Wikipedia:Reference desk/Archives/Science/2014 July 6

Science desk
< July 5	<< Jun | July | Aug >>	July 7 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

July 6

Household LED lighting

Some LED lights for household use to replace incandescents have a similar frosted, globular appearance, and a similar size, to the bulbs they are replacing. An LED drawing about 12 watts is intended to replace an incandandescent 60-watt light bulb or thereabouts. My question is, how many individual light-emitting diodes are inside that LED "bulb"? Or more generally, what are the specifications of the largest single LEDs now in common use? --98.158.139.69 (talk) 07:54, 6 July 2014 (UTC)

There are a large number of incandescent replacements of various quality and using a large variety of LEDs types and brands so I don't know if your first question can meaningfully be answered.

As for the second, do you really mean largest? Or brightest?

One of the brightest single die LED is the latest version of Cree XM-L [1]. However although it can be powerful enough to beat a 60W incandescent, it isn't really intended for the purpose you describe and isn't used much for it. (Note that despite the similar name, the XM-L Easywhite isn't that similar as it's a multidie LED.) See e.g. Cree's own incandescent replacements which definitely don't use an XM-L or even a single LED of any type [2] by any resonable definition.

Then there's stuff like the SST-90/CST-90 etc which are single die and can be brighter than the XM-L although with a much larger surface area [3] [4] [5]. But again, although it can be brighter than a 100W incandescent, isn't really intended or well suited for incandescent replacements and so isn't used much for that purpose.

Of course you can also get massive LED arrays. E.g. BridgeLux make one that can output up to 16k lumens [6] [7] (probably some more details [8] on the test) [9] which is a lot more than any 60W incandescent (probably closer to a 750W halogen). I mean at that level, the LED itself is likely consuming well over 100W. (If you're really interested in size, the Vero 29 LED array is I believe ~29mm in diameter hence the name.)

I doubt this is the most powerful (or largest), many manufacturers have their own arrays, particularly the so called "chip-on-board" [10]. And depending on the quantity required, they may be able to produce something more if really needed.

Chinese manufacturers in particular seem quite willing to do that, even for relatively small quantities, although I'm not sure that the output will necessarily be much better than from a reputable manufacturer, or that it will last long or that it will be very uniform. But see e.g. these "500W" LED www.aliexpress.com/item/500W-Lamp-LED-chip-LED-chip-size-Taiwan-Epistar-45x45mil-47000-5000LM-FREE-SHIPPING/1401410645.html www.aliexpress.com/item/Free-shipping-500w-led-chip-white-color-8-4A-72-82V-3-years-warranty-high-brightness/1421216136.html www.aliexpress.com/item/500W-Cool-Cold-White-6000K-10000K-14000K-20000K-High-Power-LED-Light-45Mil-Chip/1253392737.html (which are ~ 47.6x47.6mm). Closest video I could find is this 200W [11] (running at 170W) but I don't know whether it's a Chinese LED and at 200W it probably isn't that different from the Bridgelux Vero 29 anyway.

In other words, the answer to your second question depends a lot on stuff like, what do you mean by "single LED", "common use", "largest"/"brightest".

If you're imagining use in household lighting, one of the obvious complicating factors is it's not necessarily meaningful to simply think of these in isolation. As shown by the incandescent replacement cases, using a single LED may actually be far more difficult given the specific light pattern requirement. Also it tends to complicate heat management. Of course using too many LEDs may similarly make it complicated to achieve your desired light pattern and otherwise complicate design. On the other hand, as particularly obvious in the flashlight arena, having a bright single LED (whether multiedie or single die) may sometimes be less important than having a surface brightness (i.e. the light output is spread out over a small area), one of the reasons why stuff like the SST-90 is much less popular than the XM-L. (The other is efficiency which of course related to both battery life and heat output and is particularly important in a flashlight.)

Besides all that, I don't believe most household lighting is really that well designed. For starters, until very recently the sort of lights available were very limiting, even more so if you only wanted incandescents or halogens. This and other factors means many people have a small number of bright lights which doesn't give that uniform lighting, when they'd probably be better served by more smaller lights. In other words, what's commonly used for incandescent replacements (or other replacements) may not necessarily be what would be well suited and commonly used if you can actually afford a professional lighting designer.

