Capacitor leakage anyone?

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

Hi all!

My 18 months old Epox 4PEA+ motherboard just went berserk yesterday. After
opening the chassis I spotted that capacitors near AGP port, SIL SATA
controller and TI firewire controller are somewhat inflated and that they
have some kind of yellow crust. I've heard of capacitors with low quality
electrolyte made by some Taiwan company that went in mass production, but
didn't think that a near $200 motherboard would have these installed.
Furthermore all of damaged capacitors are small greenish ones with a three
point star engraved on upper surface. On the other hand those black with
sliver stripe and letter "K" engraved on their surface look like they just
left production plant. Anyone else having the same problem?

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

Hrvoje Èièak wrote:

> My 18 months old Epox 4PEA+ motherboard just went berserk
> yesterday.
> I spotted that capacitors near AGP port, SIL SATA controller
> and TI firewire controller are somewhat inflated and that
> they have some kind of yellow crust. I've heard of capacitors
> with low quality electrolyte made by some Taiwan company that
> went in mass production, but didn't think that a near $200
> motherboard would have these installed. Furthermore all of
> damaged capacitors are small greenish ones with a three
> point star engraved on upper surface. On the other hand those
> black with sliver stripe and letter "K" engraved on their
> surface look like they just left production plant.

See www.badcaps.net for the story. Several Taiwan companies were
involved, not just one, but they all bought their capacitor chemicals
from a single source that produced a badly counterfeited a Japanese
chemical package. All of those capacitors were the low ESR (effective
series resistance) type, typically used for filtering highly pulsating
DC into smooth DC, such as for voltage regulators found on
motherboards. Other electrolytic capacitors aren't low ESR and are
used only for bypass, that is they're fed DC and are meant to keep that
DC clean when chips require sudden bursts of power. The only solution
to the bad caps is replacement, which isn't difficult but requires a
powerful soldering iron (40-50 watts) and Japanese brand capacitors
(Panasonic/Matsushita, Sanyo, Nichicon, United Chemicon, Rubycon).

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

Hrvoje Cicak wrote:

> Thanx for the info... I don't see myself as an experienced
> electronics tehnician bold einough to do some soldering on
> MB, but will try to find one, 'cause I really like this
> motherboard, and I would like to see it repaired.

Don't let the typical computer technician repair it because few of them
know how to solder (my apologies to those of you who do), but most TV
and audio repair shops should. Insist they use capacitors rated not
only for 105 Celcius but also for low ESR because not all 105C caps are
low ESR. www.digikey.com, www.mouser.com, and www.bdent.com sell
appropriate caps. Replace not only bulging or leaking one but also any
with colors and markings similar to them since they're likely marginal
as well. The caps most likely affected are typically surrounding the
CPU and any donut-shaped coils.

If you attempt the job yourself, get a 40 watt or higher soldering iron
(not gun) and practice desoldering some old, unneeded 4-6 layer circuit
boards (not 1-2-layer -- unrealistically easy). The safest way to
remove a capacitor without special tools is by cutting them off on top
with wire cutters so each wire lead can be unsoldered and pulled out
individually. This may seem drastic but is actually gentler than most
other methods, especially removing solder with a suction bulb (tiny
turkey baster). The next safest method is using narrow (2mm) copper
desoldering braid, but it needs plenty of heat, so wipe off and tin the
soldering iron tip frequently, and cut off any used braid before
desoldering the next lead. Wiggle each wire lead side to side to crack
off any remaining solder before pulling it out, but if it doesn't crack
apply fresh 60-63% tin solder and start all over. It's not unusual to
consume 1" of desoldering braid per wire lead.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

Hrvoje Èièak wrote:

> Hi all!
>
> My 18 months old Epox 4PEA+ motherboard just went berserk yesterday. After
> opening the chassis I spotted that capacitors near AGP port, SIL SATA
> controller and TI firewire controller are somewhat inflated and that they
> have some kind of yellow crust.

