SPDs !!!

[JohnW2]

Apologies for the crude drawing but hopefully it makes sense. House 1 is fed by L1 and N and has a heavy load connected. As the neutral is open circuit it will rise in potential to L1 voltage. ... House 2 has low loading and is connected to L2 and the shared neutral which is now at L1 potential which will have a voltage of up to 400V depending on loading ... House 3 has zero load as they’re on holiday. ....

Fair enough. However, with the neutral that far upstream (before it splits for the feeds to houses with the three different phases - hence affecting all properties supplied by the transformer), there will be many houses connected to each phase (and the 'open' neutral) which I would have thought, in practice, would roughly average out as ...

If each of the houses were using exactly the same power then the neutral drops to 0V and the more out of balance the phases become the higher the neutral voltage.

- i.e., not so much a case of "each of the houses ...." but, rather, "if the total load of all the houses on each phase were roughly the same for each phase, then the neutral drops to roughly 0V".

In other words ... as we know from diversity (as well as many other things in life/world!), in practice 'average' works quite well. Hence, with an upstream neutral fault such as you postulate (leaving all properties connected to the transformer affected), it's pretty unlikely that the the imbalance of loads in properties supplied by the three phases would be large enough for the neutral voltage to rise to substantially above zero.

Kind Regards, John

 

[JohnW2]

There are none. It's C3. The section about loss of life / injury and the rest is for commercial or industrial installations, where C2 could apply if SPDs were not installed or were not functioning.

That's exactly my point (although I'm not really sure about a risk of death/injury even in commercial/industrial installations).

As I said, hitherto BS7671 (or, at least, the spirit thereof) has been almost entirely about safety - indeed, almost entirely about avoidance of electric shocks to persons and electrically-initiated building fires. A regulation with which non-conformity can never be greater than C3 seems to have moved to goalposts to things that are not related to 'safety' in any normal sense.

Reduces the distance between the conductors and the SPD, relying on the N-E link at the origin would significantly increase the length of one of the conductors if only the N or E was connected to the SPD and reduce it's effectiveness significantly.

I thought that eric's point was that, on the basis of his belief that "the unit has two modules, one line to earth, one neutral to earth", the 'neutral to earth' one would achieve nothing if the "neutral" and "earth" connected to the SPD (say, in a CU) were two substantial conductors which were joined together (usually) a metre or two away. Can you clarify?

Single phase supplies to dwellings are typically from a 3 phase transformer with a 3 phase cable in the street, one phase and NE tapped off of that cable for each house. .... A break in the NE part of that cable before it connects to a selection of dwellings is disconnecting the neutral in a 3 phase system that has unbalanced single phase loads (each house), and the neutral point will shift depending on the loads connected. ...

See my response to RF's similar (nearly identical) comment. As I've said, I would have thought that, in practice, if the break in the PEN were so far upstream as to affect all installations connected to the transformer, it would then be unlikely that there would be sufficient imbalance to result in substantial neutral voltages. Is that not the case?

The primary use case for SPDs is to protect equipment within the installation from damage. LED lighting being one of the main items that's relatively easy to destroy with transients, along with other installed devices such as heating and lighting controls.

Indeed - and, as I've been saying, that seems to be a completely new twist to the raison d'être of BS7671 - which I am far from convinced is what 'people' necessarily want.

Most of what's in BS7671 regarding the use of SPDs is next to useless, including the risk assessment for lightning strikes and the nonsense about excluding domestic dwellings that have nothing of value in them.

I agree - although, as you presumably realise, I would probably go a fair bit further than you.

Kind Regards, John

 

[bernardgreen]

The primary use case for SPDs is to protect equipment within the installation from damage. LED lighting being one of the main items that's relatively easy to destroy with transients, along with other installed devices such as heating and lighting controls.

Manufacturers of cheap and/or fake equipment that will be destroyed by transients are passing the buck to others to provide the protection necessary to prevent the equipment from being damaged/destroyed.

Equipment that is properly designed will have protection from transients built into the equipment.

 

[JohnW2]

Manufacturers of cheap and/or fake equipment that will be destroyed by transients are passing the buck to others to provide the protection necessary to prevent the equipment from being damaged/destroyed. .... Equipment that is properly designed will have protection from transients built into the equipment.

That's all true, but I have no problem with consumers wanting to be able to choose very cheap products and to accept this means that they might, probably once in a blue moon, be more likely to fail (in the absence of customer-provided protection) than would be the case with much more expensive products.

Flameport mentioned LED lamps/bulbs, which is a good example. Cheap, and many not-so-cheap, ones rely on just a capacitor to limit the current through the LED element, and it therefore follows that a high frequency (rapid rise time) transient can result in a high (potentially 'lethal' to the LED) transient current to flow through the LED element, even if the voltage of that transient is low (see **).