Nil Einne (talk) 16:42, 6 July 2014 (UTC)

Thanks for that! You asked "what do I mean by single LED". I have here at hand a flashlight that uses 4.5 volts and has a clearly visible array of 16 LEDs. When I look at LED turn signals on cars (presumably using 12 volts) or LED traffic lights (presumably using 120 volts or more), each light similarly displays an array of LEDs, but those LEDs are obviously larger than the ones in the flashlight. So I was presuming that the LED "light bulb" I talked about would contain a similar array of LEDs, using 120 volts, and I wondered how many were likely to be in there. (Yes, I realize that the voltage to the individual LEDs may be lower if they are wired in groups in series; I just mention the voltage incidentally.) Your second link refers to 80 LEDs in a "bulb" of similar style and size, so that's an answer.

As to the second question, I didn't have any specific interpretation in mind, but yours (or something less detailed) is the sort of answer I was expecting. Thanks again. --50.100.189.160 (talk) 05:38, 7 July 2014 (UTC)

You said "incindentally", but it's worth pointing out that LEDs must run at specific (low) voltages which correspond to the photon energy (and thus vary by color). --Tardis (talk) 06:11, 7 July 2014 (UTC)

I tend to think that a single bright light source, pointing in all directions, suspended from the center of the ceiling, is the most efficient placement of lighting, regardless of the light source (LED, CFL, incandescent, halogen, etc.). You want two things with your light, for it to spread everywhere in the room, as evenly as possible, but also to reflect only once before you see it, to limit absorption. Those two goals tend to be in opposition, though. For example, placing lights in corners tends to light the room more evenly, but then you get more reflections off walls and thus more absorption and lower efficiency. If you have white walls the absorption is a less of a factor, than, say, dark wood paneling.

Oh, and also you want to avoid blindingly bright light. This is less of a problem if it's on the ceiling, as you don't normally look in that direction, whereas light sources placed lower in the room need to be pointed towards a wall to reduce this problem, but this again increases absorption. Here technologies that spread the light out over a larger area, like traditional fluorescent tubes, might have the advantage. If they won't blind you, then you don't have to point them towards walls or add a frosted cover to spread the light out (which also decreases efficiency). I think ultimately we may have true LED TVs (as opposed to those sold as LED TVs which are just LCD TVs with an LED back-light) that can be left on white (or whatever color we prefer) when not in use, to light our rooms. StuRat (talk) 14:13, 9 July 2014 (UTC)

People who are immune to addiction

For understanding any biological process it is invaluable to look for mutants that fail to undergo it. It would seem like the same should be true for addiction. Simply doing an online search, I see people who claim to be able to stop smoking cigarettes whenever they wish ( [12] [13]). But I'm having trouble finding the inspiration to think of a way of phrasing the search to pull out serious research papers, indeed if any exist. Has anyone tried to find these people and look for a genetic basis? Wnt (talk) 18:03, 6 July 2014 (UTC)

There is some evidence that people who are genetically predisposed to very low levels of novelty seeking (such as myself) also have a low (but not zero) risk of developing addictions, but the evidence is not conclusive. 24.5.122.13 (talk) 22:12, 6 July 2014 (UTC)

Identifying a unique group of people who are (intrinsically rather than situationally) immune to (substance or behavioural) addiction in order to probe what genetic mutation might be responsible for their immunity is a research idea that runs into ethical and practical challenges. Genome-wide association study (GWAS) is in its infancy and needs a massive number of statistical tests to eliminate false-positive results. One problem is that genes identified in GWAS for drug dependence may be involved either in adjusting brain wiring prior to drug experiences, subsequent to them, or both. On-going research into drug, alcohol and tobacco addiction focusses on Dopaminergic pathways of which the Mesolimbic pathway is believed to be a "reward" pathway. The Substance dependence article gives over 100 references in this field. 84.209.89.214 (talk) 00:05, 7 July 2014 (UTC)

You're absolutely right about the difficulty of GWAS even for far more easily diagnosed disease. I was hoping for some sort of gift from the gods here - an old history that carefully analyzed the opium addiction of hundreds of Chinese nobles in a few pedigrees who are now preserved safely in a crypt, or a lucky hit from the reverse genetics of a particular allele of something like dopa decarboxylase. I suppose one of the countries like the UK that DNA test all arrestees might conceivably cross-reference that with blood data on addictive drugs and then see which required treatment for or showed some visible sign of withdrawal, but as you say that is an ethical line I'm not at all eager to see crossed. Wnt (talk) 14:28, 7 July 2014 (UTC)

