Driverless LED Lighting

Herewith some of the up-and-coming players shipping products based on AC, or driverless, LEDs.

Many of the posts and discussions here are around leading LED technology and how it is changing lighting technology, and most of the technology discussed is LEDs powered by DC current. Since a lot of lighting is connected to AC power, this means most LED lamps need a driver circuit to power the lamp. It has been mentioned a number of times here that the driver circuits currently limit the lifetime of a lamp assembly. There is another approach to constructing LED lamps, which is variously referred to as driverless LED or AC LED technology.

Over the last year, a number of articles have appeared on this approach, such as on EETimes Europe's LED Lighting page in September 2012 and October 2012, and an article on LED Professional last June.

Companies getting a lot of press include e:lumix Technologie AG based in Augsburg, Germany, and Lynk Labs based in Elgin, Ill. In researching this blog I also looked at GT BiomeScilt GmbH based in Ritterhude, Germany, and Enilight based in Shropshire, UK.

Figure 1

e:lumix T8 fluorescent replacement lamp data taken from its x:enia™ data sheet. The bulbs with higher color temperature have significantly higher efficacy. Some outputs have low outliers, such as the 2400 lm bulb at 2700 K and the 2900 and 4500 lm bulbs at 8500 K. The difference in efficacy between low and high CCT is obvious.

e:lumix makes LED chips in a factory in Shenzhen, China, using "high performance machinery from Western Europe" according to its English language website. It does R&D, assembly, and all QA operations in Europe, and invested €70 million (US$93 million) in new facilities in Augsburg last year. In the article on LED Professional, it was also revealed e:lumix engaged AT&S in Austria in an R&D contract, working on insulated metal substrate (IMS) technology.

e:lumix claimed to EETimes it has a patented LED chip technology that is inherently self-protected from overload. In the same interview e:lumix revealed it is using silicon substrates instead of sapphire. In a previous discussion here, we talked about some US DoE data that indicates packaged LEDs on silicon are more than 20 percent lower in cost than on sapphire (at 150 mm wafer size).

Although the e:lumix website refers to applications from street lamps to public buildings, the current product line seems limited. (Note: I have seen some indication in web browsing that e:lumix sells and licenses products largely to Asia OEMs, so there may be more product not branded with the company name.) In Europe at least, the company has a line of T8 replacement bulbs, which it backs with a five-year warranty and sells under the brand x:enia™ T8.

Figure 2

x:enia™ lamp data efficacy vs. CCT. All of the low-efficacy lamps are either 25.8W or 38.7W.

At a color temperature of 8500 K, e:lumix claims it's achieving as high as 120 Lm/W. These bulbs can be used as true "drop in" replacements in fluorescent fixtures -- the ballast and starter can remain in place. However, e:lumix notes, "For dimming, removal of the existing ballast and replacement with a wired bridge may be required." Data from the T8 bulb datasheet is charted in Figure 1.

I charted the x:enia data against CCT to understand the performance better. In Figure 2, you can see that the efficacy is correlated to CCT, and the unusually low values are for two wattages of lamps. This is perplexing since the lamps are essentially LEDs in series, and higher wattages use more lamps. I was unable to learn from e:lumix what causes the outliers. Perhaps one of our readers can explain.

Figure 3

Lynk Labs Tesla™ TS651WXK-12 LED SMD array. This part is made for direct drive using low voltage AC (12V) and attains about 65 lm/W at an output of 55 lm. Image and values (calculated) from thedatasheet. Lynk Labs also offers various single chip packages for direct-drive AC at 100V and 120V.

Lynk Labs positions itself as the leading provider of AC LED technology. It states, "Lynk Labs has built a broad patent portfolio of technology surrounding various forms of AC LED technology from chip level to drive method that simplifies the challenges of integrating LED technology into the existing AC powered lighting infrastructure." Its focus seems to be on providing building blocks to simplify final design and production of AC LED lamps by OEMs. Figure 3 shows one of its products.

