How a Philips light bulb uses blue LEDs to produce white light

Figure it out yet? Any idea what’s going on in the image above? Here’s a hint: that’s a Philips LED bulb in action.

There’s no two ways around it: many potential buyers have been turned off by the yellow cap pieces on some LED bulbs. These bulbs might be efficient, increasingly affordable, and last for upwards of 25,000 hours, but what the heck is with that day glow yellow?

Those three yellow pieces on the top of certain LED bulbs are the remote phosphor… and they are not there just because some misguided designer thought they looked cool. These are a crucial part of the design for one reason, which the image above makes abundantly clear: the bulb uses blue LEDs. What the remote phosphor does is convert that blue light to a shade that is more acceptable to what we expect from indoor lighting. In this case it’s 2700K, or warm yellow.

So, as you probably guessed by now, the image shows a bulb that is missing one of its remote phosphor panels. The phosphors that are in place are doing exactly what phosphors do — emit light through the process of luminescence. What this means, functionally, is that the yellow pieces get the blue light and convert it into exactly the shade of light that Philips is looking for.

As for the “remote” part, that just means that the phosphor is not connected to or built into the LEDs. This is done for a number of design reasons, including the placement of the highly directional LEDs and to prevent multiple shadows from forming.

When one of the yellow pieces is removed — be careful of your eyes if you try it at home — you can clearly see that the light from that third of the bulb is a sort of royal blue. The rest of the light goes through the phosphor and is emitted as a shade of white.

And those neon yellow plastic pieces? When the bulb is on you can’t see them at all, which is why the remote phosphor isn’t as big of a problem as you might think when seeing a bulb on the shelf.

As you can see in the smaller image, Philips’ AmbientLED uses three sets of six blue LEDs, plus the remote phosphor, to generate its light. The company’s LED Prize bulb, in order to take efficiency to the next level, uses a different mix. That bulb uses three red LEDs, two blue, and then another shade of blue. It’s a more complex operation but it helps the L Prize winner get to an excellent 94 lumens per watt, where the normal model gets about 65.

And why use blue LEDs at all? Why not just use LEDs that produce a color that we like? This gets pretty complex, but basically it comes down to something predictable: blue LEDs are more efficient than other colors.

If you were interested in this post, you might want to check out my book, LED Lighting: A Primer to Lighting the Future. It’s available from O’Reilly (DRM-free), Amazon, iBooks, and others.

Source

The fuel cell that turns poop into power

Bioengineers at Oregon State University (OSU) have developed a microbial fuel cell that can treat waste water — and generate significant amounts of electricity at the same time.

The microbial fuel cell (MFC) works much like a normal fuel cell, but it uses waste water as a fuel (instead of hydrogen or ethanol), and specially-crafted bacteria act as a catalyst (instead of platinum). In the case of this fuel cell, developed by Hong Liu and her OSU colleagues, waste water comes into the fuel cell (at the anode), and bacteria oxidizes the organic compounds, producing spare electrons that flow to the cathode — creating electricity. The anode and cathode are separated by a membrane that only clean water can pass through, purifying the waste water.

All told, the MFC produces two kilowatts of power per cubic meter of bioreactor volume — not a huge amount, but apparently 10 to 50 times more power electricity than other MFCs on the market.

A microbial electrolysis cell — not the same as OSU’s fuel cell, but the concept is the same

Compare this MFC with conventional waste water processing plants, which for almost 100 years have used activated sludge to clean the water. Activated sludge was developed way back in 1913, and it also uses bacteria to break down organic compounds in waste water — but unlike the MFC, the process consumes a lot of electricity, rather than producing it. According to the OSU, around 3% of the United States’ power consumption is spent on treating waste water.

 

Yup, it’s poop.

Moving forward, Hong Liu says the next step is to scale the fuel cell up to commercial use. Instead of waste water treatment, which deals with vast volumes of effluent, Liu says the first commercial pilot study might be in a food processing plant, which also produces a lot of delicious waste water. The other problem is cost: For now, it’s very costly to build high-volume MFCs — but OSU reckons that the initial outlay can eventually be brought in line with activated sludge.

 

Suffice it to say, if the treatment of waste water actually produced electricity, we’d be onto a very good thing indeed. Not only does the MFC provide a very green source of electricity, it would also be hugely helpful to developing countries that don’t have reliable access to electricity or lack water treatment plants. Even if MFCs can only produce 1% of the US’s power requirements, that would still be a total swing of 4% — a truly vast amount of money/oil.

