LED Light Bulb

Yes, the power supply will accept 240V.

You do know that LED bulbs are DC powered, not AC powered, so you'll need a transformer, unless the bulb has one built-in.

Actually, a full wave bridge and some current limiting device is all that is required. I can't imagine what you mean by needing a transformer.

Most people don't know that a fluorescent lamp is a DC device. It just happens to be bi-directional. I like to call it a bi-polar DC device. The only reason AC is used on a fluorescent lamp is so that all the mercury (which throws out the photons) isn't driven to one end.

Yes, that's true. And you're right that most people don't know that.

Well put. Many here do not understand any of what you posted.....but its all true....

On DC RN ships in my time, an electrician had to go round the ship each day and operate all the lighting polarity reversing switches to keep the Fluorescent tubes from burning at one end....of course on AC RN ships, there was no need!!!

I must confess that that is news to me! What voltage is used to operate a fluorescent lamp on DC? What type of current limitation do they use? I've been under the impression that a significantly higher voltage is required to start a fluorescent lamp than to run it. If so, how do they do that with DC?

Then how do they handle arcing at turn-off?

Its actually much the same as an AC Fluorescent light, you have a ballast to create the high voltage to strike the tube and a starter that "breaks" the circuit to create the high voltage "strike".

The circuit is non polarized, so will work no matter how the connection is made to the 220 vdc.....in fact, I cannot remember a difference to the 220 VAC circuits, other than the polarity changer that has nothing to do with the way the circuit works.

The reason that the polarity must be changed each day is to even out the wear on the end filaments, otherwise one will burn out before the other, rendering the tube useless. In a hot cathode tube, the filaments are used as heaters to make the strike easier and quicker.

I found a good primer for you or anyone else who is really interested in learning how such circuits work here:-

Fluorescent lamp 101

From this I took this which should convince you:-

Fluorescent lamps can run directly from a DC supply of sufficient voltage to strike an arc. The ballast must be resistive, and would consume about as much power as the lamp. When operated from DC, the starting switch is often arranged to reverse the polarity of the supply to the lamp each time it is started; otherwise, the mercury accumulates at one end of the tube. Fluorescent lamps are (almost) never operated directly from DC for those reasons.

Though I am pretty sure the RN used inductive and not resistive ballasts and a starter, in a similar way cars used to produce high voltage sparks for the plugs using contacts and an inductance.

I hope this post was useful and helpful.

Thanks. That Wiki article is quite extensive!

It looks like much of that information will become of mostly historical interest; I'm well on my way to eliminating both fluorescent and incandescent lamps in my home. as existing ones fail, they are being replaced with LED lamps.

Good job.

I have a few places where I have not found affordable LED versions, so I am stuck a bit longer with CFLs in some situations. I am not willing to invest $25 in a single LED version.....

The only place where we have normal tubes is in the cellar, seldom used.

Here the utility companies find it cheaper to subsidize higher efficiency devices than to build new power plants, so the actual prices paid may be as little as 1/3 of the MSRP; occasionally even less. Unfortunately, the biggest subsidies seem to be on the CFLs (which I am phasing out), rather than on the LEDs.

The ones I have not yet found in LEDs are the common Edison base 'flame' lamps for our chandeliers. Those I have found are too wide just above the connector to fit our sockets.

Also, I doubt if an LED unit could replace the Incandescent inside an oven, without re-designing the oven to bring the light in through some form of window.

One interesting point: we have one hallway light that of course is controlled by two switches, and it is wired such that one side or the other of the lamp is always connected to a hot wire. When the lamp is switched to the Off position, the wire going from the other side of the lamp to the disconnected switch has enough capacitive coupling to the neutral to make the LED lamp just barely glow. In the daytime it appears off, but at night it makes a nearly perfect night light (I prefer very dim night lights)! I haven't taken the time to measure how much power it uses in the Off position, but I bet it is less than a Watt. I also have not taken the time to check if the wiring could be changed to eliminate the off-glow if not desired, but I think it could (switching neutral rather than hot).

Running LED for lighting can save up to 78% in cost. If the run to the light is longer than 30m use CAT 5 cable, (LAN computer cable). If 1,5mm2 is used make sure you have the pair twisted and this should eliminate any after glow. Make sure you have a seperate LED lighting circuit and no other light sources attached to the circuit, (fluorescets, energy saving bulbs, out door incandescents etc). Keep all LED circuits separated.

It may be possible to have a back feed from the negative to earth and this could possibly cause the glow.

Try a twisted pair of wires first.

Please keep me posted of any changes that I may log them to find the solution. Busy testing this now.

Thanks

IQ

OK, you made me take the time to figure it out. Yes, the wiring can be changed to avoid the nightlight effect, if desired. In fact, we like the nightlight in the hallway, so I won't change it, but here it is:

Hot wires are shown in red, although the colors don't show well here. Neutral wires are shown in black, and floating wires are shown in (beige?). The regions of capacitive coupling are highlighted in yellow. I only showed one of the four possible ON diagrams. If the lamp is connected to either unswitched wire, the wire on the other side of the lamp will be floating when in the off position, and therefore subject to capacitive coupling. If the lamp is connected to either switched wire, both sides of the lamp will be connected to the same side of the line when in the off position, and there are no floating wires.

I have plenty of Cat5 cable on hand, and I'm aware of the different twist rates, but I'm not clear on which wires one would use and why. The hallway is only 4 meters long, so uses something like 7 or 8 meters of cable.

Why do you say to keep LED circuits separated? I'm not aware of any problems with any mixture of Incandescent, CFL, and LED lamps. I think I've used all possible pairings, and quite likely used all three on the same circuit.

