Posts Tagged ‘Home lighting’

In my previous entries regarding the Cree LR6, I’ve noted the good and bad sides of the product in some detail. I’ve noted my dissatisfaction with the brightness of the diffuser, which has caused me to first apply plastic trim rings to add a little cutoff, then later, to simply not turn them on. Dimming performance over the years has been disappointing as well. No dimmer I have found has dimmed them satisfactorily, most cause them to flicker. The latest from Lutron, designed specifically for LED/CFL sources, did not fix the issues, so I simply gave up. Rather than remove these expensive retrofits ($65.00+ each), I chose to do what many of us do when caught in a  quandary – stopped using them at all. Estimating these were not used more than an average of 1 hour a day for the last 4 years, total operating time is less than 1,500 hours. I’ll give them 2,000hrs, assuming that when they were first put in place, I used them more than I did as we grew tired of their glare and flickering under dimmer control.


LR6 installed in kitchen 2010. Note that all of the units appear to be of the same neutral color and uniform to one another. The color from the fixtures initially was quite nice, even though the brightness was always objectionable.

It became quite obvious that Cree chose to design this product to excel in efficacy, at the cost of aesthetic appearance and visual performance. Over the years they have made excuses for this by stating it was their intent to produce high angle brightness to “fill” rooms with light. This is the old “Volumetric Lighting” dodge that others have used (Acutity uses this term now) as an excuse for  the appearance of brightness in diffused optics, stating it removes the cave effect and improves visual brightness in a space. I’m not a proponent of this, nor do I agree with its basic logic. Visual brightness is uncomfortable, reduces visual performance, and creates a sense of institutional aesthetics to space. I see no reason to use high angle brightness throughout a space to light the perimeter walls, which is what is proposed in the volumetric argument. If you want the walls lighted, then use wall washers or perimeter wall lighting. In the rest of the space, brightness above 45 degrees must be controlled to reduce glare and veiling reflections on work surfaces and displays. The argument that this is no longer necessary due to the anti-reflective nature of modern displays is simply a marketing game of distraction. Veiling reflections can appear on any surface, including smart phones, magazine pages, desk surfaces, even ink on mate paper. There is no case to be made for accepting high angle brightness, period. Volumetric lighting is a sham that should be called out for what it is – lazy design targeted at achieving high efficacy (fixture efficiency) by eliminating proper optical and brightness control features that negatively impact total lumen output.

In an effort to reduce the offensive brightness of the LR6, I created a simple plastic ring trim for the LR6, which was presented in the 52 in 52 project back in 2010. It helped, but was a band-aide job to be honest.

That all said, I have recommended the LR6 retrofits to others over the years, specifically in commercial applications where the brightness issues would be less irritating or problematic. They work well in commercial truck-stop restrooms, and the local Walmart toilet, for example. I’ve also seen them in retail dressing rooms, with some success, and a couple of gift shops in hotels, as well as a few hospitality applications. Some look fine, and the product well suited, others have been awful. That’s the nature of products sold to save energy first – at all costs.

Now to the Fail
This leads to the next installation of the LR6 saga. After less than 2,000 hours of on-time, I noticed s decided shift in color of two of four units. One shifted to pink, the other to green. The remaining two remained steady and neutral. This condition worsened so quickly that it appeared to happen day-by day. Considering how little these fixtures were used, the change and rate of shift was somewhat alarming.


LR6 in Kitchen 2014. Note shift to green and pink in 2 of 4 units after just 2000 hours in service.

Now, anticipating there will be an argument made that the plastic glare shields I added have caused the issue, I tested the fixtures with and without the shields at 4 points within the fixture, including one at an LED in the array center using thermal probes, and found no difference with or without them, so am certain these had nothing to do with this issue.

