Fascinating failure mode of metallised film capacitors.

[@Tocsin-Bang]

If I were still teaching electronics, I would set watching Clive's videos as homework. The amount of seriously practical electronics is brilliant.

[@flyingporker100]

I thought I knew some electronics, and I taught it as well. Then I watched Clive and realised I knew next to nothing!

[@johnrehwinkel7241]

I was familiar with the self-healing effect and how it eroded the electrode material, but I had never SEEN the damage. Thanks for unrolling it and taking pictures!

[@mind_onion]

You helped me save my work 10 grand by helping me to notice what was wrong with a computer we had to replace, along with the super expensive software that was paired to the computer. 10 grand for software, or 10 dollars to replace some popped capacitors. All because I was watching your videos! Thanks!

[@tennicktenstyl]

what software? if it's from a company that respects its clients it should be somewhat tied to your account, also I don't think motherboard replacement would suddenly deactivate all the stuff

[@mind_onion]

Bartosz Olszewski We have a several 10,000 dollar per annum deal with them to fix everything except the computer. The software runs a specific piece of X-ray diffraction equipment. Very niche, very high tech (but the software is passable at best, tbh, these guys are engineers, not programmers). But yeah, they basically told us the software was paired to the machine, and as far as we could tell, that was indeed the case. Our IT guys couldn't seem to figure out a way pass the copy protection, anyway.

[@tennicktenstyl]

magnusorion0 oh, that's awful. Like, even though you pay so much they still want to rip you off.. Douchebags, I use Adobe and Autodesk products with plugins and stuff, and have never had problems with license transfer..

[@mind_onion]

Bartosz Olszewski They have a very small market with a high skill set requirements, so its not all that surprising. But yeah, it is a lot of money.

[@Arckivio]

I have now watched loads of Clive's vids & I've learned loads!!!! The explanations & descriptions are brilliant & always keep me interested. By no stretch of the imagination am I good with electronics but love to learn how things & Clive is better than How it's made!

[@mjkmojofool]

You have a very interesting mind coupled with a knack for explaining in clear language, the workings of various electrical principles and devices. I really enjoy and learn a lot from your videos - thanks for sharing.

[@SuperFredAZ]

when I first started to buy led lamps for my home, I was expecting 10+ years of life, most of my led lamps last only 2 or 3 years. Thanks for the explanations. The leds are surviving but the drivers last only a few years due to effects as you show.

[@N1RKW]

Wow! I've been playing around with caps of all types for most of my life, and I never knew about this. Thanks, Clive! You've given me something else to look for when testing/repairing electronics. :)

[@Derundurel]

I've seen this too. A friend's central heating timer failed in such that you could have hot water or central heating, but not both! The series capacitor had degraded to the extent that there was not sufficient current to hold in both relays at the same time. Changing the capacitor resolved the problem, though it did recur some months later. In hindsight, I should have replaced the spike suppression components ahead of the capacitor. NOTE: - when replacing these capacitors, be sure to use the appropriate safety type as a short could very easily result in a fire!

[@johnfrancisdoe1563]

Merlin Skinner And note that the capacitor safety rating is very much about having this gradual failure mode as a major safety feature.

[@chrismni8316]

You would be a great Professor - certainly your students would love you. Keep up your great and also entertaining work!

[@craignehring]

I love understanding failure modes of things, it helps to see the big picture. Good show, thanks for this

[@opera5714]

Noticed that was a X2 capacitor. In the KEMET data sheet years ago it stated NOT FOR LINE DROPPING. I talked to Jim Lewis about this and he said that X2 capacitors are actually very poor capacitors and after years they absorb moisture and corrode the metalization layer. Also, they are only designed to get through EMI testing. No one cares if they work years later. The construction that makes them fail safe also makes them fail. I get a chuckle when people post that you have to use X2 caps for line dropping. So if that happens with quality capacitors, think what happens with Ying Yang brands.

[@johnfrancisdoe1563]

opera Does this mean they won't even do the suppression job in long living devices, where a loss of capacitance would ruin the noise suppression without obvious device failure?

[@whitesapphire5865]

And years ago when we had a capacitor failure on our 20kVA generator, one Mr clever clogs boldly proclaimed "Capacitors do not change in value".... Well that one did! We replaced it with a self healing motor run capacitor. Touch wood, it's been okay so far. Thanks, Clive, for explaining how those capacitors can lose their capacitance.

[@azzym312]

Long ago i was the engineer in a capacitor factory ( Siemens AG). Later also with Philips. We found that capacitors which have a DC bias across them fail in this mode because of moisture penetration and electrolytic corrosion. Power supplies using a capacitor as a dropping impedance always have a DC bias across them ( superimposed upon the mains voltage). If the epoxy dip is not good the life is greatly reduced. Capacitors were tested (sampled from a lot) for life by dipping in warm water for a few days and then testing for DC leakage . Some of our best capacitors MKL , MKH, Mepolesco and the like often showed this failure and the lot had to be destroyed. Your capacitors come from rather shifty sources. They skip on this test. Trusting to luck instead. This is the most common mode of failure. Any plastic film always has holes, or nearly holes in it. After capacitors are assembled and before potting they are subjected to a short HV pulse which clears any short circuit during rolling . Some capacitors do not recover from the pulse and are automatically discarded. ( sometimes quite a few actually). Some manufacturers use a very low grade film which is very permeable to water. These capacitors usually fail very soon.