Glad to have your answer, appreciated. Ok here goes, an LED can only be based on a forward manner. It conducts anode to cathode only, It cannot go the opposite way. This we know as reversed bias. If you switch the live feed in and an earth is connected to any fitting on the circuit, a backfeed seems to be produced which them allows a flow in the neutral via the earthing. So a small current flow will keep the LED glowing after power is disconnected. If the fitting are plastic, they are not earthed, if they are metal clad, a current is present and it can only be introduced into the circuitry via the earth. I am trying to prove this as induction and capacitance does not seem to play any part in the afterglow. As stated, I have induced 863A at 1.3 Tesla, a massive magnetic field and zip, nada, nothing happened. I switch of the power and the LED goes out. I also have a strong feeling that if the house has fuses rather than a earth leakage unit, (S Africa, or an RCD, UK) the back feed from the fuses assist the afterglow as if a RCD or earth leakage unit is used any out of balance between the live and neutral will trip the device, hence no circuit to feed the neutral no current flow, no afterglow. But with fuses, the live is dead, the neutral feeds via any earth, i.e water pipe, sewerage pipe of metal construction, TV aerial earth etc. Sio, in essence the only way to kill afterglow is to have a double pole switch to break both live and neutral hence no current flow in any circuit. However, I also noted form experimenting, that installing a twist in the live and neutral kills the afterglow and this has me puzzled as UK and SA have single phase connections. If you had a two phase connection the after glow will be present and only a very fast pulsed light should be emitted from the negative cycle of the 2nd phase, as the positive cycle is disconnected by a switch, The pulse then comes in via the second cycle and earth connections, i.e TV, water pipes, etc. I also noted that many suppliers and installers now use CAT 5 cable and this has a very short, tight twist to each pair of wires. The CAT 5 can handle the low milliamps needed to drive the forward bias to obtain light from the LED and for some strange reason no afterglow is every found. This is why I am busy testing and looking for comments form people to find the common anomaly. It seems to be on older type houses and I don't know if it is a fused DB board or a circuit breaker type DB board that people are having hassles on, as it does seem to make a difference. So the more info I can gather the sooner I can set up another test board in various configurations and see what I can find that is causing the problem. It does not seem to be common in the USA, Canada or Australia, so far it is the UK, Norway, S Africa and only on isolated homes. So I need to find the oddity. If you put a twist in the feeder and neutral and it stops the after glow.let me know, as I cant see why an inductance or capacitance would make the after glow, but, rightly, a twist will cancel out any capacitance. (Transposing the wires at any given point will cancel out magnetic fields, cancelling any capacitance and inductance, but this should only occur at higher voltages, not at 10V. The longer the cable the higher the capacitance and this could occur on a 30m parallel twin wire of smaller cross section. Any info you have may help so keep it coming, Thanks, IQ My quote: I like walking in the forests.Just because I am dragging a body along should not be cause for concern. Enjoy, I need to laugh. Life is getting too serious or I am getting to stuffy!

See K Fry's post. You seem to be talking about a single LED. Please separate your post into paragraphs! I can't understand it as posted...

Well done, your post has been accepted as being accurate and valid.

I should have indicated in the drawings above that in the 2nd and 3rd drawings from the top, the LEDs are technically switched off, but do glow dimly (again, I believe due to the capacitive coupling between the floating and adjacent wires).

In the 4th and 5th drawings, the LEDs will be truly off.

Hi dkwarner,

I was looking for solutions to the nightlight issue and stumbled across your interesting ideas (post #26). In particular, diagrams 4 and 5 from the top seem to solve this issue. However, you also mentioned in a later post #71 that both sides of the lamp being hot when off, as in diagram 4 poses a safety issue. Yes, if someone were to change the bulb and touch the terminals thinking the switch is off will get a nasty shock. I was wondering if you have any other safer options. Thanks for your help.

Best regards,

homediy

Yes. I can quickly think of at least two possible solutions for maximum safety, still using LED lamps.

1. Install a transformer and new sockets to convert to 12 V(or other lower voltage) AC operation.

2. Using the full mains voltage, install a 2-pole relay inside the lamp base connection box, so the leads from the relay to the lamp are only a few cm long:

As indicated in the right-hand drawing, there will still be capacitive coupling in the off positions, but not enough current will flow to activate the relay and the lamp will be truly off.

Excellent! I like the 2nd idea.

I vaguely recall that when one is using a relay, there needs to be a capacitor, diode (?) or something in parallel so that when the relay opens, the back emf will not cause problems.

I don't know enough about electronics to further comment on that though.

Thanks for sharing your ideas!

Cheers!

In DC circuits, a reverse diode is indeed commonly used to absorb the magnetic field energy at turn-off. That can't be used on AC relays, as the diode would short the input half the time, and there is no way of predicting which direction the current would be flowing at the time of turn-off.

Fortunately, at AC line voltage, the coil has sufficient turns of fine wire that the impedance of the coil limits the current at turn-off, and the transformer that provides the line voltage can easily absorb that bit of energy, so that's not a problem.

I haven't tried this, but I suspect this needs to be a true magnetic relay; I haven't seen a two-pole solid state relay with line voltage input (of course that does not mean they don't exist).

Oh yes, I was referring to DC and we are talking about AC!

I was wondering, if it's AC, the relay would have constantly switching N-S poles meaning during the switch over, it might turn off for that tiny moment making the light flicker? I'll google what's a true magnetic relay.

Thanks again.

Cheers!

If the run to the light is longer than 30m use CAT 5 cable, (LAN computer cable).

I wonder if you read the thread so far. The OP has an LED bulb of the type commonly available for replacing incandescents in homes. Such bulbs are powered by "mains" voltage, and include built-in power supplies and an array of LEDs in a package that can be screwed into any typical light fixture.

Cat 5 cable is not usable for home AC "mains" wiring, if you want to keep your insurance intact.

Hi. I have to agree with you here. I only realised this when I finally went to bed at 2.30am. I have mistaenly 'assumed', (and I hate myself for doing so and hate the word), that we are on two different LED lighting types. I appreciate that this is nopw pointed out to me.

The replacement LED bulbs have capacitors in them. so capacitance can be the issue. I have been working with LED lights that run off a driver unit that only provides to a max of 20W, 24V. The mains input being 240V. This is why the LED's can be driven on CAT 5 cable as it is sufficinet for the voltage and wattage.

So, I now have a new test to run and finally found the direction I was looking for.

My apologies to all for 'assuming' that you were running the LED's through a driver unit. The system I am testing is, HideAlite. Model. Jolly Pro, This system offers a 'master driver' and 'slave driver' and all LED's are required to be connected in series.

However, the lights you are replacing, are all parallel connected. Now this may have an influence on the LED lights and the after glow.

Again. my apologies to all for my 'assumption'. Back to the drawing board and testing with alternate LED's.

IQ

OMG!

Thanks anonymous, I understand that OMG means 'Own My Goat'. Lucky lad, I still owe the cow. Thanks for the education. My field is HV, that is anything over 11kV up to 765kV and there we don't worry about LED's. This LED afterglow is interesting. It would be great to find the solution. Anyway, look after that goat. Thanks for the support, Cheers for now. IQ

On this side of the pond, "OMG", in my experience at least, means "Oh My (insert preferred deity here)"

Thanks for the correct interpretation, but it still goes over his head I feel.

He writes as if he lives on another planet........almost no contact with the human race up to now maybe?.....

KAK. So I am still on earth? OF. You are HUMANS? I thought I was HOME. This IS my 1st encounter with HUMANS. Scarry.

I am 98% sure you dont like me. I am 100% sure I dont care!

Lighten up! Most of us have been given assignments outside our comfort zone at one time or another. IQ has clearly stated that this is one such case.