Note also that the difference in image color is due to two factors. First, the original image was taken during the day (see the window), so there was a fair amount of daylight added to the space. Second, the camera used to take the first image has long been replaced. The second image was taken with the camera in my Samsung Note 3. This changes nothing in regard to the results being shown. The green and pink shift issue is very real. In testing the color performance of these fixtures, from the baseline of 2700K when new, the neutral fixtures were now 2750, the pink ranges from 2249 to 2399 depending on how long it is left on, the green is stable, but at 3350. The neutral fixtures were now 95CRI (92 CRIe / 94CQS), the pink 91CRI (90CRIe / 91CQS), and the green 87CRI (81CRIe / 89CQS), while the Du’v’ was as expected – the neutral was +.001 (anything under .002 is acceptable, the pink  -.0045 to -.0061, and green +.008, which, combined with the CCT shift, explains the movement from neutral to green and pink hues. In other words, the difference in color was not just subjective, it was measurable and significant, with the two shifted products falling well outside the ANSI bin standard for the 2700K CCT baseline. It was also interesting to note that after the neutral units had also shifted cooler, indicating they were likely on their way toward similar failure as the other obvious pair.  This was a disappointing outcome from a $275 investment and completely reverses my impression of this product from exceptional to unsatisfactory.

The shift from the black body line tells the whole story with the pink and green movement clearly visible.

The shift from the black body line tells the whole story with the pink and green movement clearly visible.


Overlay of the three colors shows how the original white light characteristic (the green line) is shifted both up and down to produce the pink variation (yellow line) and the green (red line).

The differences between the outputs is also clearly visible when compared against the 2700K standard center for the McAdams Ellipse, showing a variation well outside 3 steps, making it clearly, objectively verifiable as being a visible difference.

The differences between the outputs is also clearly visible when compared against the 2700K standard center to the McAdam ellipses, showing a variation well outside 3 steps, making it clearly, objectively verifiable as being a visible difference.

The Fix is In
Rather than replace these products with the newest version of the same – which to me has the same glare and brightness issue – regardless of the new lower price, I decided in stead to remove them and replace them with a more conventional PAR20 retrofit lamp from Philips. This allowed me to re-install the baffles in the downlights to create a proper appearance to the fixtures. I chose this as an opportunity to install new nickel silver baffles to fit the metalwork in the kitchen. The new lamps use 8W, instead of the 13W of the LR6, cost only $14, and produce more light on the surfaces I care about – the counter tops. Better still, the glare bombs are gone. While the end product efficacy is a bit lower due to the baffle absorbing some of the light, I just don’t care. I can now use the overhead lighting again. Hazah!

New Philips PAR20 LED lamps in new nickle baffles ends the glare, reduces offending brightness and reduced energy consumption by 45% - and eliminated the color shift issue.

New Philips PAR20 LED lamps installed with new nickel baffles ends the glare and offending brightness while reduced energy consumption by 45% – and eliminated the color shift issue.

Still Not Perfect
One issue remains, however. Dimming of the new “Dimmable” retrofit lamps remains pathetic. The new lamps only dim to about 50% of full brightness, using the newest Lutron wall box dimmer designed for dimming LED/CFL retrofit style products. However, at least don’t flicker while doing that, so I’ll live with it. The reduction in fixture brightness, coupled with dimming to whatever level they can deliver is a big improvement in both visual comfort and overall appearance in the space. Concerned that we were placing these lamps in a downlight, when the label for the product specifically states “not for fully enclosed fixtures” I checked operating temperatures and found that in the large, ventilated 6″ recessed can in an un-insulated space, their was no indication these lamps are under any stress. I am also not a fan of 2700K LED color, as it tends to look a bit artificially yellow-orange, but it is what it is.

Still Not a Fan of the Retrofit After all these Years
I remain solidly against retrofit solutions, as I believe they present compromises in performance and appearance, and apparently in reliability. Yes, I recognize it is the cheapest and easiest solution to deploying LED technology and saving energy. So are adding dimmers and automatic lighting controls, often for a lot less money.  While we spin our collective wheels selling ourselves short with cheap fixes, the actual potential and performance liabilities create reputation issues for the technology that will take years to overcome. The CFL died a PR death more than a technical death, LED retrofit products have the same potential for harm to the ultimate goal of integrating LEDs into the mainstream.