To have a more sophisticated means of current limitation would considerably increase cost, and it is (should be) for the consumer to decide whether that cost is worthwhile in order to reduce the (I think extremely small) risk of a 'transient-induced failure'. There must be something approaching a total of around 50 LED lamps/bulbs in the 'commonly used' parts of my house, and I nearly always buy 'the cheapest I can find'. In the days of incandescent bulbs, I was probably, on average, having to replace at least one every week but, with LEDs I probably only have to replace one every 'few months'. Even if the ones I replace had all been killed by transients (which is probably extremely unlikely), I personally would not see that as a reason for installing something to 'protect' them, and certainly not as a reason for buying, say, 50 much more expensive lamps/bulbs.

As I see it, issues of 'safety' (as normally perceived) and other things are getting confused. The PRS legislation specifically asks for "ensure that the electrical safety standards are met during any period when the residential premises(2) are occupied under a specified tenancy", and goes on to define 'electrical safety standards' as "electrical safety standards means the standards for electrical installations in the eighteenth edition of the Wiring Regulations, published by the Institution of Engineering and Technology and the British Standards Institution as BS 7671: 2018(3)". Apart from that legislation, I imagine that nearly everyone commissioning an EICR (which, again, assesses things in relation to BS7671 regulations) does so in the belief that it is identifying issues in relation to the safety of the installation. However, BS7671 has now been extended to include regulations that go beyond what would usually be regarded as issues of 'safety' - so I think that a lot of people are probably being misled.

I have no problem with people (whether landlords or not) being told that some of the equipment connected to their installation may be at very slight increased risk of damage due to voltage transients because they don't have SPD protection, but I personally don't think that should be recorded on a report relating to a 'safety inspection' - and nor do I think that it is helpful to report it as a 'non-conformity' with regulations which most people believe relate to safety. For example, I personally do not think that people contemplating a CU change should be told that one without an SPD would be 'non-conformant' (let alone 'non-compliant') with regulations - at least, unless the actual (non-safety-related) reason for that statement is carefully explained.

[ ** I haven't looked into how SPDs work, but do I take it that they provide protection again 'low voltage' (less than normal peak voltage) rapid-rise-time transients (such as, per above, could kill something like an LED), I wonder (i.e. do they 'filter')?? For example, I imagine that this could easily kill a cheap LED lamp/bulb which relied on a capacitor for current limitation ...

Kind Regards, John

 

[ericmark]

I did post @flameport video showing him testing the SPD, he used a wind up mega to show how at a set voltage they start to conduct, so it is clear from his demo they only start to work at around 500 volt, so would not stop the spike you show.

I did open one LED lamp, wish I had taken pictures, but found why it did not work, dry joint so repaired and went back into service, however it starts with a capacitor to limit current, in series, then a full wave rectifier, then a smoothing capacitor in parallel, so there is no strobe effect and a small leak resistor so stray currents will not cause it to flash when switched off.

Clearly with that current limiter frequency is rather important, but can't see spikes like you show affecting them. I have had one other LED lamp fail, since it replaced a florescent tube and the old ballast could be removed or left in place I would assume a pulse width modulated regulator as clearly it needed a large voltage variation depending is ballast removed or not, and that unit worked one day, then not the next, and I assume the regulator failed, it had another florescent on the same switch, so does seem likely the back EMF on turning off was cause of failure. However the replacement unit then ran until my son who now lives in the house replaced it with GU10 down lights.

 

[JohnW2]

I did post @flameport video showing him testing the SPD, he used a wind up mega to show how at a set voltage they start to conduct, so it is clear from his demo they only start to work at around 500 volt, so would not stop the spike you show.

As may be apparent, that was my implied 'fear'.

I did open one LED lamp, wish I had taken pictures, but found why it did not work, dry joint so repaired and went back into service, however it starts with a capacitor to limit current, in series, then a full wave rectifier, then a smoothing capacitor in parallel ..... Clearly with that current limiter frequency is rather important, but can't see spikes like you show affecting them.

It will obviously depend upon the arrangement in the lamp in question, but the only thing you mention which would reduce (perhaps even eliminate) the risk of a spike such as I showed killing the LED element would be the 'smoothing capacitor'. Do you happen to remember (if you ever knew) what were the capacitances of the series and parallel capacitors in the one you dissected?

However, I don't see how one can 'have it both ways' with this argument. With a lamp such as you describe, I would have thought that by far the most crucial determinant of whether the LED element would survive a 'spike' is the ratio of the capacitances of the two capacitors you mention, with the voltage of the 'spike' being very much a secondary consideration. Hence, if the values of the capacitors were such that the LED would survive a very rapid rate-of-rise 'spike' of about 180V (such as I illustrated), then that ought to go a long way to enabling the LED to survive a spike with the same rate-of-rise which was of much greater voltage, wouldn't it?

