On 2007-03-31 02:31:56 +0100, daytripper <day_trippr (AT) REMOVEyahoo (DOT) com> said:

Perhaps I was not clear enough for you: the issue is whether the lower
reliability of electrolytics on the system boards of desktops (go and
read the ortiginal article if you have forgotten) is a significant
factor in desktop reliability. Don't bring up Stratus or other HA
systems in your answer.

Or to put it another way: if I wanted to spend a given amount of money
to make a desktop more reliable would I do it by buying a motherboard
with no electrolytics or (say) buying a system with redundant power or
mirrored disks. Don't bring up Stratus or other HA systems in your
answer.

I at no point used the word "sudden". I did use the word "death" to
mean "failure".

Never mind, I'm not going to waste more of comp.arch's time on this idiocy.

Tim Bradshaw <tfb (AT) tfeb (DOT) org> wrote:
: On 2007-03-31 02:31:56 +0100, daytripper
: <day_trippr (AT) REMOVEyahoo (DOT) com> said:
:
:: Ok, so you got confused along the way. Mostly your own doing,
:: as the point has always been: use cheap, low-mtbf caps and
:: you run a higher risk of a system failure, compared to the
:: use of high-mtbf components.
:
: Perhaps I was not clear enough for you: the issue is whether
: the lower reliability of electrolytics on the system boards of
: desktops (go and
: read the ortiginal article if you have forgotten) is a
: significant
: factor in desktop reliability. Don't bring up Stratus or other
: HA
: systems in your answer.
:
: Or to put it another way: if I wanted to spend a given amount
: of money
: to make a desktop more reliable would I do it by buying a
: motherboard
: with no electrolytics or (say) buying a system with redundant
: power or mirrored disks. Don't bring up Stratus or other HA
: systems in your
: answer.
:
::
:: For what ever misguided reason, you got caught up in the
:: phrase "sudden system death",
:
: I at no point used the word "sudden". I did use the word
: "death" to
: mean "failure".
:
: Never mind, I'm not going to waste more of comp.arch's time on
: this idiocy.

Okay then: [Moe to the boys: "Settle down".....'SLAP']

Hey everyone, I didn't post this question as a troll or to start an
argument. It was just an honest question about whether or not this hardware
(ie, solid caps) can make a significant difference in the overall "lifespan"
of a consumer-grade, desktop machine. As I originally posted, my homebuilt
machine (motherboard) is about five years old now, and is starting to
misbehave a bit. I finally found the time to pull the guts out of my
machine the other day only to find that yes, one of the capacitors is indeed
showing signs of leaking and I guess imminent failure. The machine still
works, but I'm definitely now on borrowed time.

I just simply wanted to know if the new solid-cap technology from Gigabyte
was worth the few extra euros (oops, I mean dollars) that I'm sure they will
ask. I find this thread so far quite interesting but I simply don't
understand all the "mala leche." ;-) Thanks to everyone that's offered
their thoughts.

/TLG

In comp.sys.ibm.pc.hardware.chips The Lone Gunman <lone (AT) invalid (DOT) net> wrote in part:
I don't disagree with any of the previous posters, but wish to
point out this depends entirely on what you mean by "lifespan",
and what the solid caps are replacing -- there have been
several epidemics of bad caps (often unstable isolation oil).

When the choice is between bad caps and solid, then yes, solid
will give longer desktop life. When the choice is between
good electrolytics (not always an oxymoron!), then the life
extention is probably in the out-years beyond normal lifespan.

Five years is above the historical "lifespan" of most
MS-Windows machines (3-4 usually considered max). OTOH,
I've used the same mobo continuously powered for 8 years.

You may be seeing abnormally bad caps. Try
http://www.badcaps.net

-- Robert

"The Lone Gunman" <lone (AT) invalid (DOT) net> writes:
What often helps in such cases is to push all socketed stuff a little
bit (or alternatively, pull it out, and reseat it).

If you are using liquid cooling or some other exotic cooler that does
not provide airflow to the CPU capacitors, electrolytic capacitors get
hotter and age faster than with normal air cooling (especially if you
also overclock); in these circumstances the solid capacitors may have
a significant advantage.

- anton
--
M. Anton Ertl Some things have to be seen to be believed
anton (AT) mips (DOT) complang.tuwien.ac.at Most things have to be believed to be seen
http://www.complang.tuwien.ac.at/anton/home.html

Robert Redelmeier <redelm (AT) ev1 (DOT) net.invalid> wrote:
: In comp.sys.ibm.pc.hardware.chips The Lone Gunman
: <lone (AT) invalid (DOT) net> wrote in part:
:: It was just an honest question about whether or not this
:: hardware (ie, solid caps) can make a significant difference
:: in the overall "lifespan" of a consumer-grade, desktop
:: machine.
:
: I don't disagree with any of the previous posters, but wish to
: point out this depends entirely on what you mean by "lifespan",
: and what the solid caps are replacing -- there have been
: several epidemics of bad caps (often unstable isolation oil).

To be honest, I don't know what I mean by "lifespan." For me personally, I
would think it would mean a mobo that can last at least five years before
going belly up. <shrug>

:
: When the choice is between bad caps and solid, then yes, solid
: will give longer desktop life. When the choice is between
: good electrolytics (not always an oxymoron!), then the life
: extention is probably in the out-years beyond normal lifespan.

Uh, ok. Not entirely sure what you mean here, but I think I get the idea.

:: As I originally posted, my homebuilt machine (motherboard)
:: is about five years old now, and is starting to misbehave a
:: bit.
:
: Five years is above the historical "lifespan" of most
: MS-Windows machines (3-4 usually considered max). OTOH,
: I've used the same mobo continuously powered for 8 years.

