New 'plasma lamps' to replace fluorescent bulbs, LEDs

Boffins hope to be rolling in cash after finding a new way to use plasma to make light bulbs. The research team, which was funded by the Air Force Office of Scientific Research, has made wafer-thin, flexible sheets of light by putting plasma in an array of micro-cavities in a sheet of aluminium foil, and now hopes to build a …

[dr2chase]

Efficiency/heat claims don't seem right

If you're flat-out more efficient than fluorescent, you claim that. They didn't claim that, they claimed that they were "directed". So they're not more efficient than fluorescent. At the same time, they claim to produce less heat than LEDs. That means more efficient than LEDs -- which are more efficient than fluorescent, and are also directed.

This does not add up. Lumens per watt, please, and talk to us about CRI and spectrum.

Cheaper than LED, and directed, and nearly as efficient, that would be worth hearing about.

[BristolBachelor]

I don't know what the latest is, but when I talked to someone in the industry about this a few years ago (3?), he told me that the end-to-end efficiency of LEDs was still not up there with florescents. This is comparing an LED together with the PSU required to operate it from 240V (quoting Lumens/watt for a device running from 4V doesn't really count when your supply is 240V), and a T4 florescent with electronic control gear. However he did say that LEDs were improving still, so maybe they have caught up.

As for CRI & spectrum, they sound like the same tech as florescent tubes, so probably the same phosphors, so depends on which mixture of phosphors they use; CRI can be anything from only green to pretty good.

[dr2chase]

I think that's old info now

At least from US 110/60 Hz, I can run 9+ watts of LED from 11 watts at the wall (with a somewhat crappy power factor, but it's an old converter). If you're willing to live with welding-arc blue ("cool white") you get about 125 lumens per LED watt; derating by 20% still gives you 100 Lumens/watt, which is well above standard fluorescent. "Good" light (with better CRI) is probably at or above fluorescent efficiencies, with much better quality. White LED spectra are not spiky; white LEDs use phosphors just like fluorescent tubes, but the starting color is a range of frequencies, not a single color spike.

I've played with this some at home -- for me "good" light is a mix of warm/neutral/cool LEDs. You do have to heat sink them a bit to keep them in their happy place. Higher efficiencies at lower currents, so to get more, you need to spend more.

[Giles Jones]

LEDs have a few issues.

Viewing angle, you need to reflect the light to spread it widely and a good lens/reflector costs a lot to develop.

They run on DC voltage, so you need a transformer to drop the voltage and a bridge rectifier to convert AC to DC (then all the capacitors to smooth it out). A transformer isn't very efficient, especially the square ones.

[Duncan Macdonald]

PSU efficiency can by very high - cheaply

A very simple design works well for smaller LED lamps. A mains rated capacitor in series with a bridge rectifier which feeds the LEDs. (For the UK 240v 50Hz mains, a 0.33 microfarad capacitor gives a current around 20 to 25 ma depending on the number of LEDs.)

This design is currently used in a number of the low wattage bulbs available on the UK market.

A dropper capacitor does not dissipate significient power and the poor power factor (high reactive component) does not matter for low power devices.

[A J Stiles]

Rectifiers?!

LEDs are *diodes*. You can actually use them as their own rectifiers, if you can live with only half of them illuminated at once.

The reverse breakdown voltage (when they act like Zener diodes) is greater than the forward voltage drop for all LED chemistries I have seen. You can wire up a nice long series chain of back-to-back pairs (blue or white LEDs need about 3V to light them, and Zener at about 4V, so they will only be dissipating any power half the time) and drop the excess voltage with a capacitor (isolation is not necessary). You can safely leave the leading power factor of such an arrangement uncorrected, as it may help to cancel out the lagging PF of another appliance. (And anyway, on most tariffs, you don't pay for reactive power.)

Also, "A transformer isn't very efficient, especially the square ones" is just plain wrong. Transformers are reckoned to be the most efficient machines ever invented. If you use enough copper and decent steel, anyway.

[Full Mental Jacket]

Area

Neither does the report mention the area of the material needed to produce the same amount of light as the other methods. It won't be yet practical in a house to replace a bulb with a panel a meter square, for example.

