Yes it is faster,,,, And?

That's what I've been trying to get across.

I might be completely off in left field but as yet no one has been able to show me there is enough difference between aqueous and non aqueous to give a hoot about.
If there is evidence to that effect I'd be glad to see it.

.

[gg1978]

Re: If you're looking for a solid 550W PSU, check this out.

Thanks to the person who posted pics.. I have never gotten around to it, and i had resized them before i attempted to post them here. Just timed out everytime.. Anyway, it's quite an impressive PSU, caps nonwithstanding.. Even so, i've been running it for over 6 months with no nastiness.. You'll also notice, it has independant regulation on +3.3, +5, and +12V.. Just look at the 24 pin ATX connector, and you'll see an extra "sense" wire in place on one of each of them, sensing the voltage at the motherboard level.. Don't see that very often either...

[Spacedye69]

Re: If you're looking for a solid 550W PSU, check this out.

Sorry KC8ADU.

PCBONEZ, the abused comment was meant to be a joke. Bad form on my part.

The original post I had was to help answer the question on what would be the best choice. I believe there is enough discussion and anecdotal evidence on this forum to, if not prove, at least strongly infer that aqueous cap do not fare as well in high temp situations.

My second post was to counter you statement that you did not care what was best, but that you had a buttload of MBZ. Not really the answering the original question of what would be best.

My third post was to reiterate my original statement. Somehow you got the impression that I said "Never use aqueous in a p/s". That was not my point.

I have used MBZ in power supplies before, and am sure I will again in the future, if needed. I would never say they are the best choice, but sometimes the only choice.

You have said in another post that MFR rating is not an indicator of actual life expectancy of a cap. I would agree with this statement. I don't see how quoting Panny's hour rating on caps would be any indication of which would last longer in a p/s.

You have also stated before that one should "do the math" before replacing components with lower esr or higher capacitance at random. I also agree. My first choice in a power supply would be something along the lines of Nichicon P series or Rubycon Z series(which is aqueous!), both which are designed for the high temp/high ripple p/s environment, IF AVAILABLE! Sizes like 3300uf 10mm are not available in a long life series from any good mfr, so if absolutly needed, and ultra low esr 2000hr cap would have to be used. The OST RLS that are in this p/s are not ultra low esr 2000hr caps.

I'll shut up and crawl back in my cave now....

[weirdlookinguy]

Re: If you're looking for a solid 550W PSU, check this out.

Finding long-lifes in 10mm looks impossible right now. Are OST's and JPCE-TUR's really THAT bad? I know they aren't good but my 5 year old HP has a bestec PSU loaded with OST's and JPCE's and it's still going strong after 5 years. I know Bestec sucks, OST sucks, etc etc but is OST really THAT bad? How long do you guys think this PSU will go with the stock caps?

[JEWilson]

Re: If you're looking for a solid 550W PSU, check this out.

Just a quick comment to add FWIW

These O/P caps

Surely if they are being used in multiples for the rails, then
take two of these - get the total capacitance of two, I would envisage
you would be able to get a cap of the total value with of course a
larger form factor, specifically diameter.

Solution - remove two 10mm - insert a cap double the value
of those two removed with larger diameter with a consideration for the
height and span the leads of the larger cap across the former
locations of the two 10mm removed? This would leave a set of via
unpopulated.

Perhaps this way you might be able to accomodate better quality
caps in the space available. Of course, you might end of with the
odd 10mm remaining but, increasing the O/P filter capacitance
would do no harm here.

Just a bit of lateral thinking - monkey around with the physical
constraints - what's the problem?

Am I being clear? Fair suggestion?

[kc8adu]

Re: If you're looking for a solid 550W PSU, check this out.

Just a quick comment to add FWIW

These O/P caps

Surely if they are being used in multiples for the rails, then
take two of these - get the total capacitance of two, I would envisage
you would be able to get a cap of the total value with of course a
larger form factor, specifically diameter.

Solution - remove two 10mm - insert a cap double the value
of those two removed with larger diameter with a consideration for the
height and span the leads of the larger cap across the former
locations of the two 10mm removed? This would leave a set of via
unpopulated.

Perhaps this way you might be able to accomodate better quality
caps in the space available. Of course, you might end of with the
odd 10mm remaining but, increasing the O/P filter capacitance
would do no harm here.

Just a bit of lateral thinking - monkey around with the physical
constraints - what's the problem?

Am I being clear? Fair suggestion?

no.
look at a psu(not a gutless wonder)and note the inductors between each pair of o.p caps.