Research paper: DOI: 10.1039/C2EE21964F

Source

Zeolite thermal storage retains heat indefinitely, absorbs four times more heat than water

Hold onto your hat/life partner/gonads: Scientists in Germany have created small, zeolite pellets that can store up to four times more heat than water, loss-free for “lengthy periods of time.” In theory, you can store heat in these pellets, and then extract exactly the same amount of heat after an indeterminate amount of time.

Zeolites (literally “boil stones”) aren’t exactly new: The term was coined in 1756 by Axel Cronstedt, a Swedish mineralogist who noted that some minerals, upon being heated, release large amounts of steam from water that had been previously adsorbed. For the last 250 years, scientists have tried to shoehorn this process in a heat storage system — and now, the Fraunhofer Institute, working with industrial partners, has worked out how to do it.

I will try to explain how this works, but the science is fairly complicated: When Fraunhofer’s zeolite comes into contact with water, a chemical reaction adsorbs the water and emits heat. When heat is applied to the zeolite, the process is reversed and the water is released. Because the heat is locked up in the chemical structure of the zeolite, the material never actually feels warm — which is why this is a “loss-free” storage method.

These two processes can be kept separate — so first you charge the balls up with heat, and then later you can just add water (!) to release the heat. This reaction occurs all along the surface of the zeolite — and because zeolites are porous, a single gram of the material has a surface area of 1000 square meters (10,700sqft). It is for this reason that Fraunhofer’s zeolite can store up to four times more heat than water.

While the hydration/dehydration process is well understood, the main technical challenge was building an actual heat storage system. “First we developed the process engineering, then we looked around to see how we could physically implement the thermal storage principle — i.e. how a storage device has to be constructed, and at which locations heat exchangers, pumps and valves are needed,” says Mike Blicker, the group manager. As you can see in the picture on the right, the setup is fairly complicated. The team has now successfully built a transportable 750-liter storage tank, which is currently being wheeled around Germany to test the storage system in real-world situations.

Moving forward, this could be huge news for almost every technological and industrial sphere. Currently, there are very few options for storing heat other than water, which can’t store much heat for a given volume, and it loses heat relatively rapidly. Power plants, biogas plants, steel mills, factories — these all produce vast amounts of heat that could (and should) be reused. They wouldn’t even have to be used on-site, either: charged-up zeolite balls could be distributed to nearby homes and offices. In the future, Blicker suggests that we could eventually replace house water tanks with zeolite systems, too. “It would be ideal if we were able to devise a modular system that would allow us to construct each storage device to suit the individual requirement,” says Blicker.

Personally, I’m hoping for a module small enough to put inside each of my seven computers. I wonder if that’ll be enough to heat my shower in the morning…

Read more at Fraunhofer, or check out Microsoft’s solution to waste heat: Data furnaces

Source

Philips’ L Prize bulb is efficient, expensive, and available

Perhaps you recall that, a few years ago, the Department of Energy put together a competition called the L Prize. This was designed to promote LED lighting by giving a big cash prize, and some great PR, to the lighting manufacturer that was first able to produce a bulb capable of meeting certain stringent criteria. In August 2011, lighting giant Philips won part of the L Prize and received a cool $10 million for their trouble. What they produced was an “A” bulb, which fits in a standard household socket, yet is light-years ahead of the incandescents that it’ll be replacing. That bulb is finally available for sale and one can be yours for the low, low price of $60.

The L Prize winning bulb (yes, it’s an self-congratulatory name, but it’s better than the 10A19/LPRIZE-PRO/2700-900 DIM) might be initially expensive, but it has some outstanding specifications that will make it cheap over the long term. Consuming just 10W, it’s able to produce 940 lumens, making it a good deal brighter than your average 60W, and it should last for 27.4 years if you use it three hours a day. The warranty alone is good for 36 months or 15,330 hours, whichever comes first. It’s a 2700K bulb, which means it’s a “warm white” shade, not like those ugly blue-tinted CFLs you bought a few years ago.

