X2, AC Line Capacitors

I just noticed in the latest issue of "Banana Skins" that some X2 AC Line Capacitors lose 10 % of their capacitance per 1000 hours of operation.

"...well-known problem on X2 capacitors (across the line) X2 capacitors are safety capacitors designed to fulfill ... UL specs, CSA specs....
Means the cap is allowed to do everything, but not to fail in an unsafe way!

“To be here on the safe side such caps will be produced in a special way with special design and process parameters.
Disadvantage is here that such caps are not really stable concerning capacitance!
The capacitors are designed to fulfill in minimum the requirements of the IEC which say, that such a cap is allowed to lose max 10% of its capacitance during 1000h of operation.”
X2, X1, Y2, Y1 capacitors are designed to be self healing. That means that if an over voltage pulse is absorbed, the plates and dielectric around the punch through are damaged but become non conductive once the "flash" has passed.

Mains power is riddled with HV artefacts. Some of these HV pulses will damage the capacitors. This happens throughout the life of these capacitors. They are designed to absorb the HV pulses before they can seriously affect the performance of the following circuitry.

All X & Y rated will be repeatedly self healing on an hourly basis and as a result the accumulation of repaired patches must reduce the capacitance. It's inherent in the design.

If the spec says >90% for 1000hrs, then expect your caps to be better than that. But they will gradually lose capacitance and that loss of capacitance does not stop them doing their job of absorbing HV spikes.


2006-09-08 2:04 pm
If you use X capacitors as intended, this kind of drift is unimportant. The difference in suppression effectiveness between a 220nF and a 200nF is negligible.
If you use them for other applications, requiring accuracy, that's your responsibility....


Paid Member
2007-09-15 8:14 am
Very interesting... the loss of capacitance.

One of my first projects, an electronic fluorescent light starter (Elektor from about, ohhh.... 25+ years ago) which has been in daily use all that time has recently been under suspicion of late in that it's possibly damaging the tubes... all the resistors, all 3 of them, are OK, which leaves the caps. Once I would have had the enthusiasm to go and measure and test and check, but now.... a new conventional one for £0.21 is much easier.
Last edited: