Maximum voltage drop on SWA runs to outbuildings, am I over engineering the solution

There seems to be another misunderstanding here. I wonder how many have noticed that White LEDs are slow to turn off the light when the power is removed? This is due not to the power supply, but to the fluorescent material that actually gives the light you see, the LED being UV. Then there is the power supply in the lamp. Well designed ones use a switch-mode circuit that gives a constant current whatever the input voltage, so no flicker even if the mains dips a lot. Cheap lamps give cheap results, and dimmable ones are of course not mains voltage resistant! Realistically flicker from voltage drop is not a problem with LEDs, They are better than all other lamp types, at least those without electronic power supplies. My battery work lights continue to give light (dimming of course) for 30 seconds or so, 10W LED. The charger might give a single reduction in light if there is a volt drop, but does this matter, it is not really a flicker and not really noticeable. Power savings from oversized cables  make a lot of assumptions, and copper gets more expensive every day. Don't go over the top!

BS7671 is quite "out of date" as you say, but one should always make your own assessment before believing everything you read. It has many gaps once electronic loads are encountered, the biggest one being that these loads often take "constant power" so when the mains voltage falls the current increases, the opposite of ohms law!

It`s also surprising how many folks think that electric showers do similar

I have to admit my assumption was that an electric shower had a fairly simple resistive heating element and thermostat arrangement with current increasing with voltage, do they actually operate at a lower voltage with switched mode power supply?

Do not get worried, showers really are just mains rated resistors in water.Some of the clever ones have something akin to a triac based lamp dimmer to give variable heat output, but the simpler ones are just mains  and then the water flow is modulated to set the temperature.

(and once you leave the land of the euronorm, the resistors are not always in an earthed metal casing either.)

I think Ebee meant that some folk imagine that showers are constant power devices in some way - and as there are a mix of 240 /230 and 220V elements from different makers at different times, at the level of  specmanship a '10kW' shower may be nothing of the kind on your local mains..

For what its worth our quite new not quite 10kW shower flickers the our  LED lights in the living room but not the dining room, despite the two being lamps of the same part number from  well known manufacturer. It's quite noticeable - the light goes off totally for 100-150ms and then comes back on.  (I think that clause about designs being subject to change without notice is very true in the world of domestic electronics.)

Mike

Some info on rise and fall times of phosphor based LEDs is here 

They are certainly slower than the other kind of  'white' LEDs that are 3 colour LED chips in a common package, but not that much slower. A few hundred nanoseconds. Most of the afterglow of LED lamps relate  to DC smoothing caps.

A graph from that paper, note the graph horizontal unit is 1*10^7 i.e.100nsec /square

Mike

Some current regs. and considerations might still be based upon the old filament lamp, where a drop is Voltage caused a definite noticeable loss of light output..

http://www.lamptech.co.uk/Documents/IN%20Voltage.htm

Z.

Very interesting graph. So for a filament light a 10% drop in supply voltage gives a 25% drop in output. I would assume that for an LED light with switched mode power supply there would be no impact, maybe even some improvement because the output from the power supply would be closer to a standard square wave. Even with a basic circuit I would expect to see very limited difference with an LED.

Just found this article 

https://www.electronicspoint.com/forums/threads/running-led-bulb-on-lower-voltage.275364/ 

Based on this with a very simple driver a 10% difference in voltage has a very limited impact on output but as we approach 20% to 25% drop the impact on output is considerable.

Therefore there is a risk of problems and the question will be what voltage range has the device been optimised for. If it's centered on 230V there is probably quite a lot of wiggle room but if it's 240V there is much less. If I was designing a low cost LED light I would certainly be thinking about reliability as well as performance over the required voltage range and may design for a higher voltage to give better reliability, sacrificing light output at lower voltages, which are unusual in the UK at least.

Indeed the lamps that take  'universal' mains - really meaning terrestrial" , but lets not quibble, from say 85 to 250V have inside an SMPS inside that is regulating for constant LED current, not LED voltage, and these are pretty bombproof.

But the simpler ones (230v only)  are closer to the internals of these ones, which are a Phillips special for Dubai, but not that different from their  euroland  offering, except being a bit better engineered.

I have the Osram variants of this LED filament design, and one of these is  flickery, exacerbated I think as the LEDs have a very steep forward voltage current curve, so a small change in voltage is a large change in current, and as electrons are  almost proportional to photons, a large change in light output too.

However the fact that 2 'identical' lamps are different suggests that the design is not exactly fixed yet.

(and I have only cut one open, the other may not be the same inside.)

Mike