Likewise, when reading any post, we all naturally make assumptions based on our personal experience or lack thereof. I have certainly made some erroneous assumptions on CR4, and again, IQ has admitted doing that, so let's see if we can be of any help!

Why thank you, but you spoiled my crescendo. Ah well. Some news. I have a DC powered LED light system. It operates from a 240v AC, single phase supply, wired into a driver. The driver produces 10-24V DC output. The LED clusters, each of 3 LED's, is wired in series to the driver. (Not parallel). My mission is to discover why certain homes using the LED lights, have an issue with one or more clusters staying lighted when de-energised from the mains power. One or more clusters will have an after glow, much like a night light. There is no apparent reason for the LED cluster to stay on. Now, I have tried your advice with capacitance. The lighting circuit is indeed a parallel circuit, which in every house, any where in the world, is NORMAL. (For 240V or 120V AC ligthing circuits). These circuits are run on 1.5mm2 conductor. NOTE: Not all conductor insulation is PVC. Most nowadays is XLPE insulated which is thinner and offers better insulation characteristics, (This you may not have known). However, XLPE does create a capacitance charge on any conductors, be it Aluminium or copper. The capacitance charge on a 1.5mm2, 3 core indoor cable is 100nF/Km. If I use 30m of cable, there is not sufficient energy, capacitance to ignite the LED's. (The above figure is fact). Therefore I cannot subscribe to capacitance causing this after glow on DC power LED light cluster.So Mr. Warner, I thank you for your drawing which makes a possible reason, but when taking the facts in, it is not possible. Next solution: Inductance, An MV cable, 3 core XLPE, all cores powers and loaded to produce 90 degrees Centigrade temp, at 863 Amps, producing a magnteic field of 1.3 Tesla. with the LED cluster directly placed on the cable, should have wiped out the LED's, or the driver. Result. Not a thing, the light operate and go of when switched off. No afterglow. Next: the entire circuit moved to a 500kV capacitor bank, inside an HV lab. With this to add further assault on the LED clusters and driver unit, still energised, I energised an HV cable, 1000m long, loaded (wound) on a steel drum. I energised the cable via the cap bank to 69kV, the circuit mounted on a board and placed directly under the cap and between the cap and energised cable on the drum. This was done for 25 minutes. Now I have assaulted the circuitry with abuse here, it switches off and no afterglow. If you are wiring in perallel on 120V or 240V AC, on short wiring runs which are single wires drawn into conduits or trunking or what ever you use in your home, why are you experiencing an afterglow on LED lights? Not all, perhaps some. Your home will never experience any of the abuse I have thrown at these LEDS with acpacitance or Inductance. So I have to rule out any of this two in your home. Issue: this seems to happen on older wired houses. Possibly different earthing is the culprit. Back feeds, but one cannot reverse bias an LED light and make it glow. The phases are single phase and 2 phase systems at 50Hz. On 2 phase I can justify the leakage via the negative cycle to earth, but agin, this is reverse bias on the LED cluster. no afterglow! On speaking with the supplier, they have installed a modification or a capacitor in the driver unit. The effect still happens. Now as you say, I am seeking help, I have no idea now, why some of the clusters in some houses stay on after being switched off. One client has 3 clusters on the same circuit, he switches off and 2 extinguish and the third stays on with no power on. OK, it could now be a capacitance charge from the driver circuit, but, this would bleed off rather quickly and the LED would go out. It stays on. My question, WHY? There is no earth as the driver unit is double insulated, only two wire connection, pos and neg in from the mains of 240V AC. I have no inductance, no capacitance, no physical connection of any nature external to this LED cluster on the same circuit as the other two. Why does one LED cluster stay lighted and not go off. If anyone can makes sense of this for me, I am happy to listen as I cannot figure this out. It needs a current to operate the LED in a forward bias. If there is no earth attached, only a negative return to the driver, what is powering only 1 LED cluster? Now that is the problems and the floor is open for comment, info, experimental ideas and If Anonymous can be bothered, he may actually be very helpful here. So that is my problem, any ideas at all as, as I stated, my filed is anything from 1kV and upwards. I play with big voltages both AC and DC. But this lighting has me baffled along with a few of my colleagues, so we are here, I am here for you expertise, as I am happy to share my knowledge of cables, overhead lines, sub stations and marine and wind power with you lads. Fair trade, no robbery. The floor is open and thanks for your comments. Anonymous, I shall miss the banter. Thanks guys. if there is anything that is not clear, simply ask me for more info. Thanks IQ

First, if you haven't already seen it, take a look at my post #61, showing my experiment and results.

Once again, PLEASE break your posts into comprehensible paragraphs, so each item can be addressed separately.

Since most household devices in the US are powered by 120 VAC, they require more current than similar devices powered by 240 VAC, and that means our standard wire is larger than that used in 240V regions. The size 12 wire I've been using is very close to Ø2mm, so has an area of just over 3mm^2. For a given wire spacing and insulation material, the larger wire will have greater capacitance.

My home was built in the early '60s, and the coil of cable I've been using was purchased somewhere in the '70s, so both probably have PVC insulation. I just checked the box; it says nothing about the insulation. I'm not sure what XLPE is, but I suspect it may be related to UHMWPE (Ultra High Molecular Weight PolyEthylene). In any case, thinner and better insulation characteristics both would lead to greater capacitance between parallel conductors located in a single sheath.

Your figure of 100nF/km means 0.1nF/m, or 100pF/m, which agrees fairly well with my cable. My meter indicates 3.3nF between black and center wires, 3.1nF between white and center, and 2.3nF between black and white wires for my roll of roughly 40 m of cable. Here is a photograph of the 120V LED lamp connected only by a 3.3nF capacitor (Labelled 332k, meaning 33 followed by 2 zeros and a k for kilo, or 3300kpF) through the meter indicating 0.104 nA.

When you say "If I use 30m of cable, there is not sufficient energy, capacitance to ignite the LED's.", I can only come to two possible conclusions: either

1. You are using a 'driver' that is really a power supply intended to be used with much heavier loads. As such, it can't do anything with a few nanoAmperes of current. You MUST use the same kind of devices that are causing the problem. or

2. You are trying to use the cable on the low voltage side of the driver. It MUST be on the AC high voltage side.

Just in case, I have done this experimentation with a 'Lights of America' model 2526LED-LF3, for 120V-60Hz 4W 85mA, 3000K 200lm.

Obviously you can't use the same exact unit, since you are working with 240V.

That model etc. is quoted directly from the side of the lamp. My measurements indicated a full power current just over 70mA, and our power line voltage today has been 119V, so the power would calculate at 8.3 Watts. On the other hand, this is a medium-quality meter, and does NOT say true RMS. I have not gotten out my 'scope to look at any waveforms. If my meter is responding to peaks it could be 70mA for a fraction of the cycle, which could conceivably give a true power of under 4W.

