Electrolytic Capacitor Life Span.

[ducatiman583]

Hi All,

I have a problem with a certain type of electrolytic capacitor currently being returned from our customer because of leakage. It is a 1000uF, 10Volt type. They all have the same "N3" (March 2001) date code and were made by the same Company in the same factory. They have been in use since the end of 2001. I do not know what operating conditions they have experienced whilst with the customer.

My question is: Am I seeing the natural end of the life span of such a capacitor? Or am I dealing with a bad batch of components?

All help much appreciated.

 

[ScottyUK]

You can't determine that without knowing what the service environment has been. Specifically I'd be looking for service temperature, ripple current and frequency, service voltage. On the face of it six years isn't very good for a small capacitor, but if the cap has been working near its rating then it is entirely possible. Big electrolytic caps used in high power converters tend to be on their way out after about 5 years unless the manufacturer has been unusually generous with the rating or they've been working way below their ratings. In those applications it tends to be heating due to high ripple current which kills them.

What does the manufacturer say?


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[Comcokid]

Beware! There has been a rash of bad electrolytic capacitors from about 2000 until about 2005 (this is my guess on the dates - research it). I recall seeing a article in in EE times about 2 years ago that the problem originated when the proprietary formula for the electrolytic of one cap manufacturer was either copied or stolen by another cap manufacturer in the Far East, and the formula spread to other manufacturers. However, they did't get the formula right (or they didn't have all the proprietary data).

The end result is that capacitors have been failing in computer motherboards made approximately around these dates.

I don't fool around with computer mother boards, but I've seen two with this failure mode already. The capacitors leak, and the capacitance goes down before the capacitors fail with a short. A board with these capacitors will show several capacitors of the same type that are visually leaking.

Normal capacitor wear/tear has the capacitor decreasing in capacitance with no leakage visible.

A quick google even brought up this website:

http://www.badcaps.net/

I'm sure more internet searching will find more data.

This was previously mentioned in a reply by me in another post: thread248-173627

Normal electrolytics are typically rated for 85C or 105C. A check of the data sheet will show a rating at these temperatures for so many thousands of hours.

 

[ScottyUK]

Has anyone posted the brand names used by the companies who stole the formula? There are certain brands which I favour for certain applications, typically high frequency, high ripple applications, but they are all European. These look like relatively little ones - I 've tried to see who made them but it's not apparent in the photos.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[itsmoked]

ducatiman583; Leakage? Chemical or electrical?

Remember that AL caps have generally lousy lifetimes. Like 1,000 hours... These lifetimes are at the rated temp with the rated ripple. If you actually use them at some much lower temp and with lower ripple the lifetime expands out exponentially. What this means is small operating changes can greatly change the expected lifetimes, so you need to understand how your caps were utilized.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com