That's an amazingly common problem nowadays. I'm starting to wonder if
it's "planned obsolesence"

I had a similar problem with a year-old Epox board, recently, but it's
not just Epox.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

On Wed, 08 Jun 2005 12:06:07 -0400, Bob Horvath wrote:

HI all,
I had a FIC board go out with the capacitor issue a few months
ago.
Bob

>Hrvoje Èièak wrote:
>
>> Hi all!
>>
>> My 18 months old Epox 4PEA+ motherboard just went berserk yesterday. After
>> opening the chassis I spotted that capacitors near AGP port, SIL SATA
>> controller and TI firewire controller are somewhat inflated and that they
>> have some kind of yellow crust.
>
>That's an amazingly common problem nowadays. I'm starting to wonder if
>it's "planned obsolesence"
>
>I had a similar problem with a year-old Epox board, recently, but it's
>not just Epox.

 

[jad]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

soyo here , and many

"Bob Horvath" <bhorvath13@comcast.net> wrote in message
news:ei6ea1pk9po2nppm5dnvssa5au279urni3@4ax.com...
> On Wed, 08 Jun 2005 12:06:07 -0400, Bob Horvath wrote:
>
> HI all,
> I had a FIC board go out with the capacitor issue a few months
> ago.
> Bob
>
> >Hrvoje Èièak wrote:
> >
> >> Hi all!
> >>
> >> My 18 months old Epox 4PEA+ motherboard just went berserk yesterday.
After
> >> opening the chassis I spotted that capacitors near AGP port, SIL SATA
> >> controller and TI firewire controller are somewhat inflated and that
they
> >> have some kind of yellow crust.
> >
> >That's an amazingly common problem nowadays. I'm starting to wonder if
> >it's "planned obsolesence"
> >
> >I had a similar problem with a year-old Epox board, recently, but it's
> >not just Epox.
>

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

do_not_spam_me@my-deja.com wrote:

>
> Hrvoje Cicak wrote:
>
>
>>Thanx for the info... I don't see myself as an experienced
>>electronics tehnician bold einough to do some soldering on
>>MB, but will try to find one, 'cause I really like this
>>motherboard, and I would like to see it repaired.
>
>
> Don't let the typical computer technician repair it because few of them
> know how to solder (my apologies to those of you who do), but most TV
> and audio repair shops should. Insist they use capacitors rated not
> only for 105 Celcius but also for low ESR because not all 105C caps are
> low ESR. www.digikey.com, www.mouser.com, and www.bdent.com sell
> appropriate caps. Replace not only bulging or leaking one but also any
> with colors and markings similar to them since they're likely marginal
> as well. The caps most likely affected are typically surrounding the
> CPU and any donut-shaped coils.
>
> If you attempt the job yourself, get a 40 watt or higher soldering iron
> (not gun) and practice desoldering some old, unneeded 4-6 layer circuit
> boards (not 1-2-layer -- unrealistically easy). The safest way to
> remove a capacitor without special tools is by cutting them off on top
> with wire cutters so each wire lead can be unsoldered and pulled out
> individually. This may seem drastic but is actually gentler than most
> other methods, especially removing solder with a suction bulb (tiny
> turkey baster). The next safest method is using narrow (2mm) copper
> desoldering braid, but it needs plenty of heat, so wipe off and tin the
> soldering iron tip frequently, and cut off any used braid before
> desoldering the next lead. Wiggle each wire lead side to side to crack
> off any remaining solder before pulling it out, but if it doesn't crack
> apply fresh 60-63% tin solder and start all over. It's not unusual to
> consume 1" of desoldering braid per wire lead.
>

The two biggest things to watch out for is overheating, causing traces to
peel up, and pulling the feedthrough out along with the lead.

The cap is bad to begin with so I rock the can till the leads break loose
inside and then pull the can off, leaving the leads in the board. That then
give access to one lead at a time to heat and pull, with a nice piece to
get the needle nose on.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

<do_not_spam_me@my-deja.com> wrote in message
news:1118234258.439792.307430@o13g2000cwo.googlegroups.com...


Hrvoje Èièak wrote:

> My 18 months old Epox 4PEA+ motherboard just went berserk
> yesterday.
> I spotted that capacitors near AGP port, SIL SATA controller
> and TI firewire controller are somewhat inflated and that
> they have some kind of yellow crust. I've heard of capacitors
> with low quality electrolyte made by some Taiwan company that
> went in mass production, but didn't think that a near $200
> motherboard would have these installed. Furthermore all of
> damaged capacitors are small greenish ones with a three
> point star engraved on upper surface. On the other hand those
> black with sliver stripe and letter "K" engraved on their
> surface look like they just left production plant.

