Electrolytics blow after 1 hr

J

Jeff Liebermann

Guest
Another day in bad caps hell. I fish a Dell GX520 out of the pile at
the local recycler and proceed to try and raise the dead. I've done
this many times with good success, but not this time. Upon initial
inspection, I find the usual bulging electrolytic caps. There were 4
identical caps, 2200uF 6.3VDC 105C. I replace them with 2200uF 10VDC
105C caps (Panasonic ECA-1AHG222). Note the increase in working
voltage. These are brand new caps from Digikey. I run the unit
overnight and find that two of the caps now have bulging tops.

Assuming that I may have had some marginal caps in the bag (unlikely
but worth testing), I replace the two that bulged with 2200uF 10v 85C
caps (Panasonic ECA-1AM222) also from Digikey. This is somewhat of a
downgrade, but should work. I also replace the power supply, for no
better reason than I had a spare.

Same problem as before. The two caps blew their tops within a few
hours, leaking caustic goo.
<http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg>
The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.

Using my finger and a thermocouple thermometer, I determined that the
bulging caps are not getting warm or at least are not any warmer than
other nearby components.

I've replace literally hundreds of caps but I've never seen this
before. It really makes me wonder if the bad caps problem might be
caused by something other than defective caps and/or marginal
capacitor voltage ratings.

Duz anyone have a clue as to what is causing this problem? I haven't
had time to probe the board for power supply bus noise or excessive
current quite yet. What I could use is some speculation as to the
probable culprit.

--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558
# http://802.11junk.com jeffl@cruzio.com
# http://www.LearnByDestroying.com AE6KS
 
Jeff Liebermann wrote:
Another day in bad caps hell. I fish a Dell GX520 out of the pile at
the local recycler and proceed to try and raise the dead. I've done
this many times with good success, but not this time. Upon initial
inspection, I find the usual bulging electrolytic caps. There were 4
identical caps, 2200uF 6.3VDC 105C. I replace them with 2200uF 10VDC
105C caps (Panasonic ECA-1AHG222). Note the increase in working
voltage. These are brand new caps from Digikey. I run the unit
overnight and find that two of the caps now have bulging tops.

Assuming that I may have had some marginal caps in the bag (unlikely
but worth testing), I replace the two that bulged with 2200uF 10v 85C
caps (Panasonic ECA-1AM222) also from Digikey. This is somewhat of a
downgrade, but should work. I also replace the power supply, for no
better reason than I had a spare.

Same problem as before. The two caps blew their tops within a few
hours, leaking caustic goo.
http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg
The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.

Using my finger and a thermocouple thermometer, I determined that the
bulging caps are not getting warm or at least are not any warmer than
other nearby components.

I've replace literally hundreds of caps but I've never seen this
before. It really makes me wonder if the bad caps problem might be
caused by something other than defective caps and/or marginal
capacitor voltage ratings.

Duz anyone have a clue as to what is causing this problem? I haven't
had time to probe the board for power supply bus noise or excessive
current quite yet. What I could use is some speculation as to the
probable culprit.

Try using Low ESR caps in that circuit... Most switching signals need
Click to expand...
caps with virtually no R in them.. Otherwise, they'll heat up in side
a bit and vent, or partially get there.

Jamie
 
On Wed, 23 Feb 2011 16:42:44 -0800, Jeff Liebermann <jeffl@cruzio.com>
wrote:

Another day in bad caps hell. I fish a Dell GX520 out of the pile at
the local recycler and proceed to try and raise the dead. I've done
this many times with good success, but not this time. Upon initial
inspection, I find the usual bulging electrolytic caps. There were 4
identical caps, 2200uF 6.3VDC 105C. I replace them with 2200uF 10VDC
105C caps (Panasonic ECA-1AHG222). Note the increase in working
voltage. These are brand new caps from Digikey. I run the unit
overnight and find that two of the caps now have bulging tops.
(snip)
Click to expand...

Duz anyone have a clue as to what is causing this problem? I haven't
had time to probe the board for power supply bus noise or excessive
current quite yet. What I could use is some speculation as to the
probable culprit.
Click to expand...
Jeff, I've been reading your posts long enough to conclude you know
what you're doing, generally in spades.