[@teebosaurusyou]

Can a new replacement capacitor be dipped in an extra layer of epoxy to provide extra protection against moisture?

[@ryanjofre]

I would buy you a cup of Coffee or a Beer for this comment if I could. God bless you and thanks so dang much for chimming in.

[@timbdotus]

Keep in mind that most epoxy is hydroscopic by nature. You’d be better dipping it in some sort of lacquer or conformal coating designed for sealing out moisture while at the same time being non-conductive. A liquid Teflon might work as well, as it’s hydrophobic, oleophobic, heat resistant and a very good electrical insulator.

[@TheTardis157]

The more I watch these videos. the more I want to take a course in basic electronics to understand what you are talking about more in depth.

[@harryshector]

Really good video. Suddenly, many previously puzzling failures have a reasonable explanation.

[@sweburner]

I had this happen in a water kettle, one of those that keeps the water at a set temperature by switching the element on and off via a relay. Not knowing much about capacitive droppers, but having seen you do the calculations on here gave me the thought of replacing the cap... And it worked! It was a properly rated cap though, so it really shouldn't have failed.

[@bigclivedotcom]

+Knutte Söderberg Even the good ones fail gradually. Search my videos for the word "fascinating" to find why.

[@3butalcomp3]

Hie BigClivedotCOM, an excellent video sharing and explaining a huge reason for electronics failure. With almost 40 years on my bench I really liked this video. Since when it comes to capacitors many overlook Metalized caps and focus on electrolytic caps unless some metalized cap has a tiny burn hole in it's casing. Great job.

[@jgildert]

It's so good when you can easily repair expensive items with such a cheap component! I repaired a dumpster dive plasma TV with a 20 pence capacitor :) Just a shame it got knocked over shortly afterwards by a clumsy family member. Ah well. One of my best, cheapest, albeit short lived repairs.

[@koraypekericli]

Very interesting find. I never thought film capacitors would degrade gradually in such fashion. Great information for troubleshooting. Thanks. K.

[@theevilone]

I had a remote control plug which failed like that quite quickly, progressed to the point where it doesn't engage at all now. That is a very possible and interesting explanation for what happened!

[@incum12]

Great explanation! These are the scourge of all consumer electronics if you ask me :( Turned out they used them also in Senseo coffeemakers (from 2009-2011 or something) causing all kinds of weird problems like not hot enough water, turning off at random times, and half cups of coffee. And the biggest problem was sometimes there was no problem at all wich ofcourse makes you think it was an incident/you did something wrong. Gradually over time it gets worse, but by then the device is no longer under warranty.... These are great for the manufacters, making sure the device will die so you have to buy a new one, but not as fast that they will have to replace it under warranty. The Senseo worked without problems for about 2 years, then the trouble started and after 2,5 years I had to replace it :( Tip if you have devices with these in them: do not use them on the same wallsocket as other devices (like a lamp) and always pull out the plug if you do not have to use it all the time to make it last as long as possible.

[@richardsandwell2285]

Another great video, so interesting, we maybe need to fit larger more rugged capacitors, maybe ceramic, and just make things much larger. When LEDs fail on me, it always seems to be driver related.

[@khronscave]

Several series of Yamaha audio-video receivers (ie. home-theater amps and such) are prone to failure precisely due to this. They use a capacitive dropper to supply the stand-by circuitry, and when that's done, they won't power up anymore. I've revived one or two with this issue.

[@channelsixtysix066]

Excellent video, Clive. The common thought was metal film capacitors were pretty much indestructable. You have shown otherwise and they degrade like anything else.

[@Electester]

Great video, I learned much from it. Now I know why my 24x5050 LEDs GU10 lamp has reduced its power, and it is possible to see right with the naked eye- because it shines less. And guilty one has decreased from 680 to 320nF after 3-4 years and instead 3,5W we've got approx. 1,8W. I got to the center by scraping resin/glue around the edges, and pry the PCB off. Time consuming work... but now I can repair my lamp.

[@frankwebb7507]

Many thanks bigclivedotcom for making this video, the premature "failure" of metalised film capacitors due to the capacitance decreasing in value has been a major problem in the world for well over 15 years now but is largely unknown even by many Engineering professionals. I think that the proliferation of poor quality products has fueled the trend but even reputable manufacturers have their problems. The links below reference an article that I prepared back in 2012 about the failure of a KARL DUNGS gas fire controller due to the 0.68uF ballast capacitor decreasing in value. www.productreview.com.au/r/cannon-canterbury-in-built/307583.html And these are the photos on my Photobucket site: s1189.photobucket.com/user/Gantos1/library/DUNGS%20CONTROLLER?sort=4&page=1 This link is about a DRAYTONS gas controller in the UK that was failing also due to the 0.68uF ballast capacitor decreasing in value. www.howtomendit.com/answers.php?id=49588 The worrying trend here in Australia is that imported products from China are failing in very short time of only 2-3 years due to very poor quality metalised polypropylene film capacitors decreasing in value, I have this year alone repaired two room cooling fans due to the split phase induction motor phase shift capacitors decreasing rapidly in value from around 1.5uF to about 0.3uF after only two summers in use and you have to remember that these fans would be only used intermittently for about three months in a year so the change is rapid. You are absolutely right to be concerned about mains spike suppression applications, the governing standard here is IEC 60384-14, the usual film capacitor lifetime formulas based upon operating voltage and temperature do not predict the rapid failures that I have investigated in Australia, we do have severe high voltage mains transients during electrical storms due to poor conductivity soil so this may be a factor but I don't think it is the major cause. This report is also very instructive and contains photos of a failed capacitor similar to yours cds.cern.ch/record/2038610/files/45-56-Gallay.pdf.

