Why do caps blow? ie, swell

Discussion in 'General Electronics Chat' started by tpny, Dec 29, 2013.

  1. tpny

    Thread Starter Member

    May 6, 2012
    216
    0
    Hi, this seems like a common enough problem in diagnosing a faulty board, in my case an lcd tv's power supply board. But my question is general, what causes caps to blow or swell up, can we make a list of any possible reason is appreciate! Thanks!
     
  2. Dr.killjoy

    Well-Known Member

    Apr 28, 2013
    1,190
    156
    Heat.dry out. Voltage.esr. problems with circuit
     
  3. strantor

    AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    I think the dielectric gets a pinhole, starts leaking electrons. This turns the cap into a resistor, which turns into a small space heater, which turns into popcorn. Not sure though.
     
  4. DerStrom8

    Well-Known Member

    Feb 20, 2011
    2,428
    1,328
    Yup, as Dr.killjoy said, the heat tends to dry out the electrolyte and the increased resistance causes them to bloat and change its capacitance. That is why you should NEVER mount an electrolytic capacitor against a heat sink. Also, try to find ones that are rated for 105*C with a low ESR. This helps present them from drying up so easily due to heat and keeps the ESR reasonably low.
     
  5. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    heat from ripple current
     
  6. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,982
    3,713
    Two reasons

    One, as said above, the electrolye (usually a water based salt solution) drys out and inreases resistance. Also generating heat because the ESR is too high for your application drys our electrolyte - can even directly lead to thermal expansion and pop/ swelling.


    Or two,
    Connecting them in reverse polarity at more than ~2 volts for a long enough period will reduce the Al2O3 dielectric and allow current to flow through the aluminum and electrolyte. Resistance will be high just at breakthrough but fall as more surface area is reduced - allowing more and more current to flow - effectively applying more and more power to the water-based electrolyte solution. The thermal expansion of the heating electrolyte causes the can to swell and leak or sometime burst and pop.

    Non-polar electrolytes have the oxide coating on both sides. Of the aluminum foil. The foil is essentially inert at that point - completely insulated except were the wire lead is connected.

    Bonus
    Very old electrolytic a can also just dry out on their own an appear as a short (very low capacitance) without generating heat.
     
    Last edited: Dec 30, 2013
    tpny likes this.
  7. tpny

    Thread Starter Member

    May 6, 2012
    216
    0
    what can we expect the lifetime of an electrolytic cap on a board to be? If all conditions are normal, not overheat, etc..
     
  8. DerStrom8

    Well-Known Member

    Feb 20, 2011
    2,428
    1,328
    There is no "normal" when it comes to circuit operation. Atmospheric conditions and even height above/below sea level affect it. There's really no good way to answer this question. Just look for good capacitors--Rubicon, Panasonic, etc--and avoid cheap Chinese or Korean ones--Samxon, Capxon, (pretty much anything with an X in it is garbage, in my experience). Also go for low-ESR types, rated for 105*C, as I mentioned before.
     
    tpny likes this.
  9. strantor

    AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    Estimates are given in the datasheets.
     
  10. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    Maybe 10,000 hours if you are lucky.

    I remember Jim Williams the genius at Linear Technology wrote an article explaining how the electrolytic caps failing was the reason most PC's only lived a couple of years.

    The article was eventually taken down because it did not reflect kindly on LT's big customers on the Pacific Rim.

    But even the best aluminum electrolytics have a poor life span. The cheap junk used in PC's is like russian roulette.
     
  11. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    Operating temp affects it most and most people incorrectly assume the cap maker's data sheet specs reflects AMBIENT temperature which is not correct. Ripple current causes the capacitor to run significantly hotter than ambient... hence their short life.
     
  12. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,982
    3,713
    Then again, I have two HP6228 bench power supplies from the mid-1970s with four big Sprauge "Compulytic" capacitors in each and they all work fine. The unit still meets all specs. They are the original spec replacement part and date coded mid-70s so I assume they are original. Quality, quality, quality.

