/ Home & Energy

Are your LED light bulbs burning out too soon?

LED light bulb

LED light bulbs have a tendency to make grand claims about their lifespan – it’s common to see manufacturers promise bulbs will last 25,000 or even 50,000 hours. But we’ve found many failing well before this.

One of the attractions of LED light bulbs is that they’re supposed to last a long time. And so if you’re shelling out for these bulbs – typically more expensive than other types of light bulb – you’ll want to be sure that they’ll live up to those claims.

But our tests show that not only do many LED light bulbs stop working before the end of their promised lifespan, some don’t even reach the soon-to-be-implemented EU minimum lifespan of 6,000 hours. We discovered bulbs from both Ikea and TCP that failed to reach the 6,000 hour mark for the majority of samples we tested.

Ikea bulb among failures

In the tests – which were carried out by Which? and our European partner organisations – we took five samples each of 46 different bulbs. The bulbs were switched on for two hours and 45 minutes, then switched off for 15 minutes, in a continuous cycle until they burned out.

Five different bulbs stopped working before the 6,000 hour mark for the majority of samples we tested, though the TCP and Ikea bulbs were the only ones which were sold in the UK. Both have since been discontinued.

New EU regulations which will come in from 1 March 2014 say that 90% of any batch of LED light bulbs should last at least 6,000 hours.

Another five bulbs stopped working before the 10,000 hour mark for the majority of samples we tested, despite claiming lifespans of at least 25,000 hours. None of these bulbs were sold in the UK.

In total, 66 of the 230 samples we tested failed before the 10,000 hour mark, though they all claimed they would last at least 15,000 hours.

Has your bulb burned out early?

Ikea said the bulb had passed its own tests and those in a third-party lab. It’s looking into why the bulb failed our test and has removed it from sale in countries where it was still available.

TCP said it was already aware of the problem with this bulb and withdrew it from sale when they discovered the problem. TCP added that it no longer deals with the supplier of that particular bulb and now make their LED bulbs in-house.

We’re in the process of testing the life span of many more LED bulbs, and we’ll update you if we find others that burn out prematurely. But we also want to hear from you – have you bought bulbs that haven’t lasted as long as they should?

Comments

LED,s are CURRENT devices requiring a set current of usually low voltage from a SMPS supplying a regulated current . YES you can get direct connection but it requires complicated series/parallel combinations . There is no starter

Sandra says:
24 January 2018

Phillips 9290002296A maybe a year or so of infrequent use.
GE 3YD8 about the same.
We replaced every bulb in our house about 3-4 years ago and noticed NO change in electricity usage, and quite a few have burned out. I’m going back to incandescent!

Ugh, now I’m reading you can’t even buy incandescent bulbs any more! I guess it will be CFLs then, mercury be damned.

CFLs are fine. They contain very little mercury.

Very little indeed. I suggest going for LED bulbs, which have several advantages and there is now more choice. Perhaps buy one and see if you are happy with it.

I have noticed a growing number of LED light fixtures that obviously have the LEDs separate from the control electronics. Hopefully this will end the problem of the heat from the LEDs affecting the control electronics and very long lifetimes will be achieved. Electronics can fail for other reasons, so it would be best to adopt a standard design for the LED drivers, but I wonder if the industry will do this.

There are standard design,s Wavechang3 its more a case of fitting cheap rubbish components for the biggest profit . I could build an analogue transformer fed unit in less than an hour . All you require is a set current for the LED,s – simples .

LED drivers deliver a pulsed supply rather than a constant current, presumably because this produces less heat for the same amount of light.

Did I not say ANALOGUE Wavechange no pulsed / switching at HF of the power supply and the power supply is remote that cant affect the LED,s which operate more effectively with a continuous constant current . Thats basic LED principles / Pulsed is not natural to LED design its only there for cheapness ans ease of installation . Not only are LED,s constant current devices they are used to supply a fixed current to many high end power amps

I understand, Duncan, but all the LED bulbs I’ve checked emit radio frequencies* and surely that shows that the LEDs are fed from a pulsed supply. With a light fixture where there is not the same shortage of space as in a bulb it would be practical to use an analogue supply, though I suspect it would either produce less light or use more power, or both.

*I have checked using a battery-operated transistor radio on the LW band. It’s crude but effective.

Patrick Taylor says:
4 February 2018

Seems to me that somebody or some organisation should be working on a lampholder tester that will report on some potential circuit problems and any spiking.

I am non-technical but I see from the IoM guru that he traced two failed lamps to a corrosion on a fairly recent MK light switch. I do know that within six years of having my house rewired I had loose cables in two light switches which would produce I imagine the same results as established by the guru.

This is a sample of his work
youtube.com/watch?v=RdqzdMUO6QE

Patrick if you have loose cables in two light switches then,if I was you , I would borrow an ELT -earth loop tester to see if all your sockets have had the earth connection tightened . One of the sockets in a flat I owned , after a rewire had infinity on the ELT as the earth screw hadn’t been tightened up. Light switches, I am afraid arent built to the same robust standards as those way back in the 50,s which had thick heavy duty brass contacts , now they are thin and flexible . I am sure somebody will come on quoting BSI etc Standards but what cuts it with me is actual -REAL – down to earth performance and PRACTICAL engineering technology , manufacturers seem to hide behind modern “standards ” . I have in my hand a 50,s light switch and a modern one big differences and I am not talking about the plastic design.

I have a couple of Martindale EZ150 socket testers that provide a quick way of checking sockets and extension leads for problems of the type that Duncan mentions. In the early 80s I lived in a rented flat and used my home-made socket tester and found no Earth on the socket for the socket for the electric fire in the lounge. It’s important to use a socket tester that tests earth loop impedance and most don’t, which is why I mention one that does. They are so simple to use that a child can use them and I sometimes produce one when visiting friends and let the nearest child have fun checking for problems.

Anyone who is having repeated problems with LED lamps failing could usefully have their wiring checked. Problems may already be apparent if switches are erratic or lamps flicker. LEDs take little current compared with old bulbs and that can sometimes cause a problem. Dimmers should be replaced with ones designed for LED lighting.

I’m concerned about how cheap mains sockets and switches have become in recent years.

I wired my house and after 35 years have had a couple of light switches where a wire came loose. I’d suggest poor installation is far more likely than dodgy switches and sockets, although I did buy MK stuff throughout. If a screw connection is tightened onto a properly stripped cable it should not come loose but, if not done tightly in the first place heating and cooling at the terminal can make a poor connection worse.