[@teebosaurusyou]

Thank you Frank for posting your information and the time you spent. Knowledge is power!

[@rich1051414]

Seems it would be a good idea to use caps with higher voltage ratings to avoid the problem?

[@SeanBZA]

The most common effect is also visible on those lamps where they use a small converter, where you find the lamp flashes once a second or so as the capacitor fails. Otherwise the cheapies will dim and eventually flicker. There are some near me in a new install, about one year on and most are a disco light display. 50W downlighter lamps, all 3 led types so have a small SMPS inside.

[@EllAntares]

I saw homemade device of so called Chizhevski ionizing lamp. it's air cleansing device (if you know "plasma cleaner" inside ACs, that's the same thing). it was a voltage stepping UP circuitry, with lots of film caps. Ones from old TV were used. One of them failed in weird way that in plastic between foil films was created a channel, from which a flame-like electrical discharge emerged

[@MasterIvo]

Mighty interesting. I was wondering about overcharging a capacitor and letting it internally discharge, what type of capacitor would have self healing capability to sustain prolonged discharges?

[@johnfoth]

The metallised film capacitors do have the self-healing mechanism that you talked about. However, you'd have to have an awful lot of local failures for that to happen. I also note that there weren't loads of holes in the film - so this probably isn't the mechanism causing the failure in this case. There's another mechanism in which moisture getting into the capacitor (especially in a warm humid environment) causes the aluminium or zinc electrodes to oxidise - effectively making them into a non-conducting transparent, very thin glass - which you can't see. I suspect this is what is happening in your case.

[@DPortain]

This was very enlightening. Previously, I thought only dielectric caps could fail from age.

[@Milosz_Ostrow]

+DPortain Perhaps you meant "electrolytic caps". All capacitors have a dielectric layer.

[@Graham_Langley]

Hi Clive. This is a well-known failure mechanism of X2-rated caps as used in capacitive droppers. Some makes are apparently worse than others. I've fixed loads of them for others over the years. The last was a cooker clock/timer with a 1u0 cap had dropped to around 600nF according to my notes. Saved someone around £250 a replacement module would have cost. When given such a device to fix first check the output voltage of the supply. If it's much lower than it should be then 10:1 it's going to be the dropper cap that's gone low value.

[@teebosaurusyou]

Can a new replacement capacitor be dipped in an extra layer of epoxy to provide extra protection against moisture?

[@ryanjofre]

ALLOT of these failures all over the planet are because of greed and inferior quality(garbage) "Lytic" and Film capacitors that should have never been produced or sold(I'm not even talking about counterfeits here just junk). On the other end of that coin though is selfish and greedy consumers(especially in first world countries) that aren't willing to pay for quality and wanna have their cake and eat it too!!!

[@ryanjofre]

teebo - Failures are generally caused by garabge or low quality capacitors. The electronics corporations don't give a sh+t about a $35.00 Blu-Ray player sold @ Wal-Mart. It's gonna have garbage caps that can absolutely buldge and leak after ten years, not premium Japanese caps that can last for 30+. Just like allot of these mundane electronics devices like he's showing here. LOTS of our household devices(especially nowadays) have horrible or mediocre quality caps. Dipping your caps in epoxy to protect against moisture would be very interesting in extreme environment use(100% humidity) but it's not going to do anything to help a cap that's junk to begin with. Namaste

[@listerdave1240]

I've repaired lots of such devices, mostly timer an remote control plugs, almost all of which failed due to the dropper capacitor losing most of its capacitance. I think the main problem in all these devices is that they ignore the fact that there will always be voltage spikes on the mains and they have no protection against them. It is not only high voltage spikes that damage them but even modest spikes having a sharp voltage gradient. The spikes result in a huge spike of current through the capacitor which can overheat weak points and damage the insulation even without exceeding its voltage rating. Such spikes can be caused by airconditioners switched with mechanical relays and other heavy current appliances. I think there is a simple solution and that is to put a modest value resistor in series with the capacitor such that it drops about 10 volts in normal operation. When there is a voltage spike the capacitor will behave as a short circuit but now the voltage will be developed across the resistor limiting the current to well below what the capacitor can handle. I have tried this using resistors of about 270 to 560 ohms depending on the particular device. They still function normally despite the slightly lower voltage now available. Now only time will tell whether they last longer. The oldest modified one I have has now been working for about three years which is beyond what seems to be the average lifetime of these devices but not enough to be sure as I also have at least one (unmodified) timer that worked well for eight years. An MOV would certainly also help but I guess it is generally considered too costly for such cheap devices. A resistor on the other hand costs next to nothing.

