Use of Class Y Capacitors in SMPS

Discussion in 'General Electronics Chat' started by PlasmaT, Mar 13, 2015.

  1. PlasmaT

    Thread Starter Member

    Feb 19, 2015
    40
    6
    Just to give a background to the question:
    A few days ago I got a nasty shock from a output of a 12V power supply adapter (SMPS) . when I checked resistance between a mains supply pin and output pin, the resistance read about... hmmm not sure the exact values, but it was in the kilo ohm range. So I opened up the unit and checked the isolation stage. The resistance between the capacitor connecting the primary and secondary stages (where the so called Y capacitor, blue color normally), read about 12 kohms.

    The capacitor was removed and checked and it read 12 kOhms. So the problem was indeed the capacitor. Later I replaced it with a similar capacitor and now things are back to working condition with isolation from mains.

    My question is, why did the manufacturer use a class Y capacitor for the isolation stage. Class Y capacitors conduct at failure, this is a safety issue. Dont you think a class X capacitor should have been used so that the user would be protected in case of a failure?
     
  2. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,803
    1,237
    Avarice, lax enforcement, ignorance (in no particular order) --- à la self-igniting CFLs :rolleyes:

    Best regards
    HP
     
  3. RamaD

    Active Member

    Dec 4, 2009
    254
    33
    Failure Mode of both the X & Y caps should be open.
    X caps when they fail shorted, as they are across the line, there is the fire risk.
    Y caps when they fail shorted, there is risk of shock.
    Y caps are impulse tested for higher voltages than X caps.
     
    PlasmaT and Hypatia's Protege like this.
  4. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,803
    1,237
    Having Um... (only now) done my 'home work', I must concur...
    So... I hereby redirect my little whine (post #2) toward the component manufacturers and the relevant (internal and regulatory) compliance chain...

    This to be filed under 'Shameless face-saving ploy No. 31822':cool:

    Unabashedly
    HP:D
     
  5. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    I have a shorter method. It goes like this:

    Oops, missed that one. :oops:
     
    R!f@@ likes this.
  6. PlasmaT

    Thread Starter Member

    Feb 19, 2015
    40
    6
    Thanks RamaD, I did not know that both class X and class Y capacitors (should) fail by open mode only.
    I was in the view that X capacitors fail by "Open" mode and class Y capacitor fail by "short circuit mode".
    So was I wrong about the capacitor class ?, Does it mean that normally, class Y is safer than class X capacitors?
     
  7. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    In round numbers, X caps are line-to-line, line-to-neutral, etc. Y caps are line-or-neutral-to Earth. Y caps have greater high voltage reliability requirement because a 110 VAC line can momentarily be 1000 V above Earth while still being 110 V line-to-neutral. This comes into play more with three-phase power systems, but the ideas are the same. Both classes usually are of a film construction that is called "self-healing." A momentary short burns open rather than consume enough energy to set something on fire. The caps are UL rated for fire safety, but that doesn't mean everything around them is. X and Y caps can be ceramic, but I've never seen a ceramic X cap in use because they are much larger than films for the same capacitance. Ceramic Y caps are common, as in your case. The cap bridges the isolation barrier as part of the EMI noise control. Usually a smaller value than the Y caps at the power supply input filter.

    All of that was from memory. It's close to reality, but a true compliance person might chime in with some actual facts.

    ak
     
    PlasmaT likes this.
  8. RamaD

    Active Member

    Dec 4, 2009
    254
    33
    AnalogKid has given the reason for Y Caps testing voltage being higher.
    Again, rephrasing the words, both X & Y Caps should not fail 'shorted' is desirable. Not sure if it is certified. Again, out of memory, Y Caps of ceramic types are designed for this not to fail 'shorted' condition. Due to safety current limits of 1/4 to around 3mA in different standards, these tend to be small.
     
    PlasmaT likes this.
  9. PlasmaT

    Thread Starter Member

    Feb 19, 2015
    40
    6
    So basically, structurally, class X and Class Y are the same. Class X and Class Y are named based on the usage. Class Y is more safer than class X.
    I think I got it. Thanks.

    So I conclude : the use of class Y capacitor at isolation stage is a good design approach.
    The reliability of the capacitor was the problem.
    All caps should only fail by "open" condition.

    Thanks.
     
  10. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,348
    Hello,

    Here is a quote from the capsite:
    And here the full page:
    http://my.execpc.com/~endlr/line-filter.html

    Here are some more pages with the explanations:
    http://www.justradios.com/safetytips.html
    http://www.ehow.com/info_8649387_differences-x2-y1-y2-capacitors.html

    Bertus
     
    PlasmaT likes this.
  11. PlasmaT

    Thread Starter Member

    Feb 19, 2015
    40
    6
    A while ago, I thought I understood everything about class X and Y caps. But it seems that this is a very complicated topic.
    I did not notice in any of the above reference specific suitability of a class X or Y caps to be used between isolation stages; where galvanic isolation is required. I presume that using a capacitor between primary and secondary winding is not a correct thing to do.

    Could you kindly enlighten me on this please?
     
  12. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    I've seen it, especially in low power supplies, but I've never known why there is that bias. I think Power Integrations shows it in some of their app notes. In general, it is an EMI noise control technique, and because it is across the isolation barrier I assume it must require Y rated caps. It shunts secondary side high frequency spike noise from flyback transformer secondaries to the power line Neutral in supplies with no Earth Ground connection.

    ak
     
Loading...