I’ve been using this kitchen as a test bed for applying retrofit solutions in a typical down-light condition. This is the last time we will apply this approach, as I am still unsatisfied that this is the best looking, or best performing solution. Now that down-lights can be found for under $100, with proper optics and external, truly dimmable drivers, my next round will be to do what I know to be the best solution – replace the fixtures completely and be done with it. For the meantime, I can now at least enjoy the kitchen lighting again.


The recent article: LED Bulb Efficiency Surges, But Light Quality Lags states very well the findings of the DOE and others reviewing LED retrofit lamp performance. While well stated, there are severl missing dynamic issues in the conversation that need to be included if LED is to overcome the failure of the CFL to capture the consumer market it so desperately seeks to dominate.

While efficient, there has been no great interest in the consumer market to lamps with poorer quality at higher prices.

While efficient, there has been no great interest in the consumer market to lamps with poorer quality at higher prices.

The CFL lamp has failed in the consumer market for these reasons:

  1. Light quality is poor in comparison to the far cheaper incandescent lamp. This includes color quality, distribution (photometric) pattern is poor (flood type products)
  2. Appearance and fit of the product into existing fixtures – i.e. ugly to look at, stick out of fixtures, create dark spots in shades and fixture diffusers, etc.
  3. They did not last as long as advertised. When switched frequently, the life of a CFL screw base product can be shorter than a long life incandescent. In outdoor cold climate environments, some fail within a few months. In down-lights and enclosed fixtures most fail even more quickly.
  4. They cost too much compared to incandescent of higher quality
  5. They save some energy, but have so many other liabilities the consumer does not take this seriously.
  6. Flickering starts, flicker under dimming, and 120Hz strobe effects from cheap ballast designs
  7. Slow to warm to full brightness – often taking longer to get up to full light than many products are on for in many rooms (pantry, closet, hallway, etc.)
  8. Mercury disposal concerns for some


To set things off on the proper foot – I do not like complexity when it is not necessary. I’ve noted many times that if energy were free and maintenance was not a consideration, the perfect light source is the tungsten halogen lamp. This technology delivers a very attractive white light, is very easy to control, provides optical focus, and is as simple as it can get. The low voltage versions of this technology are equally attractive, accepting that transformers were a horrible thing to tag onto an otherwise neat little light source. I have made hundreds of lights using halogen lamps, mainly 12V versions, starting back in 1985. It was my go-to light source. I still have boxes of transformers and sockets, acquired over years of making lights for myself and others.

Applying LEDs in efficient lighting designs is no more complex than use of any other source, just more productive.

Applying LEDs in efficient lighting designs is no more complex than use of any other source, just more productive, and attractive than CFL or other conventional “efficiency” improving sources.

That said, there is no escaping that energy is an issue, and maintenance is a pain. The cost of operating halogen technologies is simply impossible to bear. This is why we have HID sources with all their ugly liabilities, and the fluorescent lamp.  While I get HID technology as a giant super-power halogen device, it has always been a clumsy, heavy, messy engineering gadget that sets aside the art of lighting for raw lumen energy. Fluorescent lamps have are a source you are forced to live with, in an uninspired, just-get-lumens-in -the-box sort of way. There is very little to love about their scale, lack of focus-ability, ballast hardware, delicate tubes, and ghastly glow. I’ve specified millions of these lamps into existence, wishing every time there was a better way. I never made a single art light using fluorescent lamps, not because itsn’t possible, but because I never liked them enough to give them that part of my time.

The emergence of solid-state lighting, specifically LEDs, hit me in two ways. One, I get the small controllable source I had with 12V halogen. Second, I get the efficiency and raw lumen potential of fluorescent that made it indispensable. Because of this, the last time I made a light using halogen technology was in 2004, and that product was converted to an LED sources in 2006. For my own use, every halogen light I made from 1993 to 2004 still in use around the house, has been converted to LED. Every new fixture made since 2005 has incorporated an LED light source, without exception. I do not use retrofit lamps. I either tear down and rebuild products to utilize LEDs properly, or design them around LEDs in the first place. (more…)

The Replacement Dichotomies

Side One: It is acceptable, if not desirable, for LED luminaires to be replaced at the end of their service life. This is a common position among a wide range of LED product manufacturers. They make the case that extracting performance and costs from LED products requires a level of integration that cannot be accomplished using modules. This further forwards to concept that modules restrict design freedom, that integrated products are free to create light source forms to suit the intended end-product design, without restriction of standardized sockets or modules. Therefore, it is proposed, that the highest performing SSL products will be integrated units, replaced at the end of their life with the next generation of even higher performing product. The model often used to illustrate this approach is that of televisions, where the entire units are replaced, rather than serviced, with newer generation products.