But ... did you check the polarity? I ask that because there are
cases where the PCB screen legend was arse-about, causing grief to
cap-replacers who didn't record the "outgoing" cap polarity and match
that against the screening.

Otherwise, if heating isn't an issue then what WW said is a
possibility.
 
"Jeff Liebermann"

http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg

The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.
Click to expand...

** Errr - I hate to tell you this pal ........

but the pic shows all 4 caps installed with reverse polarity.

A smaller cap ( C310) is clearly installed correctly and so are all the
others I can see.

The markings on the PCB are a bit unusual with the white triangle indicating
positive.



..... Phil
 
On Thu, 24 Feb 2011 09:43:24 +0800, who where <noone@home.net> wrote:

Otherwise, if heating isn't an issue then what WW said is a
possibility.
Click to expand...
Typo alert! Would you believe "what WS said"?
 
"Phil Allison" <phil_a@tpg.com.au> wrote in message
news:8sls4jFn6iU1@mid.individual.net...
"Jeff Liebermann"

http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg

The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.


** Errr - I hate to tell you this pal ........

but the pic shows all 4 caps installed with reverse polarity.

A smaller cap ( C310) is clearly installed correctly and so are all the
others I can see.

The markings on the PCB are a bit unusual with the white triangle
indicating positive.



.... Phil
Click to expand...
I think that triangle is actually a "+" sign, and if so, I agree

Arfa
 
"Arfa Daily"
"Phil Allison"

"Jeff Liebermann"

http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg

The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.


** Errr - I hate to tell you this pal ........

but the pic shows all 4 caps installed with reverse polarity.

A smaller cap ( C310) is clearly installed correctly and so are all the
others I can see.

The markings on the PCB are a bit unusual with the white triangle
indicating positive.


I think that triangle is actually a "+" sign, and if so, I agree
Click to expand...
** I see a white circle outlining each electro, then a solid white
triangle making a chord across the same circle plus a large white "+" sign
right next to the triangle.

The bulging electros are in all in backwards .....


...... Phil
 
"Jeff Liebermann" <jeffl@cruzio.com> wrote in message
news:0e9bm6h2ikhm2pfmdr0ubgbl9ituco944l@4ax.com...
Another day in bad caps hell. I fish a Dell GX520 out of the pile at
the local recycler and proceed to try and raise the dead. I've done
this many times with good success, but not this time. Upon initial
inspection, I find the usual bulging electrolytic caps. There were 4
identical caps, 2200uF 6.3VDC 105C. I replace them with 2200uF 10VDC
105C caps (Panasonic ECA-1AHG222). Note the increase in working
voltage. These are brand new caps from Digikey. I run the unit
overnight and find that two of the caps now have bulging tops.

Assuming that I may have had some marginal caps in the bag (unlikely
but worth testing), I replace the two that bulged with 2200uF 10v 85C
caps (Panasonic ECA-1AM222) also from Digikey. This is somewhat of a
downgrade, but should work. I also replace the power supply, for no
better reason than I had a spare.

Same problem as before. The two caps blew their tops within a few
hours, leaking caustic goo.
http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg
The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.

Using my finger and a thermocouple thermometer, I determined that the
bulging caps are not getting warm or at least are not any warmer than
other nearby components.

I've replace literally hundreds of caps but I've never seen this
before. It really makes me wonder if the bad caps problem might be
caused by something other than defective caps and/or marginal
capacitor voltage ratings.

Duz anyone have a clue as to what is causing this problem? I haven't
had time to probe the board for power supply bus noise or excessive
current quite yet. What I could use is some speculation as to the
probable culprit.

--
Click to expand...
The caps are in backwards.

tm
 
On Thu, 24 Feb 2011 12:44:58 +1100, "Phil Allison" <phil_a@tpg.com.au>
wrote:

"Jeff Liebermann"

http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg

The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.

** Errr - I hate to tell you this pal ........

but the pic shows all 4 caps installed with reverse polarity.

