From Newsgroup: sci.electronics.repair

I have a PC PSU that went defunct some weeks back, and I finally got the
cover off to look at the component side of the circuit board. Looks
like two electrolytics are faulty, with one having a really large
"dome." Rather than completely disassembling the circuit board (and
also having to discharge the large electrolytics), I'm wondering if I
could cut and remove the existing one and use the existing leads to
solder in a replacement? Problem is that there's no extra lead length
between board and cap, but thought someone might have a method. I was thinking of just maybe cutting through the center of the cap as its
small and then there would be enough lead length left to solder in the
new one, but not sure if that would work. Thanks for your help.
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On Fri, 17 Apr 2026 14:59:18 -0400, Nick Sebul <nsebul@nowhere.net>
wrote:

I have a PC PSU that went defunct some weeks back, and I finally got the >cover off to look at the component side of the circuit board. Looks
like two electrolytics are faulty, with one having a really large
"dome." Rather than completely disassembling the circuit board (and
also having to discharge the large electrolytics), I'm wondering if I
could cut and remove the existing one and use the existing leads to
solder in a replacement? Problem is that there's no extra lead length >between board and cap, but thought someone might have a method. I was >thinking of just maybe cutting through the center of the cap as its
small and then there would be enough lead length left to solder in the
new one, but not sure if that would work. Thanks for your help.

Most PC psu are too crowded to do anything fancy from the
component side and there often isn't sufficient headroom to
allow for added 'pigtail' detail on a suitably-rated replacement.

Just forcefully pulling the body of the cap, to separate from
its leads (still soldered in the board) will reveal what is
available to work with. The leads will break internally in the
capacitor and the copper/Al bonding lug will be pulled through
the rubber seal. Only the lead wire portion of the remeins
will be solderable.

It's worthwhile going to the trouble of getting at the solder side.
This will also allow testing of other aging caps that may not show
obvious signs of catastrophic failure and ensure a proper solder
joint on suitable or superior replacements.

Just keep track of the screws and mechanical details, as competent
and careful reassembly is critical to later, safe operation. Hipot
barriers should be retested on the completed reassembly, prior to
re-use.

RL
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On Fri, 17 Apr 2026 14:59:18 -0400, Nick Sebul <nsebul@nowhere.net>
wrote:

I have a PC PSU that went defunct some weeks back, and I finally got the >cover off to look at the component side of the circuit board. Looks
like two electrolytics are faulty, with one having a really large
"dome." Rather than completely disassembling the circuit board (and
also having to discharge the large electrolytics), I'm wondering if I
could cut and remove the existing one and use the existing leads to
solder in a replacement? Problem is that there's no extra lead length >between board and cap, but thought someone might have a method. I was >thinking of just maybe cutting through the center of the cap as its
small and then there would be enough lead length left to solder in the
new one, but not sure if that would work. Thanks for your help.

If you can't unsolder 2 wires at a time, using a solder sucker, or
desoldering braid, maybe flush cutting diagonal water cutters might
work. They're narrow enough to fit under the electrolytic and allow
you to cut both leads (one at a time). Then, you can unsolder each
wire individually.

<https://www.google.com/search?udm=2&q=flush%20cut%20wire%20cutters>

If you're going tool shopping, also buy some PCB liquid soldering flux
(not plumbing flux), a solder sucker, 91% alcohol, flux brush, etc.

<https://www.google.com/search?q=solder%20sucker&udm=2> <http://www.chipquik.com/store/index.php?cPath=350> <https://www.google.com/search?udm=2&q=flux%20brush>
--
Jeff Liebermann jeffl@cruzio.com
PO Box 272 http://www.LearnByDestroying.com
Ben Lomond CA 95005-0272 AE6KS 831-336-2558

--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

Thank you both, I have decided to remove the board, carefully do the discharges, and then proceed with a proper replacement. A question
though: what recommended electrolytic type and brand, preferably
something good enough that I won't have a repeat in a year. I have junk
box electrolytics, but no way I'm going to use those. Thanks again.
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On Sun, 19 Apr 2026 17:06:52 -0400, Nick Sebul <nsebul@nowhere.net>
wrote:

Thank you both, I have decided to remove the board, carefully do the >discharges, and then proceed with a proper replacement. A question
though: what recommended electrolytic type and brand, preferably
something good enough that I won't have a repeat in a year. I have junk
box electrolytics, but no way I'm going to use those. Thanks again.

Don't judge a book by its cover and don't estimate the lifetime of an
aluminum electrolytic by its brand name. The operating life of the
typical aluminum electrolytic capacitor found in a typical PC power
supply is determined by the ripple current and operating temperature.

