It's not just electrolytics that can go bad...

[japlytic]

As I said earlier, AC and X-class capacitors can lose their capacitance over time, depending on usage conditions.
But I also found that Y-class and high voltage ceramic units can also lose their capacitance over time (I found this during a check of units I pulled out of equipment being scrapped).

[tom66]

Re: It's not just electrolytics that can go bad...

And poly-film or polypropylene.

[kc8adu]

Re: It's not just electrolytics that can go bad...

when crt monitors were king i replaced nearly as many film caps as lytics.
these were tuning caps in the h-out stage.
got an old mitsubishi made thomson i use as a bench monitor that i need to replace one in now.it is so old that it uses a mess of relays to switch them when switching scan rates.

And poly-film or polypropylene.

[ben7]

Re: It's not just electrolytics that can go bad...

The ones in floro lamp ballasts tend to go quite often. (Catch on fire)

Run caps from fans, etc, do this too. Well, they usually melt down too.

[pol098]

Re: It's not just electrolytics that can go bad...

In this and a previous thread
https://www.badcaps.net/forum/showth...hlight=x-class
there's discussion of X-class and other capacitors losing capacitance, and a comment that it's rarely noticed as they're basically for EMI compliance and nobody notices (or cares).

They can be used for more critical purposes. I have a thermostat remote control (Digistat SCR) which failed; I discovered that a 680nF X-class capacitor losing capacitance is the prime culprit, with lots of these units failing for the same reason (symptom: indicator lights come on normally when thermostat should trigger, but no heat. Giveaway: no relay click). In mine it measured 290nF. Resistance (at low measuring voltage) was fine; ESR reading was about 680/290 times that of a good capacitor, which looks like an ESR meter responding to the reactance, no significant real ESR. I searched here to see how a film capacitor can lose so much capacitance; repeated flashover and self-healing is what's suggested here. Sounds reasonable, but I can't be bothered to unwind it...

[tom66]

Re: It's not just electrolytics that can go bad...

And I've recently discovered leaky ceramic caps, not shorted... just leaky... and temperature dependent... FUN.

[kc8adu]

Re: It's not just electrolytics that can go bad...

found 2 shorted tantalums and 1 disc ceramic in a midland 220mhz radio yesterday.not sure why but thats all that was bad.they were supply bypasses in the pa.the pa transistors survived whatever happened.

[turbozutek]

Re: It's not just electrolytics that can go bad...

Yep, two dehumidifiers and an air conditioner I had both failed to produce enough power to click the relay over; with the problem being a failed 1uf cap on the mains!

Last thing I suspected, but after a day of replacing transistors and resistors I got there in the end!

Chris...

[pol098]

Re: It's not just electrolytics that can go bad...

Thread revival for Googlers: re X2 mains-rated film capacitors losing capacitance there's an extremely interesting article at
http://www.rutronik.com/index.php?id...3464919ce30c1c
C-stability in X2 capacitors. I won't quote the entire article, but essentially it is as surmised on this site, repeated local flashovers which gradually reduce the area of metallisation. This is by design, a fault condition merely causes a tiny change in capacitance rather than a serious short-circuit. IEC regulations allow up to 10% worst-case loss over 1000 hours (there are about 9000 hours in a year). A "weak" batch of capacitors may approach this limit while still meeting the standard. One technique to avoid flashover is to use sections in series, so that the voltage across each section is too small to be able to flash; but this increases cost and bulk. The Vishay F1772 series (the subject of the article) can avoid this corona effect. Presumably this loss of capacitance when operating at mains voltage is probably mainly a problem of the 230Vac mains world.

By the way, I mentioned an X2 capacitor that had caused a central heating controller to fail by dropping from 680nF to become an otherwise good (low ESR, leakage not measurable with simple instruments) 290nFcapacitor. The controller circuit uses a primitive capacitive voltage-dropper to generate ultimately 24Vdc from 230Vac mains; as the capacitance drops, the output voltage drops until too low to operate the relay. The physical size of the capacitor is much smaller than the Vishay F1772 equivalent. As often happens in the real world, improving the circuit with a better capacitor is impossible - no room.