[JEWilson]

Re: If you're looking for a solid 550W PSU, check this out.

To be fair PCBONEZ does suggest something similar (or the same)

https://www.badcaps.net/forum/showpo...1&postcount=40

If it were possible (and emphasise if), these may well be good psu(s) post
recap. Shame (for me) you seem to have them all in the States.

[gonzo0815]

Re: If you're looking for a solid 550W PSU, check this out.

You can do this only, if both caps are at the exact same location.
If there is a inductor or coil between them, this is an filter circuit, where the coil essentially force the ripple voltage trough the first capacitor (due to it`s higher resistance for AC currents).
If you are replacing the first capacitor, the ripple voltage will certainly raise beyond the spec`s.
Voltage and transient regulation could be significantly altered too.

What a pi Filter will do, you can leanr from :

http://www.bonavolta.ch/hobby/en/audio/supply.htm

Scroll down the page and don`t let you be fooled by the linear power supply, the AC / DC filtering is always the same, regardless which PSU topology is used.

[PCBONEZ]

Re: If you're looking for a solid 550W PSU, check this out.

Spacedye69

I think it's largely a matter of taking isolated posts out context and running with it ...
... x both ways.

Baring aqueous vs non-aqueous it seems we are actually of like minds.
[Look up the old ones, do the math, all that.]

-

As to aqueous vs non-aqueous,

yeah, I've seen a lot of discussion and opinion that says non-aqueous is better but saying it over and over doesn't make it true.
-
I haven't seen any hard evidence that it is in fact true.
(But that doesn't make it untrue either... now does it?)

** I'm actually trying to find out if it's true. **
** That's the whole point of my argument. **
** If it's true... Prove it to me!!!!!!

The closest thing to a hard numbers I have to look at to compare aqueous vs non is the Endurance ratings in the data tables.
-
Those numbers do not support a blanket statement that non-aqueous is better.

This may not have been true 10 or 15 years ago but I wasn't looking at caps 10 or 15 years ago. I only know what the tables say right now.

.

[gdement]

Re: If you're looking for a solid 550W PSU, check this out.

I now have 2 of these Acbel power supplies, and they are indeed impressive.
The issue of the 4-pin motherboard connector was resolved by building an adapter. I just desoldered the P4 connector from a dead board, and snipped the plug from a junk Powmax. Soldered them together to make a short P4 extension cord, which solves the problem of the 8-pin plug bumping into components on the 4-pin board.

Unfortunately, I don't think it's working out with my motherboard.

Here's the hardware it's been hooked up to:
ABit AN7 (late nForce2 socket-A)
Barton mobile at 2100/200mhz (1.65V I think)
ti4200 128MB video
2x 512MB ddr400 (each double banked)
1x hdd, 1 cdrom drive.

I was previously using a HiPro 300W. It worked with voltages in tolerance for about a week, then I had a spontaneous reset. I checked the voltages again and found that my 3.3V line had dropped to about 3.08V. So apparently it was overloaded and got ruined.

I switched to a Sparkle/FSP 300W, voltages were within tolerance but on the low side. It seemed to work fine for a few days but I wanted something with more guts.

Finally got my Acbel 550W, and voltages were very good. All rails were on the high side of tolerance, except the 3.3V line which was 3.28V.
It's been running for a few days, now I see my 3.3V reading has dropped to between 3.20-3.22V. The "3VDual" voltage is worse, at about 3.17V. All other rails are still as strong as ever.

I compared the labels on the AcBel and the HiPro, and I guess I can see why I'm still having a problem:

HiPro 300W:
3.3V 18A
5V 25A
3.3+5V max 175W

AcBel 550W:
3.3V 22A
5V 26A
3.3+5V max 160W (I knew it was optimized for 12V... but damn)

So as great as this 550W psu is at the 12V side, it could be disappointing for older boards.
I guess I need to unplug this tank before I ruin it, and go back to that old Sparkle. I'm disgusted.
Is there any simple improvement that would beef up the 3.3V line on this thing? I could probably turn up a pot, but I don't think that's going to keep it from dying.

[PCBONEZ]

Re: If you're looking for a solid 550W PSU, check this out.

Build a 12v to 5v regulator IC into the 4 pin lines to your drives. (1 amp at 5v per drive.)
Clip the 5v wire to that bundle coming from the PSU and power both the 5v and 12v at the drive end with the 12v from the PSU.
This way you power both the +5v and +12v to the drives off the 12v rail and unload some work from the 5v line in the PSU.
You take about 4 watts/drive off the 5v rail that way and you are still pulling less than 2 amps/drive off the 12v wire.