Shop OBEX P1 Docs P2 Docs Learn Events
Popped Capacitors — Parallax Forums

Popped Capacitors

xanaduxanadu Posts: 3,347
edited 2012-06-02 03:32 in General Discussion
Popped caps have been plaguing electronics for what, a decade now?

pcpsu.jpg


I've read a lot about popped caps and all the stories between stolen electrolytic formula to sabotage. I just want to know, when can we expect to stop seeing popped caps? If it were really a problem that existed so long ago, why are we still seeing new caps pop. The 'solid' aluminum electrolytic pop too!

I've taught a bunch of 'techs' how to replace them. Sometimes I don't think they should be replaced though. In some cases it seems like there might be some other problem causing the cap to go bad, and the new one might lead to frying something else.

So, what do you think? Is it just the electrolytic formula that 'goes bad' and in most cases replacing them will fix it? Or, is it possible that heat or current is the cause, and the poor innocent cap is the victim here?

In the 500+ motherboards I've repaired I've had a few come back popped all over again, so now I'm going to ask the pros here for advice. Thanks.
«1

Comments

  • kwinnkwinn Posts: 8,697
    edited 2012-05-30 18:19
    I am sure the problem has more to do with the operating conditions than the chemistry. The main causes of capacitors failing are the electrolyte drying out and reducing capacitance to a fraction of the original value, or “wiskers” growing through the electrolyte creating a short or high leakage.

    High ambient temperatures or self heating caused by large current peaks at high frequencies are are a common cause of failure, particularly in switching power supply and PWM circuits.

    Operating a capacitor close to its maximum voltage rating increases the risk of shorts due to wisker growth, and operating at higher temperatures makes the problem worse.

    Ultimately the problem comes from using the lowest cost components that will do the job instead of one that has a bit of a safety margin in it's ratings but costs a little more.
  • RDL2004RDL2004 Posts: 2,554
    edited 2012-05-30 18:37
    Most non-electrolytic capacitors have very long lifetimes, but with electrolytic capacitors a normal lifetime may actually be only a few thousand hours. There are many factors involved as Kwinn has pointed out, but excessive heat is probably the most common problem. I read of a certain model TV that had problems with early failures because the way the power supply board was mounted exposed some caps to more heat than was expected.
  • ercoerco Posts: 20,244
    edited 2012-05-31 11:17
    Seems like caps have been problematic from the start. In the OLD days (vintage radios & TVs, 40's-60's) the original paper capacitors dry out and/or get leaky. Cap values change and they become fire hazards inside those hot, high voltage environments.

    http://www.antiqueradio.com/Dec04_Cantelon_capacitor.html

    And that's just the start of the fun/danger. Some old TVs (and probably some radios) have a "hot" chassis. No isolation transformer from the AC line, in fact the chassis is directly connected to one side of the AC line.

    http://www.techlore.com/article/16182/Getting-Started-to-Repair-Your-TV-Set----The-Ins-and-Outs/
  • sam_sam_samsam_sam_sam Posts: 2,286
    edited 2012-05-31 15:23
    This seem to be a real problem with bad caps in power supply LCD monitor LCD TVs and the like I replace all of the caps with Panasonic FC series

    The gutless, bloated, and fried power supply hall of shame http://www.badcaps.net/forum/showthread.php?t=8490

    This is a very help full site when working on power supply or LCD monitor and LCD TV
  • xanaduxanadu Posts: 3,347
    edited 2012-05-31 22:26
    Thank you for the advice and information.

    I guess the problem is a combination of factors. I didn't know capacitors went bad, I mean did but not bad as in leaking. I have electronic equipment from the 80s and the caps are fine, however these devices are not on 24/7/365 and also don't use switching power supplies. It seems like linear power supplies may be less hard on the caps?

    Anyway, I broke down and replaced the caps in the pictured power supply, all 1000uf 6.3v, replaced with 1000uf 12v with a higher tolerance low ESR rating. Not much choice in the matter waiting for another PSU even overnight wasn't an option. I did order one anyway as a spare.

    I love the ATX spec, but these mini computers seriously need to start using ATX spec power supplies. It's complete BS manufacturers are allowed to produce 1000 different versions of one power supply that could be interchangeable. I'm writing a letter lol...
  • localrogerlocalroger Posts: 3,451
    edited 2012-06-01 07:52
    You are not properly initiated into the wonders of the electronic world until you have seen a capacitor explode.
  • GadgetmanGadgetman Posts: 2,436
    edited 2012-06-01 11:18
    No, no ,NO!