A smaller cap ( C310) is clearly installed correctly and so are all the
others I can see.

The markings on the PCB are a bit unusual with the white triangle indicating
positive.

.... Phil
Click to expand...
AAAAAAARGH! I can't believe I did that. I usually assume that the
silk screened marker was matches the negative stripe on the
electrolytics. Very embarrassing.

Thanks for catching my mistake (as I crawl back into my hole and
sulk).


--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558
# http://802.11junk.com jeffl@cruzio.com
# http://www.LearnByDestroying.com AE6KS
 
Whats the problem? It all fits into the URL:

http://www.LearnByDestroying.com





"Jeff Liebermann" <jeffl@cruzio.com> wrote in message
news:rfkbm6hkov35ue2vbdnn8c1im1bju99vk4@4ax.com...
On Thu, 24 Feb 2011 12:44:58 +1100, "Phil Allison" <phil_a@tpg.com.au
wrote:

"Jeff Liebermann"

http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg

The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C caps.

** Errr - I hate to tell you this pal ........

but the pic shows all 4 caps installed with reverse polarity.

A smaller cap ( C310) is clearly installed correctly and so are all the
others I can see.

The markings on the PCB are a bit unusual with the white triangle
indicating
positive.

.... Phil

AAAAAAARGH! I can't believe I did that. I usually assume that the
silk screened marker was matches the negative stripe on the
electrolytics. Very embarrassing.

Thanks for catching my mistake (as I crawl back into my hole and
sulk).


--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558
# http://802.11junk.com jeffl@cruzio.com
# http://www.LearnByDestroying.com AE6KS
Click to expand...
 
Maybe not very commonly known anymore, but its always been a standard design
practice through the decades, when a board is built, that all the polarities
of electrolytic caps face the same direction. Have seen a few cases the last
couple of years where this is NOT followed though.
 
The caps are in backwards.
Click to expand...
Frightening. I had the right idea ("AC getting through" is a subset of
"wrong polarity"), but I never would have guessed that. How can a
manufacturer so blatantly ignore a standard that's been around for
I-don't-know-how-many-decades?
 
On 2/24/2011 6:58 AM, William Sommerwerck wrote:
The caps are in backwards.

Frightening. I had the right idea ("AC getting through" is a subset of
"wrong polarity"), but I never would have guessed that. How can a
manufacturer so blatantly ignore a standard that's been around for
I-don't-know-how-many-decades?
Click to expand...
That's easy, there's so many standards to choose from.

Jeff
 
The caps are in backwards.

Frightening. I had the right idea ("AC getting through" is a subset
of "wrong polarity"), but I never would have guessed that. How can
a manufacturer so blatantly ignore a standard that's been around for
I-don't-know-how-many-decades?

That's easy, there's so many standards to choose from.
Click to expand...
Well, "many" standards is equivalent to "no" standard! <grin>

Years ago, when I was working in Ecuador at a NASA site, we had to replace
some blown power rectifiers. I -- and everyone else -- assumed that power
rectifiers always had the cathode on the bolt end. Uh-uh. These were
backwards -- anode-to-bolt -- so our replacements kept blowing.
 
Jeff Liebermann <jeffl@cruzio.com> wrote in
news:rfkbm6hkov35ue2vbdnn8c1im1bju99vk4@4ax.com:

On Thu, 24 Feb 2011 12:44:58 +1100, "Phil Allison" <phil_a@tpg.com.au
wrote:

"Jeff Liebermann"

http://802.11junk.com/jeffl/crud/GX520-bad-caps.jpg

The two large black caps in the foreground are the surviving 105C
caps, while the blue caps in the background, are the bulging 85C
caps.

** Errr - I hate to tell you this pal ........

but the pic shows all 4 caps installed with reverse polarity.

A smaller cap ( C310) is clearly installed correctly and so are all
the others I can see.

The markings on the PCB are a bit unusual with the white triangle
indicating positive.

.... Phil

AAAAAAARGH! I can't believe I did that. I usually assume that the
silk screened marker was matches the negative stripe on the
electrolytics. Very embarrassing.

