Possibel earthing issue?

[JohnW2]

Surely that is an oxymoron. Even if it were alowed, it could hardly be described as Class II. I.e with no insulation.

I'm not so sure about that. As I understand it, the requirement for Class II require that there is adequate (double or 'reinforced') insulation between 'live parts' and the outside world - so I'm not sure that would theoretically preclude electrical connection of some outer conductive part (outside of the insulation) being electrically connected to something which was not 'live' during normal operation.

However, you're pursuing something which I only wrote 'as an aside'. My primary point was that, in relation to the suggestion that filter capacitoirs might be connected from L and N to the unearthed casing of a Class II item, that is something which I have (a) never seen, (b) never heard of, (c) not been able to think of a reason for and (d) am not sure that it would even be allowed. As the 'aside', I merely added that, as a generalisation, one does not expect (maybe 'not allowed'?) anything (whether a filter capacitor or anything else) to be (electrically) connected to the unearthed casing of a Class II item.

Kind Regards, John

 

[plugwash]

When such equipment has a "3-wire mains lead" it is, indeed, common to have filter capacitors between L & E and between N & E, resulting in the situation you describe. However, I've never noticed filter capacitors being connected to the unearthed case of a Class II item and, at first thought, I can think of no reason why that would be done. I'm not even convinced that it would be 'allowed'.

Lots of AV and IT equipment has exposed metalwork, sometimes including a metal case, connected to the ELV "ground".

Basically every switched mode power supply has filter capacitors between input and output, they would never meet EMC regs without them. On a class 1 power supply the mains ground can be used as a barrier between input and output but on a class 2 power supply there will inevitablly be leakage between input and output caused by the filter capacitors.

 

[JohnW2]

Lots of AV and IT equipment has exposed metalwork, sometimes including a metal case, connected to the ELV "ground".

Yes, but that "ELV ground" is not earthed.

Basically every switched mode power supply has filter capacitors between input and output, they would never meet EMC regs without them. On a class 1 power supply the mains ground can be used as a barrier between input and output but on a class 2 power supply there will inevitablly be leakage between input and output caused by the filter capacitors.

That all makes sense - but if the exposed metalwork is connected to "ELV ground" (a.k.a. one side of the output of the SMPSU), then one has unearthed exposed metalwork connected to the LV supply through just a capacitor (which could fail S/C), and I'm rather surprised that such is 'allowed'.

Kind Regards, John

 

[plugwash]

If the PSU builder is doing their job properly they shouldn't just be using any old capacitor, they should be using special safety capacitors (class Y1 IIRC) that are extremely unlikely to fail short.

 

[JohnW2]

If the PSU builder is doing their job properly they shouldn't just be using any old capacitor, they should be using special safety capacitors (class Y1 IIRC) that are extremely unlikely to fail short.

That sounds reassuring. Are you suggesting that such a capacitor would qualify as having "double or reinforced insulation" - since that would seem to be what it would be required to have if it were connected between a live LV conductor and something touchable in the outside world?

Kind Regards, John

 

[Harry Bloomfield]

Thanks for that quick response.
Yes, I'll have to look for a double square symbol; I hadn't noticed one before, though. I must also check, as I said, for an earth wire.

Is that actually true?

When such equipment has a "3-wire mains lead" it is, indeed, common to have filter capacitors between L & E and between N & E, resulting in the situation you describe. However, I've never noticed filter capacitors being connected to the unearthed case of a Class II item and, at first thought, I can think of no reason why that would be done. I'm not even convinced that it would be 'allowed'.

Kind Regards, John

It would be pointless to have the filter capacitors, without an earth - idea is to discharge any interfering components to earth. No Earth connection, no filtering.

 

[JohnW2]

It would be pointless to have the filter capacitors, without an earth - idea is to discharge any interfering components to earth. No Earth connection, no filtering.

I'm not sure about 'pointless' - a capacitors between the incoming L & N which provided a relatively low impedance path for high frequency 'interfering components' would keep much of those components out of the LV wiring, wouldn't it?