    Initiation is done by charging one up, then tossing it to the newbie(while he's standing with his back to you) and shouting 'Catch!'...
  • PublisonPublison Posts: 12,366
    edited 2012-06-01 15:35
    I have replace a lot of Caps in Samsung LCD TV's. They always put under rated caps in the Power Supply. Change out the 6.3VDC with 10VDC and 10VDC with 20VDC.

    Instant gratification of an exploding Cap is to put a Tantalum in backwards. :)

    Jim
  • kwinnkwinn Posts: 8,697
    edited 2012-06-01 22:57
    Publison wrote: »
    I have replace a lot of Caps in Samsung LCD TV's. They always put under rated caps in the Power Supply. Change out the 6.3VDC with 10VDC and 10VDC with 20VDC.

    Instant gratification of an exploding Cap is to put a Tantalum in backwards. :)

    Jim

    Good suggestion but you should also look for a better quality capacitor with a lower ESR rating as well. This is needed most for high peak current circuitry like switching supplies and PWM circuits.

    As for capacitor lifetime being short, that is not necessarily true. A line of equipment made in the mid to late seventies came with electrolytic caps on the 5V supply (55,000uF, 16V) that rarely lasted more than a year. I replaced them with a 55,000uF 25V "CGS" (Computer Grade) capacitor and as far as I know none of the replacement capacitors died before the equipment was retired. I still get the odd service call on one that I installed in 1977. In most cases this equipment was left on 24/7.
  • xanaduxanadu Posts: 3,347
    edited 2012-06-02 01:42
    Equivalent Series Resistance. I'm determined to figure out why this is such a problem, are manufacturers not aware that switching power supplies wreak havoc on higher ESR caps? These higher freqs lead to dielectric breakdown or possibly internal heat that breaks it down?

    I cannot comprehend how a manufacturer can choose the 'wrong' capacitor, after years of failures. That does not compute. Case not closed. When I find non 'low-ESR' caps in a device with a switching power supply I'm gonna go to work on the manufacturer with a screw driver and blow torch.
  • GadgetmanGadgetman Posts: 2,436
    edited 2012-06-02 01:57
    The problem is that GOOD capacitors are expensive.
    Higher voltage rating? = Costlier.
    More Farads = Larger and costlier.
    Lower ESR = Even more expensive.

    Of course some companies will try to 'balance' the expected lifetime of the unit with the assumed lifetime of the components. They always expect a few to break before the sytem is obsoleted or out of warranty, but the cost of replacing those units can be much smaller than the cost of using good quality components throughout.
    So in a sense it makes for 'good business' to use the cheapest stuff they can get away with.
    If they calculate right, they save a bundle, with only a neglible loss of reputation.
    If they miscalculate, or the components were a bit worse than anticipated...
    (Anyone ever heard of an Bang & Olufsen system suffering from blown caps?)
  • xanaduxanadu Posts: 3,347
    edited 2012-06-02 02:09
    Gadgetman wrote: »
    The problem is that GOOD capacitors are expensive.

    Thank you Gadgetman. If that is true then I'm going to become a consumer advocate. Bang & Olufsen prices are a far stretch from a $0.03 cap vs a $1.30 cap. If I were to replace all the caps on a motherboard with low ESR it would raise the price by a couple of bucks, not hundreds or even thousands of dollars. I'm still not drawing the connection between saving money and producing a product guaranteed to fail.
  • GadgetmanGadgetman Posts: 2,436
    edited 2012-06-02 03:32
    The price matters when you push 100.000 or more of the same unit...

    Bang & Olufsen aren't expensive just because of good caps, but because they don't cut corners anywhere.
  • ercoerco Posts: 20,244
    Another popped cap story that ends well. I was given a Peltier mini-fridge that works great, but the 12V 4A switching power supply brick was well... bricked. The 12V no load voltage dropped under load. I ordered a 5A replacement unit off Ebay but that was a joke and a fire hazard, so I cracked open the original PS today. It's a switching power supply, lots of components, inductors and two big aluminum heat sinks. I instantly spotted one big 2200 uF electrolytic cap near the output which had vented. I just happened to have a same-sized 3300 uF cap in my drawer (hoarding pays!) so I put it in. I also checked a few diodes near the output which tested fine. I didn't expect that replacing one bad cap would solve the problem but it totally did, we're back in business. Another Festivus miracle! Love those easy little victories. I'll be celebrating with a cold beer whenever this li'l guy can chill it down.
  • Heat has always been and still is the grim reaper of electronic equpment.

    In my early days of field service contracts required regular Preventive Maintainance visits to clean filters and lubricate fans.