"Calculating the Operational Life of an Aluminum Electrolytic
Capacitor" <https://forum.digikey.com/t/calculating-the-operational-life-of-an-aluminum-electrolytic-capacitor/45517>

Various manufacturers have online calculators for estimating the life: <https://www.google.com/search?q=aluminum%20electrolytic%20capacitor%20life%20calculator>

One problem is that the calculators assume an appropriate product
life, proper component selection, too low a voltage rating, and
undersized capacitance (which increases ripple current and therefore
self heating. If the design and component selection is by "cost
reduction", with the goal being to minimize production cost at the
expense of literally everything (including capacitor life), it will
not matter what brand or value capacitor you use, the expected life
will probably be low. If you want to play it safe, buy a capacitor
that will fit in the PSU, use capacitance and voltage ratings that are
the same as the original, and maybe buy a 105C rated capacitor instead
of 85C. I can't provide specific numbers and specs for a likely
capacitor because you did not provide a PSU maker and spec details on
the capacitor you're replacing. If you had provided numbers, you
would likely get better answers.

Incidentally, don't discount using your junk box electrolytics. You
can test the capacitors using an ESR (equivalent series resistance)
tester. The ESR value (in ohms) will give you a clue as to the
"quality" of the capacitor.

<https://www.google.com/search?udm=2&q=esr%20meter>

Use one of these charts, preferably the one that comes with whatever
tester you own, to determine if the capacitor can be used. Also,
measure the capacitance and compare with the markings on the
capacitor.

If you really want to dive deeply into the abyss, you can measure the
ESR at different temperature extremes. Here's what I did: <https://www.learnbydestroying.com/jeffl/Electrolytic-cap-test/>
Note that capacitor "B" was bulging and in the process of failing. It
was acceptable at high temperatures, but fails at room temperature.
This explains the common phenomenon of a PC PSU functioning then warm,
but refusing to operator or restart when the PC is turned off.

Good luck.
--
Jeff Liebermann jeffl@cruzio.com
PO Box 272 http://www.LearnByDestroying.com
Ben Lomond CA 95005-0272 AE6KS 831-336-2558

--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On Sun, 19 Apr 2026 20:45:16 -0700, Jeff Liebermann <jeffl@cruzio.com>
wrote:

Incidentally, don't discount using your junk box electrolytics. You
can test the capacitors using an ESR (equivalent series resistance)
tester. The ESR value (in ohms) will give you a clue as to the
"quality" of the capacitor.

<https://www.google.com/search?udm=2&q=esr%20meter>

Use one of these charts, preferably the one that comes with whatever
tester you own, to determine if the capacitor can be used. Also,
measure the capacitance and compare with the markings on the
capacitor.

Oops. If forgot to include a link to the capacitor charts: <https://www.google.com/search?q=esr%20meter%20chart&udm=2>
--
Jeff Liebermann jeffl@cruzio.com
PO Box 272 http://www.LearnByDestroying.com
Ben Lomond CA 95005-0272 AE6KS 831-336-2558

--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On Sun, 19 Apr 2026 17:06:52 -0400, Nick Sebul <nsebul@nowhere.net>
wrote:

Thank you both, I have decided to remove the board, carefully do the >discharges, and then proceed with a proper replacement. A question
though: what recommended electrolytic type and brand, preferably
something good enough that I won't have a repeat in a year. I have junk
box electrolytics, but no way I'm going to use those. Thanks again.

Same size, voltage and temperature rating (or better).

Digikey delivers small packages for ~8$

RL
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On Sun, 19 Apr 2026 17:06:52 -0400, Nick Sebul <nsebul@nowhere.net>
wrote:

Thank you both, I have decided to remove the board, carefully do the >discharges, and then proceed with a proper replacement. A question
though: what recommended electrolytic type and brand, preferably
something good enough that I won't have a repeat in a year. I have junk
box electrolytics, but no way I'm going to use those. Thanks again.

I would use Panasonic 105C capacitors from Digikey or Mouser.
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On 4/20/26 9:52 AM, legg wrote:

Digikey delivers small packages for ~8$

RL

Yes, I know, I just got off the phone with them. Very disappointed.
For a hobbyist like me, they were ideal as I could mail in my order and
save on shipping cost when I only needed a small handful of items. I'd
print out their mail-in PDF and fill out the info then send check or
m.o. When I couldn't find the form on their website this time around, I contacted them only to find out they did away with the mail-in option.


--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: sci.electronics.repair

On Mon, 20 Apr 2026 15:11:10 -0400, Nick Sebul <nsebul@nowhere.net>
wrote:

On 4/20/26 9:52 AM, legg wrote:

Digikey delivers small packages for ~8$

RL

Yes, I know, I just got off the phone with them. Very disappointed.
For a hobbyist like me, they were ideal as I could mail in my order and
save on shipping cost when I only needed a small handful of items. I'd >print out their mail-in PDF and fill out the info then send check or
m.o. When I couldn't find the form on their website this time around, I >contacted them only to find out they did away with the mail-in option.

So the PC supply you're repairing is in 'the PC' you need to
place an order on-line? Most public libraries will book time
on their machines - and ISPs offer some kind of 'web-mail'.

List dimensions, values and destination in an e-mail and
I'll send you a packet.

RL
--- Synchronet 3.21f-Linux NewsLink 1.2