I am a bit confused by your use of the term 'afterglow'. To me, afterglow is what I see immediately after turning the lights off. That fades away within a minute or two. The dim glow that I've been referring to lasts as long as the circuit has power, but the power is switched off. By the way, If I turn the circuit off at the breaker panel, the lamp does go out completely.

The bottom line:

I am confident that your high voltage (kV+) has nothing to do with the problem. That is easy to prove: if the lamp goes completely out when the entire circuit is de-energized (as at the breaker panel), then the nearby HV lines have nothing to do with it; If the lamp still glows when the entire circuit is de-energized, then the nearby HV lines may well have something to do with it.

Rather it has to do with how the units are wired; any time one of these units is wired in such a way that turning it off leaves a wire connected to the lamp but nowhere else (a floating wire), that wire becomes a plate of a capacitor (or you can think of it as an antenna) to obtain energy from nearby wires.

I've shown in an earlier post how to solve the problem for lights controlled by two switches. For simple on-off circuits, the simplest way to avoid the problem would be to use a two-pole switch, so the otherwise floating wire would always be connected, either to neutral or to hot. If you need a diagram, I can provide it.

Good luck on your assignment!

TKWarner: Thank you, appreciated. I believe you have proved to be correct on AC circuits and residual capacitance found on wires. This leakage is also present over the switches as leakage and being adjacent to other circuits and their capacitances, which will very with loads on the circuits, is the cause of the afterglow.

I cannot simulate this problem on DC circuits, having tried today, it just does not occur.

So this issue cannot be solved, in fact it will increase and the reason follows:

XLPE; Cross linked Polyethylene or as it was know in the USA, PEX; Polyethylene cross linked,

It is a very high value insulation, used mainly on HV cables up to 500kV. White in colour.

XLPE is being used on LV cables and wires, replacing PVC and rubber. The problem which is occurring now is that the high value of insulation of XLPE can allow the manufacturers to reduce the insualtion thickness on wires. This will cause higher capacitance leakages.

XLPE cable, once discharged and left for a period of time, will build back up to the initial voltage applied as a capacitance charge. It is therefore very appicable to attach an earth to the conductor(s) on MV\HV cables of XLPE construction and leave it applied until ready to re-energise the cable. Otherwise, if the earth is left off and one touches the condutor, one will get the full intial voltage dishcharge, which can kill you. So anything over 0.5kV with XLPE insulation which has been energised and now disconnected, should immediately have an earth applied to the conductor and the earth left connected, while you work around the cable end.

Now before anyone wishes to attack what I say here. . This does not happen on PILC (Paper insulated lead covered cables) or Oil Filled cables which both have paper insulation. (This is my filed of expertise).

This only happens on XLPE insulated conductors. This will not happen on household wires at 240V or 120V AC. But the capacitance charge leakage, due to thinner insulation, will increase. So this means that when your LED's are off, the capacitance will be higher with XLPE insulated wires and the LED on mains supply will glow brighter.

Now that expalined, I thank one and all for the information, in particular Mr. DK Warner, for his attention to detail and explanations. And to all others who's assistance and advice I have taken and tried.

There is clearly an issue with wiring on the older houses and clearly an issue with the type of lighting systems used, i.e. 120-240V AC mains, versus DC driven LED's (via a driver unit). So, apples and oranges were being compared. Coupled with different earthing on older houses, different wiring and insulations and several other factors, the after glow will remain. On DC powered LED's via the driver with an output of 10V, I doubt the afterglow will occur and until I can establish how just one house is wired, and has this problem, I can go no further with testing.

Thanks to all who have assited me with this task, I shall now get back to where I like being and that is anythng over 11 000V and upwards. I shall ensure that when my report is finalised, that all person contributing information, are credtited in the report.

Thanks again One and All, appreciated the help in your field.

IQ

Thanks for your summary! Many who come here never come back to let us know when their problem is solved, or at least understood.

Now as to there being no solution; in practicality that may be true depending on the way houses were wired, the way the actual lamps are mounted, and the effort the homeowner is willing to exert to solve the problem.

But there are several possible solutions, depending on the circumstances:

1. The simplest solution is to replace either the lamps or the drivers with higher power units that don't exhibit the problem.

2. In some cases, the switches may be rewired or replaced and rewired. I've shown in an earlier post how the wiring can be changed for the hall-type 2-way switches. Below are possible solutions for simple ON-OFF circuits.

In the first set of three drawings, the switch is wired in the neutral wire, so the socket is always hot (NOT a recommended wiring procedure, but undoubtedly done somewhere). The floating wire can then have capacitive connection either to a ground wire or to a metallic conduit (shown in yellow as before). One solution (circuit 3) would replace the existing switch (probably SPST) with an SPDT, and connecting it with the common to the lamp, so the switch connects either to hot or to neutral. There is a safety issue here, because when the lamp is OFF, both sides of the lamp socket are hot. The better solution would be to connect the switch on the hot side, but that might be difficult in a retrofit.

The second set of three is a bit far-fetched, as it depends on capacitive coupling to the hot wire of some other parallel wire/cable, but I know there are bundles of cables loosely tied together in my attic, so it is possible. Again, the solution is to replace the SPST switch with an SPDT and connect it with the common to the lamp, as shown in circuit 6. There is no safety issue here, as both sides of the socket are neutral when turned OFF.

One I got started on this, I couldn't stop! I spent essentially an entire day on it. If you have any use for the drawings in higher quality than they show here, send me a Private message with your email, and I can send any of several formats (PDF, JPG, DXF, etc.)

Dick Warner

I will drop you my work email and if it is OK with you, I shall include your drawings and discoveries in the report. (Ericsson Cable Co, Sweden). And No, I am not Swedish in case you wondered. I deal with what I call 'sensible sized wires',rated up to 500kV, per phase. Each phase having a workable leakage voltage on the cable sheath, which is regulated to 150V Max, on the outer sheath of a cable by use of Link Boxes and transposing sheath voltages every 1000-1500m This is my field. Building these systems and involved in Technical requirements/solutions on T&D systems. LED's? I place my faith with you lads. Thanks again for the help, appreciated. IQ

<...the switch is wired in the neutral wire...>

That is not permitted in BS7671 and is certainly not a good idea.

Correct! I was showing possible ways that a single LED lamp with two switches could be wired such that it would glow dimly when "turned off". I should have mentioned that in the post.

I have two chandeliers (with multiple LED lamps connected in parallel) in my home that glow dimly when "turned off", but only for some, not all, 120V LED lamps. I have NOT investigated how they are wired, although one has a single switch, while the other has switches at each end of a short hallway.

GA.

That makes complete (and common) sense to me personally.....

I like your methods, thoughts and findings!!!! Well done and thanks.