[@bigclivedotcom]

The simplest LED lamps use a series resistor. Not just to protect the capacitor, but to prevent that high current spike from damaging the LEDs.

[@lumpyfishgravy]

Fascinating! I have only one product in the market that really depends on a cap this way, and I'm fairly sure I specified Class X. Do you think it's possible you have a lot of electrical fast transients in your local mains? Motor drives etc. Around 1-2kV, but short (50ns) and frequent. I generally reckon those things get well dented making their way down the mains cable, but your cap demo has made me think again.

[@opera5714]

Noticed that was a X2 capacitor. In the KEMET data sheet years ago it stated NOT FOR LINE DROPPING. I talked to Jim Lewis about this and he said that X2 capacitors are actually very poor capacitors and after years they absorb moisture and corrode the metalization layer. Also, they are only designed to get through EMI testing. No one cares if they work years later. The construction that makes them fail safe also makes them fail. I get a chuckle when people post that you have to use X2 caps for line dropping. So if that happens with quality capacitors, think what happens with Ying Yang brands.

[@scottjohnson7774]

what voltage rating are they? I do tube amp repair and caps are always an issue especially when run close to their rated voltage . UK mains rectified voltage is around 340 so the cheap units use 400 volt caps . I have found that they don't have enough headroom for any mild spikes. a 600 volt unit is much more robust in these cases

[@crocellian2972]

Scott Johnson - Do you know which way to install them? See Mr. Carlson's Lab.

[@exaisle]

Thanks Clive.....Just noticed this video....and I've just found arcing in the relays of a gate opening controller....the resistors in the snubber (?) circuit intended to prevent arcing are open circuit...I'm wondering if this was caused by failing x2 rated capacitors...having read this I'm going to replace all three components. Unfortunately, my multimeter can't check capacitance....

[@TerryClarkAccordioncrazy]

It's a common failure mode of Senseo coffee machines which also have a capacitor dropper PSU. I thought it was due to mains spikes causing the self-healing effect and thus tiny holes in the metallisation but from what you see here it could be another effect of the plates degrading. Interesting. I wonder why the capacitor manufacturers can't stop this happening.

[@bigclivedotcom]

+Terry Clark Probably because they are expected to make the capacitors as small and cheap as possible. They probably have specifications about using with a MOV transient suppressor, but that probably gets cut for cost savings too.

[@thestalkinghead]

very interesting to know, they don't look like they have failed so it would be hard to (visually) diagnose a problem, but most definitely something to keep in mind for broken electrical items

[@alleycatjack4562]

I'm a little surprized you hadn't found this earlier. I work with repairing a lot of old electronics, so capacitor failure is always present. They are wonderful components, but in the end usualy the culprit on circuit failures. One thing to note however is that you aren't always able to get a proper reading of a capacitors capasitence while it is in circuit. Sometimes it is required to remouve it, or at least one of it's leads to get a proper reading.

[@Milosz_Ostrow]

A few years ago I had bought a multi-pack of LED night lights and one by one, over a course of about two years, they gradually faded out. They should have lasted ten times as long. I had opened one up, and seeing that it used a capacitor dropping scheme to regulate the current through the LED, assumed it was a crappy design that allowed the white LED to see peak currents on each cycle that exceeded its rating and greatly accelerated LED aging. Since the night lights were a snap-together design that required great force to pry apart, I deemed it too much trouble to work on them and decided to toss them when they became too dim or quit altogether. It never occurred to me that the polyester film capacitors might have been degrading in the manner demonstrated in this video. If I had known about this phenomenon, I would have tried measuring their capacitance. Polyester film capacitors normally last decades; they're among the most reliable capacitor types made, but only if they're rated conservatively, e.g., picking a capacitor with 3-4 times the DCWV rating of the mains RMS AC value.

[@AnsweringAtheism]

Thanks for the info. This reminds me of the tendency of multi-megaohm carbon bootstrap resistors in switching power supplies to slowly fail open even though they are operating well below their rated power. That's caused by microscopic arcing that gradually eats enough carbon to open the resistor.

[@daShare]

The first thing I check when any mains powered item (with a capacitive dropper) goes bad is that series cap. I've seen them fail in all sorts of things, from water tank pumps to hi-fi gear. The "turns off immediately after turning on" is due to the PSU not being able to sustain the rail when the relay coil starts to draw current. It can also result in erratic operation in general as the PSU fluctuates with load.

[@ke7eha]

The self healing Nair of the coil calls is a designed in feature. this is for safety, particularly with y rated safety caps. the idea is to never fail into a shorted configuration.

[@reddragon27284]

I've come across this a couple of times, one with a fan that had a remote control and the other with a wireless outdoor socket. Both times it's been the yellow X2 type cap that has severely reduced in value.

[@jellyfishjelly1941]

Never trust capacitors you should. Always check them you must!

[@MC-Racing]

Regarding the capacitors in LED lamps, when it reduces power to the LEDs, one need to keep in mind that the voltage across the capacitor will go up, and therefore the power dissipated in the capacitor will possibly go up and burn it out. i have had that kind of failure in a cheap LED lamp.