If you’ve purchase lighting recently you might have noticed the Philips AmbientLED 12.5-Watt A19, which looks a whole lot like the L Prize winner, but is available for $24. The L Prize bulb is immediately recognizable because of the three neon yellow sections (the remote phosphor) on its top, but past that the two look a whole lot alike. What’s sort of difference does the extra $36 get you? The AmbientLED consumes 12.5W and produced 800 lumens, meaning that the more expensive bulb is over 33% more efficient in terms of lumens-per-watt. Of course the AmbientLED is a perfectly fine bulb so you might think of this as paying more for the performance model, the way you would with a luxury sedan.

Sooner or later some rebates will kick in and this $60 bulb will be in the sub-$30 range, plus its efficiency gains will trickle down to the more affordable models. Until then, it’s a nice bulb and a (relatively) cheap thrill if you are into lighting or the energy market.

And when I write that it’s a cheap thrill, I’m speaking from experience — I actually purchased one and have been using it. I wasn’t able to do much scientific testing but I did find it to be noticeably brighter than a normal 60W-equivalent bulb. I measured the power consumption to be about 8W, so either my testing tool is off or that 10W number is just reference point that none of the bulbs will go over. I’ve just started using it though, so I won’t fully know the color temperature (it seemed a bit cool) or power consumption until the bulb has had some time to break in, which normally is considered have happened by the 1000 hour mark.

This tiny piece of lighting history is available today from places like Bulbs.com and EFI.org. It will be more widely available on Earth Day (this coming Sunday).

Source

How to cool a PC with toilet water

 

Hot on the heels of news that Google uses toilet water to cool one of its data centers, it has emerged that an enterprising hardware hacker had the same idea some seven years ago. As you will see in the following pictures, though, Jeff Gagnon’s computer is much more than a toilet-cooled rig — it’s a case mod tour de force.

For a start, it doesn’t actually have a case — instead, it has a lovingly hand-cut and -painted (while topless; see below) Tux penguin backboard, which the motherboard, hard drive, and power supply are all nailed to.

Then there’s the CPU waterblock, which has been handmade from a lump of copper and, I presume, an arc welder or a soldering iron. But where does that tubing go, I hear you ask? Where’s the water reservoir, the pump, the radiator?

Well, it just so happens that there’s a toilet on the other side of the wall — and so Gagnon, who is obviously a fan of doing things as gracefully as possible, simply drilled through the wall and put a water pump in the cistern of a toilet. There’s no sign of a radiator: It looks like it’s actually just an open system, with water freely flowing in and out of the cistern. Here’s a quote from Gagnon, on the actual functionality and efficiency of the system:

The more we go to the toilet, the more the CPU heat is going down. With this setup, the CPU heat is about 19 Celsius — very nice! After a weekend out of home (not flushing the toilet for a while), the toilet reservoir is warm and the CPU heat is about 29 Celcius…

The completed system runs ClearOS, a small and medium business-oriented Linux distro based on CentOS and Red Hat. As far as we can tell, the server is affixed to the wall of an office somewhere in Quebec, Canada. For more photos of the toilet-cooled, caseless masterpiece, click through to the next page.

Short of computers that are immersed in Fluorinert and cooled with liquid nitrogen, this is probably the most extreme case mod we’ve ever seen. If you’ve seen anything more extreme (or insane), do let us know.

A better shot of the water cooling pump in the toilet (though I’m not sure where the heat exchanger is)

Real case modders hand-solder motherboard power cables to the PSU…

Likewise, real case modders paint their Tuxes while topless

The finished computer, in situ

 

One more photo of the water pump in the toilet
See the full gallery of images at the Clear Foundation website

MIT stacks solar panels like pancakes, increases their power output by up to 20x

What’s better than one pancake? A whole stack of pancakes! Using the same logic, a team of MIT researchers have stacked a bunch of photovoltaic solar cells together to produce up to 20 times the power output of conventional solar power installations.

Normally I’m the first to drop my jaw in awe at MIT’s latest and greatest innovations, but this one really is a bit of a no-brainer. Basically, photovoltaic cells themselves aren’t all that expensive — according to MIT, they’re only around 35% of the total cost of a solar power installation. The main issue with solar power (and its main cost) is its low energy density, and thus the sheer surface area required to generate a sizable amount of electricity. This is why you need to cover your whole roof with cells to power your light bulbs, and why solar power plants would have to occupy tens of square miles of desert to produce as much power as a nuclear power plant.