At last some clear thoughts and actions....

Now take a look at my post #65.

I tried several other screw-in LED lamps and one unit that has a 'wall wart' to convert the voltage down to 12V to power the LED assembly. When first connected using the 3.3nF capacitor, this unit would give out a brief flash, then nothing. All the others were higher powered than the original '25W replacement', and NONE of them would glow with the 3.3nF capacitor in series with the power.

So it appears that this problem only exists for very low-power units, and your testing must be done on such units.

<...Running LED for lighting can save up to 78% in cost....>

• What level of circuit protective device is used on that <shudder>?
• What steps are taken to identify the <...CAT 5 cable...> as carrying power for a lighting circuit <...wince...>>?
• What measures are in place to stop some unsuspecting soul putting a meaty 500W incandescent load onto the circuit in place of the <...LED...> <wince, duck>?

No qualified local Electrician would do such a thing!

I can not find any LED bulbs that cheap.

Around here they are about $60 for a LED bulb which puts out nearly what a 60W incandescent would.

How many years would a LED bulb have to last to make the initial price a good deal?

And I have had some bad luck already in this area. I bought 6 LED GU10's for my kitchen. One was DOA, two more burnt out in the first 2 months. The wholesaler would not refund or replace either. Dont buy at EWEL!! Luckily these were only $15 each.

But I went to Walmart and bought a couple flourescent GU10's for the same $15 price each and they are still going strong 4 years later.

LED has great potential but I am not going to be the lab rat anymore.

On a side note my TV uses RGB LED's and the picture is as bright and colorful as the day I bought it in 2007. But this is PHLAT light technology which seems much better thought out and effective.

So yeah what brand are you guys buying that cost $25 each? Where do you get them? Online? The $60 ones are Phillips and pretty much the only ones I have seen in local wholesalers.

For $25 I would take the plunge.

When you install these incandescent replacement LED's are you installing them on a feeder that is parallel wired or series wired? May sound a tad daft a question, but I have a good reason, as if one utilises a driver with ac240V (120V) input and the output is 10-24V DC, then the LED lights need to be series wired. Just for comparison, check out Orsram Sylvannia for LED lights. I am currently in Sweden and the price for a 2.5W LED light is 145 Swedish Krona. This is where I am trying to run this testing and research on after glow of LED lights. They have singe phase and 2 phase systems here. So any offers of information or advice from your experiences is appreciated that I can simulate the problems and find a solution to the problem. I shall run tests on 240V AC on a parallel wired and a series wired system and see what happens about afterglow. It has me puzzled and the installer say a simple twist in the conductor run eliminates the after glow. {If parallel wires are used, (untwisted pairs)}. I cant even simulate the afterglow issue as of yet, on twisted or untwisted wiring. Has me puzzled. And I am listening to your experienced advice. Parallel or series wired chaps, which one do you find is right? Any advice please!

Sorry I think you believe I am using LED's with seperate drivers.

The lighting I refer to is a complete unit and suited for AC circuits.

All residential AC circuits are of course wired in parallel.

I would think the LEDs may stay excited due to an induced voltage in the wiring.

This would indicate a circuit conductor in very close proximity to the wiring for the LEDs. This interfering circuit would likely have a large current draw and/or have a large surface area in contact to the sheaths of the LED circuit.

I would also buy the capactive effect story as just as likely but I have never seen it.

It is very common to see disconnected circuits showing 120V on them due to inductance from another circuit. This is misleading as you can stick these wires in your mouth and not get a shock. The voltage will say 120V but the current capable of being transmitted this way is generally not enough to power anything.

But we are talking about LED's here, which at full power require less then a Watt generally, so a small glow could be powered by a few milli-Watts more then likely.

To recreate this effect, find a 30-100Amp load, run this load on conductors about 100-200ft long. Also run the LED circuit wires the same length and get them in very close proximity, as close as you can for the entire length of the runs. You may have to seperate the conductors in each circuit so they dont cancel each other out.

This is basically a single turn transformer which can reproduce the source voltage on the induced line but with a very limited capacity to pass current.

Again I am only talking about AC circuits here, not DC.

On a DC circuit I dont know what would cause the effect if it is prolonged.

The glow effect lasts indefinitely I assume and so residual capacitor charge is not very likely.

Thanks for this. I tried the capacitive and it can happen but it bleeds away quickly so a short duration afterglow can occur. As for the inductance I actually ran a cable to 863 amps, had a magnetic field of 1.3 Tesla and nothing happened. So being adjacent to induction filed is not causing after glow. The lights I am running are DC, wired through a driver. I have induced the circuit with 69kV, nothing, no afterglow so HV circuits don't seem to cause any effects. However, I now need to try the incandescent replacements types as they have capacitors installed in the circuitry and this may have an effect by induction from other circuits. If a longer length of wire is used for a long LED light run, capacitance plays a part in the afterglow due to the long wire run and capacitance of the copper wire, and this can be eliminated by a simple twisting together of the live and neutral, over short lengths, longer lengths a twisted pair is better. So I shall try a parallel 240V AC circuit, rather than the 240V, driver/10VDC circuit on test now. Look up Rosewoods LED Lighting, they have pretty good info on LED for architecture and suggest CAT 5 for controls and lighting runs over 30m long. And Hidealite suggest no runs over 45m on DC LED's, wired in series. Osram was not bad and Phillips was not bad either for information and light fitting. Progress Lighting was not great for info, but good on product range. www.LED.alibaba.com seemed to be reasonable on prices Thanks for your advice and help, appreciated. I will let you lads know what the outcome is, when I get a result. Cheers & Thanks IQ

Sorry but just checking was that 863Amps AC or DC? And was that Voltage the same or higher then the LED circuit?

Only AC creates inductance and the induced voltage current will also be AC so if your LED circuit is DC I'm not sure what the induced AC would do to it, if anything.

If the 863Amps was from a DC circuit it would not effect the other circuit I do not think.

Only when both circuits are AC do I have experience with disconnected lines showing Voltage, and in theory could keep a very small load partially powered up.

Also how did you set that up so quickley?

This was 863 Amps AC. The voltage was 59kV. I work in a High Voltage test lab for a cable Company. My field of profession is HV up to 765kV, testing at voltages of up to 20MV. The big stuff. I have this project on LED lights to run, not my forte. But, it does seem I have been given wrong data and probaly the wrong lighting system is being tested for this afterglow..

My electronics knowledge is outdated, its to small to enjoy for me. I like the higher voltages to play with.

This is why I came on this site as I realised there are people who know better than myself about these LED lights and there peculiarities. Hence I sought there help.