[@andiyladdie3188]

Many thank's for sharing your knowledge!

[@theirisheditor]

This probably explains the early failure of two LED bulbs, which I'm fairly sure used capacitive droppers based on their noticeable stroboscopic light. A 7W generic 'GLS' went down to about 3 watts and a MiniSun 4W candle type bulb went down to 1.5 watts going by the watt meter. I don't remember what the GLS type bulb used when new, but the MiniSun bulb started off about 4.5 watts when purchased a year ago and both were used a few hours a day.

[@JLK89]

I think it would be interesting to compare different types of film caps. PE vs PP vs PS vs polyester, etc. My general understanding is that polyprop makes the best film caps. I have a new design where I use a 2kV pp cap to form a high power 125khz resonant transmitter where the caps generally reach 1-1.5kV. It would be quite detrimental to me if these caps' values deteriorates, as the resonant frequency changes with it...

[@andygozzo72]

Maybe the ac ripple current through them is causing it?? You sometimes had special high ripple capability capacitors in crt tv scan circuits..maybe these modern tiny things just arent capable enough to be used in this way but ok for signal coupling/decoupling.

[@mjouwbuis]

Most manufacturers abuse "accross the mains" capacitors (X class) for series power supplies such as the ones Clive shows. However, capacitor manufacturers often have different capacitors rated especially for series usage. Presumably they are better at handling ripples.

[@andygozzo72]

yep, the scan coupling ones would be better, or maybe 'motor run' types, but are much larger than the normal ones!

[@electronicspamore590]

We have an led lamp that uses this metalized film capacitor and a couple of years, the brightness of the lamp degrades until it was really dim. And also in single phase motor in fans and some washing machines, this was the main cause why these devices fail.

[@zx8401ztv]

I never realised how crap that type was, thanks clive :-D

[@MushVPeets]

The mystery of the magical shrinking capacitor! I'm no electronics guy, but this's an interesting thing to know nonetheless.

[@thomasivesdal-tronstad4648]

My Samsung LCD monitor from 2007 failed after only 3 years. I opened it up and replaced the filtering capacitors on the secondary side of the switched power supply (crappy cheap ones) with more expensive ones of a decent brand made for that exact purpose (smoothing high frequency ripple). Same capacitance, higher voltage rating, physical size not too big. The monitor is now 10 years old and is still working perfectly. Not even dead pixels! I've found that in most cases it's that exact component that tends to fail in most modern electronics. The high frequency and power conditions of the secondary side of a switched power supply really stresses the caps and so it often dies first.

[@andygozzo72]

capacitor plague, as its known....2007 would be right in the middle of the afflicted years.... crappy 'copied'/'counterfeit' electrolytics... loads of pcs were affected with motherboard caps blowing their ends and innards out and even catching fire...psu's also...

[@andygozzo72]

and Samsung tvs/monitors were notorious, so unfortunately you were double stuffed!

[@johnfrancisdoe1563]

Thomas Ivesdal-Tronstad Those are electrolytic capacitors. Completely different problem!

[@nodriveasusephotos8019]

i have those cheap led lights. they just stop working full stop. no dimming. maybe they are susceptible to surges? what if one of the 4 diodes failed?

[@bigclivedotcom]

+Alan Wolstenholme It's usually an open circuit LED. Often with a visible black dot inside.

[@nodriveasusephotos8019]

thanks

[@gregzeng]

Thank you. When you say :CHEAP", in isolated places, it is not initial cost that matters. Isolated: polar regions, deserts, remote farms, etc.

[@AndyHullMcPenguin]

I think this applies to motor capacitors in domestic appliances too. Kenwood Chef food mixers for example are almost indestructible, except for the motor capacitor. As it gradually fails (probably due to arcing from the motor brushes), it tends to also kill the triac. If you want an almost indestructible food mixer for next to nothing, pick up an old Kenwood Chef A901 and replace the caps and triac. It will be good for another 30 years.

[@meowingmono]

what's triac

[@AndyHullMcPenguin]

Good question, in this case an 8A 600V Triac. en.wikipedia.org/wiki/TRIAC is used to chop the 240V mains voltage, and supply only a percentage of the waveform, in a similar manner to pulse width modulation. This is a pretty common usage for triacs, and there are a number of simple ways to control them. The older Kenwood Chef mixers use a form of variable resistor and the modern ones use a small microcontorller to switch the triac. If you want to see some more examples of speed controller schematics based on triacs, follow the link encrypted.google.com/search?q=triac+speed+controller+schematic&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwj0ieLqtKrKAhWD9HIKHTGhDkoQsAQIJw but do bear in mind that if you intend to try this at home, you are switching mains voltage, so take care.

[@meowingmono]

Andrew Hull​ thank you :)

[@AndyHullMcPenguin]

No problem. If you are wondering why you would use a microcontroller rather than just a potentiometer to control the triac, It allows you to do a "soft start" (build up the speed more evenly), and also in some cases to react to a stall and speed sensor on the motor shaft, to cut the current if the user jams the motor. You can do some pretty useful things in software that would take a lot of complex hardware to achieve, and microcontrollers these days are very cheap.