[Robert Carnegie]

@Area

A disc the size of an old LP record, that plugs into a regular light socket hanging from the ceiling and itself hangs horizontally, would be useful. Or maybe you'd have some kind of fan structure, or a hemisphere dome, to cast light in more directions. It would replace the lampshade that you usually put around an ordinary lightbulb, it would be aesthetically pleasing in itself.

[Killraven]

Why a bulb?

If this stuff works, we can get rid of the concepts of bulbs altogether. How much light can be generated, then considering the power consumption, of a centimeter-wide strip running completely around the ceiling edges of a room for total lighting and elimination of shadows?

[JimC]

@area

I dunno, a metre square illuminated panel, flush with the kitchen ceiling or almost so would be just fine by me. Better than tubes or worse still the dame spotlights SWMBO has got in there.

Assuming they prodce diffuse light I can imagne home decorators of the future giving your ceiling a multi coloured panel setup with some of the panels being illuminated... come and see my Mondrian ceiling... That might work in the bedroom [grin]

[Richard Harris]

Here's hoping

For the love of God, just make them work, instantly, like the old ones used to.

[Alan Dougherty]

Low watt halogen in old style bulb surrounds? ASDA generaly have them and seem to work fine for me.

[Nigel 11]

LED

So go for LEDs, which also turn on instantly. They will last very many years and will save you several times the (somewhat high) purchase cost in electricity. Light quality not quite as good as halogen, much better than CFL. Buy "warm white" unless you really like a brighter-than-daylight effect.

[Stevie]

Bah!

"Last for many years"

The traffic lights and publicly owned vehicles (school busses etc) have had LEDs for a few years now and let me tell you that real-life experience is that the damn things break down just like proper bulbs do, they just keep giving light - sorta - as individual LEDs in the dozens-in-the-bulb start failing.

The only difference is that because there are a few glimmering elements still working, the bloody local council won't replace the buggers.

I'll take fluorescents because the current ones *do* come on instantly and they provide a decent amount of light until they blow (which hardly ever happens of course until it does). They pollute, of course, and cost a fortune to make compared to Tungsten filament bulbs, as do LEDs. Mercury or Arsenic, what's your poison?

And while it's true the Tungsten refining process is hardly clean of environmentally friendly, the fabbing of the ICs necessary to power-up your fluorescents or LEDs is way dirtier. IC fabbing plants typically leave superfund-class cleanup problems when they finally shut down.

[Ken Hagan]

Spectrum?

If this is coming from a plasma, does that mean the spectrum of the generated light is vaguely thermal, rather than the spikey fluourescent crap we're all suffering at the moment.

(Anecdote: I was at the local hospital having my eyes examined the other day and I was given a colour vision test. Lots of cards with numbers written in coloured dots. Only problem was, the local lighting was all coming from CFL lighting. Unsurprisingly, I scored miserably. It's kinda hard to see coloured dots when there's no ambient light of that colour to be reflected. More surprisingly, the doctors doing the test didn't seem to think the lighting was an issue. I expect we'll be reading reports in a few years time telling us that the nation's colour vision has nose-dived and medical experts are mystified but think there may be a connection with <insert-fad-of-the-day>.)

[The First Dave]

RTFA please - the plasma produces UV light, that is turned into visible light by phosphors: exactly the same method as the good old 6ft tube, TV tubes and even the modern coiled-up CFL tube.

[Bronek Kozicki]

CFLT turns spikes of UV light into spikes of light in visible range. There is attempt to spread the visible output but that's it - the result is still very uneven. You might be right but if there are no UV spikes to begin with, the output could be more even; although I would not expect this to be as good as halogen with appropriate coating applied (e.g. Solux)

[Nigel 11]

But there is some ...

But there is some light of that colour. It's a lumpy spectrum with too much green, but there are no black extents in it, and even if there were, you'd see the dots that colour as dark.

