2
\$\begingroup\$

I recently plugged in an old Apple IIGS that had been sitting in a basement for almost 15 years and it worked for awhile before blowing up in a smokey mess. The reason for the smoke was easy to diagnose, one of the line filtering film capacitors had exploded and shattered. The fuse was also blown in the process (so I'm assuming the capacitor managed to somehow blow in a short-circuit fashion).

Although it is straightforward for me to buy replacement parts and repair the device, is there a particular reason why a film capacitor would fail in such a spectacular fashion? My understanding is that typical film capacitor failure modes involve a gradual reduction in capacitance, rather than a sudden internal short circuit. Does a capacitor failure in this fashion typically indicate another component has also failed down the circuit? Unfortunately, I do not have schematics for the power supply, but my understanding is that "Type X2" capacitors are "across the line" -- between line and neutral. The exploded (and also the intact capacitor on the other side of the transformer) read: 0.22uF X2 and have a 250V rating. See picture

\$\endgroup\$
1
  • 1
    \$\begingroup\$ If you are measuring a short circuit, check the diodes. \$\endgroup\$ – user_1818839 Dec 30 '14 at 16:44
5
\$\begingroup\$

Going by the colour of the capacitors, they look like Rifa PME 271M types. If that is the case, they are well-known in vintage computer circles. The BBC Micro PSU was fitted with them (made 1981 or so) and they fail with a lot of smoke but no fire. The smell is characteristic, a sort of wood-smoke, bonfire smoke smell. The standard fix for the BBC Micro (and similar Acorn PSUs) is to replace them all with modern X-class equivalents.

In this forum post:

http://www.vintage-radio.net/forum/showthread.php?t=69128

There's some failure analysis, which suggests a mismatch between the capacitor itself and the epoxy coating. Rifa capacitors that have not yet failed usually have many tiny hairline cracks in the epoxy outer shell.

Do an image search for "Rifa PME271M" to see more examples of failed capacitors.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ That was exactly it! And you are right, the "working" capacitor has hairline cracks in the outer casing as well. Guess I'll be replacing all of them. \$\endgroup\$ – Zuofu Dec 31 '14 at 0:23
  • \$\begingroup\$ Yep, classic failure on older gear, these come out with extreme prejudice wherever found as it is not so much a case of will they fail, but when will they fail. I notice another two smaller parts close to the mains input that also look like they want replacement (The one good thing about these is that the colour is immediately distinctive). \$\endgroup\$ – Dan Mills Sep 22 '17 at 16:20
4
\$\begingroup\$

X and Y type capacitors are specified to fail open, withstand gross overload and self-heal the microscopic defects they accrue over time (this is done to ensure they are safe for direct connection to line voltage without any protection mechanisms - MOVs, PTCs, fuses - in front of them). In 99.999% of cases this means you should never experience what you saw. They are designed specifically against failing short. X capacitors are different from Y ones; X capacitors are connected between line and neutral whereas Y caps are connected from L/N to ground. Obviously, Y capacitors are even more stringently specified against shorts, as this could cause a dangerous situation.

That being said, the odd one out might fail short like this. It is certainly a freak accident, quite possibly because of the age of the machine. X capacitors failing short shouldn't cause too many problems in a sound electrical installation.

The X and Y capacitors are directly connected to line voltage, i.e. there should not be any components in series with them that can increase stresses to the point where they are likely to fail faster when another component has failed. Also, failure of the X capacitor probably doesn't indicate failure of any other components down the line; blue smoke isn't contagious by itself.

\$\endgroup\$
2
  • \$\begingroup\$ I think that all of a particular vintage of those old RIFAs will fail just like that. They are essentially a "replace on sight" sort of a part, when sufficiently old. Any of those made in the 80s or earlier certainly should be replaced without question. \$\endgroup\$ – Kuba hasn't forgotten Monica Dec 29 '17 at 22:12
  • \$\begingroup\$ Is there any reference for the "specified to fail open" part? This answer claims they are not designed to fail one way or the other but rather are designed to be more resistant to failure. \$\endgroup\$ – crokusek Sep 22 '18 at 23:25
0
\$\begingroup\$

Ditto what Bob said: X type capacitors are designed to fail short and blow a fuse. Y type capacitors are designed to fail open soas not to cause a dangerous shock if case ground becomes disconnected from electrical ground.

When a Class-X capacitor, also referred to as an "across the line capacitor"—the capacitor placed between line and neutral—fails because of an overvoltage event, it is likely to fail short. This failure, in turn, would cause an overcurrent protective device, like a fuse or circuit breaker, to open. Therefore, a capacitor failing in this fashion would not cause any electrical shock hazards.

If a Class-Y capacitor, also known as the "line to ground capacitor" or "the line bypass capacitor"—the capacitor placed between line and ground—fails short, this could lead to a fatal electric shock due to the loss of the ground connection. Class-Y safety capacitors are designed to fail open. A failure will cause your electronic device to be subjected to the noise and interference that the capacitor would normally filter out, but at least there will be no fatal electric shock hazard.

Source: https://www.allaboutcircuits.com/technical-articles/safety-capacitor-class-x-and-class-y-capacitors/

\$\endgroup\$
-2
\$\begingroup\$

X type capacitors are designed to fail short circuit, this is to ensure full and safe operation of protective devices (fuse/circuit breaker etc...).

\$\endgroup\$
2
  • \$\begingroup\$ No. You are wrong. \$\endgroup\$ – winny Sep 22 '17 at 14:39
  • 2
    \$\begingroup\$ @bob, you might want to provide a source for that claim. \$\endgroup\$ – Voltage Spike Sep 22 '17 at 15:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.