[@meowingmono]

+Andrew Hull :)

[@nathanlucas6465]

I've got some led lights in the house (claimed 3x1w in each bulb) that have been slowly getting dimmer. slowly enough that it went unnoticed until one of them failed completely and was replaced with a new one (same batch as the original, bought in bulk) so could this be the reason for them dimming down?

[@bigclivedotcom]

+Nathan Lucas Not so likely with the higher power LEDs. It's more likely to be phosphor or LED degradation that causes progressive loss of intensity.

[@mikecamps7226]

I'm a few years late to reply...….But did you measure the thickness of the metalized layer ? Looks as if its aluminum foil. Perhaps copper might not degrade in the same manner and last longer. This has been of interest in the last month or so, relative to the old style paper & wax units from the early era where the paper degrades....then the cross over in the era to the newer synthetic materials as polyester and the vintage polyester caps becoming leaky......considering some brands are notorious for becoming leaky and other vintage caps of the same material but different manufacturer last in comparison. Makes for an interesting situation to determine expected service life as such that goes along with the electrolytic as they age and dry out. I recently had a batch of new poly caps come to me.....and the leads ended up coming out of the body or just after soldering they would come loose...….brand new.

[@leohobbleohobb3781]

great to know.Does they degrade/change value do to heat?

[@bigclivedotcom]

Not really, for film capacitors it's just aging and voltage spikes.

[@compactc9]

How old are these devices? Basically how long do these capacitors seem to be lasting before they finally do this?

[@bigclivedotcom]

+compactc9 The one in the video did not last long. Lifespan depends on the quality of the construction and the electrical environment it is used in. A lot of high voltage transients will shorten its life greatly.

[@Milosz_Ostrow]

+compactc9 As in my earlier comment here, I've seen LED night lights begin to degrade almost immediately and fail completely within two years. In that sense, one could say the capacitors were failing from the first minute that the product was put into service, although the failure was gradual, not catastrophic and instantaneous. It is an issue of the manufacturer selecting capacitors with a working voltage rating that is far too low, so ordinary mains fluctuations and spikes on the line begin to damage the capacitors from the very start.

[@compactc9]

We just started using LED lights instead of either CFL bulbs or the now unavailable regular filament bulbs. I changed the CFLs in my bedroom out for LED bulbs, as I got tired of the CFLs having to warm up, and I honestly prefer the light they give. CFLs have a rather harsh light I think. We haven't had one of the LED bulbs go out yet, but when one does I'll have to open it up and see what I find.

[@slavric]

+Milosz Ostrow In my case a lot of LED lights were failing in my house and at first I thought, that the little PS units are failing. The LEDs begun blinking and/or loosing their brightnes almost completely. When I had a pile of them, I started exploring and found out, thai the LEDs are failing, because they are just soldered to the aluminum boards without any heat conductive paste at all. No wonder they were failing. My neighbour also had problems with LED lights in his wall. In his case also LEDs were failing, but those were the ordinary 5mm LEDs. I made him new guts with capacitor from company Iskra (Slovenia/EU). Theese wont fail, they are high quality and 630V rating. I used 9W COB LED, which is powered with only 0.9W and this will last many years. I have one 1W LED in my hallway and it dhines 12 years, also it works on ~100mW.

[@thebeststooge]

+compactc9 I purchased 12 60 watt equiv LED bulbs and one went out on me as I was standing there as it flickered and flickered more and finally POOF. I never did figure out what went out in it but the bulb wasn't even 2 months old. Now 6 other bulbs are still going...for now.

[@TrollFaceTheMan]

Very informative thanks!!

[@cmj20002]

Would have been cool if you fixed the device showing that it was indeed the failed capacitor. Most think that these type of capacitors hardly ever fail and I know that isn't true. I have watched many repair videos and some times it is the film cap that went bad.

[@grahamrdyer6322]

Good one Clive.

[@edwardcoleman9291]

What are the foil and connection made of,, for precious metals recovery

[@wb6wsn]

The metalization on the plastic dielectric is vacuum deposited aluminum. Fifty years ago, when I had a company making high-quality plastic film capacitors, I created terminations by turning a rack of capacitors on end and plasma flame-spraying a layer of tin on the end of each capacitor. An ordinary tin-plated steel wire was then soldered (or spot-welded) onto the tin terminal. The finished capacitors were then either epoxy dipped or put through an epoxy powder / thermal cure operation to form a hermetic seal over the capacitor. So sorry, not much worth recovering.

[@picobyte]

It's very rare but they can fail as A dead short! Had it once in A hand drill. The cap across the rotor failed hard.

[@wgm-en2gx]

Clive or anyone, I think capacitive droppers are uncommon in North America. I think we use step down transformers. I'm I under the wrong impression? Or are these droppers not something that would pass UL or similar safety testing?

[@westinthewest]

My guess is that the lower voltage in North America means that a dropping capacitor has to be rated at twice the size of one used at 240V.

[@jayodea2661]

Clive what area of Scotland do you hail from? cheers for the great videos btw :)

[@bigclivedotcom]

Jay Odea Glasgow, but now on the Isle of Man.

[@acmefixer1]

What is fun is to take a CD and put high voltage across it and watch the high-voltage arcing eat away the aluminized foil! It's just like the film on the capacitors!