Kind \Regards, John

 

[ericmark]

The series capacitor was a metal type and the parallel one an electrolytic type, but did not note values. Was rather chuffed when I found the dry joint so soldered it and returned to service. It was this G9 type which does not allow the use of the glass cover as it is so big, the electrolytic capacitor was nearly as big as the whole lamp it replaced no wonder the old lamp flickered and new one steady as a rock.

 

[JohnW2]

The series capacitor was a metal type and the parallel one an electrolytic type, but did not note values.

That's obviously what one would expect. If the electrolytic was adequate, it should protect the LED element against very brief rapidly-rising 'spikes', even if 'high voltage' (>500V), without the need for an SPD

I think that it is going to take a lot to convince me that I personally (and probably most other people) 'need' SPD 'protection' - whether to protect our cheap LED lamps/bulbs or anything else

Kind Regards, John

 

[bernardgreen]

If the electrolytic was adequate, it should protect the LED element against very brief rapidly-rising 'spikes'

To remove rapidly-rising 'spikes' a small capacitor should be fitted in parallel with the electrolytic capacitor.

Many electrolytic capacitors are constructed from two metal foil electrodes separated by a very thin moist paper or fabric layer and these are tightly wound into a spiral. This winding is inductive and presents an impedance to high frequency. The insulation between the electrodes is created by a non conducting layer of metal oxide ( or other salt ) on the electrodes

 

[JohnW2]

To remove rapidly-rising 'spikes' a small capacitor should be fitted in parallel with the electrolytic capacitor.

True (I should have referred to the 'parallel'/'smoothing' capacitor, not the 'electrolytic' one), but don't forget that it cannot be all that 'small' - the 'parallel capacitance' has to be quite 'large' (effective capacitance wise) in relation to the series capacitor, since they are effectively acting as a 'potential divider' across the spike voltage.

Kind Regards, John

 

[RF Lighting]

I think that it is going to take a lot to convince me that I personally (and probably most other people) 'need' SPD 'protection' - whether to protect our cheap LED lamps/bulbs or anything else

I don’t doubt that for a moment! But why would you actively avoid having one? SPDs can be picked up for as little as £20 and the very worst possible outcome is that you’ve spent £20 that you could have wasted elsewhere, but maybe just maybe one day 25 years down the line it operates and saves one piece of equipment.

I’ve had one in my CU for probably 20 years now and as far as I know it’s never been needed, but I’d rather have it and not need it that need it and not have it.

 

[flameport]

I wonder (i.e. do they 'filter')??

They do not.
They only work for spikes which have a voltage above the normal peak AC, what's shown in the graph would not be affected by an SPD at all.

SPDs are simple devices - they have a very high resistance at normal operating voltages, and a very low resistance above that.

 

[JohnW2]

They do not. They only work for spikes which have a voltage above the normal peak AC, what's shown in the graph would not be affected by an SPD at all. SPDs are simple devices - they have a very high resistance at normal operating voltages, and a very low resistance above that.

As you will presumably realise by 'reading between my lines', that was my suspicion.

So that presumably makes them even more 'useless', since they won't protect my cheap LEDs (or anything else) against rapidly rising transients unless they are 'high' voltages?

Kind Regards, John

 

[JohnW2]

I don’t doubt that for a moment! But why would you actively avoid having one?

It's not really a matter of my 'actively avoiding them'. As you say, they're not expensive, and I can't see that they can do any harm (unless, I suppose, one were to 'burst into flames'), so if I found one in my house, I would not demand that it was removed! However, I'm sure that if I really wanted to, I could fill my house with any number of things (not necessarily electrical) to 'protect' me from all sorts of very rare 'risks' - but I'm only inclined to do so if I perceive a significant 'need'.

In any event, it's not really me that I'm concerned about (I can 'look after myself' ). I'm starting to see people being told (sometimes in a sort-of disparaging fashion) that a particular CU is 'non-compliant' because it does not have an SPD - and that seems to be very likely to mislead them into thinking that it is a 'safety' issue and/or than not having an SPD could cause their installation to, say, 'fail an EICR'.

If someone fully understands what an SPD is, and does (and doesn't do), and then still feels that they would like one, then fair enough - but I don't think that's what we're going to be seeing. I strongly suspect that within a few years we'll SPDs in a high proportion of installations, but that many of the owners of those installations would have 'declined' them had they known the full facts.

Kind Regards, John

 

[bernardgreen]

Interesting reading. Does your house have a lightning protection system. ? That will affect the type of SPD(s) that should be fitted.

https://www.electrical-installation.org/enwiki/Elements_of_the_protection_system

It is an odd term "lightning protection system"

surely it should be "protection from lightning system"