I suspect if there is a problem, it may be the converse of what you have stated: that by making all the red dots (say) look duller than they would be in natural light, and the green dots brighter, it may enable a red-green colour-blind person to read what he should not be able to. (Just like using blue light to make obvious the spy-code dots printed in yellow on white by many inkjet printers).

If you were having difficulty, I suspect you are indeed colour-blind, and maybe you were getting an insufficient amount of help from the crappy illumination, just enough to make you think that you ought to be able to read the thing if the illumination were better. If you have normal colour vision, the "hidden" letter stands out like a yellow crocus in a flowerbed of purple ones. Unless, of course, you can't distinguish yellow from purple!

[Ken Hagan]

@Nigel

There's some, but since your pupil size will be determined by the overall level of illumination, there will be far less non-green than the designers of the pattern intended. That means the patterns will be very low in colour contrast and you'll need much better colour vision to get any of the patterns that don't involve green. Colour vision tests under anything less than proper "broadband" daylight are just silly.

I've had these tests before and I'm not perfect but I imagine the test is designed so that very few people are. That day I got almost nothing past the first page. I've never scored anywhere near that badly before and I can't say I've noticed any trouble outside the rather dark and monochrome surroundings of the eye clinic.

[Pan_Handle]

Liver disease?

[Graeme 7]

Ms Hilton as seen on the Hilton?

I'm quite sure there are loads of reasons not to, but could they develop this further to create low energy, light and easy to make big TVs?

My thinking isn't so much for the consumer but more for large assed TVs on the sides of buildings etc...

My list of possible issues so far are

1. Getting the right colours

2. Response time to pixels

3. Making the pixels - but on the side of a building they wouldn't need to be so small

4. Black - you wouldn't get that from aluminium

Nothing too insurmountable there, I think, but of course there are likely to be more issues I haven't thought of.

[Owen Carter]

Humm. Plasma is a type of Vapour?

Just looked through their website; lots of jargon and pretty pictures. Not one mention of how power is actually supplied to this.

Which is important since the biggest limitation on conventional plasma lighting is not the fact it shines all round (solved by correctly shaped and silvered emitters), but rather the substantial pack of power electronics that supplies the RF frequencies used. Amongst all the guff praising it's advantages I see no lines saying 'just plugs in to a 12v DC supply!', or 'no inconvenient RF-frequency inverters to carry'.

- It will fail to replace LED's in home electronics and cars/boats unless the control electronics let you just apply 3v/5v/12v DC (eg. the standard power for cars and home electronics) while keeping the price very low; though it could do well for some applications if it appreciably extends battery life compared to LEDs.

- It may be better able to compete in AC powered lighting space; especially with super low heat output.

[Why Not?]

oh super

yet more technology specific light fittings.

Like the look it may allow people to design new installations but if they want to take the existing market. Manufacturers we want cheap, efficient, daylight capable bulbs that fit in bayonet cap fittings. Its that simple.

[Nigel 11]

If the light lasts

If the light-source lasts a good many years, there is little point in making it separable from the fitting. If the light-source is compromised by so doing, it's worse than pointless.

This is the case with LED light-bulb replacements. It's hard to effectively heat-sink a GLS bulb-shaped object that's expected to illuminate in most directions. This is why you can't get an LED equivalent to a 100W tungsten bulb. Better, I think, if the LEDs are permanently attached to a metal structure that's both an attractive light fitting and a half-decent heatsink.

What they do need to invent is a ceiling-rose replacement onto which light fittings clip for both mechanical and electrical support, which does not require any electrician or DIY skills to detach and attach a new fitting. Something like lighting track but with only one attachment point rather than a slot. Would need to be a UK or Euro standard so every manufacturer adopted it.

BTW does anyone know where I can buy an LED uplighter to replace one with a 300W linear halogen?

[Will Godfrey]

"What they do need to invent is a ceiling-rose replacement onto which light fittings clip for both mechanical and electrical support, which does not require any electrician or DIY skills to detach and attach a new fitting. "

These have existed since at least 1990 when I installed some fake antique fittings in a listed building. The removable part slides sideways onto the base, and in place, the join is almost invisible.