[@jimlarsen6782]

Fascinating. Who would have thought of gradual plate ablation. Someone also said that transformerless power supplies constantly draw current with no load. Maybe in your high level tinkering you could confirm or deny this. Thanks

[@bigclivedotcom]

The capacitive dropper supplies pass more current than you'd expect for the load, but it's out of phase with the voltage. Your electricity meter should only charge for the real power. (It's quite a complex subject involving power factor.)

[@jimlarsen6782]

Would my kill a watt device factor this in? It seems dead on with led light bulbs. Aside that I have noticed my bathroom led bulbs on a dimmer buzz like incandescent filaments, and they use dropper capacitors.

[@crocellian2972]

Check Mr. Carlson's Lab for information on how best to install these beasts. Turns out, they aren't symmetrical.

[@andygozzo72]

well, they ARE in terms of polarity, BUT one foil is on the outside so susceptible to hum pickup or hum radiating if the outer foil is connected to a high signal point or 'sensitive' part of the circuit... old waxed paper caps were commonly marked with one end as 'outer foil'

[@Thecando]

Oh man…. I never realized this…. I’ve been told that film types last for a very long time. In this case, it seems like a ceramic would be the best. Is that correct?

[@bigclivedotcom]

It depends on the application. Ceramic are usually low values.

[@jbravo36]

Would this be the cause of an outdoor pir floodlight to fail?

[@bigclivedotcom]

Martin cowell Possibly, if it's the older style halogen type. But water ingress to electronics is also a common factor.

[@68MalKontent]

That's why I tend to put capacitors rated at 630VDC / 275VAC at minimum in such applications. I figure - the more insulation thickness the better :)

[@markusbrandhuber9043]

That´s interesting and now i begin to understand why my MR16-LEDs sometimes are flashing

[@inthenameofjustice8811]

Now THAT is interesting.

[@Byronjojo]

Sooo stock up on good old capacitors?

[@bobyk87]

Most electronics failures are capacitor related. Best dielectric is vacuum, glass, teflon, etc, and it's expensive. You can reuse resistors, inductors, etc. but caps one should never. Always take fresh new ones, especially electrolytics.

[@philipdowns4176]

This blew my mind Clive!!! Could paper and oil capacitors fail like this?

[@bigclivedotcom]

+Philip Downs I'm not sure. I think the issue with the metallised film capacitors is that the metallisation really is just molecules thick.

[@olradguy]

pio caps are garbage, old ones are always leaky.

[@morpheox]

This is probably a part of (or in part due to) what in the design business is called planned obsolescence.

[@slavric]

I'm well aware of this phenomenon. It is ment to be for self healing, but if the cap is abused all the time, it will fail soon. In my area is a factory for metal film capacitors named ISKRA. They make very HQ capacitors. The capacitors in this configurations should be rated AT LEAST 400V, better is 630V. Those won't fail.

[@mjouwbuis]

+Srečko Lavrič 400V is too low a rating for a mains series dropper by any definition. 630V is better but may still exhibit some self healing as peaks can be above 630V. 1000V will last forever but is larger and more expensive. X2 rating (275-315VAC) is the cheapest way to get one that will theoretically withstand mains voltage but most of these are meant to self heal quite enthousiatically. They need actually be rated for series operation to guarantee prolonged functionality.

[@stefantrethan]

+mjouwbuis Do you have any more information or any example type of series operation rated cap I might look up? I have found several such failures (coffee makers are typical) but now I have a design application where I need to detect a mains control signal, I was just planning on using an X2 cap. This will be a high volume production item and I want to avoid any problem. Are you guys sure these failures are from self healing and not oxidation or electromigration or something? I have looked at the foil and seen no holes, they might be microscopic but I don't think so. I have also tried to cause the failure with high voltage but it just burns big holes through multiple layers. What about the series resistor that is often present to limit current spikes, would it not interfere with self healing? As an aside, I have also noticed a similar "eaten up" pattern on the outer wraps of foil caps that measured OK. I do not know if this is normal from the production process or the onset of progressive failure.

[@mjouwbuis]

stefantrethan have a look at www.vishay.com/docs/28153/anaccaps.pdf

[@stefantrethan]

+mjouwbuis Thank you, much appreciated.

[@MisterJFAmbrose]

Ahhhh! This may explain why I have a LED lamp that flickers on and off. I'll have to check it out. Thanks.

[@NineToFiveGamer]

Just had one fail on a power supply for an arcade game for the second time in two months. What would be a viable replacement that won't be destined to fail??? any brand recommendations? the original uTx (China crap) won't last for shit

[@andygozzo72]

what sort? electrolytic? film? if electrolytic, try Panasonic, Rubycon, or BC/Vishay

[@mrb692]

I’m aware this comment is old, but you can replace a capacitor with one that has a higher voltage rating, so long as the capacitance is equal. If you’re experiencing consistent failures, you could try that.

[@neville3151]

Failure analysis....most impressive.

[@davidchester1612]

Hmmm.... there may be a place for high quality capacitors in clear casings or dipped in clear epoxy lacquer, not colourless mind you just see through. If they took off and clear cases were seen as the "premium" ones then the standard ones would move to this format as the cost difference would be next to zero to produce. With the clear case you could see indications that the film is degrading.