I don’t know why this attracts an implied criticism of “BSI etc Standards”; these have over many years been put together by practical engineering. The current standard dates to 1999 and is a European standard – BS EN 60669-1. It contains a large number of requirements and appropriate tests. For example, torque, flexion, pull-out tests on terminals after conditioning for 7 days at 70⁰C, use of non-creep materials and so on.

I suspect the real problem with such items is not the standard, but whether the item, even if it claims compliance (CE mark) actually has been designed. tested and manufactured to comply. It comes down to policing; are we equipped to see that appropriate standards and regulations are being observed?

Malcolm when you put all your eggs into the one basket and put total reliance on standards it can leave you open to any possible failure on their part under actual public use thats why I like US built to last standards that stand up to the abuse of your typical US consumer . It isnt just a case of saying – well the US have different designed washing machines etc , that doesn’t or wont stand up to reality if the actual practical construction of the product isn’t built to a rugged specification. Heavier materials/electric motors etc . We have the “Italian line ” beautiful to look at but given a bit of abuse it falls apart just like their electrical wiring and electric systems did in the past .Even the Japanese learned the hard way that light construction wasn’t long term reliable . We keep on being told – ah but you abused it , as an excuse for flimsy engineering . Look malcolm I rode and owned British motorcycles – big heavy steel construction which held speed back but I saw in the 60,s one of them nearly cut a mini in half in a major accident it proved a point to me. I fully realise you must have standards but when they fail -what then ? Years of debate and more years of talk until forced to take action ?

duncan, standards are, for the products we use, “safety” standards in the main to ensure that consumers are protected. Imagine if we didn’t have them – a free for all for any cheap company to send us unsafe stuff. They often, for electrical goods, emanate from the IEC – International Electrotechnical Commission – upon which I presume USA standards are also based.

I don’t put “total reliance” on standards; I simply point out the practical information and tests included. Durability is something we should assess over and above a safe product and many standards include endurance tests.

Were exactly have standards generally failed us?

What would you have standards based on? Durability? I’d certainly like to see the EU have a requirement for longer life products, but although it is less easy to specify. It can still, I believe, be reasonably estimated by looking at quality of components, build quality, brand history (and who has manufactured it), and accelerated testing, some things I consider consumers’ associations should have done, and should be doing.

How would you put a standard together for light switches?

Light switches malcolm ? I would do a practical, down to earth test on them involving continuous switching over a period till it reaches breakdown . Then if not satisfactory I would engineer heavier more robust contacts and test them till breakdown —and note the difference in length of Duty cycle. I have seen all types of switches in mains wirelesses from cheap contacts to platinum car ignition ( old school ) types . When I have ex government equipment that never breaks down including the mains switches it tells me – yes it can be done but the cost is higher thats why I suspect the lowest pass for components nowadays makes many manufacturers very happy . Public use equipment is definitely not of government military standards

Any light switch bought in the UK and complying with the relevant standard should be entirely fit for purpose regardless of its price and should last a lifetime. Some fittings exceed the standard and are preferable to work with because they have better cable entries, layout, terminals, and switch performance. Any light switch is only as good as the last person who wired it and fitted it. I suspect that most replacement light switches bought on-line or in a DIY store are not installed by an electrician and that the unit is selected more for its appearance rather than its constructional and performance quality. Both characteristics are available in the same product but at a higher price.

I would do a practical, down to earth test on them involving continuous switching
Section 19 of the standard requires the switch (up to 16A capacity) to be subjected to 40 000 operations at nominal voltage and current and then is subjected to electric strength and temperature rise tests. I estimate that could be 20 years normal use.

So no blame to the manufacturers – just blame the customer or the electrician who fitted it ?

Nominal – minimal of what is expected, as I have said before the lowest common denominator not the highest. Now there,s a thing could posters back up that light switches of the most modern sort last 20 years ? . I have 60,s light switches still going strong along with pvc covered electrical wiring using the old -red-black- green coding no problem with any of it not even a blown fuse at the distribution board and I have a maze of wiring in the room I have my test equipment that brought comment from the social work department who insisted on fitting 2 types of fire /smoke sensors in 4 rooms . The only time the fire station was called was when I burned toast in the kitchen and an automatic signal went back to the – council care /alarm people via a unit in the hall connected permanently to the telephone line. I must be doing something right.

Our previous house built in 1984 still had nearly all of its original light switches as installed when built. They were all working perfectly well when we moved in 2012 and I would expect them to be still working satisfactorily today and for the next forty years at least.

Although the fittings installed by trade electricians might not look so smart and stylish as other electrical switches and sockets on the market I would say that they are almost invariably of good quality and made to a high standard [even above the basic technical standard]. They are designed for reliable and safe performance over a long life, and are made with the trade in mind for ease of compliant installation.

As with many products there is a wide range on the market and customers are free to choose the type that suits their needs. The important point is to ensure that any light switch sold into the UK is compliant with the relevant technical standards. I am fairy satisfied that that is the case, and that the electrical engineering trade bodies keep an eye on all products and take action if they become aware of any non-conformities so that they can be removed from the market. Trade electricians tend to use products from a limited number of manufacturers based on their own experience and that of their colleagues of what is the best combination of safety, performance and price.

I cannot see that it is in the interests of any manufacturers to lower their standards because the all-important specifiers [the house builders and commercial developers] would soon drop them in favour of better products. Unlike in some consumer goods, instead of a race to the bottom there is more likely a race to the top because they are selling into a market that is not particularly price sensitive at the first level. Some house builders even go so far as to advertise the make of their electrical fittings to reassure their customers that they are buying a well-made home.

Many light switches only go on and off once or twice a day so a test of 40,000 cycles seems more than adequate to me. I am not sure that the current load is critical in testing a switch since the most essential factor in safe operation is the mechanical function. Nevertheless, I would expect the nominal voltage to be the same as the regular mains supply [230 V] and the current to be no less than the normal lighting circuit capacity [5 A]; I can’t see why anyone would go to the trouble of stepping down the voltage and current and specially configuring the test apparatus for what is not a particularly prolonged test [probably no more than 60 hours or even less]. But then I am no expert in these things and would accept Malcolm’s assessment.

If you check on the tightness of screws fastening electrical cables even after a short time you can quite often tighten them a little more maybe not much but the screws can be loose even if perfectly tight when installed .Nothing today is as good as it was in the past . ok a few things perhaps

I have not had problems with light switches except for a bathroom pull-switch that I had to change twice when I lived in my previous house. I must change the switch in my downstairs toilet because the lamp sometimes flickers.

