Isolation Transformer - A discussion

[JohnW2]

With a floating supply, the supply does not need to be disconnected to avoid a dangerous situation. That makes this point void.

I'm not sure that 'void' is necessarily the right word. I haven't suggested that ADS would be necessary/desirable with a floating supply, merely that 'earthing' (which, I continue to believe, exists primarily to facilitate ADS when the supply is earth referenced) would not facilitate/achieve this with a floating supply.

As I've said before, the point of the bonding is to ensure that a dangerous touch voltage cannot occur. .... Again, I suggest you are letting the tail wag the dog. What you have described is a situation where the bonding is inadequate. If (say) the CPC was 10mm² then the potential of the exposed cp would be about 20% of the supply voltage - i.e. less than 50V.

As I've just written, a 16mm² CPC in a "2.5mm²" cable would provide 'safe bonding', and a 10mm² one might just about (an exposed-c-p potential of about 49V with my usual supply voltage) - but I'm not sure what relevance that has to 'the real world'

EDIT: In effect, the regs considerably muddy the waters by closely tying two different functions together - effectively using a combination to allow one of the requirements to be considerably relaxed ..... it would certainly be possible to design a circuit such that the ADS took longer than the normal times provided that we increased the "strength" of the bonding to limit the touch voltage to a safe level ...

If you're talking about local 'Supplementary Bonding' with relatively short lengths of substantial cable bonding together all simultaneously-touchable exposed- and extraneous-c-ps, then, yes, that would presumably suffice from the point of view of protection from electric shock, but the ADS (disconnection time) requirements obviously do exist, and presumably relate to reduction of the risk of fire as well as electric shock?

Kind Regards, John

 

[SimonH2]

... but seem to rely on the one 'level of insulation' qualifying as "reinforced" (something which is hard to believe/understand in some cases!).
... However, as above the insulation is very commonly not "double" but, rather, single and allegedly "reinforced".

I think things would get very complicated to describe if we were to start considering the "often questionable" design/manufacturing practices of some devices

... but the ADS (disconnection time) requirements obviously do exist, and presumably relate to reduction of the risk of fire as well as electric shock?

I doubt that.

From the PoV of fire, the key requirement is that things don't get hot enough to start a fire - so essentially it's akin to the adiabatic equation when determining acceptable cable/protection combinations. So I suspect that in many cases, the disconnection time could be considerably longer - possibly measured in seconds, not milliseconds.

When considering shock, time is of the essence. Hence, I believe, the 0.2s disconnect time required for many circuits.

 

[JohnW2]

I think things would get very complicated to describe if we were to start considering the "often questionable" design/manufacturing practices of some devices

Indeed. As I've said, many/most of the Class II items I have dissected have had a single layer of insulation (hence, presumably, allegedly "reinforced") preventing one touch live parts. However, in at least some cases I personally find it hard to describe the insulation as 'reinforced" since I imagine that I could probably push my finger through it if I pressed hard enough!

I doubt that. From the PoV of fire, the key requirement is that things don't get hot enough to start a fire - so essentially it's akin to the adiabatic equation when determining acceptable cable/protection combinations. So I suspect that in many cases, the disconnection time could be considerably longer - possibly measured in seconds, not milliseconds.

There's bit of a problem there, since adiabatic calculations are only valid when adiabatic conditions exist. That is usually taken to mean that the duration of current flow is no greater than 5 seconds but, rather surprisingly, 434.5.2 talks (in relation to an adiabatic calculation) about fault currents with a duration of "less than 0.1 secs".

I find the regs rather confusing in relation to these matters.

It is true that explicit requirements in terms of disconnection time only seem to exist in Chapter 41 ("Protection against Electric Shock"). However, 411.3.2.3 & 411.3.2.4 of that chapter does impose disconnection time requirements in relation to distribution circuits (5 sec for TN and 1 sec for TT - goodness knows why that difference!), although I would have thought that such circuits, per se, represented hardly any risk of electric shock.

Chapter 43 ("Protection of Overcurrent"), which relates to "dangers due to thermal effects" (presumably including fire) does not seem to specify required disconnection times but, as you say, talks about adiabatic calculations, but only for fault currents of less than 0.1 sec. That creates a bit of a vicious circle, since it is implying that the adiabatic calculation is only applicable/valid if the disconnection times as required by Chapter 41 (for "protection against electric shock") are already satisfied!

Anyway, returning to your point/suggestion, since disconnection time requirements do exist (even for distribution circuits), we are stuck with them, even though, as you have said, one could minimise the risk of serious electric shocks by using adequate bonding, regardless of ADS.

Kind Regards, John