[@frac]

Cheap monitors almost always fail with this mode. If you Google 'monitor cap replacement', you'll get flooded with hits. If you see a broken but otherwise nice monitor dumped somewhere, and you have the inclination, it's almost always worth swapping out the caps. I've saved two friends' large screen LCD TVs from being discarded this way. Tossing a $2K investment or $20 in caps... hmm...

[@dushk0]

You don't know what you're talking about. Those were bad batches of electrolytics. This is the "self-healing" feature over time.

[@luminousfractal420]

Always shows passion when you look at he screen 2 seconds in and there's just parts and shite all over the place 🤣 wouldn't get that with Linus

[@BRATWURST1]

The failure of metallized caps is most likely due to their inability to handle even modest amounts of current.The metallized film is only a fraction of a micron in thickness.Foil types are way better but are bulkier and more expensive.

[@andygozzo72]

exactly, most likely its the ripple current, those tiny things simply cannot handle it long term... ok as decouplers, but capacitive droppers..hmmm.... genuine class X across mains types would be better...but not perfect..

[@stanmarsh14]

Ahhhhh, good old RIFA's, or as us slot machine restorers call em, Bob Marley's, with the way they love to smoke and go bang. JPM MPS based machines are especaly well known for em in the PSU, for the smoothing circuit.

[@kevywevvy8833]

+stanmarsh14 the rifas that smoke are metallised paper, they fail for different reasons. Dropper capacitors should be metallised polypropylene and these are the ones that deteriorate over time

[@gavincurtis]

Dodgy cheapo shenzhen underground market underrated self oxidizing capacitors! At least you can replace them with good Pansonic units. Or even Orange Drop safety rated capacitors if there is room.

[@artifactingreality]

Damn caps, seem to be the root cause of the good majority of faults in electronics. Though I guess something like a filter cap is taking the majority of the abuse in a given circuit so its not entirely surprising.

[@gregorythomas333]

Time to order a 1000 pcs bag of these capacitors!

[@stargazer7644]

This is why capacitors have a rated lifetime.

[@BenjaminEsposti]

It's a self-healing property, if there is a dielectric breakdown, then the thin metal layer gets zapped and melted away, "healing" the capacitor.

[@3butalcomp3]

Benjamin Esposti it may "heal" the capacitor into a working state, but this capacitor has now changed. Depending on how much is zapped, and melted away will not only change the capacitance for sure, even if just a little - but will shorten its life because now the metalized film will naturally disintegrate faster under normal current demands until it fails because of this erosion that has taken place from a current spike, or just normal usage. This capacitor may have a chance of not failing as well to be fair -depending on it's application and if it is never damaged again and also , just because it has been compromised in such a manner. We wont know until it fails.

[@BenjaminEsposti]

Jeff L Ya that's what I was saying lol.

[@3butalcomp3]

Benjamin Esposti I took your comment : "It's a self-healing property, if there is a dielectric breakdown, then the thin metal layer gets zapped and melted away, "healing" the capacitor". As if you thought the cap was healed and still good, due to this healing property. This cap may work for a while but it's not healed, it's actually damaged and waiting to fail. Under working conditions, a cap that has been damaged this way will certainly cause a change somewhere in voltage, current handling, not to mention unknown EMI as well as correct waveform or waveform amplitude even though the item is functioning. You're not wrong at all, I was commenting you with some info.

[@BenjaminEsposti]

Jeff L Ok lol, yeah I thought about typing out basically the same thing you said. But I was lazy or tired and didn't feel like it xD

[@3butalcomp3]

heh, cool. I get that way too sometimes, so much good stuff to check out and I want to comment but feel burnt out after a while of watching. and commenting. and fixing typos! Have a good one...

[@glasstronic]

Yep.

[@jonaskarud]

Hallo, Clive! Very good video, but You cant measure capacitance when the capacitor is placed in a circuit!

[@geekycow]

Depends on the circuit, and the method of measurement. But in droppers like this you often can.

[@firestarspelt]

i had a monitor thatd turn on for a second then turn off i would think that was also capacitor never was bothered to find out though dont have a capacitance meter and there was just way to many capacitors anyway

[@3butalcomp3]

firestarspelt You can simply troubleshoot for voltage, especially in this case where you had about a second of operation before it failed. Chances are the problem was is in the power supply, and not necessarily a bad cap, although very possible. You may have had a bit of cold solder or another issue altogether. No matter how you slice it, a DVM can get you by more than you may think.

[@233kosta]

I wonder if this problem could be avoided altogether by sufficiently overrating the capacitor 🤔

[@bigclivedotcom]

If the value is too high it increases the stress on the voltage regulating components.

[@233kosta]

@@bigclivedotcom I was thinking overrating its upper voltage limit, not its value. Working on the assumption that a chonkier dielectric would be less susceptible to this sort of damage and might therefore either last a lot longer, or not degrade at all.

[@markdischinger5651]

Probably getting atmosphere and moisture causing the failure. Possibly the environment these are operating in.

[@andygozzo72]

should be sealed in epoxy... if it has split, as happens with old RIFA branded ones and some others, they can definitely absorb moisture and go bang quite nicely..

[@antigen4]

hmmm - very interesting