To combat this issue, MIT has built 3D stacks of photovoltaic cells. These have the same footprint of a conventional, flat solar power setup — but as you can see in the picture above, the total surface area is much, much larger. The team built a variety of 3D designs, including a cube, and in all cases they produced between two and 20 times as much power as a flat panel. The most interesting facet of this discovery, though, is that these 3D stacks produce lots of extra power whenever the sun is near the horizon, i.e. in the morning, evening, winter, or at latitudes far away from the equator. With conventional, flat cells, it’s hard to capture low-angle light, but with an accordion structure (as pictured) the relative angle would be closer to 45 degrees.

Squeezing more power out of low-light periods is a huge, huge deal. Beyond the large footprint required, the main issue with solar power is unreliability (cloud cover) and natural fluctuations (winter, night time). Solar installations produce most of their electricity in the middle of the day, but that’s when you least need electric light or heat. As a result, you need to store the energy for later use, which usually involves a big battery of some kind. If these 3D solar cells can capture enough light that they can produce reliable electricity in the morning, evening, and winter, then this could be exactly the kick-start that solar power needs — especially when you combine it with the ion cannon that cuts photovoltaic costs in half

Source

Canon 5D Mark III full-frame DSLR: 22 megapixels, automatic HDR, costs $3,500

It’s feature-and-price creep, but who’s complaining: The Canon 5D Mark III digital SLR comes with a 22.3 megapixel sensor, 6 frames per second shooting, and a raft of gotta-have secondary features such as automatic HDR imaging and in-camera picture rating. The full-frame body is $3,500, up $500 list over the Canon 5D Mark II (both with no lens), about $1,000 above the current street price. Some much-improved accessories are jaw-droppers: a radio-control Speedlite 600EX-RT for $630 and a wired/wireless WFT-E7A file transmitter for $850.

For the last decade, the Canon 5D Mark I, II and now III have been the pro’s backup to Canon’s main battle tank DSLRs costing $5,000-$7,000. If you own a Canon Rebel, even three grand for a camera before buying a single lens seems breathtaking. If you’re a professional shooting portraits, commercial work, or weddings, or an enthusiast looking to expand your photographic reach, the Mark III porridge is just right, neither too hot nor cold. With 6 fps shooting, it’s even a passable action camera with much improved autofocus.

Nikon fanatics will gloat because their baby, the full-frame Nikon D800, rings in at 36 megapixels and costs about the same as the Canon 5D Mark III at $3,000 street. At a plodding 4 fps, the D800 is not the camera to capture greyhound racing, though. There’s even an enhanced sharpness Nikon D800e($3,300), without the low-pass filter. Canon is banking on most photographers not needing Nikon’s 33MB images. The camera should be an artistic tool, not a source of glee at Seagate and Western Digital every time you click the shutter.

The Canon EOS 5D Mark III is also the sum of a dozen nice little features. Press the Rate button while you’re waiting for the next pitch or for the bride to adjust the groom’s tie, and you can rate the previous photos 1-5 stars. There are dual memory cards, one CF, one SD. (Nikon may be one up on Canon here with the newest memory format,XQD, on some of its new cameras. XQD is faster and more rugged, if not yet standard.) Standard ISO goes up to 51,200, almost 10 times more than the Mark II. Canon says there’s a two f-stop improvement in low light quality. Want to take a high dynamic range (HDR) photo that stitches together three different exposures? The Mark III does it in-camera and also saves the separate images. A quiet mode won’t startle children, ministers, or judges and still allows 3 fps shooting.

Autofocusing is improved (Canon claims by 50%) with a 61-point auto-focus screen and Digic 5+ processor. Canon appears to be gearing its autofocus toward f/2.8 and some f/4 lenses favored by pros; not all AF points work with slower lenses (f/5.6, f/8) but the camera will eventually find focus. This is important: Autofocus speed is a hot button with some action shooters who abandoned Canon for Nikon five years ago; Canon has since made autofocus and low-light quality priority. Water and dust sealing is improved but you still need to be careful changing lenses outdoors and use a rain shield when it’s raining not drizzling.

Canon says it has stepped up the video capabilities; the 5D Mark II has been perhaps the most popular DSLR video camera among serious photographers. The 5D Mark III allows for two methods of SMPTE time-code embedding for ease of synchronizing multiple-camera footage and audio in post-production. Not that anyone expected it, but this is not the next generation of high-def video with 4K imaging, even if the sensor itself is capable of 5760×3840 pixels at max resolution shooting still photos. The Mark III shoots up to 1920×1080 progressive (1080p) at 30 fps.