On a HV DC circuits, the only higher voltage obtained is the initial switching voltage which drops down qucikly to the stable straight line of applied voltage. It is a spike, which is intially very high and it only happens on switching. It is simply a massive pulse, (in rush current), lasting up to a few seconds, especially on HV DC lines around 400kV. The spike is enough to flash to ground at midspan of any transmission line and this is approximately 10m from the line to ground, (Varies in other countries in midspan heights).

LED; My next step is to contact the manufacturer, ensure the data I have is correct, build a couple of circuits in parallel and series, and see if I can obtain the afterglow, by any means possible, when power is disconnected from the circuit. This will be on 240V , single phase thoriugh a driver, to 10V DC out with 4 LED light fittings coupled in series. Then I shall change configurations in circuitry and keep trying capacitance and inductive loads at well over what would be found in any home and see if I can recreate the after glow.

Frankly, I can only see this afterglow coming from a capactitance charge on the feeder wires, but one needs a long circuit for this capacitance build up, or from another source of current flow, such as an earth connection.

I shall keep you posted of any solution found. Thanks for all the help so far.

IQ

My hallway is wired in the standard fashion, using RomeX™ cable (3 parallel [not twisted] wires contained in a single sheath). My home was wired at a time (early '60s), when only the kitchen, bath, and utility room have grounded outlets. All others were polarized 2-pin sockets, and none of the lighting boxes are grounded (they're plastic anyway). The 3-wire cable is strung from one switch past the lamp to the other switch, with only one conductor cut to insert the lamp, so the uncut parallel conductors will be a nice capacitor the full length of the cable.

I've got most of a 250ft roll of that RomeX™ cable, and a pair of two-way switches that I just replaced with silent ones. One of these days I'll set up a test using long and short cables. I'm betting that the one made from a short cable will not glow in the Off position, or glow very feebly, while the one made from a long cable will glow brighter than my hall light in the Off position. I'll take some measurements...

GA.

Well done & tested.

What appears to be lighting the LEDs at a low level is "leakage" which is always there in AC circuits to a greater or lesser degree. There are actually major losses for power stations due to the usually long runs that are involved......I read that recently in an article, but at this time I don't know where....

I did find an article that covered the subject in depth here if anyone is interested:-

http://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=1270&context=ecetr

What would be an interesting, but difficult to measure fact, would be to find out if the loss is increased when a LED bulb is in use, but switched off and still glowing? My bet it would make no difference whatsoever.....the energy gets wasted with or without the LED bulb.

So really, it gets more efficient by being used for some illumination!!! (Almost a Joke!)

"What would be an interesting, but difficult to measure fact, would be to find out if the loss is increased when a LED bulb is in use, but switched off and still glowing?"

Actually, it wasn't that difficult; I've already done it. When I connected the roll of cable through the current meter to the outside two conductors, with nothing connected at the other end, the current was 0.105mA. When I added the LED lamp between the center and one outside conductor at the other end, the current increased to 0.150mA. So the cable was using roughly 2/3 of the wasted energy, and the lamp was using roughly 1/3.

Back when we used only incandescent lamps, 0.105mA was considered zero on a 120V circuit!

As I said at least once before, I have no intention of 'correcting this problem', as it just enabled me to remove a nightlight from that hallway. On the other hand, I prefer a totally dark bedroom, so if it were there, I would correct the problem.

That got me to thinking... I have a two-socket ceiling lamp in my bedroom, currently containing two CFLs. When I replaced one of the CFLs with the LED unit, there was no glow; but when I unscrewed the second CFL, leaving only the LED unit connected, the glow appeared (although dimmer than in the hallway, due to shorter wires). So with the LED and CFL connected together, the driver for the CFL used any available energy (without perceptible glow), preventing the LED unit from glowing.

Further experimentation has shown that the '25 W replacement' used in my experiments is the only one that is sensitive enough to have the 'problem'. I tried several other types of 120V LED assemblies powered only through a 3.3nF capacitor. One of the others would give a brief flash when connected; the rest showed no perceptible light. So the simple solution for my bedroom (when I eventually replace the CFLs with LEDs) is to use slightly higher powered lamps!

GA again, you must be getting weary of my GAs, sorry about that!!!

We don't always agree, but here you are doing a grand job while I just sit and read your test results....

So that dim glow does cost something, really interesting, Thanks.

Thanks again! Of course the GAs don't mean much unless someone else agrees, and acts accordingly...

While I was working on this, it suddenly occurred to me that virtually every cable in and outside the house is wasting a bit of energy, even when everything is turned off! Presumably there is greater waste here with 60Hz than in places where its 50Hz, but then most places with 50Hz have twice the voltage, so it may be the other way around... Of course the waste is a tiny fraction of normal usage, but multiply by millions of homes...

Absolutely correct.....

OK, a couple more (maybe last) comments:

I have been putting most of the pertinent posts together with higher quality graphics for IQ to use in his report. I found it interesting to see that when I copied from CR4 (viewed on a Mac using Firefox 18.0.2) and pasted into Word, the graphics looked better in Word than in Firefox. They still did not look as good as the original jpgs exported from my CAD program. I ended up pasting screen shots from CAD into Word. The best quality would seem to be using PDF, but that would take a bit longer.

While editing, I noticed that in Post #65 I said that the cable alone had a current of 0.105mA, and the cable with lamp had a current of 0.150mA. Then I saw that in Post #64 I said that the lamp, connected with a 3.3nF capacitor (equivalent to the capacitance of the cable), carried a current of 0.104nA. When I first saw that, I thought that one or the other was in error, but then I realized that in the case with the cable, the lamp and the cable's capacitance are in parallel, while in the second case the capacitor and the lamp are in series, so those numbers are indeed consistent.

Time to get back to work on the things I'm paid for...

In case anyone is interested, I finally found time to cut open a failed unit of the LED lamps that glow slightly when turned off. This one had the normal faint glow when switched OFF, but only flickered at a fraction of normal brightness when switched ON. I used a lathe to make cuts through the plastic bulb and aluminum heatsink to get it apart. Here is the circuit of the unit:

I've shown C1-4 as two small capacitors and two larger ones, as that is how the unit is constructed, in order to fit the round circuit board. There is clearly a direct path from the hot side of power through R1, the diodes, and R2 to the LEDs and back, so it is easy to understand how this unit glows slightly when one side of it is not connected directly to power, but to a wire that has capacitance to the other side of the power source.

The value of R2 was partially obscured by adhesive, but appeared to be 37R2, which I would assume meant 37.2Ω, although I find it strange to use a resistor with that high a precision. The Ohmmeter indicated it was open, so I soldered a 39Ω resistor in parallel and tried applying power. The LEDs glowed very brightly for about half a second, then nothing, so apparently that value was wrong and blew out one or more of the LEDs, but clearly R2 was the failed component.