I have had many problems with dimmers, albeit not in my own home. It’s common for dimmers to fail because one of the components (the triac) is often under-rated and fail when an incandescent bulb fails, due to a momentary high current.

MK brought out dimmers incorporating fast-blow fuses which seemed like a good idea but it was necessary to remove the switch plate to replace fuses – a daft design. I had to replace a couple of these expensive dimmers with cheaper but more reliable ones. Not surprisingly, the design was withdrawn by MK. I complained to B&Q and they denied selling them.

Wonder why that is? I’ve noticed it as well.

If you are referring to loose screws, the usual problem is tightening a screw onto flexible cable that has been soldered (tinned) at the end. That was a common problem when electrical products were supplied without a plug. I have no idea why the manufacturers did this. With light switches and sockets there should not be a problem if they have been properly fitted, though removing them when decorating might result in loose screws.

Wavechange – if it’s convenient and the wiring is in the right place you could fit a normal wall switch for the light in your downstairs toilet. You don’t have to have a pull cord switch in such a space as unlike a bathroom it is only in use for short periods and does not produce a damp atmosphere. The wall switch can be just inside the door although most installations are on the outside alongside the door.

When decorating around a light switch it is only necessary to back out the screws holding the faceplate by a small amount. It should not disturb the terminations of the conductors in the cables.

John – I would not be keen to install an ordinary light switch – which is unsealed – near a sink. I know not to touch switches with a wet hand but would a child?

“The cords of cord-operated switches are allowed in zones 1 and 2 and are recommended for bathrooms and shower rooms to account for the humidity and condensation that could occur.”https://electrical.theiet.org/wiring-matters/53/section-701/index.cfm

When wallpapering the switch plate is likely to be pulled well forward and this can loosen the cable connections, especially now that single strand cable is standard. It would be better if light switches were designed so that the switch plate was removable for decoration, leaving the switch and wiring safely enclosed. I have fitted Screwfix LED dimmers of this design. It would be possible to do the same with power sockets and I believe this is done in the US and Canada, albeit without the live parts properly shielded.

In parts of the States they not only have sockets in their bathrooms but TVs as well. The sockets are frequently next to the washbasins. It makes for interesting conversations.

I agree that wallpapering is tricky around light switches and sockets because it can be difficult to keep the paste away from the electrics but I have managed it on many occasions without having to pull the face plate out too far. A well made light switch will have the termination screws accessible from a side angle so they can be checked for tightness.

If space in a toilet cubicle is at a premium or if the room also contains a shower then I would certainly not put a wall switch on the inside, but there is generally no condensation or humidity problem and the space must be ventilated anyway.

nominal”” means at its full rated current. Other tests are done at 1.1x voltage and 1.25 x current. So at its highest rating, or higher, not lowest. Note that 40 000 operations still requires a fully functional switch; it is not a test to failure.

It is worth finding out what standards require before jumping to conclusions.

I have checked the offending switch and the screws are tight. I wish that pull-switches had standard mountings so that they can be replaced without moving the screws. The room is dry. Having dismantled the switch, the problem is that it operates by bridging two contacts with a strip of copper. The connection was erratic because the contacts are coated with copper oxide, which acts as an insulator (technically it is a semiconductor). Copper is only suitable for wiping contacts, so this switch arrangement is unsatisfactory, even if compliant with the relevant standards. After cleaning the switch has been replaced and is working. I have made a template in the hope I can find a switch with similar mountings.

I wonder if some of the problems with premature failure of LED lamps are due to poor switches.

Wavechange the manufacturers did it because many people just screwed down onto a splayed multi wire spreading it about so that its mechanical strength was reduced as only part of the cable was clamped , it also reduced the amperage it carried as it could blow at the connection rather than the fuse . The problem was the tightening screw wasn’t wide enough and and sometimes slipped to the side of the soldered cable not making a tight connection.

Solder cold-flows – under the pressure of a screw that initially is tight the soldered strands “retreat” from the pressure and a loose connection ban result. Metal ferrules are a good way to treat thin stranded conductors, or use a pressure-plate terminal that spreads the screw load.

We seem to be some way off the topic?

I see the point but suspect they replaced one problem with a bigger one. I’ve seen many examples when screws have come loose thanks to soldered wires. Sometimes it’s the Earth that comes adrift, which is potentially dangerous. Sadly, I have seen people solder the ends of wires before fitting plugs to copy what manufacturers have done. 🙁 I was very glad when moulded plugs arrived.

Malcolm – A wire protection leaf helps but it does not prevent the problem of cold flow of solder. Metal ferrules are excellent and used for many purposes, such as products that still have rewireable plugs fitted for the UK. The physical layout of the connections in mains plugs was never standardised so providing ferrules ready-fitted is of little use. MK introduced a plug designed for use with the three conductors cut to the same length, forgetting that the Earth must be the last to be pulled free. Crazy.

Trying to get back on topic, I wonder if the low current used by LED lamps is causing erratic switch contacts, which may result in failure. Here is a brief explanation of ‘wetting current’: https://en.wikipedia.org/wiki/Wetting_current

Many devices. particularly inductive or capacitive. take an initial inrush current much higher than their operating current – including LED lamps.

Do we know of any evidence that switches suffer from a low current problem?

I don’t know if low current or inrush current might be a problem, Malcolm. I have established and temporarily cured the problem with my poorly designed switch. BS3676 (undated)

BS3676 was first published in 1963. I wonder how old your switch is?

The house was built around 19 years ago. The switch is also labelled EN 60669-1, again undated. At my previous home I did replace a pull-switch twice because of flickering lights, but did not do a post mortem on either failed switch.

The first reference I have to BS EN 60669-1 is 1996 (the EN was 1995), revised 1999 with the latest amendment 2008.

Yes, I’d looked this up, and there is a CE mark too. I cannot be sure if the switch is as old as the house. I don’t understand why one half of the contact is made of a suitable material and the other half is plain copper. I learned long ago that plain brass and copper should be used as wiping contacts to keep them free from oxide.

I did pull apart cord operated light switches to see the failures Wavechange . Early types had spring operated changeover switches – either the spring broke or the cheap contacts broke/corroded . Later micro-switch versions were used looked smarter but again must have been cheap micro-switches when pulled apart either the push button did not function or the contacts had been sparking due probably to too little mechanical pressure therefore building up a high resistance in the contacts . You could tell quality in my day the good ones were big and round containing large heavy duty mechanics and a strong /good spring and high quality contacts . When you pulled it you got a reassuring loud click and that required a bit of pulling. A I keep saying I dont think the quality is there in new products regardless of how many BSI directives I am hit with.