Some features you might like aren’t there. The rear LCD is brighter but not hinged for over-the-head shooting. (Some sub-$1,000 Canons do that.) USB is 2.0 not 3.0. Canon refuses to integrate even a half-fast GPS module that would be always available if not quite as accurate, thus ignoring pros whose stock-agency shots would sell better if clients could search by location. Instead, the shoe-mount GPS Receiver GP-E2 will be available for $390 and injects not just latitude-longitude but elevation, time of day, and the compass heading. In theory, you could relocate the exact GPS coordinates of the photo (not camera) by adding in the focus distance-to-subject.

What else? The new Speedlite 600EX-RT at $630 blows by the Speedlite 580 EX II ($450 street) with wireless radio control (new), a bit more power, beam adjustment for 20-200mm lenses, and legacy infrared control. Infrared triggering of multiple flashes works fine indoors, not so well outdoors in the sun. Canon says the radio transmitter works out to 100 feet and controls up to 15 additional flashes in five groups, allowing for automatic flash metering; a Linked Shooting feature lets you remotely trigger 15 remote cameras simultaneously. Built-in radio control is going to narrow the appeal of the Pocket Wizard brand remotes used now by pros, especially since RF interference issues with previous Canon (but not Nikon) strobes drove Pocket Wizard users crazy. Where Canon radios work to 100 feet, PWs are good to 1,000 feet, but the mainstream PWs can’t auto-control exposure. The two brands are apparently not compatible. As a preemptive measure at Canon and also at cheaper competitors, Pocket Wizard just introduced an improved, cost-reduced Pocket Wizard Plus III for $140.

The new wireless file transmitter also has a gigabit wired connection. The new battery grip works well in portrait mode for shooting, well, portraits. If you don’t want a $630 flash atop your camera, you can swap in Speedlite Transmitter ST-E3-RT that does all the wireless controlling (but no flashtube) for $470, meaning for $160 extra you could get an entire flash.

If you’re looking at higher-end Canon DSLRs, meaning those selling for $1,000 and up, here’s the pecking order:

The Canon EOS-1D X, due this spring at $6,800, is the workhorse: 18.1-megapixel full frame sensor, 5184×2456 pixels, blazing 12 fps shooting, 400,000-count shutter life, and serious weather-sealing. It does full-HD video, uses two CF card slots, and will work with all the accessories just announced for the 5D Mark III. There’s no built-in flash, no GPS.

The Canon EOS 5D Mark III, out this month, is also full-frame, has the highest resolution of any Canon camera, and a few unique features such as HDR imaging, and separate CF/SD slots. There’s no built-in flash. Shutter life is rated at 150,000 cycles, and the dual-battery grip BG-E11 runs $490. You can buy two 5D Mark IIIs for the price of one EOS-1D X. Canon will also sell a 5D Mark III kit with the workhorse 24-105mm f/4 lens at $4,300, saving you $350. While it’s only $500 costlier on list price, the outgoing 21.1-megapixel 5d Mark II at street is selling for as little as $2,400.

The Canon EOS 7D is an 18-megapixel crop-sensor still and full-HD video camera with a 1.6X multiplier, so a 200mm telephoto lens is effectively a 320mm. It shoots 8 fps, which is good enough for most users, and ISO up to 12,800 (decent, not world-class). There’s a single CF slot. Autofocus isn’t as sophisticated as on the 5D Mark III or 1D X. Shutter life is rated at 150,000 clicks. There’s a built-in flash and wireless (IR not RF) controller for Canon strobes. The 7D, like the sub-$1,000 Canon Rebels, suffers from a lack of highest-quality wide angle prime (fixed focal length) and zoom lenses. For instance, the Canon EF-S 10-22mm ultrawide zoom f/3.5-4.5 ($800) has the same coverage as the Canon EF 16-35mm f/2.8L II ($1,500) zoom on full-frame cameras, but build quality and low light capabilities aren’t comparable. List price of the 7D is $1,700 and street price has been as low as $1,400; either way you can get two 7Ds for the price of one 5D Mark III, or four 7Ds for the price of a 1D X. Accessories are cheaper, too: $175 for the BG-E7 battery grip, for instance. The 5D Mark III, Canon EOS 60D, and 7D (but not 1D X) take the same LP-E6 batteries if you mix and match cameras.

Read more at Canon, and start saving up.

Source