All of these components are surface mount, on a single-sided circuit board 0.9" (23 mm) in diameter. The resistors are 0.06" x 0.12" (1.5 x 3 mm).

Hi. Thanks for this drawing and information. It makes sense why the capacitance of the wires is causing the glow. The twist that is proffered by some electricians makes sense too, as this is simple transpositioning of the conductors. (as is done on any HV lines to reduce capacitance over long distances). NB: it does not always work, I am told now. Clearly there is a vast difference between what you are using and what I have been asked to test here. The test report was submitted and all documents were logged in and issued. Thanks to all who assisted me, and even to those who thought me a F'wit, but, I appreciated all your inputs. My particular thanks to DK Warner for his patience and assistance. Thank you all. Regards. IQ

You still don't seem to understand that we are talking about standard 120V or 240V Edison screw base (E26 or E27) lamps. They screw into standard household lamp sockets, and are always wired in parallel. See my previous post for photos of packages.

With respect to the afterglow, many, if not most, "white" LEDs use phosphors to convert shorter wavelength light into the colors not initially produced by the LED itself. These phosphors are normally fluorescent (convert short wavelength energy into longer wavelength(s), but are also commonly phosphorescent (store some of the energy and continue to release it after the excitation has ceased). I see several things glowing when I first turn the lights off in my room on a dark night, including the screen of my old oscilloscope, a CFL or two, and some white LEDs.

Ok so the decision to wire your DC circuit will be based on the source Voltage and the Voltage of your load.

Now because each LED will vary slightly it is usually best to run them in parallel and use current limiting resistors to maintain the correct current through them.

An LED is a diode, and so will only take the Voltage it requires to be forward biased.

This would seem to indicate you could hook any LED to any Voltage and it would be fine, this is not correct. Once the LED is forward biased it behaves just like a dead short and the current will attempt to reach infinity and the heat will destroy the LED instantly.

If the supply Voltage is 120VDC and the LED's require 1V each to forward bias you could in theory run them all in series with no resistor, but the real world doesnt work that way. The LED's will vary slightly and create problems and fail.

You could in the same scenerio run 12 parallel groups of 10 LEDs in series and then use resistors to limit the current through each group.

Or you could run each LED by itself and use a resistor to limit the current to the LED rating.

Any variation on this would work as well, but the safest for the LED is each by themselves and run in parallel with all the others.

Is this clear enough?

If you give me the rating of your LED's and the source Voltage I could give a better response I'd bet.

Cheers.

I have never seen house lighting wired in series, would you be so kind as to demonstrate your opinion with a drawing please? I feel that you possibly mean something else, maybe you are using the wrong word(s)?

All house wiring in both the UK & Germany are made with parallel wiring......each light receives full mains voltage.....wiring in series would seriously reduce light output.

please see post #43.

The third sentance I state something like "All residential AC wiring is parallel"

My post which you responded to is referring to an electronic non-residential circuit proposed by IQ. If you look at my post just above your I also refer to 120VDC I believe, not AC.

But hey why buy a dimmer when you can wire all your lights in your house in series and have some nice low mood lighting! I suggest red lights.

Cheers.

I think that you answered the wrong post......check carefully!!!

Where are you located? How recently have you looked? Things have improved greatly in the last year.

Here are three photos taken recently in a nearby Costco: (I've left them sideways to save space. I have no connection to Costco beyond a membership.)

The first is your 60W replacement, which you can see is $15.

These are the 25W replacements, at $6.30 each.

I also have been extremely happy with these flashlights: I was happy when they were 3 for just over $20, and bought a dozen for home and gifts. Just last week they dropped to just over $15 for the three-pack.

Well I need to stop shopping at the Electrical wholesaler.

To be sure I just called and they gave me these prices on the Sylvania incandescent replacements. (I think I posted Phillips earlier, sry it was Sylvania)

12W = $43.99

14W = $46.08

Could not tell me the equivalent wattages but they should be somewhere around the 60W equivalent. These bulbs were over $60 about 6 months ago or so.

I bet Walmart or Home Depot or Canadian Tire would be more in line with the prices you posted.

Pretty sad when Walmart can beat my companies discount.

Thanks for the heads up.

I've had similar experiences at electrical 'wholesalers'.

The 800 lumen one says it uses 13.5 Watts.

I forgot to add that my wife likes all of these; a very significant factor!

One might believe that the wholesalers would only carry higher quality units, thus the higher prices, but I'm not convinced.

Costco has an excellent return policy, so I believe they choose their products with quality in mind. In fact, after I bought my first LED lamps there two or three years ago, I got a letter from Costco offering to take the lamps back, since they were not meeting the life expectancy advertised. I kept mine.

Now if those lamps you found are actually made in the USA, I'd be tempted to pay the higher price. The above are made in the Orient (actual country intentionally omitted).

I have only bought on ebay, some lasted only 12 months, but I still have LED lamps that are 7 years old and still in daily usage.

The money saved on ebay can be tremendous. Look for sellers with 100%.....

A compliment, you named the name, something that more of us here should do....many thanks.

How, pray tell, is a fluorescent lamp a DC device?

It may work on DC (for a period) but the modern fluorescent lamp which was created in 1934 was developed solely on and for AC.

I wouldn't confuse the modern fluroescent lamp with earlier arc-discharge lamps, which were developed with DC at first.

See Andy's recent post for an explanation.

Actually, a fluorescent lamp at low frequency (less than 1000 Hz) exhibits characteristics similar to a diode. Granted it does not work very well as a diode because the voltage drop from cathode to anode is a function of the length of the lamp. In the old tube technology days, they kept this distance very short but the original diodes were electron tubes.

On AC circuits, the current flow is briefly zero at the cross over from one polarity to the other. The electron flow reverses when the applied voltage reverses and the current and voltage inside the lamp are nearly in phase.

Above 1000 Hz, the diode effect falls off and the plasma effectively becomes resistive. The modern fluorescent is fundamentally the same as it was in 1934.

I generally get your sense of humor, but not here.

LED "bulbs" are those things that you screw into a light socket, usually 120 VAC in the US and 240 VAC in the rest of the world. The device, which consists of a power supply and usually many LEDs, runs on AC.

Although the individual LEDs inside the bulb run on DC, a transformer has nothing to do with changing AC to DC. And of course, no such change is required: all the electronics are built into the bulb.

Sorry,

You're right. Electronics is not my first language. Rectifier.

You are right. The electronics are built inside the bulb to convert the AC to DC volts for the LED. Today I plug in the bulb it blown off. Lights goes off after 5 minutes and burning smell from the bulb. I think 220vac or 240vac should not have so great impact to the converter electronics. I suspect the frequency. Can it be 60hz at 220vac ? Normally 60hz is for 110vac. My house supply is 240vac at 50hz.