See www.badcaps.net for the story. Several Taiwan companies were
involved, not just one, but they all bought their capacitor chemicals
from a single source that produced a badly counterfeited a Japanese
chemical package. All of those capacitors were the low ESR (effective
series resistance) type, typically used for filtering highly pulsating
DC into smooth DC, such as for voltage regulators found on
motherboards. Other electrolytic capacitors aren't low ESR and are
used only for bypass, that is they're fed DC and are meant to keep that
DC clean when chips require sudden bursts of power. The only solution
to the bad caps is replacement, which isn't difficult but requires a
powerful soldering iron (40-50 watts) and Japanese brand capacitors
(Panasonic/Matsushita, Sanyo, Nichicon, United Chemicon, Rubycon).


Thanx for the info... I don't see myself as an experienced electronics
tehnician bold einough to do some soldering on MB, but will try to find one,
'cause I really like this motherboard, and I would like to see it repaired.

Hrvoje

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

This guy can probably fix it for you: www.motherboardrepair.com


"Hrvoje Cicak" <hrvoje@bilanca.hr> wrote in message
news:d86ptd$luu$1@ss405.t-com.hr...
>
> <do_not_spam_me@my-deja.com> wrote in message
> news:1118234258.439792.307430@o13g2000cwo.googlegroups.com...
>
>
> Hrvoje Èièak wrote:
>
>> My 18 months old Epox 4PEA+ motherboard just went berserk
>> yesterday.
>> I spotted that capacitors near AGP port, SIL SATA controller
>> and TI firewire controller are somewhat inflated and that
>> they have some kind of yellow crust. I've heard of capacitors
>> with low quality electrolyte made by some Taiwan company that
>> went in mass production, but didn't think that a near $200
>> motherboard would have these installed. Furthermore all of
>> damaged capacitors are small greenish ones with a three
>> point star engraved on upper surface. On the other hand those
>> black with sliver stripe and letter "K" engraved on their
>> surface look like they just left production plant.
>
> See www.badcaps.net for the story. Several Taiwan companies were
> involved, not just one, but they all bought their capacitor chemicals
> from a single source that produced a badly counterfeited a Japanese
> chemical package. All of those capacitors were the low ESR (effective
> series resistance) type, typically used for filtering highly pulsating
> DC into smooth DC, such as for voltage regulators found on
> motherboards. Other electrolytic capacitors aren't low ESR and are
> used only for bypass, that is they're fed DC and are meant to keep that
> DC clean when chips require sudden bursts of power. The only solution
> to the bad caps is replacement, which isn't difficult but requires a
> powerful soldering iron (40-50 watts) and Japanese brand capacitors
> (Panasonic/Matsushita, Sanyo, Nichicon, United Chemicon, Rubycon).
>
>
> Thanx for the info... I don't see myself as an experienced electronics
> tehnician bold einough to do some soldering on MB, but will try to find
> one, 'cause I really like this motherboard, and I would like to see it
> repaired.
>
> Hrvoje
>

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

I didn't look at the badcaps.net site until after I sent the previous post.
It looks like they do repair too.

"Hrvoje Cicak" <hrvoje@bilanca.hr> wrote in message
news:d86ptd$luu$1@ss405.t-com.hr...
>
> <do_not_spam_me@my-deja.com> wrote in message
> news:1118234258.439792.307430@o13g2000cwo.googlegroups.com...
>
>
> Hrvoje Èièak wrote:
>
>> My 18 months old Epox 4PEA+ motherboard just went berserk
>> yesterday.
>> I spotted that capacitors near AGP port, SIL SATA controller
>> and TI firewire controller are somewhat inflated and that
>> they have some kind of yellow crust. I've heard of capacitors
>> with low quality electrolyte made by some Taiwan company that
>> went in mass production, but didn't think that a near $200
>> motherboard would have these installed. Furthermore all of
>> damaged capacitors are small greenish ones with a three
>> point star engraved on upper surface. On the other hand those
>> black with sliver stripe and letter "K" engraved on their
>> surface look like they just left production plant.
>
> See www.badcaps.net for the story. Several Taiwan companies were
> involved, not just one, but they all bought their capacitor chemicals
> from a single source that produced a badly counterfeited a Japanese
> chemical package. All of those capacitors were the low ESR (effective
> series resistance) type, typically used for filtering highly pulsating
> DC into smooth DC, such as for voltage regulators found on
> motherboards. Other electrolytic capacitors aren't low ESR and are
> used only for bypass, that is they're fed DC and are meant to keep that
> DC clean when chips require sudden bursts of power. The only solution
> to the bad caps is replacement, which isn't difficult but requires a
> powerful soldering iron (40-50 watts) and Japanese brand capacitors
> (Panasonic/Matsushita, Sanyo, Nichicon, United Chemicon, Rubycon).
>
>
> Thanx for the info... I don't see myself as an experienced electronics
> tehnician bold einough to do some soldering on MB, but will try to find
> one, 'cause I really like this motherboard, and I would like to see it
> repaired.
>
> Hrvoje
>