    Once the cooling system was checked out the amount of AC ripple was recorded for all DC voltages and if it was too high the first thing to do was replace the caps. Most times that corrected the problem and saved having an emergency call out because a cap went pop.
  • ercoerco Posts: 20,244
    Stop overheating, you silly circuit, or I'll pop a cap in ya!

  • Erco,

    One lesson that I learned is that electrolytic capacitors don't like to be hot.
    It's funny because I remember always hearing to place the capacitors as close as possible but guess what happens when you stick them near a heat sink.
    The electrolyte dries up, the capacitance changes, and pop goes the weasel.
    Either the capacitors were underrated or the cheaper than dirt special of the week.

    Hopefully your stash were from a reputable Japanese company.
  • ercoerco Posts: 20,244
    edited 2018-06-18 19:43
    Heat is the painfully obvious problem here. This heat sinks get too hot to touch, even with the guts exposed to open air on my bench. In use, it's actually sealed in an unventilated plastic box. I'm surprised it lasts more than a few days and doesn't melt or catch fire.
  • Oh wow. That issue again, and again.

    I once had that happen to a Dell Optiplex design that I took in as partial payment for a job. The poor thing had its power supply in a really bad position. It managed to cause not one, but all of its power supply caps to explode.

    I mentioned that carefully to the IT guy for the location regarding the whole line, it seems they were definitely aware of them, and did indeed get replacements.

    Dell is blaming its own vendor for that problem.
  • I'm okay with it as long as the AC supply eventually becomes the power for a docking station and the mini-fridge can switch between fully autonomous and remote control.
  • ercoerco Posts: 20,244
    Remember this guy?

  • RaymanRayman Posts: 13,798
    I was just reading how Sparkfun was founded when computer power supply broke and they couldn't find parts for it anywhere... Guess it's easier now...

    Electrolytic caps do seem to fail a lot... I think Tantalum ones are better...

    I think it's because electrolytic have liquid on the inside... Once it overheats, the liquid probably boils and pops the cap off...
  • ercoerco Posts: 20,244
    edited 2018-06-19 02:08
    Rayman wrote: »
    I was just reading how Sparkfun was founded when computer power supply broke and they couldn't find parts for it anywhere... Guess it's easier now...

    At New York Toy Fair last February, I met Max Bogue, who invented the 3Doodler pen. He said a 3D printer failed the night before a big job was due, so they just broke the nozzle off and finished the print by hand. Necessity being the mother of invention, 3Doodler was born! That story roughly jibes with this article: https://www.forbes.com/sites/amyfeldman/2017/03/01/from-kickstarter-to-20m-in-revenue-how-3doodlers-inventors-built-a-business-in-3d-printing-pens

    He tried selling it to the big toy companies, who all said no, so he Kickstarted it and laughed all the way to the bank. Never let someone else's lack of vision kill your dreams!

  • Erco,

    Unless it can make a profit quickly many companies will pass on it.
  • frank freedmanfrank freedman Posts: 1,974
    edited 2018-06-22 04:51
    There is a reason the tops of capacitors are scored but it doesn't always help. Had a cap blow in a GE anode starter once. Most of the can was small strip and the contents looked like cotton shreds everywhere. 300VDC in a failed cap, spectacular! Another time, fourth of July no less, I shut off my computer for the night, not quite dark outside. Bout an hour later sounds like someone lit off fire crackers; no big deal 4th of July right. Next day computer doesn't work. EVO video card popped most of its caps. Looked like a bunch of cheap new years confetti popper streams with the cans on the ends of some......... That's when I found out about the cr a ppy cap problems.
  • Voltage spike. I've only had this happen if the voltage spikes to high (or reverse polarity). It's already been mentioned but higher voltage.
  • Another EDN "tale from the cube"...
    A-bad-capacitor-story-ends-happily
  • Tracy Allen,

    Funny you should mention Tracy because I notified the product engineer immediately when I found burnt caps in a customer return.
    The engineer tells me that the customer caused it to fail and not to worry.

    6 months later I see a few more returns with the same problem and then we are told about people commenting on tech forums.
    Turns out the product works great except for with fluorescent lights.
    Anyway we ended up having to make some hardware changes and rework thousands of units.
    The capacitors weren't bad but the product was designed for ordinary light bulbs.
  • Yesterday I got a postcard in the mail notifying me that I've been identified as a potential member of a class suing a list of capacitor manufacturers. This is due to faulty electrolytic caps I may have purchased in the past. Apparently, I don't have to do anything to remain in the class, but I have to notify the litigants if I want to excuse myself from the class and pursue damages on my own.

    I threw the postcard away. By the time any damages trickle down to me from my minuscule purchases, I might have a check for ten cents to cash.

    -Phil
Sign In or Register to comment.