2 of 2 people found this helpful
In professional production processes shelf life for elcaps is limited to 3 years. It might also be the solderability which is an issue and not the deformation only. In practice I have no (or hardly) any failures due to the deformation of the oxide in an elcap. If you want to be on the safe side (but who has the time?) you have to apply the working voltage to the device and limit the inrush current (via a current source or resistor). After a while the leakage current of the device is drastically reduced. This is one of the reasons that I measure input power in standby mode of a product after several hours.
Guaranteed lifetime or basic lifetime of 2000 hours is not bad. You can calculate useful lifetime with your application conditions (ripple current and ambient temperature) and with a given use of the device (x hours on / x hours off) you get a lifetime calculation. The elcaps are most of the time the first to fail, so in fact you can calculate the lifetime of the whole product. Using Multicomp capacitors is not my first choice. You will be unsure of the specification of the device. Supplier as Rubycon, Lelon and others give formulas and equations to calculate for useful lifetime (with a certain guarantee)
Best regards, Enrico Migchels
Your 3 year figure is certainly useful guidance - I'd missed on this tip somehow! If 2000 hours "isn't bad" though, I'm a little worried that a product (working within ratings) is only guaranteed for less than 3 months. It is obviously a harsh world out there.
I am familiar with the idea of refoming capacitors if left unused for a long time (especially in repair of vintage electronics), and it certainly does improve leakage currents to a fraction of their original value. In the case of my capacitors however I suspect a production fault, reforming these every so often seems unlikely to have prevented can rupture. I cannot help wondering whether it is the same problem (rumoured to be related to industrial espionage of a component in the electrolyte formula) which caused the failure of thousands of PC motherboards a while ago.
Thanks again for taking the time to answer,
2000 hours load life is under maximum rating (current, temperature). As your application is certainly less, the lifetime jumps to normal values (many years). However you have to measure and calculate this lifetime figure. I recently had NCC elcaps in my design which could not meet the spec on ESR (they were copycats of the original products). These copycats will not meet the desired lifetime!
best regards, enrico migchels
The "lesson learned" here is stick with known names. Multicomp is a "generic" in house brand - not unlike generic food in the grocery store.
Stick with well known manufactures that back their product - they cost a little more but you won't end up throwing away parts in a few years or worse RE-installing them!
I also enjoy repairing & restoring electronic equipment for my hobby (amateur radio) too. It is amazing to see how many of those old caps spring back (reform) to life! In most cases though I change them out when they getb past 30-40 years - where they are critical (example: B+ supply).
This wikipedia article discusses a particular failure cluster, I believe related to poor industrial espionage where the secret formula is stolen, but one of the ultra secret ingredients is left out.
Several years ago there was a spate of bad capacitors on the market but I'm surprised that Farnell sold any. If the caps have failed during storage (assuming normal room temperatures) they were definately of faulty manufacture. The electrolyte must have gassed, suggesting there was no depolariser in the formula!
I design and service a wide range of audio equipment including many vintage types. I am always surprised how certain makes of electrolytics still function perfectly after 20 years or more of use. On test their value and leakage current are still well whithin spec! Nichicon, Kemet, Nippon-chemi-con, Elnet are the makes that spring to mind. On the other hand there are well known makes that always show signs of deterioration after this period of time.
Last week I was tortured again with an flood of bad capacitors. It is not enough to use only the well-known brands; you really have to secure the logistic-channel. If you buy for people who buy from people.... you might end up with unreliable and inferior components. An extra problem is that will not get the technical support you need to the root cause.
best regards, Enrico Migchels
I too have depended upon Farnell for my caps and have used the multicomp brand, though I have been strongly advised by my engineers to stick with "only the ...con brands." I tend to be a bit more trusting, so if they tell me that their product is just as good, I take them at their word... but, if they bite me, I bite back.