Thanks for catching my mistake (as I crawl back into my hole and
sulk).
Click to expand...
I reversed an electrolytic once,it blew and sprayed goo on the side of my
face,and foil in the instrument.

it's not as bad as my cutting through a power cord (missing GND pin,needed
replacement) that was still plugged in. :-(
Big flash and loud KAPOW,hole in diagonal pliers,and boss nearby was
filling out my performance review,and was working on the part about
following safety prodcedures. I'm glad the pliers had insulated handles.


--
Jim Yanik
jyanik
at
localnet
dot com
 
it's not as bad as my cutting through a power cord (missing
GND pin, needed replacement) that was still plugged in. :-(
Big flash and loud KAPOW, hole in diagonal pliers...
Click to expand...
You're lucky you didn't get molten metal in your eye. I once deliberately
shorted out an AC line with a scredriver. The result was not at all pretty.
 
William Sommerwerck <grizzledgeezer@comcast.net> wrote in message
news:ik5oho$q4c$1@news.eternal-september.org...
The caps are in backwards.

Frightening. I had the right idea ("AC getting through" is a subset
of "wrong polarity"), but I never would have guessed that. How can
a manufacturer so blatantly ignore a standard that's been around for
I-don't-know-how-many-decades?

That's easy, there's so many standards to choose from.

Well, "many" standards is equivalent to "no" standard! <grin

Years ago, when I was working in Ecuador at a NASA site, we had to replace
some blown power rectifiers. I -- and everyone else -- assumed that power
rectifiers always had the cathode on the bolt end. Uh-uh. These were
backwards -- anode-to-bolt -- so our replacements kept blowing.
Click to expand...

An instructor told me, ass-u-me anything and you will make an ASS of U and
ME
 
In article <8sls4jFn6iU1@mid.individual.net>, Phil Allison
<phil_a@tpg.com.au> writes

but the pic shows all 4 caps installed with reverse polarity.
Click to expand...
Bingo.

--
Mike Tomlinson
 
On Thu, 24 Feb 2011 08:09:46 -0600, Jim Yanik <jyanik@abuse.gov>
wrote:

I reversed an electrolytic once,it blew and sprayed goo on the side of my
face,and foil in the instrument.
Click to expand...
Only once? I think I've probably done it maybe 10 or more times in my
life. Sometimes, nothing happens. Other times, I spray smoldering
oily confetti all over the shop. Back in the tube days, I once
managed to convert a metal can electrolytic into a small rocket, that
cracked the Bakelite(?) case on the TV. With tantalums, the amount of
toxic smog that such a small device can produce is truly impressive.

Of course, after discovering I had simultaneously missed the obvious
and ruined my reputation for infallibility, I managed to make things
worse. I had visitors in the office at the time, and was engaged in a
very distracting conversation, while trying to replace the reversed
electrolytics. Of course, I had solder in two caps backwards before I
noticed that I had repeated my mistake. It's really hard to ignore
the silk screened + sign, and the square pad on the + lead but I
succeeded. Since it was obvious that nothing useful or successful was
going to happen last night, I replaced the caps, ran a few simple
tests, and made a hasty exit towards home.

In the true spirit of Learn By Destroying(tm), I measured the
temperature of the various caps with a thermocouple probe. 36C on all
4 caps whether inserted backwards or correctly. Basically, the same
temperature as most everything else in the area. It appears that
these caps do NOT get unusually hot when reversed. Some do, but these
didn't. Also, I'm rather impressed that the Dell PCB pads did not
fall apart after having the same parts replaced ummm.... 4 times.

it's not as bad as my cutting through a power cord (missing GND pin,needed
replacement) that was still plugged in. :-(
Big flash and loud KAPOW,hole in diagonal pliers,and boss nearby was
filling out my performance review,and was working on the part about
following safety prodcedures. I'm glad the pliers had insulated handles.
Click to expand...
Nicely done. I vaguely recall cutting across a hot zip cord while
plugged in, and destroying a pair of borrowed pliers. That was minor
compared to all my other mistakes and accidents, which I'll mercifully
decline to itemize.

--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top