Don't know how much they cost or where they came from. I was just paid to install them. The units were 'very' heavy with lots of glass bits and bobs. I had to prod around to find joists :(

[Jan 0]

@ Nigel 11

Good points there. I'll second that request for an LED replacement for linear halogen uplighters. I love cool white LEDs and I have lots of 3 - 5W GU10 LED lights, but the only high output LED lights seem to be foot square panels, that won't sit in the upside down dome of floor standing uplighters. I'd like to see high power LED wall uplighters too, it's a bit wasteful, but light bounced off the ceiling is very pleasant.

[Kubla Cant]

"bulbs that fit in bayonet cap fittings" - if only it were that simple.

I have bulbs with BC, small BC, ES, small ES fittings.

I have tungsten incandescent, CFL, "efficient halogen", and I soon expect to have LED, too.

I have candle, golf-ball, tall thin tubes, twisty tubes, GU10s, G4s.

Unsurprisingly, when a bulb fails (which happens all the time, despite the extravagant promises on the boxes - has anybody ever claimed on a light bulb guarantee?) my huge collection of spares doesn't include the right combination of these attributes and I have to go out and buy a replacement.

[Phil Lewis]

Yes, I always write the start-of-use date on the side of my bulbs so that I know when I put them in. I had 3 premature failures (an estimated 60-100 hours) of some £8 CFL daylight bulbs from BLTdirect. They sent me some more at no cost - even after 1 year of purchase!

It is true that the higher power CFLs seems to have a more limited life. The electronic components definitely aren't up to handling the extra heat in many of these. The GE ones used to last for years but now I've noticed that the more recent one use crap components too and also break prematurely.

Oh yes, and there seems to be no correlation between what you pay for a CFL bulb and how long it lasts. I've had cheap ikea ones way outlast their stated life.

On the bright side, I've had some success in repairing the bulbs by taking them apart and replacing components from other similar broken bulbs where the same component survived. Really only worthwhile if they are the more expensive models.

My advice: buy those Eon-subsidised ultra-cheap 11W CFLs from Homebase etc and just use more of them. (I think they were 10p each??)

[Bob Dunlop]

The plasma lights can be dimmed

Let's hope that's more than two steps between full on and full off.

And instant on would be good too.

Hording my collection of hot wires here.

[Anomalous Cowherd]

I look forward to seeing this on BBC's "Have Your Say"...

Smaller, more efficient, non-toxic - they'll be furious, and somehow manage to blame the EU.

[Torben Mogensen]

360 degrees

I don't see this as waste, as most fittings for fluorescent tubes include mirrors to direct the light.

The main disadvantage is that you need larger fittings than for sources that are naturally directed. Conversely, you need larger fittings for directed sources if you want 360 degree illumination.

But I guess that it will be relatively easy to roll the plasma sheet into a tube to get 360 degree illumination. So there is no clear disadvantage unless you need a very focused spotlight, in which case you will need lenses.

[Tom 38]

Eden Park?

Guess neither of them ever watched rugby.

Or plan to sell in New Zealand.

[someone up north]

NO new technology or science

there is no new technology or science here,

it is all old stuff nicely packaged together

just when manufacturing technology is mature!

every one know phosphors is highly efficiency

[Stevie]

Bah!

Because the PC and HiFi worlds have shown us the Way Of Things: Replace complete technologies every five minutes!

Just what we need. Another high-tech lightbulb, full of exotic things you don't want in the garbage.

Christ this is getting old. Fluorescent because incandescents are wasteful. LEDs because fluorescents are polluting. Plasma because LEDs aren't cool enough.

And the price of the lightbulbs goes up and up as more and more people are involved and more patents are lodged and no-one talks about the real costs involved, and everyone lies like stink about their favorite lightbulb. Last week a spokesman/cheerleader for LEDs was put on the spot during a rubbish-the-fluorescent interview on NPR when he said that fluorescents were polluting and then was asked "But don't LEDs have arsenic in them?"

And people are buying crates of incandescents because fat-heads were allowed to pass a law making them illegal next year for Azathoth's sake.

Pass the whale-oil.

