LED switching capacity

[SUNRAY]

I installed a 500W linear halogen mid 90's never had to replace a bulb, the fitting had suffered a lot with water ingress but still working. 2011 we had UPVC cladding installed so i installed a 50W led in a slightly different location, I went for a german make which at the time were supposed to be the bees knees and they came with a 3 year warranty. Just managed to get the second replacement in the 3 years, currently on the 2nd cheap ebay version. No I don't think they do have a longer life expectancy.

 

[88v8]

...I don't think they do have a longer life expectancy.

They should, but I suppose like most things they get hijacked by our obsession with 'cheap'.
Our halogen was the same, leaky but working.

The thing is with thatch though, can't put halogen too near because of the heat.

 

[JohnW2]

They should, but I suppose like most things they get hijacked by our obsession with 'cheap'. Our halogen was the same, leaky but working.

'On average', LED lamps/bulbs certainly do last very much longer than any sort of incandescent (albeit probably not as long as many expect**). I have a fairly obscene number of lamps/bulbs in my large house and (quite apart from the electricity bills!), in the days of incandescents I used to joke that buying bulbs was a very significant part of the running costs of the house - I certainly used to by bulbs 'by the dozens', rather than in ones or twos. Since progressing (via CFLs) to LEDs almost everywhere that has totally changed, and it is only very occasionally that I have to replace one.

I would add that, over the years, I've done a fair bit of experimenting, as a result of which I now usually buy more-or-less the cheapest LEDs I can find - since my experiments have shown me that the most expensive branded ones are not appreciably better (in terms of survival time).

As with almost any manufactured (particularly mass-manufactured) product, there are a small number of very early failures but, once one has been in use for a week or three they usually last for a long time.

** I think that some people's expectations of life expectancy (of any sort of bulb/lamp) may be influenced by a lack of understanding of the limitation of the figures quoted/claimed by manufacturers. When life is claimed to be, say, 10,000 hours, that will nearly always be a median figure, meaning that 50% are expected to fail before 10,000 h (and 50% later than that), but this tells us nothing about the distribution/spread of survival times in those which don't get as far as 10,000 h. To consider an extreme, if half of them only lasted for 1 minute (or even 1 second!) and the other half lasted 10,001 hours, that would still be a median of 10,000 h.

What we really need (but very rarely are told) is some 'centile' other than the 50% one (the median). If we were told that, say, 90% were expected to last 5,000 h, that would be more meaningful, and it would then be clear that such a product was preferable to one in which 90% lasted only, say, 1,000 hours, even though it is quite possible that both had a median expected life of 10,000 hours.

The thing is with thatch though, can't put halogen too near because of the heat.

One thing that someimes gets overlooked is that, although the front if the envelope of an LED will be dramatically cooler than that of an incandescent (and particularly a halogen), the 'other end' of an LED can get very hot (due to 'the electronics'), particularly high-powered ones (which often have a substantial heat sink for that reason).

Kind Regards, John

 

[88v8]

Ah yes, median, mean, average. Interesting observation.

In the attic I have over 200 incandescents, bought in a fit of indignation when the EU gave notice that they would be banned. Since then of course, LEDs have been developed that ape the shape and light pattern, so my incandescents are somewhat obsolete. Never mind.

Thanks for the heads up about temperature. Fortunately the On period is too short for this to be an issue, but it does perhaps explain why these low-value items tend to have alloy casings.

 

[JohnW2]

Ah yes, median, mean, average. Interesting observation.

No sort of 'average' (of which mean and median are easily the most commonly seen) will tell the consumer what they really want to know. I've illustrated the potential problem with medians (which is what the 'life expectancy' figures for bulbs nearly always are), but there would be a similar issue with means - if half fail after '1 minute' (i.e. effectively a zero life) and the other half all fail at 20,000h, the mean will still be 10,000h, but I doubt that many people would be very happy about a 50% chance that one would fail the first time they switched it on (even though the 'expected' {mean} life was 10,000h !).

Medians have an advantage for the manufacturer if they want to 'accurately determine' expected life, since they only have to wait for half of the test batch to fail (e.g. if the median is 10,000h, that {about 14 months} is how long to wait. If they wanted to accurately determine the mean, they would theoretically have to wait until the last one died - which could be many years! In practice, they are likely to estimate even median using much quicker methods than that.