I’ve not dismantled faulty light switches before today but I am familiar with other products becoming less robust. I recently fitted new light switches for a friend and even the more expensive ones were about £1.30 each, no doubt made in China these days.

What is the voltage used in those US states ? 110 v is a lot safer than our 230 v …110 v must be used on building sites because it does not kill when touched inadvertently I would not let any electric of any voltage in a bath with me

A ll switches are of better quality than of years ago when only one or two brands were fitted by electrical contractors who did a good job and wanted them to last other brands were of a poorer quality did the job but could not be guaranteed to last

I pull apart everything and anything that no longer works just to find out what is inside and how they work even things that have no apparent way to open can be usually opened easily some do break when trying but when the no longer work it does not matter that way I find the way into the next one I try I can now mend many things that the makers insist must be returned to them for repair I ask relations and friends to give me their things that no longer work

I like an inquisitive mind Bishbut ,especially when its engineering related.

I bought my LED bulbs at Poundland no failures after 2 or 3 years Cheap enough to replace in any did fail Some think the more expensive the product the better it is I have found that with many products that is wrong Some maybe better but many are not

I would not be keen to buy electrical products from pound shops but I certainly don’t believe that ‘you get what you pay for’ is a useful guide to reliability or durability.

David Hill says:
5 February 2018

I had one LED bulb fail after two months. They should be date stamped so if they fail the manufacturer s are forced to replace within, say three years.

If they are made in China David then it is up to the importers to replace them as China has a government policy of no compensation or re-import of goods sold abroad to the general public . They force the importers to sign contracts to this effect as many importers to the UK buy dirt cheap stuff for bigger profit and that makes China look bad . There are 1000,s of factories in China many of them are small scale, China is now the top industrialized nation and it only took them 25 years . China -built to a Price , but its not their fault they are only supplying to demand .

Glengavel says:
7 February 2018

A Tesco branded GU10 LED gave up the ghost in our kitchen today. It had started flickering a bit before switching on, but this morning it flickered away for a few minutes, then there was a loud POP and the lighting circuit breaker tripped out in the fuse box. Can’t remember when I bought it but it can’t be more than a few years old.

mark wright says:
10 February 2018

Osram 5.7W, bought 5 and 2 have failed after 11 months. Claimed 10000 hours, but less than 3000 hours of use. Will be returning to shop.

Buying at a shop makes it easier. A friend had a similar problem with lamps bought online but will just buy more because of the cost and hassle of sending the dud ones back.

Do LED lamps have a code on them from which the date of manufacture can be determined? With something intended to last ten years it is unrealistic that we will keep the receipts and identify them with a specific lamp that pops after eight years, or even half that time.

In the early days of CFL’s I did take a couple back to John Lewis, together with the original boxes, and they were replaced immediately without question. LED’s are a lot more expensive and now that they have been available for a few years I would imagine that retailers will be wary of just replacing any lamp that is brought back several years after being sold. Some of our LED’s were bought with the groceries and the receipts are no longer available. I don’t intend to set up a database of which lamp is in which fitting in which room.

I think if a lamp fails within a year of purchase it is reasonable to expect it to be replaced but after that I think we should just shrug our shoulders and resolve not to buy the same again – unless there is evidence that performance and longevity have improved for that make.

I have always been concerned that it has been a bit of a gamble buying LED’s; if they perform as promised there will be substantial savings on the electricity bill [and a reduced power requirement], but if they start to fail prematurely that is wiped out by the cost of replacing them. With incandescent bulbs priced at 25p there was no need to worry about their lifetime; I am glad CFL’s have turned out not to be the long term solution and that LED’s are now generally supplanting them, but they are not affordable for many people. If their general reliability is called into question it will be a setback for energy conservation and household economy.

I may be an exception but I have kept all the packaging for LED bulbs with the receipts inside. I equipped a room at a time and would find it easy to identify the relevant receipt. When the lamps are at least two years old I will dispose of the packaging and receipts. So far I have seen no failures, but others have not been so lucky and I do see some failed bulbs when out and about.

Manufacturers don’t often put dates on their products unless these are inside or encoded. It’s human nature to be put off by older stock.

Replacing incandescent bulbs including halogen ones with LEDs and CFLs has always been worthwhile financially when lamps are in frequent use, but I’m not planning to change the old ones in the loft.

My experience is with LED bulbs from Poundland, £1.00 each. They seemed fine except in one lamp where they have consistently failed within a day or two – three in a row. It turned out that this is a lamp which is on my kitchen light circuit and which is on the end of a long spur shared with a fluorescent under-cabinet tube light. The latter are known for generating dirty mains noise. The lamp on the other side, directly connected to the mains source and not on the spur, has the same Poundland bulb in it and so far has no problems at all, the same with all the other lights where I have used Poundland bulbs. So it looks as if dirty mains can be an issue affecting the lifespan of LED bulbs, drastically in this case.

Tim – A dirty mains filter is available in the UK from https://emfields-solutions.com/filters/de-filter.asp for £40 -DE2 filter . Due to high demand they may be out of stock just now , seems you are not alone Tim in complaining about dirty mains . For those of a technical disposition like Wavechange for £135 you can buy their LCD display -“dirty mains indicator ” which will indicate also the level in MV the state of the mains in your house called the deTekta, high levels over 100mV will show % difference it makes to your supply . Now why am I showing this company ? – its British and it doesnt sell Cr@p in other words -it cares . You can buy cheap rubbish , thinking I will save money , I am only vouching for this company. If you are a hi-fi nut like me good mains are essential but living in the country I am not subject to much in the way of dirty mains , only lightning strikes and mains fail due to storms.

Duncan – rather expensive to protect a £1.00 bulb. Perhaps a ferrite core around the cable would do it? Thankfully Poundland graciously kept replacing the bulbs each time they blew then refunded me in full when I gave up after 3 attempts, now back to using tungsten.

I am with you on tungsten Tim , I have used ferrite cores when I worked with BT on the main CCS (call-connect systems )switch-box containing the CPU in factories but the coil winding’s round it are in series with the supply be it low voltage or high , efficiency is improved if a high working voltage capacitor is used to “tune ” the coil to the worst interference . BT at one time had plug in mains interference protectors but they were high quality and even in my day very expensive .

Tim – I hope you won’t give up on LED lamps. I have used a number of Tesco’s own brand lamps for two or three years in table lamps and uplighters that are on time switches and lit for several hours everyday without failure. They will not beat Poundland on price but they seem to be well-made, give good illumination, are less expensive than many other makes, and come with a guarantee. Alternatively I have bought Philips lamps from John Lewis when buying new light fittings and they have been just as reliable.