 

[JohnS]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

You'll ruin that multilayer mobo trying to resolder those
caps. If that thing is out of warranty, I'd toss it and get
another brand. If not, RMA it. Plus, I kind of suspect
that you have a voltage reg blown, and that is what took
out the caps. It may have been the psupply that took out
the Vreg ... or it may have been just heat in the case,
but it would have been around 95 F to do that. If the
box was in sunlight, and turned off, that would do it.

johns

 

[JohnS]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

I am an etech. You can' t go on a multilayer mobo with
a soldering iron ... at all. I have a rework station that
evenly heats all the leads at the same time to exactly the
right temp. That is what it takes, and it works about
half the time.

johns

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

"johns" <johns123xxx@xxxmoscow.com> wrote in message
news:d882ma$1o9t$1@news.fsr.net...
> You'll ruin that multilayer mobo trying to resolder those
> caps. If that thing is out of warranty, I'd toss it and get
> another brand. If not, RMA it. Plus, I kind of suspect
> that you have a voltage reg blown, and that is what took
> out the caps. It may have been the psupply that took out
> the Vreg ... or it may have been just heat in the case,
> but it would have been around 95 F to do that. If the
> box was in sunlight, and turned off, that would do it.
>
> johns

If it's out of warranty, and he's just gonna chuck it, he has nothing to
lose.

MC

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

"Tweek" <shawnwingetNOSPAM@hotmail.com> wrote in message
news:%LFpe.17517$_w.418@trnddc01...
> This guy can probably fix it for you: www.motherboardrepair.com
>
>
> "Hrvoje Cicak" <hrvoje@bilanca.hr> wrote in message
> news:d86ptd$luu$1@ss405.t-com.hr...
>>
>> <do_not_spam_me@my-deja.com> wrote in message
>> news:1118234258.439792.307430@o13g2000cwo.googlegroups.com...
>>
>>
>> Hrvoje Èièak wrote:
>>
>>> My 18 months old Epox 4PEA+ motherboard just went berserk
>>> yesterday.
>>> I spotted that capacitors near AGP port, SIL SATA controller
>>> and TI firewire controller are somewhat inflated and that
>>> they have some kind of yellow crust. I've heard of capacitors
>>> with low quality electrolyte made by some Taiwan company that
>>> went in mass production, but didn't think that a near $200
>>> motherboard would have these installed. Furthermore all of
>>> damaged capacitors are small greenish ones with a three
>>> point star engraved on upper surface. On the other hand those
>>> black with sliver stripe and letter "K" engraved on their
>>> surface look like they just left production plant.
>>
>> See www.badcaps.net for the story. Several Taiwan companies were
>> involved, not just one, but they all bought their capacitor chemicals
>> from a single source that produced a badly counterfeited a Japanese
>> chemical package. All of those capacitors were the low ESR (effective
>> series resistance) type, typically used for filtering highly pulsating
>> DC into smooth DC, such as for voltage regulators found on
>> motherboards. Other electrolytic capacitors aren't low ESR and are
>> used only for bypass, that is they're fed DC and are meant to keep that
>> DC clean when chips require sudden bursts of power. The only solution
>> to the bad caps is replacement, which isn't difficult but requires a
>> powerful soldering iron (40-50 watts) and Japanese brand capacitors
>> (Panasonic/Matsushita, Sanyo, Nichicon, United Chemicon, Rubycon).
>>
>>
>> Thanx for the info... I don't see myself as an experienced electronics
>> tehnician bold einough to do some soldering on MB, but will try to find
>> one, 'cause I really like this motherboard, and I would like to see it
>> repaired.
>>
>> Hrvoje
>>
>

Homie is a good dude. you can trust him as he has been in the business for
a while


--
so many jackasses, so little ammo

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

>See www.badcaps.net for the story. Several Taiwan companies were
>involved, not just one, but they all bought their capacitor chemicals
>from a single source that produced a badly counterfeited a Japanese
>chemical package.

