[mdlug] [Fwd: Re: bios battery]

[Jeff Hanson]

On 1/16/07, Raymond McLaughlin <driveray at ameritech.net> wrote:
> Really? I'm surprised that an exact match for the specs above would'nt
> be good enough.
>
> These caps 15 mm high, 9mm diameter cylinders, with leads 5mm apart and
> black jacket with copper colored printing. The only markings are
> (parallel to the axis):
>   a brand name(?) of  'I.Q'   then below that

I.Q. is the brand.  Unfortunately they seem to have disappeared and I
can't find a datasheet.

>   '6.3V 1500uF',              then below that
>   '+105[deg]C',               then below that

Without a datasheet I have to guess.  With a datasheet it is possible
to find an exact match.  There isn't a whole lot of standards spec
wise but many manufacturers copy each others product lines.  1500uF
isn't common and dimensions are odd (if your measurements are
correct).  The height doesn't matter unless there is an overhanging
heat sink.  The diameter usually does but some MBs have
dual-footprints allowing more than one size of cap to be used.

> a arrow-stripe, containing what looks like a '-' inside an ellipsoid,
> and ending in a downward pointing triangle. I take this '-' to indicate
> the negative terminal. At first I had been looking at the + sign above
> this, but now I think it is part of the temperature spec.

Correct.

> Examining the board I can see that the six capacitors share a common
> pair of fairly wide traces. The capacitor leads are spaced 11mm apart.
>
> I'm kind of skeptical about the importance of the particular brand of
> capacitor. You mention "switching power supplies", so I must make sure
> you are clear on the fact that we are talking about capacitors on the
> computer's motherboard, not in the power supply unit it self.

It's a switching regulator.  Commonly 3-phase.  Generally they require
low-ESR caps else they lose efficiency (more heat) and the output gets
ugly.

The brand only matters as far as quality and reputation.  Most of the
vendors responsible for the bad caps were not directly at fault.  It
was the dielectric supplier that was the culprit but the manufacturers
should have caught the problem in QA.

> The fact
> that the motherboard is populated by over two dozen other electrolytic
> caps (the other caps are blue and branded Sanyo or Taepo rather than
> I.Q), and countless other components leads me to question how much
> effect these caps have on the over all impedance of the motherboard as a
> whole. If I get 6.3V 1500uF electrolytics of more or less the same
> physical dimensions, shouldn't I be good in this case?

The requirements for each circuit differs.  Many of them don't matter
much but the switching supply circuits are rather critical.  The
density of the design and high ambient temperature around the CPU
increases the stress on them which is why they usually fail first.  On
the MB that I fixed, I replaced all the ones around the CPU with exact
equivalents and any others of the same brand.  Luckily, most of the
other caps were Japanese and didn't have the problem.

> If you know different I'm willing to listen, but I have the impression
> that these electrolytic capacitors are pretty much a commodity items.
> It's good to know the limits of such a generality too. :)

My best guess:  UCC EKZE6R3ELL152MJ20S
The size isn't an exact match.  10 x 20mm.

digikey.com:  565-1633-ND

Coincidentally, this is the same series I used in my MB.

Mouser Electronics and Newark don't have 1500uF low-ESR caps in stock.

Hopefully electrolytics will be phased out:
http://www.extremetech.com/article2/0,1697,2080041,00.asp