A 60 hz transformer running on 50 hz will heat up, so if it is has a very small margin for safety it could fail on 50hz. The power supply is fairly likely to incorporate a transformer or at least an inductor in a buck converter, so frequency is an issue. However, most power supplies are built with the possibility of different frequencies in mind, and it is not unusual to see power supplies that work from 47 to 63 hz.

Eventually, LED bulbs will come down in price, but yours was probably fairly expensive, and you might want to take it back (intact) to get your money back. But if you want to experiment, it would be interesting to take it apart to know what actually failed.

Yes, you should be able to use it. The utility company goes to great length to provide a nominal voltage level with a 10% tolerance level. In the USA and other places, the voltage level of 208 qualifies as a 220 VAC supply near the lower end of the tolerance spec. There are several ways to achieve the required voltage. Frequently, a supply transformer has "taps" that allows for adjustment of the output voltage in order to compensate for line losses. And the story goes on and on....

Also common in the USA electrical grid is a split 240 VAC supply. The smaller loads are usually fed to one half of the 240 VAC transformer, thereby providing 120 VAC, and theoretically reducing the shock hazard. The other half of the smaller set of power loads is fed from the other half of the 240 VAC. In this way, the theoretical load is spread over the whole coil. Larger loads use both halves or all of the 240 VAC.

It would not be in the best interest of the lamp manufacturer to make such a restriction. For a system designed to provide 240 VAC the lower limit would be 216 VAC and the upper limit would be 264 VAC. Your product should work anywhere in that range or they will soon go out of business.

If the bulb is rated for 220 VAC then he should be able to screw it in and be done.

I have several 120 VAC bulbs with Edison bases and they work fine on 120 VAC (except for causing noise on some FM radio stations).

Yes.

I like the simple and direct answers...

YES, the simple one.

It is a current controlled one. It has to have a wide band of control to do that.

"Can I use this light bulb."Since your later post indicates that this one gave out the magic smell, Not for lighting purposes.

Can you use another identical light bulb? Almost certainly! In a similar manner to CFLs (Compact Fluorescent Lamps), LED lamps are powered with fairly complex electronics, and there are several components in those circuits that are prone to failure. I have a box with quite a few dead CFLs waiting for appropriate disposal, and I have had 2 Edison base 120VAC LEDs that failed; on one of them half the LEDs still work.

As with any new technology, over time the quality will go up and the price will come down. In any production run, there are usually some marginal units that do get sold. Follow K Fry's advice: take it back and get another.

You don't say where you are located, but since you indicate 240V 50HZ, it's probably not in the Americas. Make sure you buy your lamps from a local distributor; that should ensure that the lamps are appropriate for the local power, and also makes it easy to return defective items.

Finally, if you are not on a major power grid, or if there is major industry nearby, there may be voltage spikes on your line. The electronics of these lamps is much more sensitive to short spikes than are incandescent lamps, which inherently increase their resistance at higher voltage. If you have intermittent power, some form of surge protection may be required to use LED's successfully.

You ask to share experience: I'm about 1/3 to 1/2 of the way to converting all my lighting to LEDs. So far, I'm much happier with LEDs than I was with CFLs.

I started converting to LEDs and decided to wait until things matured a little. My first LED bulb lasted about 1 month, and cost about 5 times as much as the CFL it replaced. I also have had many CFLs fail prematurely. Aside from bad quality control, my perception is that early bulbs were often not marked "base down" when appropriate.

Multiple CFLs in one fixture (such as a string of five in an over-the-bathroom-mirror fixture) seem problematic. The switch for my fixture would let out an ominous and fairly loud pop every time it was turned off, after installing CFL's. The steady-state current was substantially less for the CFLs, but on turn off, they created a large inductive spike.

I hope your experience goes well... I'm not too far behind.

Yes you can use the LED bulb in your house, but certain key body parts will (1) fall off, or (2) become excessively large, depending on whether you are facing toward or away from the light.

It is only recently that utilities have 'upped' the voltage to a more stable supply of 240V. The LED Bulb you have probably fits straight into a light bulb socket with a bayonet fitting. If so, it is safe for 240V as all suppliers have a voltage range to work with. If it is NOT an bayonet fitting as you would find on a normal light bulb, then recheck the voltage rating, the amperage but I believe if based on what you say it is a LED bulb then it should be safe for use. It will have a rectifier and transformer build into the holder. IQ

I don't know where in the world you are located...

Do you really mean bayonet? All the bayonet lamps I remember were either low voltage (like 24V or lower voltage indicator lamps) or high current (like projection lamps).

The standard sockets here in the USA are screw bases. We do have an advantage or many other places, since our standard 120V is somewhat easier to convert to the lower voltages required by LEDs. That conversion is NOT done by a simple transformer and rectifier in any 120V LED lamp I've examined. It's more like a miniature switching power supply.

Yes, I mean BAYONET fittings. The ones with the 2 little studs that stick out. They are still widely used in the world. Edisson screw is not popular and frankly is used as Industrial applications. And Edisson screw has two sizes, large or small.

I am on planet earth, I hope. However thaks for the info and please see my apology to all. I now have a different test to run on bulbs, (globes), that diretly connect to the 120V/240V circuit rather than through a driver unit converting the output to 10V or 24V to feed the series connect LED lights.

IQ

Yes you are right. The rectification from AC to DC to operate the Led is built inside the bulb. The LED light bulb that burn off is using the E27 socket. I open up the bulb I found several burned components.

Who's right?

I have a few E27 240V lamps (incandescent - many years old) that do fit our E26 (commonly known simply as 'medium') sockets here in the USA. Obviously they put out very little light when powered by 120V, but I have occasionally used them for effect...

I just carefully checked both the packaging and the website for a Feit 800 lumen LED replacement that I purchased recently, and it gives no information on acceptable voltage range; it only says "Use only on 120 volt 60 Hz circuits".

I believe it is standard practice that electrical devices should operate reliably on the specified voltage ±10%, so a 120V device should work acceptably on anything from 108 to 132V, and a 220V device should work acceptably on anything from 198 to 244V. A unit intended for 220V but operated on 240V would be pretty close to the upper limit at 240V, so it wouldn't take much of a spike to exceed the +10%. If you bought the unit locally, it should be correct for the local voltage and frequency.

I've had 2-120V LED lamps fail so far. On the first one, several of the LED's failed to a short, but that is almost certainly because the driver electronics failed first. Reverse current will almost instantly destroy any individual LED. On the second, one half of the LED's went out, while the other half continue to work. I haven't taken it apart yet, but I'm betting that is from a bad solder joint (quality control issues, exacerbated by the elimination of lead from solder).

I have disassembled a couple of failed red LED stoplights, and they both failed due to bad solder joints.