Bad Electrolytic capacitor looks OK

[Excavatoree]

Sorry to bore everyone with an almost trivial question, but up to this point, all but two of the aluminum electrolytic capacitors I've ever changed showed obvious physical signs of being defective.  How common is it to find capacitors without any physical sign of a problem?  (Stop reading now to avoid excavatoree longwindedness)

I'm staying home today for some contractors are working on the house, and my cable TV device (called a "box" in the US) started malfunctioning.  I quickly determined it was the power adaptor, and cracked the thing apart and was surprised to not see a single bad looking capacitor.  However, I checked the output filters and found one that had an ESR of 8 ohms or so.  I changed it and all is well.   

The only other capacitor like that I changed was a small one in a large monitor.  It just made me wonder. Since I really haven't seen that many bad capacitors, I thought I'd ask.

[mcinque]

You can have bad caps without any visible sign of bulging or leaking. With just some increase in ESR and in some circuits they don't assolve anymore their scope.

[cvanc]

Oh yeah, it's entirely common to find an electrolytic that still looks good but measures bad in some way (off-value, or high ESR, or...).

If it's an upright thru-hole style sometimes you can't see any visual problem until after you remove it from the board.  And sometimes there just isn't anything visually wrong to see.

Especially prone to this, in my experience, are physically smaller caps.

[ModemHead]

I'd say about 80% of the ones I replace have visible signs. Either bulging on top, or electrolyte leakage from the bottom bung.  The latter is not always visible without de-soldering the part.  But I have seen all kinds of faults like high ESR, low capacitance, and even dead shorts without the slightest hint from the outside.

[WastelandTek]

Yep, not the norm, but not at all uncommon either

[Shock]

They can dry up without leaking and be open with no visible indication as well. 

[retiredcaps]

Fixing obviously bad capacitors is how I got started in learning more about electronics almost 7 years ago.

I think I have only run into one or two caps that were bad that were not bloated.  I still don't have an ESR meter and don't feel need for one.

The smaller startup 10uF 50V caps in the primary side of the SMPS sometimes dry up and have high ESR with no signs of obvious distress.  I just replace them without testing them when I repair a SMPS.  I use low ESR caps from Nichicon, Rubycon, United Chemicon or Panasonic.

[wraper]

Most of small capacitors don't even have a vent. And if they have bulging is very rare. Larger capacitors wich fail without bulging is not that rare as well. Especially if not LOW ESR types in some switch mode PSU.

[retiredcaps]

Fixing obviously bad capacitors is how I got started in learning more about electronics almost 7 years ago.

In fact, I'm staring at the monitor that I recapped more than 7 years ago.  It gets daily use and still works fine despite using a $2.50 USD soldering iron with questionable soldering joints (my first try).

I have now fixed more monitors and TVs than I need and give them away to friends and family.

[Specmaster]

I think I have only run into one or two caps that were bad that were not bloated.  I still don't have an ESR meter and don't feel need for one.

You surprise me, I wouldn't want to be without mine now, I have come across many caps that showed no signs of distress either physically or when measuring the cap for value but check the ESR and you can see the problem straight away.

[tautech]

Without an ESR measurement you are flying in the dark some but you can get by by just setting a personal limit of how much the measured capacitance varies from label values. Somewhere between 5 and 10% below labeled value sends a strong warning to me that it's time to swap a cap out but it much depends on what type and the age of the circuit. Old designs were very conservative and cap value drift better tolerated than modern designs.....but we didn't have much SMPS back then either.
Experience counts for a lot as to which caps will be OK a little low and which caps won't.

As retiredcaps says the local VCC rail IC cap in SMPS is a primary suspect.
Secondary side SMPS can show signs of swelling OR NOT, they only quick check if you haven't got an ESR meter is to pull and value measure/check them and in most cases if you've gone that far.........replace them anyway.
I never used to but now I consider electrolytics as consumables and keep a selection of common values on hand.

[BBBbbb]

I've seen intermittent issues caused by low capacitance, although no visual changes of the caps.

[Specmaster]

I never used to but now I consider electrolytics as consumables and keep a selection of common values on hand.

Like you, I too have a stock of common sizes and caps are are first on my list of checks, after doing a thorough visual check first in case there is something obvious like a blown fuse or blown / burnt out parts.

[wraper]

I think I have only run into one or two caps that were bad that were not bloated.  I still don't have an ESR meter and don't feel need for one.

Yeah, and left many bad caps in the circuit. If the cap is not directly in the signal path, quiet often circuit still works when some capacitor is bad. Usually with degraded performance, like some hum appears in the sound if it is audio device. Reduced stability if it is some digital stuff and similar.

[wraper]

Somewhere between 5 and 10% below labeled value sends a strong warning to me that it's time to swap a cap out but it much depends on what type and the age of the circuit. Old designs were very conservative and cap value drift better tolerated than modern designs.....but we didn't have much SMPS back then either.

Strange practice, considering most electrolytic capacitors are rated +/-20% when new. Actually many of large capacitance are pretty close to the minimum of acceptable capacitance straight out from the factory. Generally actual capacitance is either around as nominal value or lower (within acceptable tolerance) but only rarely higher. Also, unless you measure at 120Hz with LCR meter it is not a reliable measurment at all. Multimeters show different value because don't use sine wave (as specified in the cap datasheets) but usually charge/discharge it with constant current and similar techniques that supposedly should do the same job. However this does not work well with electrolytic caps with their imperfections. As example, when measuring polymer cap, decent multimeter and LCR meter @120Hz will show almost the same reading. Repeat this with electrolytic cap and readings easily can be 10% different.
BTW, when cap is old and insulating oxide layer degrades, capacitance actually rises, not drops.

[tautech]

Somewhere between 5 and 10% below labeled value sends a strong warning to me that it's time to swap a cap out but it much depends on what type and the age of the circuit. Old designs were very conservative and cap value drift better tolerated than modern designs.....but we didn't have much SMPS back then either.

Strange practice, considering most electrolytic capacitors are rated +/-20% when new. Actually many of large capacitance are pretty close to the minimum of acceptable capacitance straight out from the factory.

We all have our ways to do tests based on the gear we have and what has worked for us in the past.
I agree that DMM capacitance readings can be misleading and I've learnt to trust my ST3 Smart Tweezers more but it does ESR readings by default too. 

[Naguissa]

It happened to me with a refrigerator's main board. I brought to service because it seemed caps, but looked perfectly immaculate. He changed them and its working again (this was January 4th). So it seems to be frequent.

Enviado desde mi Jolla mediante Tapatalk

[Specmaster]

ESR is a critical factor and should not be overlooked. It is in effect like having a cell go bad in battery. The internal resistance goes so high that it seriously impedes the caps ability to smooth the supply rails as the resistance chokes its ability to release its stored up energy fast enough to reduce the amount of ripple. The very thing we all do is to check volts and a cap in a power supply with a high ESR can make a dramatic effect here.