GT BiomeScilt says this about its patented technology: "An illumination device comprises a full wave rectifier for converting the AC supply into a direct current (DC). A current limiting diode module, comprising at least two current limiting diodes coupled in parallel, is coupled in series to an output of the full wave rectifier and a light emitting diode module [comprising] a plurality of LEDs is coupled in series between an output of the current limiting diode module and the full wave rectifier."

In a 2011 press release, the company claimed "GT BiomeScilt Centum presents the world's first driverless T8 LED tube." The company started manufacturing in Penang, Malaysia, in 2012. GT BiomeScilt also offers modules similar to those from Lynk Labs, in a product line called Airglow™.

Figure 4

GT BiomeScilt Delta Home® replacement bulbs. Note the linear arrays of the LED chips typical of AC LED luminaires. In its application notes, the company says these bulbs are not dimmable.

You can see the main features of the GT BiomeScilt T8 replacement bulbs in Figure 4, an image from an application note.

The company offers a few lines of drop-in replacements for fluorescent tubes. The two lines targeted at commercial applications offer 100 lm/W efficacy throughout the product families. Its Delta Home™ series is targeted at residential applications, and the bulbs have lower efficacy at 85 lm/W. The product line performance is shown in Figure 5.

Figure 5

Performance of GT BiomeScilt Centum® and Delta Home® replacement bulbs for T8/10/12 fluorescent tubes. Data (calculated) and figures from the datasheet. The Centum line requires replacement of the starter with a special CentumStart® replacement if the tubes are installed directly in existing fixtures with ballasts. The company recommends rewiring for direct supply. In that case, the hot/neutral are connected to pins on each end of the bulb, and the other two pins are directly connected together (shorted), as shown in the right figure above. For the Delta series (which are noted in the product literature as not for the US market), the bulbs can be installed by simply removing the starter, or can be direct wired as shown above. All of the bulbs are available in what GT BiomeSclit calls Warm White, Commercial White, and Daylight. Searching its website reveals that these correspond to CCT of 3000 K, 4500 K, and 6000 K, respectively.

Enilight in the UK appears to offer the most form factors in AC LED bulbs. Its line ranges from screw-in and GU10 (2-pin) base bulbs to high-bay fixtures producing as much as 27,000 lm, along with T8 replacement bulbs and several other options. It wasn't immediately clear whether Enilight is more than a distribution and marketing company. In a press release it notes that it is the distribution partner for Vestel, a Turkish company that makes and sells home appliances, TVs, and lighting products. According to the banner for the Enilight website, it's obtaining (at least some) LEDs from Panasonic. For comparison, I charted the efficacy of its T8 replacement bulbs.

Figure 6

Enilight T8 replacement bulbs. These are the lowest efficacy of the products I looked at. On its website, Enilight says its new line of iViTi® lamps will use Samsung LEDs and have efficacy of 65 lm/W, still far below the competition. For the T8 replacements, although not explicitly stated, because it says the bulbs do not require a ballast or a starter, I believe these bulbs require rewiring standard fixtures for direct AC drive. Enilight is silent on the topic of dimming, and I suspect these builbs are not dimmable.

So we have several manufacturers using various intellectual property (IP) making a range of bulbs. The efficacies are reasonably in line with other LED luminaires available. There is a chance these companies can meet their stated lifetimes (figures vary from 50,000 to 100,000 hours) with a direct AC drive configuration, if they don't suffer failures from voltage spikes and the like. There are a lot of fluorescent tubes out there, making the replacement market a big target. (According to anIndependence LED Lighting whitepaper, there are about 2.4 billion tubes in use in the US, about 600 million are replaced each year, and lighting from fluorescents consumes about 16 percent of the energy in the industrial sector.)

What do you think? Will the AC LED promoters win the day, or be a niche? Please share your opinions on the message board.