I know that the rule of thumb is 3 years... however, let's be realistic here.... Since caps do not come with an "expiration date" on the label, and I have absolutely no idea how long they have been sitting on the warehouse shelf to begin with, there is no way to know that my inventory of "new" to "6 month old caps" are not already 5 years old when I received them.
I also seem to be surrounded by machines that are FAR older than that, which were assembled of premanufactured, capacitor laden, circuit boards, that sat on a warehouse shelf for God only knows how long before were they shipped to a manufacturer and installed in these machines, which were then warehoused again, before being shipped to a distribution center and warehoused again, etc etc etc... before they were finally sold... and yet after more than a decade of full time operation... they still function fine on their original elcaps... and they were used machines when I procured them.
So... sitting in a drawer, in a climate controlled environment, there is no reason that those caps should not have lasted 5 years without a problem.
For that reason alone, I demand better customer service than what you have described.
That said, however, in your shoes, with their product being kept in those conditions, If they will not replace them when I ask, I would not argue or be rude, I would simply order a new batch of caps, wait two months, and return the defective caps... as... well as defective. Without any evidence of solder having been applied, and no evidence of abuse or physical damage, there is no reason that they should reject them.
If they do, then you are out a little shipping, if not, then you got your replacements.
2 of 2 people found this helpful
I am an hardware engineer & owner of the switching power supply company.
The Electrolytic capacitors have a twice the life if the ambient temperature drops 10 deg C.
The 2000 hours at 105 deg C rated capacitor is quite common and its life doubles to 4000 hours at 95 deg C. It doubles to 8000 hours if it drops to
85 deg C. It doubles to 16000 hours if it drops to 75 deg C. It drops to 32000 hours at 65 deg C.
The normal equipment rarely sees the temp of 50 deg C.
So the life time issue is not much to worry about.
The price of the high temp capacitors is only slightly more than 85 deg C now so many electronic equipment manufacturer use the 105 deg C caps.
Of course, in some high temp equipment they can select 130 deg C or higher caps so called, long life caps.
In my experience the lifetime of electrolytic capacitors varies widely. I have had some last decades, others a year or two. I suspect one of the main factors is temperature, with a high temperature possibly causing evaporation of the electrolyte and leading to failure. I have seen a number of small electrolytic capacitor failures in small power supplies/battery chargers such as the internal charger in those round robotic vacuum cleaners (brand name withheld). I have seen failures in well known brand old oscilloscopes and frequency meters (like 35 to 40 years old). You have to consider that the electrolytic capacitor is an entirely different animal than solid state electronics and other capacitor types in the sense that it is basically a little (hopefully) sealed bottle with some small amount of gooey or runny stuff inside. If you are buying a bunch of these, take one apart and kind of do an analysis on what you find inside. You may find differences in construction quality between manufacturers. Just an afterthought, don't try this at home, and do use sturdy protective gloves.
I wonder if anyone has any idea (or even a gut feeling!) what the expected shelf life of a "typical" electrolytic capacitor might be? Approximately 5 years ago (cannot remember exactly) I purchased a bag of ten 1000uF 25V pcb capacitors (Farnell 921 270) and when I came to use one every single capacitor had "blown" (some even leaking electrolyte.) They had been stored in normal office conditions (and were still in the bag in which they had been supplied.)
Farnell were unfortunately not interested in replacing them, but I'm very glad that they never actually got into service where presumably they would have failed even sooner. In another life I repair vintage valve radio receivers; while some electrolytics do fail after 50 years of use, many definitely survive! Is this the price of miniaturisation? Obviously there are definite implications for the stock levels I should be attempting to maintain if this is commonplace.
I'd be fascinated to hear any comments other users may have!
PS I've just found the details for these capacitors via the Farnell Australia site! They are "Multicomp High Temperature MCRH series general purpose electolytics" and apparently have a guaranteed lifetime of 2000 hours at 105 deg C. I wonder how long it was before mine failed?! Maybe I should have checked them every six months!