Tim has identified what is likely to be the cause of his LED lamps failing and I am not convinced that others will have the spike suppression needed to provide protection from high voltage spikes. I would suggest keeping the under-cabinet fluorescent lights and trying LED lighting elsewhere in the house. I’m not sure about buying electrical goods from pound shops.

With all the reported LED lamp failures with lamps bought from stores all over is it the lamps that are faulty or another cause some of which have been suggested and discussed When £1 ones last for some and expensive ones fail for others it must be for other reasons than faulty lamps A Which investigation is needed maybe

Wavechange – I was wondering whether changing the under-cabinet lights to LED tubes might be a remedy for Tim. It depends what is causing the voltage spikes.

Tim refers to a single under-cabinet light and if that’s the case your solution is probably the best option. If there are half a dozen of them then it would be a fairly expensive exercise.

The reason for the voltage spike is back-emf (counter-wmf). Older style fluorescent fixtures contain a choke ballast. When the light switch is turned off a large voltage spike can be produced by the collapsing magnetic field in the choke. If a LED (or CFL) lamp is on the same circuit electronic components can be destroyed instantly.

LED circuitry can be protected from voltage spikes but there may not be room in the lamp cap to do this. I doubt that high quality LEDs would fare better than the cheap ones but I have not done the experiment.

If you dont mind me butting in Wavechange? Some old circuits using LED low voltage had relays in series with them , as you say back EMF blew them due to the inductance of the mechanical relays . This can be cured in the Florescent light choke circuit by fitting REVERSE emf protection diodes this BEMF can also kill digital chips like the old 555 timer , keep in mind also that the spur mains length between the florescent light and the LED lights also has inductance increasing the voltage spike use 1N5000 range diodes from 400 V upwards . I have plenty of old test equipment with reverse protection diodes.

No problem with butting in, especially on matters to do with electronics! Using reverse protection diodes in parallel with relay coils is standard practice to deal with the back emf they produce, but I’m not sure how this could be used in an AC circuit. If put in parallel with the choke, surely this would negate the ability of the choke to restrict the current through the fluorescent lamp.

Wavechange a parallel protection diode allows the current to go on circulating in the coil , it restricts the Reverse Voltage across the coil so that the voltage doesn’t get too big (transient peaks ) its not perfect due to parasitic capacitance which may cause ringing , it gets more complicated after that.

Hmm. I’m not convinced. It’s standard practice to put diodes across DC relay coils but I can’t find examples of what you are suggesting and I have never seen any diodes across the chokes used in older fluorescent lighting. Some fluorescents have power factor correction capacitors across the L & N connections and I expect that this would help minimise the size of spikes. When I was a young lad and did not know about power factor, I assumed that the capacitor was for spike suppression.

Okay Wavechange how about gas tube transient suppressors fitted to old BT master sockets used as line suppressors for transients voltages or MOV,s -Metal Oxide Varistors acting like back to back zenners they were fitted to those expensive BT mains interference suppressors I posted about .

John: you can get strips of LEDs which simply stick to the underside of the cabinets and work off little power-plugs. We have them and they’re excellent.

Yes, Ian, I have seen such installations, but Tim has an existing fluorescent tube fitting and it would be easy to just replace the fluorescent tube with a new LED one. As Wavechange says, it would be expensive to replace a number of fluorescent tubes with LED tubes, and fluorescents already have very long lives, but even then it could be done progressively to see if it is the answer. The part that seems to be at risk is the LED lamp in a separate light fitting or downlighter [this is not clear from the information available] whereas the cause of the present problem seems to be the fluorescent tube kicking up a voltage spike. In Tim’s shoes, I would seek to eliminate the fluorescent tube from the equation as a first step.

There is no doubt the high price of LED lamps – even if offset against their longer life – makes trial and error an expensive exercise and the retailer would not be obliged to replace the blown lamp or refund the purchase price. In the days of incandescent lamps I would carry a stock of spares of all the lamp types in use in the house but I only have two spares now – one in each cap type – and will buy new as and when required, which hasn’t been necessary yet. If a spotlight or downlighter lamp fails it will have to wait until I can replace it, but there are others in the same space providing adequate illumination temporarily.

Duncan – I agree that a varistor would be useful for spike suppression but that’s different from simple diodes. I gave never seen one in a fluorescent lamp fixture or seen them advertised for use by the public.

John makes an important point about retailers not being obliged to replace or offer a refund for failed lamps. Neither the customer or retailer is likely to know that it’s not a good idea to have fluorescent strip lighting (rather than CFLs) and LED lamps on a circuit controlled by the same fuse. I have never seen a warning on the packaging of LED lamps.

In the last Which? review of LED lamps we were warned not to use them alongside incandescent bulbs in the same fixture because the heat produced by the bulbs could cause premature failure of the LEDs. That’s sensible advice but, again, I’ve not seen any warning on the packaging.

We are being let down by manufacturers. Tim could not be expected to know that the way he used his LED lamp could lead to premature failure. Poundland was generous in offering free replacements. In most cases the customer would have put it down to experience, especially if the lamps had been bought online.

Its not just “dirty mains” that occurs with some fluorescent,s they can generate large amounts of RFI requiring suppression.

I’m with Ian. I have 4 LED strips under kitchen cabinets that are quite unobtrusive, light the worktops uniformly and liven up the kitchen. Never had a problem, including any radio interference even with the radio very close. It is the only place in the house I’ve used LEDs, and I did so simply because of ease of installation, their very shallow depth, and light distribution. Hang the expense.

We do not know why Tim’s LED lamp failed, only speculation, so a bit difficult to blame the manufacturer. Would I use Poundland? Presumably the lamp was cheap, and therefore made very cheaply. A risk – if it works then fine, but if not maybe try another source?

Is there any evidence that more expensive LED lamps are protected against voltage spikes whereas the ones sold by Poundland are not?

The fact that fluorescent light fixtures can produce voltage spikes is well established, so maybe we have more than speculation.