There are certainly people who aren't prone to specific addictions, but not being prone to any addictions would be quite rare, if it exists at all. You might have better luck looking at meds which cause a wide range of addictions as a side effect. See Dopamine receptor agonist#Side-effects for an example. If we can figure out what it's doing, that should help us to develop meds which do the opposite. StuRat (talk) 04:09, 7 July 2014 (UTC)

This is a good lead. Searching d3 agonist addiction pulls out some less than compelling information about buspirone being used to prevent reinstatement in rats,^[14] slightly better evidence for effectiveness of an ingredient in Chinese herbal medicine (levo-tetrahydropalmatine),^[15] and various other mysterious effects.^[16][17] But I find these studies frustrating - you're seldom sure what protein is really being targeted, the authors end up trying to explain why certain doses have a small effect and larger doses have none. Numbers are small, error bars overlap, etc. No doubt there is something to it but I can't shake the impression that they're looking at something like the correlation between whacking a mouse with a rolled-up newspaper and how long it takes to solve a maze afterwards. Wnt (talk) 14:59, 7 July 2014 (UTC)

It's not even clear that the current use of the term "addiction" lends itself to such distinctions. During the period when cigarette smoking was still being studied as being carcinogenic, there was a distinction between habit-forming and physiological addictions. Cigarette smoking fell into the habit-forming arena whereas opiates were physiological. Heroin metabolism changes liver function, for example, and could cause death if withdrawn. Nicotine did not. To elevate the dangers of cigarette smoking, there was a campaign to change the definition of addiction. It's current form is something along the lines of "any voluntary behavior that a person chooses to do despite negative social or physical impacts" (paraphrased). So alcohol drinking that doesn't damage social relationships or health would not be considered "addiction" but smoking would if it continues despite the person knowing the risk of lung disease (this redefinition was the objective sought to combat smoking and elevate treatment options). A person with end stage cancer that is on Fentanyl and would suffer severe withdrawal symptoms and possibly death if Fentanyl were withdrawn is not considered "addicted" though they would be classified as "physically dependent." Since the definition is so entwined with political correctness and human behavior, it would be difficult to create an ethical and scientifically meaningful experiment. Gambling, sex, heroin, methamphetamine and tobacco are all potentially clinical addictions with very different mechanisms on behavior. A confounding question like "can someone be addicted to second-hand smoke?" or "why is the cancer patient using 200mg of Oxy a day not addicted, but the 40mg recreational user an addict?" These are political-social constructs of the definition and not a scientific distinction. --DHeyward (talk) 10:29, 7 July 2014 (UTC)

And knowing that behavior is ultimately driven by brain chemistry means that all addiction could be described as physical addiction, such as to the dopamine released by engaging in that behavior. Therefore, the distinction of whether they will suffer measurable physical harm if they go "cold turkey" seems like a more meaningful one, to me. StuRat (talk) 14:49, 8 July 2014 (UTC)

Geomagnetism

How are the earth's internal currents (that cause its magnetic field) generated? Is there a big battery in there?--109.149.211.41 (talk) 18:27, 6 July 2014 (UTC)

Take a look through Earth's_magnetic_field#Physical_origin and see if it answers your question sufficiently. Mikenorton (talk) 18:32, 6 July 2014 (UTC)

The key is dynamo theory. I believe that I have seen a diagram of a cleverly constructed mechanical dynamo, containing no magnets but only a conductive disk and electromagnets powered by the disk, which I was told would produce electricity and a magnetic field when spun above some velocity. But I haven't seen it in quite some time. Wnt (talk) 18:35, 6 July 2014 (UTC)

Not a battery but a giant magnet in pictorial terms. The Earthcore is believed to be composed primarily of iron aka ferromagnetic material. --Kharon (talk) 22:44, 7 July 2014 (UTC)

Hehe, you're barking up the wrong tree. So far as I know (and I'm no expert either) the Earth's core is vastly hotter than the 1000K Curie temperature of iron, i.e. it's not ferromagnetic. It's a conductor, moving, that has electric currents because it is moving through its own magnetic field, and maintains a magnetic field due to its own electric currents... it is, in short, as fearsome a knot of diffy-q's as you would ever cross paths with. Wnt (talk) 00:29, 8 July 2014 (UTC)

If you think the Earth's magnetic field is complicated, look at stellar magnetic field. StuRat (talk) 14:52, 8 July 2014 (UTC)