Kind Regards, John

 

[Harry Bloomfield]

I'm not sure about 'pointless' - a capacitors between the incoming L & N which provided a relatively low impedance path for high frequency 'interfering components' would keep much of those components out of the LV wiring, wouldn't it?

Kind Regards, John

Yes, but filtering by fitting them between L & E + N & E would not have the desired effect, or much effect, unless the E was actually connected to earth. Like you, I have never come across such a situation where the case has not been earthed, but yet with filters connected to E. Often a metal case provides a screening effect too, to restrict interference escaping or entering.

 

[JohnW2]

Yes, but filtering by fitting them between L & E + N & E would not have the desired effect, or much effect, unless the E was actually connected to earth.

The 'E' you're talking about is not earth - it's simply one side of the ELV side of the PSU (which, it is postulated, is connected to exposed metal parts). Given no true earth, the (L-'E' & N-'E') filtering capacitors are, as far as filtering is concerned, simply two capacitors in series between L and N - which, as I said, will reduce the leakage of high frequency components into the supply cable.

Like you, I have never come across such a situation where the case has not been earthed, but yet with filters connected to E.

Indeed. Despite what Simon suggested, I still do doubt that thus is ever done (even if it is 'allowed').

Often a metal case provides a screening effect too, to restrict interference escaping or entering.

True, but that really requires the metal case to be (really) earthed, and we are talking about a '2-wire supply'. An unearthed metal case might do some good, but it might also act as an antenna for radiation of HF interference!

Kind Regards, John

 

[bernardgreen]

Given no true earth, the (L-'E' & N-'E') filtering capacitors are, as far as filtering is concerned, simply two capacitors in series between L and N - which, as I said, will reduce the leakage of high frequency components into the supply cable.

and they also act as a voltage divider whose mid point is 115 volts AC ( assuming equal values of capacitance ) hence the 'E' will be at 115 volts AC above Neutral. Only because the capacitances are very small does the mid-point have a high impedance which limits the current that can be drawn from it ( through a person ).

 

[JohnW2]

and they also act as a voltage divider whose mid point is 115 volts AC ( assuming equal values of capacitance ) hence the 'E' will be at 115 volts AC above Neutral. Only because the capacitances are very small does the mid-point have a high impedance which limits the current that can be drawn from it ( through a person ).

Indeed - and that's another reason why I seriously wonder whether what Simon suggested is even 'allowed' - in addition to my primary thought that, no matter how 'special' they are, I don''t think that the capacitors could ever be guaranteed to never fail s/c.

Kind Regards, John

 

[SimonH2]

Well the same question arises with filter components between input and electronics (eg, electronics 0V rsil).
The latter is exported via the connections, and they have to assume the device could be used where mains polarity is indeterminate (eg where mains plugs ate reversible).

 

[JohnW2]

Well the same question arises with filter components between input and electronics (eg, electronics 0V rsil). The latter is exported via the connections, and they have to assume the device could be used where mains polarity is indeterminate (eg where mains plugs ate reversible).

That point has been raised, and discussed, above.

As I said, unless those capacitors qualify as providing the equivalent of "double or reinforced insulation" (and, regardless of how 'special' they are, are 'guaranteed' never to fail s/c), I doubt that it would be 'allowed' to connect any touchable metal to either side of the ELV (or anything else within the electronics).

Kind Regards, John

 

[Harry Bloomfield]

That point has been raised, and discussed, above.

As I said, unless those capacitors qualify as providing the equivalent of "double or reinforced insulation" (and, regardless of how 'special' they are, are 'guaranteed' never to fail s/c), I doubt that it would be 'allowed' to connect any touchable metal to either side of the ELV (or anything else within the electronics).

Kind Regards, John

From memory the X rated ones are designed so when they develop a fault, the current then blows the defect clear and they self seal. Given enough such incidents, the capacity value declines. There is still that inherent risk though, of them not blowing clear.

 

[JohnW2]

From memory the X rated ones are designed so when they develop a fault, the current then blows the defect clear and they self seal. Given enough such incidents, the capacity value declines. There is still that inherent risk though, of them not blowing clear.

I don't know about the details, but I'm aware that some our designed not to fail s/c - but, as you say, there clearly cannot be any guarantees.

Kind Regards, John