Side Two: The single most active market in solid state deployment is that of the direct lamp and fixture replacement space. This includes screw based lamps made to imitate the light output and distribution of obsolete technologies, and extends now to bi-pin linear forms to replace fluorescent sources. Oddly enough, the one lamp form that is not addressed, is the one most universally despised in commercial and residential markets alike – the plug-in CFL lamp – but let us not be distracted by this obvious and blatant oversight.  This replacement lamp direction appears to make the statement that the existing infrastructure of sockets is not replaceable, that demanding building owners and end use customers to replace existing fixtures is a burden beyond acceptable limits. This also forwards the concept that the existing socket forms within compromised products, is acceptable, regardless of its severe negative impact on SSL product performance, design freedom and appearance. (more…)

When LEDs first emerged, I was one of the many who expressed the opinion that a lighting system that could dim to a warmer CCT, imitating incandescent lamps, would be desirable. I want to take this opportunity to retract that original opinion and thought. I’ve played with it, seen the products available that do it, and have experimented with the approach… and can say unequivocally that I really don’t like it at all.

One of the problems with incandescent dimming has always been the patchwork of CCTs one gets through a space from different dimmer settings for the various products in a room. This has never been a good thing. Further, the change in CCT of an old school incandescent lamp is significantly different than halogen lamps, as it the character of the color. I for one have fallen out of love with the old incandescent lamp long ago. Over the last 20 some-odd years, I have come to use halogen sources over all incandescent forms, preferring the cleaner white color over that yellowy dinginess of the incandescent lamp. Incandescent lamps (non-halogen) produce a decidedly ugly color that I personally feel is misrepresented by their high CRI rating. The fact that the CRI formula will show a dimmed incandescent lamp with the same high CRI number, even when it very noticeably distorts color in a space, is a condemnation of our poor color performance metrics, not an indication of this lamps superior color performance. (more…)

I thought a lot about what to focus on in 2012 for this series, and decided that I had plenty to share from regular activities of Lumenique, LLC and Tasca. So, the plan is to select something completed in each of the 12 months of 2012 and feature them here. This will generally be products or projects completed for customers, but may also include a report on research work in process, when it adds value.

January Feature – TASCA Renovar Floor Lamp

This is a refitting of a Dazor table lamp, applying the TASCA lighting head, and adding an extension stand to convert a desk lamp to a floor lamp. The product was commissioned by a customer who provided the table lamp, purchased used. From the GSA and other government markings present on the original, it was obviously from a government facility. The table lamp made by Dazor has been around since the late 1930’s, where the fluorescent lamp version graced the GE display at the Worlds Fair.

For more details on this project, check out the summary and full technical specs at Lumenique 12 in 12 for 2012 – January.

For additional details on how you might procure a similar product from one of your own favorites, visit TASCA.

Also, as an update, the Lumenique Product Center now accepts all major credit cards, making your purchase experience easy and secure.

Stay tuned for additional news and updates on the 12 n 12 for 2012 review, and other interesting SSL information.


In an effort to create the highest possible performance in a portable lighting product, assembling the right combination of components is essential. Obviously the process begins with an efficient LED suited to the lighting effect desired. The LED must then be matched with an efficient driver. Finally, the driver must be fed power from an efficient power supply that converts incoming AC line voltage to clean DC power. Efficiency is generally found in matching the load of the LED to a driver designed for that load with no necessary over-capacity. Then, mating the driver to an efficient power supply matched in size to the driver’s operating load is necessary to produce the highest combined efficiency. (more…)