Not really Wavechange but didnt I bring up the fact ( after pulling one apart ) that they depend on reliability/quality of the solid state components inside them ? You also commented on it . I was joined by several electronic engineers/electricians who confirmed it . Importers have a lot to answer for, they only pay when its the biggest profit therefore China will “Build to a Price ” . I have proved personally that electronic design engineering is only as good as the parts fitted and if they are working at near design limits or have inferior heat-sinking then reliability goes out the window . I have looked at graphs shown massive escalation of radiated and induced interference when near their limits . Do you remember the scandal of unmarked chips etc being fitted to public electronics ? I have seen many but the US took action on it. Big US electronic companies are obliged to imprint their chips with a lot of detail and name themselves as manufacturers. I have bought electronic parts direct from the USA and can vouch for their top quality they had to be declared by the Homeland Security as allowed technology exiting the USA so were pretty dear but worth it as I couldn’t purchase them in Britain.

I know what’s going on Duncan, and what concerns me is that you cannot be sure that you are getting better components and build quality if you are prepared to pay more. Even the best components can fail if you push too much power through them.

Getting back to the topic of LEDs, there is so little room in the cap of an LED lamp that I would be surprised if they contained varistors or other spike suppression components. I did not see any in the CFLs I dismantled to look at circuit design and reasons for failure.

I’m not disputing what may be the cause, but we don’t know, do we, whether there is enough energy in the spike that switching off an MCF circuit can produce to damage an LED power supply some distance away. It may be better quality LED lamps have circuits more resistant to voltage spikes. Just speculation. Perhaps someone might ask a manufacturer to comment?

In the graphs I have seen hundreds of amps are produced in voltage spikes and very high voltages 800V AC for example for very short periods .

Spikes in the mains can be a serious problem, whether caused by inductive loads such as motors and chokes or by lightning. I presume that the surge protection devices sold for computer equipment use varistors but the ones I have looked at could not be dismantled without breaking them.

John Ward: just to be definitive about the situation, it is one of an L-shaped kitchen worktop with a sink just off of the corner and SES clip-on spotlights on cabinets either side of the sink to light the area. To the left of the sink (and around the corner of the ‘L’) is a hob with an extractor fan covered by a fused wall switch – the lighting has been piggy-backed off of this circuit through a junction box so that this switch controls the lights. Here there are three under-cabinet fluorescent lights (two short under-cabinet ones – 6W, one longer over the hob – 500mm T4 16W), plus one of the LED spotlights for the sink, this being connected directly to the junction box. Power for the other cabinet is run down the left of the left-side cabinet, under the floor-standers and up the right side of the right cabinet, keeping it hidden as much as possible but giving it a length of about 8m. This feeds a long flourescent 500mm T4 16W tube and the other LED spotlight. Hence all the lighting is off the same switch.

The first two 6W LED bulbs worked fine but then the right hand one went after two days. This was replaced, running with the old tungsten but still keeping the LED bulb in the left spotlight until I was able to go back to Poundland and get another a few days later, but then the same happened again with the replacement. This was again replaced but this time the bulb which had been working in the left spotlight was swapped to the right side to see if I just happened to have one good bulb from Poundland when generally they failed quickly, not being able to see how the identical cliplights could have bulbs fail repeatedly in one but not the other, the replacement going in the left side this time. The “good” bulb from the left side now failed in the right side within a day, now back to tungsten, the one on the left side (i.e. the second replacement) is still going strong after quite a few days.

As far as I can see the only possible cause is that long 8m spur, it is clear that flourescent tubes can cause dirty mains but with the ones on the left side coming straight off the ring main perhaps that “sinks” any mains noise / spikes while the 8m spur is unable to. I don’t really want a 8m spur feed for each of the lights on the right hand side but it looks like it may be the cheapest option and would certainly be the proof of the pudding assuming there is no pickup between the two feeds.

Hi Tim – Thanks for coming back with more information. You have created plenty of discussion.

Would it be practical to run the LEDs and fluorescent strips from separate switches? That would protect the LEDs from voltage spikes.

I don’t know why you are having a problem with one circuit but not the other, but it might be down to one of the fluorescent fixtures.

Tim says:
1 March 2018

Well I added a whole new spur for the LED lights only, the fluorescent tube having one of its own, put in a new Poundland bulb, and hey presto that blew again within a day. The two spurs follow the same path so perhaps there is some crosstalk between them but surely it should be much less than the amount of noise when being directly connected to the same spur.

Flummoxing. The other eleven Poundland bulbs in the house are fine.

Tim -the spurs wont make any difference the radiated interference will stay the same as will the induced interference what you require is what I said a suppression circuit inline (series wiring ) with the LED light be it built in or a plug type with the spur plugged into the suppression plug socket.

Thanks for coming back, Tim. Knowing that other LED lamps are fine, this more or less confirms that spikes from the fluorescent strip lights are the problem. I don’t understand why a spike should affect the LED if it was on its own switched circuit. That flummoxes me too.

As Duncan says you could try protecting the LEDs. I have no experience but my approach would be to go for an inexpensive suppressor across the input connections of each the fluorescent fixtures, for example: http://uk.farnell.com/c/circuit-protection/tvs-transient-voltage-suppressors/tvs-varistors

LED strip lighting is coming down in price, but I’m not impressed by the flimsy stick-on strips of LEDs, even thought they seem quite popular. I would prefer something more substantial.

I wasn’t confident in the self-adhesive pads supplied to fix LED strips and wiring to the underside of my cupboards; the surface was melamine (Contiboard). That was 5 years ago and its all still there.

Tim says:
1 March 2018

Duncan – I’m not convinced by what you say, to me it seems that the evidence provided by the working spotlight is that it is the ability of the local mains supply to “sink” the interference which is important, not forgetting that that spotlight has *three* fluorescents right next to it yet it keeps working, hence the idea of the additional spur for the other lamp which keeps blowing bulbs.

You dont convince me either Tim,sorry to say , I have had plenty of experience in noise/interference suppression either radiated or induced into the ring mains . What you are saying is that going by your “evidence ” of the ring main “sinking ” the interference that the problem is elsewhere . Well thats a first , what would need to be the case would be that the ring main had exactly the right frequency in combined -inductance-capacitance and resistance to cause a high impedance to either the fundamental frequency OR the harmonics thats a tall order a self achieving passive filter , can you prove that ? As you still have the problem and your spur is a continuation of the ring main which ,you say, “sinks ” the interference then a short bit of spur hasnt made any difference .

I used to repair TVs like and still have some metal oxide varistors that might be effective for spike suppression. They just look like ceramic disc capacitors. I have fluorescent under-cabinet lighting in my kitchen and if I had not disposed of my oscilloscope I would have checked the effectiveness of these components at dealing with spikes.

