Possibel earthing issue?

[aptsys]

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.

The filter capacitor in a Class II appliance is between the incoming and outgoing sides of the transformer. When you have your high switching frequency with fast edges, the interwinding capacitance becomes an easy path for the high frequency noise. By conducting this back to the return side through a Y2 capacitor you close the loop and reduce the EMC effects.

It's used on every Class II SMPSU and stops your DC lead becoming a large antenna for noise. Y2 capacitors are fine for this purpose because as they fail the capacitance decreases, ultimately ending as open circuit thus fail safe. They should be meeting EN60384-14

 

[SimonH2]

The filter capacitor in a Class II appliance is between the incoming and outgoing sides of the transformer.

Between which point specifically of the input ? Neutral connection ? Live connection ? Both ? Something else ? Specifically, many countries have a reversible plug (Germany(Shuko), France, Italy, USA to list just those I have first hand experience with) and so there is no way to know which side of the input will be "hot" as our friends in the USA call it.
I've just looked at a couple of our TVs and a DVD player, which all have 2 core mains leads. While they have fixed leads with brown and blue cores (I assume, moulded on plugs) I can't believe they don't use the same model with just minor changes (different mains lead, software, manual) across the whole of Europe or even the world. All have no earth connection (2 core mains lead), all have the square in square DI logo, but ...
All have I/O connections (antenna, analogue (eg VGA, 3.5mm jack), HDMI, USB etc) - and on one TV all these are on a large exposed metal plate, while the DVD player is mostly a metal case with no plastic covering. With the exception of the antenna input, I cannot see how the I/O could be isolated from the DC supplies internally.

So logically, it's hard to see how they wouldn't need two filtering capacitors - between PSU output and both sides of mains input (reversible supply in more countries than not). And it's hard to see how everything exposed could possibly be isolated from the DC busses internally (note that both HDMI and USB (and satellite on some kit) provide power to attached devices). So we really are down to the "chassis" being connected to a (weak) potential divider formed by two caps across the mains input.

In our house, the TVs (and everything else) will be incidentally earthed via connections to other stuff that is earthed. I do know that as I'm sat here with my laptop, I'm not currently feeling any of the "buzz" from light contact with the laptop. But if I grab hold of the end of the HDMI lead from the TV while touching the laptop, there's "rather more than just a buzz" At one point I got out the multimeter, but couldn't get any significant readings for voltage or current between the two.

I can only assume that such an arrangement is deemed to meet regs - hard to imagine every major manufacturer making equipment with this arrangement if it didn't. But I'm with John on this - I'm "a bit uneasy" about the whole concept.
I have noted over the years that some equipment does have a spare screw with an earth symbol

So I suppose to understand why we get this sort of arrangement, we need to consider what would drive a manufacturer to use it. In the UK, and I guess most of Europe and many other countries, we generally take having an earth for granted. From various stuff I've read and seen over the years, I think we need to assume that having a reliable earth is far from universal. So does a manufacturer make something that needs an earth, and make it the users' fault if they kill/injure themselves by using a non-earthed supply arrangement ? It might be technically valid, but I could imagine there being a certain amount of moral and reputational incentive to not do that. So once you are down to not assuming a reliable earth, then makes sense to design for DI/Class II for those markets - and then you might as well do it worldwide and save the expense of two different designs of power supply.
That's how I read things anyway.

 

[JohnW2]

...but ... All have I/O connections (antenna, analogue (eg VGA, 3.5mm jack), HDMI, USB etc) - and on one TV all these are on a large exposed metal plate, while the DVD player is mostly a metal case with no plastic covering. With the exception of the antenna input, I cannot see how the I/O could be isolated from the DC supplies internally.

I suppose that, in theory, almost anything can be isolated (e.g. by optical isolation) (as it often is with medical equipment) - but I very much doubt that such is done in 'consumer' equipoment.

I can only assume that such an arrangement is deemed to meet regs - hard to imagine every major manufacturer making equipment with this arrangement if it didn't. But I'm with John on this - I'm "a bit uneasy" about the whole concept.

Agreed - on both points.

.... So once you are down to not assuming a reliable earth, then makes sense to design for DI/Class II for those markets - and then you might as well do it worldwide and save the expense of two different designs of power supply.
That's how I read things anyway.

Yes, that makes sense - but, as we/you have been discussing, even 'Class II' does not remove the issue of I/O connectors etc. which are connected to one side of the ELV supply. ... and don't forget that SELV, if that's what one has, requires that the ELV supply is not connected to earth.

Kind Regards, John

 

[aptsys]

Between which point specifically of the input ? Neutral connection ? Live connection ? Both ? Something else ? Specifically, many countries have a reversible plug (Germany(Shuko), France, Italy, USA to list just those I have first hand experience with) and so there is no way to know which side of the input will be "hot" as our friends in the USA call it.
I've just looked at a couple of our TVs and a DVD player, which all have 2 core mains leads. While they have fixed leads with brown and blue cores (I assume, moulded on plugs) I can't believe they don't use the same model with just minor changes (different mains lead, software, manual) across the whole of Europe or even the world. All have no earth connection (2 core mains lead), all have the square in square DI logo, but ...
All have I/O connections (antenna, analogue (eg VGA, 3.5mm jack), HDMI, USB etc) - and on one TV all these are on a large exposed metal plate, while the DVD player is mostly a metal case with no plastic covering. With the exception of the antenna input, I cannot see how the I/O could be isolated from the DC supplies internally.