Yes but that was far longer than 18 months ago. That can't be the same
problem.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

"johns" <johns123xxx@xxxmoscow.com> wrote in message
news:d882ma$1o9t$1@news.fsr.net...
> You'll ruin that multilayer mobo trying to resolder those
> caps. If that thing is out of warranty, I'd toss it and get
> another brand.

Now why the hell would you toss for fear of possibly ruining it? I soldered
new capacitors on my motherboard to save money and it worked just fine.
Paying someone to do it was going to cost as much as replacing the
motherboard. Point being not that it's foolproof, but it obviously works
sometimes.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

"jeffc" <nobody@nowhere.com> wrote in message
news:f%Mpe.22$Bp.36235@twister.southeast.rr.com...
>
>
>>See www.badcaps.net for the story. Several Taiwan companies were
>>involved, not just one, but they all bought their capacitor chemicals
>>from a single source that produced a badly counterfeited a Japanese
>>chemical package.
>
> Yes but that was far longer than 18 months ago. That can't be the same
> problem.
>
>

I belive that it is the same problem since board was build around same time
that happened. I just recived it 18 months ago.

Pozdrav, Hrvoje

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

"johns" <johns123xxx@xxxmoscow.com> wrote in message
news:d882ma$1o9t$1@news.fsr.net...
> You'll ruin that multilayer mobo trying to resolder those
> caps. If that thing is out of warranty, I'd toss it and get
> another brand. If not, RMA it. Plus, I kind of suspect
> that you have a voltage reg blown, and that is what took
> out the caps. It may have been the psupply that took out
> the Vreg ... or it may have been just heat in the case,
> but it would have been around 95 F to do that. If the
> box was in sunlight, and turned off, that would do it.
>
> johns

If that happened I'd expect to see capacitors near CPU slot inflated... but
they are all normal. As I said in original post these capacitors are near
secondary controllers (SATA and FireWire) and AGP. Those that take care of
processor voltage (most commonly damaged) looks OK. Plus, I've had mobo with
vreg blown away, and you could see the damage - there was a hole on one of
"block resistors" (I don't know right English phrasing for that part).
Bottom line is that I belive that circuitry on this mobo is still in working
state, and that only capacitors are ruined... and I've seen quite a few
malfunctioning boards.
As for multilayer design. I too fear that "fixing" mobo with soldering iron
would actually ruin MB, but I ran out of solutions because similar mobo
can't be purchuased in Croatia, and I do need additional PATA and SATA
controller this MB have.

Pozdrav, Hrvoje

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

David Maynard wrote:

> The two biggest things to watch out for is overheating,
> causing traces to peel up, and pulling the feedthrough
> out along with the lead.

And overheating is more likely with an underpowered iron than a 40-50W
iron because it takes longer to melt the solder. Lifting of the copper
and pullout of the feedthroughs also happens more with an underpowered
iron since the solder isn't melted as much.

> The cap is bad to begin with so I rock the can till the
> leads break loose inside and then pull the can off, leaving
> the leads in the board. That then give access to one lead
> at a time to heat and pull, with a nice piece to get the
> needle nose on.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

johns wrote:
> I am an etech. You can' t go on a multilayer mobo with
> a soldering iron ... at all. I have a rework station that
> evenly heats all the leads at the same time to exactly the
> right temp. That is what it takes, and it works about
> half the time.