However, it remains the case that I think it would be much more helpful/meaningful to buyers if they quoted the number of hours that, say, 90% of the bulbs are expected to achieve (or, alternatively, the percentage of bulbs expected to achieve at least a certain life - say 5,000 or 10,000 hours).

Kind Regards, John

 

[plugwash]

I suspect many outside lights in domestic properties see very low duty cycles.

I have lights on the outside of the house in case I want/need to do something outside after dark, but in practice I very rarely do. Even the PIR-triggered light that lights the way to my front door is probably only lit for a few minutes a day.

I'm sure in indoor applications and high duty cycle outdoor applications LEDs have a longer life than incandescent, but i'm not at all convinced the same is true of low duty cycle outdoor applications. In such applications I suspect the dominent failure mode is related not to "wear out" but to moisture ingress and complex electronics are likely going to be more more sensitive to moisture ingress than a couple of wires connected to a glass bulb.

 

[Harry Bloomfield]

I suspect many outside lights in domestic properties see very low duty cycles.

I have three high power discharge lights around the eaves of the house, I doubt they have seen more than ten hours use in the twenty years since I originally installed them. They work via a remote control, there mostly for emergencies - my garden is large and very dark, with lots of possible places to hide. Neighbours have called the police in the past, when they fancy they have had trespassers - so the lights were to help address that and for my own convenience.

Apart from those, I have a single small lantern fitting with a 9w LED in it, giving some basic light on my drive to get to the garage and the bins in the blackness. That is powered by a solar clock, on at dusk, off at midnight, winter and summer. This is now on its third LED lamp, since I installed it with LED around three years ago - a corn cob LED, followed by a normal LED, then another normal LED E27. My suspicion is that the LED's run too hot in the enclosed space, limiting their life, so this time around (three weeks ago) I added some holes to the lantern for ventilation - we'll see how the last now.

Other than that problematical LED, the rest of the LED's have been good - originals still working from fitting five/six years ago. One regular place for lamp replacement was the oft used kitchen tungsten downlighters lamps. I originally changed for similar LED's, but neither tungsten nor LED provided a quality light, so rather than due to failure, I replaced them with dome LED lamps 18 months ago, which nicely solved the inadequate light problem.

I have been quite surprised at how much difference a fitting type, rather than wattage, can make to the quality of light in a space. I found that 3.5w LED is quite adequate for hall, stairs, bathroom and downstairs toilet. 18w LED was far from adequate as kitchen downlighters, but 18w was great in a dome light fitting in the same location.

 

[JohnW2]

I'm sure in indoor applications and high duty cycle outdoor applications LEDs have a longer life than incandescent, but i'm not at all convinced the same is true of low duty cycle outdoor applications. In such applications I suspect the dominent failure mode is related not to "wear out" but to moisture ingress and complex electronics are likely going to be more more sensitive to moisture ingress than a couple of wires connected to a glass bulb.

An interesting theory, but the 'bottom line' certainly doesn't correspond with my personal experience. I have quite a lot of outdoor lights, of various types and variously controlled by PIRs and/or time switches and manually. In the days of incandescents, I was (as in my house) constantly having to replace one or other of them. For a good few years now they have all been LEDs and, although I may have forgotten one or two, I don't actually recall ever yet having to replace any of them (which, in view of the relatively low duty cycles, is what one would expect in terms of the 'expected lifespan').

There are at least a couple of considerations which work 'the other way around'. Other than for things like PAR38s (a few of which I have), which are designed to tolerate it, any 'standard household bulb' with a very thin glass envelop will usually instantly 'go bang' if water hits the envelope whilst it is 'on' (and hence hot). Secondly, even with a lot of attention to positioning and adjustment, most of my PIRs result in the lights flashing on and off rapidly in the presence of fairly high winds - and that is likely to shorten the life of an incandescent much more than an LED.

I'm sure you're right that the failure mode of LEDs usually relates to the 'electronics', but it isn't usually all that 'complex' (commonly just a bridge rectifier, a capacitor and a 'fuse' made out of a bit of foil), and I would have thought not all that susceptible to moisture ingress.

Kind Regards, John