So logically, it's hard to see how they wouldn't need two filtering capacitors - between PSU output and both sides of mains input (reversible supply in more countries than not). And it's hard to see how everything exposed could possibly be isolated from the DC busses internally (note that both HDMI and USB (and satellite on some kit) provide power to attached devices). So we really are down to the "chassis" being connected to a (weak) potential divider formed by two caps across the mains input.

In our house, the TVs (and everything else) will be incidentally earthed via connections to other stuff that is earthed. I do know that as I'm sat here with my laptop, I'm not currently feeling any of the "buzz" from light contact with the laptop. But if I grab hold of the end of the HDMI lead from the TV while touching the laptop, there's "rather more than just a buzz" At one point I got out the multimeter, but couldn't get any significant readings for voltage or current between the two.

I can only assume that such an arrangement is deemed to meet regs - hard to imagine every major manufacturer making equipment with this arrangement if it didn't. But I'm with John on this - I'm "a bit uneasy" about the whole concept.
I have noted over the years that some equipment does have a spare screw with an earth symbol

So I suppose to understand why we get this sort of arrangement, we need to consider what would drive a manufacturer to use it. In the UK, and I guess most of Europe and many other countries, we generally take having an earth for granted. From various stuff I've read and seen over the years, I think we need to assume that having a reliable earth is far from universal. So does a manufacturer make something that needs an earth, and make it the users' fault if they kill/injure themselves by using a non-earthed supply arrangement ? It might be technically valid, but I could imagine there being a certain amount of moral and reputational incentive to not do that. So once you are down to not assuming a reliable earth, then makes sense to design for DI/Class II for those markets - and then you might as well do it worldwide and save the expense of two different designs of power supply.
That's how I read things anyway.

It's connected to the non-switched side of the transformer in the case of a basic flyback topology - so after the bridge rectifier. Orientation of the AC input doesn't matter as the impedance of the capacitors at mains frequency is high enough to limit current to less than 100uA. On a synchronous driver, it's connected to the most suitable return path from an EMC point of view, typically the centre of the switching elements.

 

[bernardgreen]

and don't forget that SELV, if that's what one has, requires that the ELV supply is not connected to earth.

An ELV supply to a door bell push button on a rain soaked front door ? Puts a lot of trust in the insulation between the mains and the ELV supply. If the ELV is from a split bobbin transformer then the risk of insulation breakdown is very low as primary and secondary windings are in separate compartments on the core.

ELV supplied from a single bobbin transformer is dependant on effective insulation between the windings. Most single bobbin transformers do have a non conductive insulating layer between primary and secondary windings. Some transformers include an earthed conductive screen between two layers of insulation between the windings ( best option ). A few transformers have no insulation layer and depend entirely on the enamel coating on the winding wire to keep mains from reaching the ELV secondary windings. ( high failure risk )

Grounding one side of an ELV supply will in almost all cases prevent or reduce the hazard of electric shock from the ELV equipment. If the transformer insulation breaks down then earth currents will trip an RCD on the mains supply to the failed transformer

With an SELV supply failed insulation in the transformer could make the entire SELV system Live and thus a significant hazard of electric shock. With no current path to ground the RCD supplying the failed transformer cannot be tripped by the faulty transformer.

Why is an SELV supply even necessary ? In my experience the only justification for SELV is when the SELV will be supplying equipment that has it's own earthing of the ELV. Earthing that might conflict with the earth of an ELV supply earthed at the source of the ELV.

 

[JohnW2]

Grounding one side of an ELV supply will in almost all cases prevent or reduce the hazard of electric shock from the ELV equipment. If the transformer insulation breaks down then earth currents will trip an RCD on the mains supply to the failed transformer

Indeed.

With an SELV supply failed insulation in the transformer could make the entire SELV system Live and thus a significant hazard of electric shock. With no current path to ground the RCD supplying the failed transformer cannot be tripped by the faulty transformer.

Indeed #2.

Why is an SELV supply even necessary ?

You tell me - but there is an awful lot of it about. Of course, in practice, it very often arises in the context of a Class II item which has no access to an earth connection.

In my experience the only justification for SELV is when the SELV will be supplying equipment that has it's own earthing of the ELV. Earthing that might conflict with the earth of an ELV supply earthed at the source of the ELV.

I don't think what you are describing would qualify as "SELV". The definition of SELV (at least, the BS7671 definition) requires that the ELV system is "electrically separated from Earth..." - and that presumably means any earth.

Kind Regards, John

 

[plugwash]

Between which point specifically of the input ? Neutral connection ? Live connection ? Both ? Something else ?

My understanding is it's usually one side of the rectifier output.

At one point I got out the multimeter, but couldn't get any significant readings for voltage or current between the two

I just measured a random "phone charger" type power supply and got about 67uA between output and mains earth and 4uA between the output and my body.

You do need a pretty good multimeter to measure this though, many cheaper ones either don't have AC current ranges at all or don't have enough precision/accuracy to meaningfully measure it. I've also found it surprisingly difficult to get a good contact between the earth terminal of a 13A socket and a multimeter probe.

I also recently stuck a neon screwdriver in a live hole and measured from the contact on the back to ground. Got about 150uA there.

Edit: I screwed up my measurements by using the wrong socket on my meter.