Regular electrolytic capacitors aren't nearly as difficult to desolder
as surface mount chips are, but even those chips can usually be removed
with just an ordinary iron and low-temperature solder, such as Chip
Quik. I've used just an iron and desoldering braid to replace several
surface mount chips (but with fewer and thicker pins than what you
probably work on), mostly to fix or upgrade memory modules.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

> You'll ruin that multilayer mobo trying to resolder those
> caps. If that thing is out of warranty, I'd toss it and get
> another brand. If not, RMA it. Plus, I kind of suspect
> that you have a voltage reg blown, and that is what took
> out the caps. It may have been the psupply that took out
> the Vreg ... or it may have been just heat in the case,
> but it would have been around 95 F to do that. If the
> box was in sunlight, and turned off, that would do it.

The last motherboard I bought was a 300-350 MHz Socket 7, while every
board I've obtained since then came from trash cans and usually needed
nothing but new electrolytic capacitors or voltage regulator diodes or
transistors (usually ruined by bad caps). Caps and regulator
components are among the easiest components to desolder and also some
of the cheapest.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

johns wrote:

> I am an etech. You can' t go on a multilayer mobo with
> a soldering iron ... at all. I have a rework station that
> evenly heats all the leads at the same time to exactly the
> right temp. That is what it takes, and it works about
> half the time.

Whatcha talking about, Willis?

I'm poor and get by just fine with an adjustable soldering iron. I set
it to 50W and use copper braid, but for SMD I use a much lower power
rating. I haven't done big SMD parts, just capacitors, buffers, and
those tiny chips on Maxtor HDs.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

jeffc wrote:

> >See www.badcaps.net for the story. Several Taiwan
> > companies were involved, not just one, but they all
> > bought their capacitor chemicals from a single source
> > that produced a badly counterfeited a Japanese
> > chemical package.
>
> Yes but that was far longer than 18 months ago. That can't
> be the same problem.

My 300W Antec is only slightly older than that, but late last year its
Fuhjyyu brand output caps on the +12V rail started to bulge. It had
been running in a 500 MHz K6-2 system that drew only 60-90W and only
1-2A of that from the +12V. OTOH the JGE or some unknown brand caps in
my 300W Powmax made over 5 years ago looked and measured fine with an
ESR meter.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

"johns" <johns123xxx@xxxmoscow.com> wrote in message
news:d882tt$1oa5$1@news.fsr.net...
>I am an etech. You can' t go on a multilayer mobo with
> a soldering iron ... at all. I have a rework station that
> evenly heats all the leads at the same time to exactly the
> right temp. That is what it takes, and it works about
> half the time.

If you're careful and use a temp controlled iron with
a fine tip you can easily replace capacitors, or any
through-hole components. Chips and micro SMDs are
another matter however.

I heat one side and tilt the cap away from the side I
just heated so that lead comes out a little. I then
swap to the other lead and work the capacitor out
little by little. Once I have a little space I snip
one lead allowing the cap to be removed when the
uncut lead is heated. I then use tweezers to remove
cut lead while heating it. You might need to add a
little solder to the pads initially to help transfer
heat into the joint so it melts properly.

I have a vaccum desolder gun which I use to clean out
the holes before fiting a new capacitor. Apparently
de-solder wick also works to clean up the holes. I
also have a jewelers magnifying lens which comes in
handy for inspecting the hole to make sure you haven't
damaged it.

Pop in the new capacitor and use just the right amount
of solder, snip the excess leads and you are good to go.

What ever you do don't attack a PCB with a soldering
iron more suited to plumbing that electronic work.

Plan ahead, don't apply heat for too long, once the
solder melts do the work and remove the iron quickly.
If you keep the joint heated while you are thinking
what to do you are likely to lift the copper traces
off of the board. Plan to spend 15 minutes per
capacitor until you get the hang of it.

I did my ABit KT133 RAID board over 2 afternoons as
it was too much for a single stretch.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

do_not_spam_me@my-deja.com wrote:

>
> David Maynard wrote:
>
>
>>The two biggest things to watch out for is overheating,
>>causing traces to peel up, and pulling the feedthrough
>>out along with the lead.
>
>
> And overheating is more likely with an underpowered iron than a 40-50W
> iron because it takes longer to melt the solder. Lifting of the copper
> and pullout of the feedthroughs also happens more with an underpowered
> iron since the solder isn't melted as much.

Yep.

>
>>The cap is bad to begin with so I rock the can till the
>>leads break loose inside and then pull the can off, leaving
>>the leads in the board. That then give access to one lead
>>at a time to heat and pull, with a nice piece to get the
>>needle nose on.
>
>