It's a well-known fault, but has anyone ever known one fail and actually
cause any damage (other than to itself)? AIUI, they're there to reduce
Not really. Maybe a blown fuse. I have had the live-earth capacitor in a filter
fail and trip the RCD in my consumer unit (meaning I was in the dark until I
reset t). Does damage to my nerves count :-)
If one of these capacitors in a mains filter circuit fails I tend to replace the
lot (for all I don't like shotgun debugging). If I was repairing something for
somebody else then I'd probably change them [1]). I might change them if I
was rebulding the PSU section of a machine I own. Otherwise I just leave them
until the fail.
[1] Otherwise when they fail a few weeks/months/years later I might find I have
to replace them for free.
Incidentally, some machines have those sealed metal cans containing filter
capacitors, inductors and sometimes discharge resistors [2]. Has anyone ever
had the capacitors in one of those fail?
[2] I wish the manufacturers would print a schematic on the can, or at least in
an available data sheet. If I measure a 1M leak to earth from the live pin (say),
I want to know if it's a deliberate discharging resistor or a leaking capacitor.
-tony