Tim says:
1 March 2018

Duncan – I don’t wish to cause any indignation as part of the exploration of this bizarre problem. Thank you for your comments. I think the next step has to be to risk putting the LED bulb from the good spotlight in the bad one with the local fluorescent light’s tube removed, so the only light on the righthand side being the LED one and hopefully no mains noise, and to see if it suddenly keeps working. If it doesn’t then I’m heading south.

I think you are right Tim. Often the best way to solve a problem is by a process of elimination. Once the source is found you can begin to address the remedy. Maybe you could replace your under-cupboard fluorescents with LEDs, or use electronically ballasted MCFs instead. I can’t recall if you have only used Poundland LEDs, but another elimination option is to try another make, say from B&Q, or even a make such as Osram or Philips, and see if that helps. We’ll all be interested to hear the outcome.:-)

Wavechange for a kosher , real British company look no further than Stewarts of Reading a fully licensed government contractor of ex WD electronic equipment . I have bought much electronic test equipment from him he is no cut price cheapskate selling dodgy gear this is the real article. All high quality tested before dispatch , if its only say up to 50Mhz bandwidth they are under £100 , he sells many types of testers some may interest you. No callers though.

Apology accepted Tim its just something I have spent many years on , if I had lived near you I would have visited you to help you as I am that kind of a guy , I have much test equipment ·

Duncan – Thanks but I’m not wanting any more test equipment because I don’t use it much nowadays. Thanks to the complexity and non-repairability of modern electronics gear I’m back to doing work that is electrical rather than electronic. Like you, I am happy to help out and I’m being asked for help with purchasing LEDs, installing replacement dimmers, etc.

I have discovered that swapping LEDs with Halogens (same shape and size) often fails. Specifically in kitchen cooker hoods.

Were the original bulbs LED Ian ?

There could be a problem with 12V lamps (e.g. MR16) or if they were controlled by a dimmer. The latter is not going to apply for a cooker hood.

The originals were halogen 20w and I switched them for LED 3w. Once in, they started flashing on and off and I discovered it’s a well documented issue and is something to do with current demand. Plenty about it on Google.

I have replaced under-cabinet fluorescent tubes with LED tubes and been very satisfied with the results. They are also available in different colour renders unlike the stick-on LED strips.

I waited for someone to reply to you Ian .The circuit being 12V DC – 20W would = 1.7 A (approx ) . To obtain 3W in LED circuits means that multiple LEDs are used , each LED requires a steady state current at whatever the total working current of EACH of them might be this could reach 2A hence the flickering . If LEDS were just one big one it would be easier to calculate. The transformer OR SMPS would be overloaded .

I’m assuming that the original lamps are G4 halogen 12V/20W or similar and the SMPS is not sufficiently loaded to work properly. In that position would either replace the driver with one designed for LED lamps or rewire the cooker hood to take 230V LEDs. The alternative is to buy some halogen lamps while they are still available.

I replaced the miserable little incandescent lamp in my Bosch cooker hood with a 470 lumen LED, which is far better. I don’t know why it’s marked 2x60W when there is only one lampholder. I’ve recently put a couple of candle lamps in a Rangemaster range cooker. Fortunately they were mains voltage.

Some 12V drivers do work OK with LEDs. I was dismayed to find that I had MR16 halogen lamps in the bathrooms, each on a separate driver. I replaced one with an LED and that was fine, so I have replaced the lot. My surveyor had warned me of the risk of overheating with a thick layer of insulation on top of the halogen lamps.

One problem with replacing halogen capsules with LEDs is that the LEDs might not be as bright. That’s certainly the case when replacing G9 (mains voltage) capsules with LEDs.

John – Did you replace the tubes or the whole fixtures? I know it can be done by replacing the fluorescent lamp ballast with an LED driver but I wonder if there are LED tubes that can be used without any modification. If so, it would be an easy option for Tim.

I overlooked the fact that we replaced the whole fixtures because we wanted some additional ones. I bought the complete fittings including tubes from Screwfix and our electrician installed them including additional fused isolation switches where they spur off the power circuit. The fittings themselves were not particularly expensive.

Thanks John. I did consider replacing my under-cabinet lighting with LEDs but the fluorescents are working fine and improving the ceiling lighting made a great improvement. My under-cabinet lighting is fed from the lighting circuit but it’s common to use the power circuit and as you say, it’s vital that the circuit is protected by a fuse. I suggest you check that you have a 3 amp fuse in the isolation switch to give greater protection against fire in the unlikely event of a problem.

The original lights were supplied from the lighting circuit but it was simpler to spur the additional ones from the nearby power socket and I did make sure that they were protected by a 3 amp fuse [the electrician had already fitted the correct fuse].

This has the advantage that if the other lights go out because a circuit breaker trips you will not be left in darkness. The lamp in my cooker hood would stay on if there was a problem with the downstairs lighting.

This LED light bulb thing is getting too technical most including me do not understand much of what is being posted now nobody know of a simple explination or is it all to complicated to understand

In a world where Which? doesn’t seem to want us to fit our own 3-pin plugs, all the above talk of fitting additional components chokes and parallel diodes seems rather academic (said the “armchair” electrical engineer).

I think John Ward proposed the most practicable solution when he asked “I was wondering whether changing the under-cabinet lights to LED tubes might be a remedy for Tim”. However it is not even clear to me whether or not the lamps in question are supplied by ac or dc.

Derek-If the lamp is supplied by a DC circuit there are plenty of suppression components fitted in the circuit to stop reverse injection of interference into a mains spur and many of those circuits dont use a choke . To elicit knowledgeable replies on any subject detailed comments+investigations take place on Which in many categories why is Engineering being singled out when other subjects go deep into law and many other issues and answers are given by the regulars ? Why single out Engineering ? as for “the Technology ” forum its been a site for social conversation for quite a while now and subjects of no relevance to technology so why keep on calling it “Technology ” why not give it a heading of “General Topics “

Without looking at the technical issues we will not understand why some people can suffer failure of their LED lamps. The manufacturers could help by stating that their LED lamps should not be used on the same circuit as fluorescent strip lighting.

I realise that last year, I decided not to install additional LED lighting in my garage/workshop because it would have been on the same circuit as three fluorescent strip lights.

Your first sentence gives the “lie ” to all those that criticise you Wavechange+myself for going deeper into faults in Engineering . We have University graduates here , how do you think you gain knowledge on a subject unless you study it more than just a surface glance , never heard of a thesis ? This modern society is developing into MT clones who say-Engineering /Industry – DONT understand it , whats going on here- dumbed down Britain where the height of “intelligence” is wondering which type of pizza to buy ? Yet other convo subjects go deep into investigation .