That's where I differ. My machine is almost never powered on 24/7. To the
contrary, it get's powered up/off sometimes three or four times a day. I
fully realize that repeated cold starts (vs running continuously at constant
state) is **much** harder on electronics, especially HD's and PS's.

: You may be seeing abnormally bad caps. Try
: http://www.badcaps.net

I KNOW I'm seeing a (singular) bad cap. Didn't you read the final paragraph
of my last post? <snigger> ;-)

/TLG

Anton Ertl <anton (AT) mips (DOT) complang.tuwien.ac.at> wrote:
: "The Lone Gunman" <lone (AT) invalid (DOT) net> writes:
:: As I originally posted, my homebuilt
:: machine (motherboard) is about five years old now, and is
:: starting to misbehave a bit.
:
: What often helps in such cases is to push all socketed stuff a
: little bit (or alternatively, pull it out, and reseat it).

Been there, done that. But thanks anyway.

:: I just simply wanted to know if the new solid-cap technology
:: from Gigabyte was worth the few extra euros (oops, I mean
:: dollars) that I'm sure they will ask.
:
: If you are using liquid cooling or some other exotic cooler
: that does not provide airflow to the CPU capacitors,
: electrolytic capacitors get hotter and age faster than with
: normal air cooling (especially if you also overclock); in
: these circumstances the solid capacitors may have a
: significant advantage.

Just using "normal", standard run-of-the mill cooling here. I DO NOT
overclock...that's either for dummy's or people who don't care about
reliability. I mean, Geez! With so much horsepower available nowadays, why
on earth would anyone consider OC their machine with it's concomitant
instabilities??

/TLG

On Sat, 31 Mar 2007 11:05:30 +0200, "The Lone Gunman"
<lone (AT) invalid (DOT) net> put finger to keyboard and composed:

IMO, the most important characteristics of capacitors used in
switchmode PSUs are their equivalent series resistance (ESR) and
temperature rating. In fact the most important tool in my toolkit,
even more useful than a DMM, is my ESR meter. A capacitor with high
ESR will experience ohmic heating by ripple currents, resulting in
premature failure. The "solid" caps are claimed to have half the ESR
of typical high grade low ESR aluminium electrolytics.

See the range of "Aluminum, Organic Semiconductor" types here:
http://www.vishay.com/capacitors/aluminum/radial/

Here is one 105degC OS-CON example:
http://www.vishay.com/docs/90009/94sv.pdf

It boasts "approximately two times the capacitance of existing
capacitors and less than half the ESR".

Here is a tech page that talks about a popular ESR meter kit:
http://members.ozemail.com.au/~bobpar/esrmeter.htm

This page describes the mechanism of capacitor failure:
http://members.ozemail.com.au/~bobpar/esrtext.htm

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

On Fri, 30 Mar 2007 20:31:56 -0500, daytripper
<day_trippr (AT) REMOVEyahoo (DOT) com> put finger to keyboard and composed:

Aluminium electrolytic capacitor failures do not usually result in
"sudden system death". I have replaced *thousands* of electrolytics in
all manner of equipment and the prevailing failure mode is an
intermittent one, or a thermal one. Marginal capacitors often come
good after being allowed to warm up. You can sometimes see this on an
old TV where the image suffers from reduced height and has retrace
lines at the top until the set stabilises. This is caused by dried out
caps in the vertical deflection circuit.

An example of caps that *do* fail suddenly and catastrophically are
tantalum electrolytics. These often go short circuit and/or catch
fire.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

Franc Zabkar <fzabkar (AT) iinternode (DOT) on.net> wrote:
: On Sat, 31 Mar 2007 11:05:30 +0200, "The Lone Gunman"
: <lone (AT) invalid (DOT) net> put finger to keyboard and composed:
:
<snip>

: IMO, the most important characteristics of capacitors used in
: switchmode PSUs are their equivalent series resistance (ESR)
: and temperature rating. In fact the most important tool in my
: toolkit, even more useful than a DMM, is my ESR meter. A
: capacitor with high ESR will experience ohmic heating by
: ripple currents, resulting in premature failure. The "solid"
: caps are claimed to have half the ESR of typical high grade
: low ESR aluminium electrolytics.
:
: See the range of "Aluminum, Organic Semiconductor" types here:
: http://www.vishay.com/capacitors/aluminum/radial/
:
: Here is one 105degC OS-CON example:
: http://www.vishay.com/docs/90009/94sv.pdf
:
: It boasts "approximately two times the capacitance of existing
: capacitors and less than half the ESR".
:
: Here is a tech page that talks about a popular ESR meter kit:
: http://members.ozemail.com.au/~bobpar/esrmeter.htm
:
: This page describes the mechanism of capacitor failure:
: http://members.ozemail.com.au/~bobpar/esrtext.htm
:

Thanks for the links, Frank. I'll get to reading them today.

/TLG

In comp.sys.ibm.pc.hardware.chips The Lone Gunman <lone (AT) invalid (DOT) net> wrote in part:
Ouch! I don't power-down even MS-Windows machines (which need frequent
reboots to recover memory leaks) unless the anticipated powerdown
is greater than 8 hours. If you want to save energy/environment,
please don't suboptimize: consider life-cycle costs.

Certainly I read. I think you should check out your mobo to be
able to put your current experience into context. Is it one of the
known bad-actors? Then a normal "good" mobo should be sufficient.
Is it not on the list? Then either it's a random failure _or_ your
usage is extreme, and you'd really best get a mobo with solid caps.

-- Robert