[John Smith 19]

*amazing* materials science *if* it works

Shrunk a large *rigid* glass fluorescent tube into a flat *flexible* plate.

Eliminated Mercury from the process.

In principle then can leverage the *decades* of work on phosphors and activator chemistry and tailor the output precisely.

Actually as an *idea* it's surprising this has not been tried *decades* ago.

Of course the driver hardware needs work but that is the same with all current CF technology.

*Cautious* thumbs up on this.

[Duncan Macdonald]

Plasma TV anyone

Hardly new tech - been used in plasma TVs for years.

[Kevin McMurtrie]

Zap!

This is interesting technology but I wonder about longevity. Plasma isn't known for being kind to its surroundings. Can they seal the the micro-cavities so well that a flaw doesn't result in a tiny moving burn that slowly consumes the panel? If so, I have a lot of unreliable CCFLs that I'd like to banish from my house.

Comparisons to LEDs are always a tricky topic because the efficiency drops rapidly as the input power is increased. For a required amount of light there's a tradeoff between adding more LEDs to reduce heating or adding more cooling to reduce the LEDs. 120+ lumens per watt would be trivial if the high cost of LEDs wasn't influencing that tradeoff.

[Goat Jam]

What about LCD panels?

Instead of surrounding the panel with CFL tubes or banks of LEDs you could simply put one of these behind the LCD and shine it through. Considering a large part of the dreaded "bevel" area is necessary to accommodate the lighting requirements of the device, it seems to me that the use of this could go some way towards achieving the legendary "bevel free" screens we all lust after.

[BristolBachelor]

Even better than that; you could have one plasma cell per sub-pixel that you want. Change the phosophors between sub-pixels to Red, Green and Blue, and just switch each sub-pixel on and off. And when the sub-pixels are off, they will be much darker than an LCD pixel trying to be almost dark.

Maybe I could patent the idea and claim back licenses on all those TVs? :)

[Simon Neill]

But...

What if they leak?

[Will Godfrey]

Then you find out if cold fusion is a reality...

Oh, erm, that's a different plasma.

[All names Taken]

Psst - anyone wanna buy a 150 watt lightbulb?

A fiver each.

[All names Taken]

Bring back the filament lightbulb I say.

These new fangled things supposed to last many years never seem to.

I've had two go bust (brown charring near vent in base unit) and all the rest go dimmer, duller and take longer to reach optimal lighting conditions by their second winter.

In either case the damaged or poorly performing energy efficient light sources get binned and so far (over about 5 years) that makes for a 100% fail rate.

[Richard 12]

Tungsten is already back!

Look for Tungsten Halogen - all the warm goodness of a white-hot filament, coupled with much longer life and a higher efficacy* than those old GLS lamps.

In some cases, Tungsten Halogen is actually more efficacious than LED, as many LED light sources have really poor drive electronics** and optics.

* Light-source "efficiency" (Wi/Wo) is hard to measure and not very useful. The common term is "Efficacy", usually measured as: (Useful lumens out) / (Watts put in).

For some reason very few of the energy-efficient light source manufacturers make these figures available.

- Do not believe the name-plate rating on an LED fitting, that's almost always the rating of the diode(s) used and not the power consumed by the entire lamp.

** The PSUs in a lot of domestic (and industrial!) LED lights are very poor - some actually dissipate double the diode power.

Even the good ones draw a lot more current than the rating might suggest, as it's drawn in 'bursts' rather than continuously like resistive loads.

- Household supplies must be getting rather nasty with harmonics these days as almost everything uses a switch-mode PSU. Excluding the oven, I've only got about 250W of resistive load in my house these days. Less than my computer alone...

[bart]

How soon they forget

Oled, Oled . . .

[Richard 12]

Not really. It's just that OLED efficacy never got high.

They're far more useful in displays.

[jennyg]

At last! Hopefully, we will actually see this on our shelves sooner or later.

Here in Australia, we seem to have missed most of the LED revolution in lighting.

And I have not had much luck with the fluorescents. Some have died within minutes of being plugged in.

So much for trying to save energy (because we are certainly not saving money)!