    Cheers.
     
  13. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    My home built bench supply was built by me in about 1980 and still has the original sprague caps. Seriously derated, significantly oversized, and very high quality (super low ESR designed for HF switching) means internal heating is non existent, hence very long life. Manufacturers no longer derate or design to last, they have to compete with the Chinese junk selling for cheap. Their target life for a product is about six months longer than the one year warranty period.
     
  14. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    A bit part of the problem is that the rubber seal on the electros dries out and gets hard. Then the seal fails and the electrolyte dries out and the cap dies.

    Before we had ESR meters in the TV repair industry one popular test was to pull the cap and poke the rubber seal with s tiny screwdriver, and if the rubber was hard the cap was dead or would soon be dead.
     
    PackratKing likes this.
  15. Metalmann

    Active Member

    Dec 8, 2012
    700
    223
    One of my old Akai Reel to Reel tape decks has two large, bulging capacitors, one has a slight leak.
    Still plays, but the left channel has gotten weaker over the years.

    Not too bad, considering it was bought in Vietnam in 1967.

    Were the capacitors made better back then?
    I'd say yes...:confused:
     
  16. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    This is fairly common in SMPSU reservoir/filters, the ripple current is at a high frequency - and being pretty frequent, heats up as losses in the cap's ESR. In the past this could be pretty dangerous as the boiling electrolyte builds up a head of steam, people have been blinded by the can as it flies across the room like a bullet - that's why, nowadays the manufacturers crease the tops of the can to deliberately weaken them, so they can vent safely without blowing the can across the room.

    It can happen in linear supplies too, if the voltage rating is exceeded leakage current rises sharply - again possibly boiling the electrolyte. Sometimes the leakage current can become excessive just because its a crap quality component - like the one that vented in my Hitachi DAB radio.
     
  17. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,982
    3,713
    When are you going to fix it?
     
    Metalmann likes this.
  18. #12

    Expert

    Nov 30, 2010
    16,248
    6,745
    My 1978 Onkyo amplifier runs 24/7. The "on" LED has burned out, the selector switches are intermittent, and the volume control is scratchy. I'll fix it when the PS caps blow.:D
     
    Metalmann likes this.
  19. Little Ghostman

    Member

    Jan 1, 2014
    294
    97
    I posted a few days ago in the other place about a bad experience with caps (i do have a point and a question, but first i have to waffle), I won on ebay a large box of new and used caps! many in the pics were huge! So I bid my 99p and I think I eventually won them for around £1.10 plus postage.
    The box arrived and inside were several of those very large jiffy bags! I tore the top off a bag and dived in. pulled out a huge cap of 400V 2200 uf (some the caps are 22000 uf!), dad walks into the lab at this point, just as I am diving back into the bag. He grabbed my arm and took the cap from my hand, funny kind of look on his face, then he took the DDM and measured across the used caps screws terminals! 400V @ 2200uf, my bottom twitched like a bunnies nose (his saying).
    We carefully slit the the jiffy bags down the sides and extracted the caps, many were fully charged at over 100V with more than a handful at 300-400V.
    So we screwed resistors across the terminals and discharged them. Anyway the main brand (and my point) was sprauge!! I have looked all over for datasheets but looks like these one are too old and the sheets are long gone :(
    I have tested the cap values and all seem good, ESR seems pretty good as well! and i like the screw terminals :D
     
  20. Metalmann

    Active Member

    Dec 8, 2012
    700
    223

    Good question.....whenever I get my music, computer, workshop, laboratory; remodeled.;)
    Right now, I don't have enough flat bench space, to tear into the bigger projects.
    Have several small piles of projects, all over this room.

    Need lots of drawers and bench space.

    In the Spring, I also have to lathe up a new capstan adapter. Pinch roller.//
     
Loading...