Convos often receive requests, in effect, for possible solutions to problems. Those solutions may not be straightforward and some people who have the knowledge attempt to explain the problem in detail. Just because the majority do not have the background to understand what they are saying does not mean it is not worth saying. A minority, even small, may pick up on the thread and help it develop into something that does produce an outcome. It is what Convos can do.

We had a discussion fairly recently and I don’t think anyone suggested avoiding technical discussion.

We seem to have become a nation where many people have an encyclopaedic knowledge of films, music, books, etc. but precious little understanding of more practical matters or wish to remedy this. No-one needs to be a university graduate to learn about technical matters.

If you watch a quiz programme the questions invariably seem to include pop music, football, films and other leisure-based questions. It seems a shame that we don’t encourage people to use their brains to store knowledge about stuff that matters in real life. What is the purpose in knowing who had a no.1 in 1962, or who had a bit part in Love Actually? I know, I’m an old grump. Just because I don’t know this stuff.

Another “old grump ” joins you malcolm , what I want to know is the state of this country after we are all dead and gone ? I get it —a University degree course in spotting the best advertisement on TV instead of a diploma you get to eat at MacDonald,s for a year with an advertising crew following you around to compile a US style series .

The reason that quiz programmes focus on leisure-based questions is that, from the audience’s point of view, they are likely to produce the most entertainment. I don’t watch any myself since I am still waiting for “Light Bulb Moment” with a starter for ten. I am hoping Alfa might have a preview somewhere in her library.

I found “Brain of Britain” and “Round Britain Quiz” entertaining. Not intelligent enough for the latter, but then I cannot fathom the Private Eye crossword either.

“Only Connect” and ‘Mastermind” are probably the most adequate quizzes on TV at the moment. OC calls for more than simply general knowledge; it requires intelligence.

This would be an interesting subject for The Lobby. Maybe we could discuss the technical reasons for premature failure of LED lamps here. 🙂

Alexander Armstrong and Richard Osman are probably highly intelligent. How can they bring themselves to preside over Pointless? I seem to keep catching the tail end of each episode while waiting for the six o’clock news; the audience even applauds the amount of the next day’s prize. Not just pointless, but mindless as well.

Your introduction of “ an encyclopaedic knowledge” lead on to this. The way many off-topic discussions are started. Perhaps, as I suggested, someone might ask those in the business about the effect of voltage spikes? Expert guidance might be useful.

What is (a little) interesting is the low value of the elusive jackpot – starts at £1000 and rarely gets to £5k. Once upon a time quiz shows offered cars, exotic holidays,”Who wants to be a millionaire”, ……. Cheap entertainment now – but rather pointless, I agree.

Pointless is so repetitive and predictable to the point of boring. I wonder how much they pay the so-called celebs to appear when they win a measly £500 each for their charities?

We much prefer ‘The Chase’. The lighting is much better too.

I turned this up earlier and although it does mention spikes, it does not refer to inductive loads such as traditional fluorescent lighting.

Are you physic malcolm – a meeting of minds over a long distance at this very moment I have that actual PDF on another browser .Do you notice the proof that an LED can be destroyed by a voltage spike I hope that puts to bed any thoughts of the LED,s from Poundland on a spur not being affected by the florescent light and for the record the majority of faults on a florescent light are caused by the choke . Notice the black rings on the end of florescent lights ? caused by RINGING in the choke circuit . Ringing is like a transient an irregular waveform that is roughly AC its action greatly reduces the life of the florescent light .

Quora.com- after a lot of info– the inductance of a choke causes a Voltage Spike when the starter switches off , this is due to the sudden change in current flow and helps achieve the spark that ignites the tube which has a low impedance at start up but high when running at
20Khz .

Blackened ends on fluorescent tube are indicative of the electrode coating that aids starting being sputtered off, and will make starting more difficult. It is approaching end of life.

I have not disputed that a current spike might be damaging the LED, but querying whether the LED circuitry is adequate to withstand it. I don’t know. I am endeavouring to find out more about LED lamp protection in their circuitry.

Its ringing that causes the sputtering Wavechange . I will check up on LED protection too this is very interesting !!

There is a fair amount of information about commercial lighting, Malcolm. Here, lightning can be more of a concern than in the home. In the cap of a commercial LED bulb there is little room for commonly used spike suppression devices – generally referred to as MOV and TLS – though these become practical in LED lighting where the LEDs are separate from the control electronics.

Traditional fluorescent lighting could be with us for a few years yet because it is efficient and durable. When I moved home I fitted new fluorescent lighting in the garage because it was a fraction of the price of the LED equivalents available at the time. I was disappointed to find nothing across the power input of the fixtures to remove spikes. I might fit my own if I want to use LED lighting above my work bench.

Duncan – The normal reason for darkening of fluorescent tube ends towards the end of the life is loss of the coating on the tube heaters that promotes emission of electrons, as in a thermionic valve. It is necessary to maintain a discharge – not just to start it. Under some circumstances, fluorescent lamps can darken for other reasons, which is presumably what you are referring to. I know only about normal operation.

http://www.ee.co.za/article/analysing-causes-blackening-ends-fluorescent-lamps.html. See emission mix and sputtering. Emitter material on the electrodes is lost particularly during starting..

as far as I know a lightning strike on a public supply can cause problems in your home.

It can, and can cause several items of electronic equipment fail simultaneously. I don’t believe that it is a common problem.

BT were so worried about it they installed gas tubes in the master sockets to stop lightning strikes melting their phones and wiring.

This Wikipedia page covers these gas discharge tubes and other ways of dealing with high voltage spikes: https://en.wikipedia.org/wiki/Surge_protector#Gas_discharge_tube_(GDT)

Here is a site covering the history of our phone system and the use of gas discharge tubes in the master socket: http://www.britishtelephones.com/lightng.htm

My advice to anyone installing LED lighting would be:

1. Choose dimmable LEDs if you intend to use a dimmer or might want to in future. Replace dimmers with ones intended for use with LED lighting.

2. It’s best to avoid overheating of LED lamps and that is more important with brighter lamps that consume most power. Avoid using them in enclosed fixtures alongside halogen lamps, which produce a great deal of heat.

3. To reduce the risk of damage by voltage spikes, LEDs and fluorescent strip lighting should not be controlled by the same switch.

Teresa says:
3 April 2018

Diall LED bulbs from B&Q. 8 bulbs fitted in two separate kitchen lights new in August last year. Dark kitchen so lights on for approx 16 hours a day. 3 of the 8 have failed in past two weeks = 3000+ hours only. Very disappointed.