Indirect Lighting for the living room...$$$

ddrueding

Fixture
Joined
Feb 4, 2002
Messages
19,729
Location
Horsens, Denmark
My new apartment has no in-ceiling lighting in the living and bedroom. That's great, as I hate the ugly fixtures usually stuck up there in apartments. This is my opportunity to do something I've always wanted to do; install dimmable indirect lighting into a picture rail around the entire room.

I had no idea it would be this expensive.

At first I was thinking LED; the geeks choice for...well, everything, really. But with pre-fab solutions costing well over $25/ft for the perimiter of a 20'x15' room; that gets really expensive.

Finally I decided to look at incandecent solutions (flouecent is nice, but not dimmable). These guys have a nice Xenon solution that is very compact and more affordable than the others. After deciding that I only needed to light one of the long walls and the two short ones, it came out at just under $1400 including the 24v transformer and the dimmer capable of the 1200W load.

Really, this is a plea to JTR (now that he has returned): Am I looking in the right direction?
 

Bozo

Storage? I am Storage!
Joined
Feb 12, 2002
Messages
4,396
Location
Twilight Zone
Be careful with the dimmer.
In the US the electric frequency runs at 60Hz. Some of those 120v/24v transformers boost the frequency to 400Hz. Changing the input voltage with a dimmer switch usually results in the transformer dieing in a very short time.

Bozo :joker:
 

ddrueding

Fixture
Joined
Feb 4, 2002
Messages
19,729
Location
Horsens, Denmark
Actually, one of the things that makes this so expensive is that the dimmer is on the 24v side and designed to handle 1200W. The dimmer alone is $250+ :(
 

jtr1962

Storage? I am Storage!
Joined
Jan 25, 2002
Messages
4,375
Location
Flushing, New York
I saw this thread yesterday but haven't had time to give it a decent response. I'll try now. First off, yes, you can indeed dim flourescent lighting. The downside is that dimming ballasts for a pair of lamps run at upwards of $75 (when you can even find them). For the size room you mentioned with indirect light on the perimeter I figure 4 four-foot T8 tubes along each of the longer walls and 3 along each of the shorter ones. That's 14 tubes and 7 ballasts, plus the associated control switch (dimming ballasts all use a special switch which is manufacturer specific to control the dimming). Lutron seems to be the big player in dimming ballasts right now. More of their product is geared towards 277 VAC commercial lighting rather than 120 VAC home lighting. However, they do have 120 VAC dimming ballasts. Sometimes these come up for sale on eBay. Besides the cost, another drawback to dimming ballasts is you get dimming to at best 1%, more often 10% or more. Because of the way your eye works (logarithmic response) reducing the light levels to 1% will mean an apparent brightness of 10%. Reducing to 10%, which will more likely be the case with most dimming ballasts, will mean an apparent brightness of around 30%. In other words, not a great amount of dimming. If your budget is really tight you can sometimes dim fluorescents on magnetic ballasts using a standard incandescent lamp dimmer. I did this in my bedroom for years before buying the 4-tube T8 fixture a few years ago. You won't get a great amount of dimming (maybe to 30% at best). You run the risk of prematurely destroying the tubes if you run them at too low a power level without heating the cathodes. Also, I don't recommend magnetic ballasts anyway because of the flicker (and this gets worse when you dim them with a lamp dimmer).

All is not lost, however. If you don't care about dimming, or can live with a couple of different light levels instead of continuous dimming, then fluorescent strip lighting is still a viable option. Perhaps you can use the 14 32W tubes I mentioned earlier in tandem for your "high" setting. This would give you roughly 40,000 lumens which would be about right for a 300 square foot room. Additionally, you could perhaps rig switches so you can have half the lights on each switch. This gives you the option of a 50% "medium" setting. For the "low" setting maybe have one 17 watt T8 tube centered on each wall controlled by a third switch. The respective lumens on high/medium/low would be roughly 40000/20000/5500. Apparent brightness levels to your eye would be 100%/71%/26%. Again, this isn't a spectacular amount of dimming but if you can live with it it can be implemented very cheaply. A big drawback would of course be slightly uneven light distribution when some of the tubes are off. Another drawback would be uneven wear on the tubes since all wouldn't necessarily be on at the same time. On the plus side you have high efficiency and wide choice of color temperature.

OK, that takes care of fluorescent so on to LED. LED interior lighting isn't yet mainstream so the cost will be high and the options limited. Still, if I were to do this I could use about 200 of the new Cree XR-Es. This would give about 135 lumens each when driven at 700 mA, or 27000 lumens total. This is not too far off of the fluorescent solution. Power consumption would be a bit more (about 500 watts counting driver losses instead of roughly 420). This is obviously because although the Crees are currently the most efficient power LED available (I've independently tested them at efficiencies of 75 lm/W when driven at 350 mA) they are still less efficient than the 90 or so lm/W of T8 tubes. However, note that in both cases we are talking bulb lumens. When you put fluorescent tubes in an indirect lighting setup most of the light from the side of the tube facing downwards will be lost. At best you'll probably end up with the same 27000 lumens hitting the ceiling as you would with LEDs. Since the Crees direct 100% of their lumens into a 180° cone towards the ceiling, all the light goes where you want it. In other words, you're taking less of an efficiency hit than you thought at first. A major advantage of using LEDs is of course infinitely variable dimmability down to whatever light level you desire. You can go to 0.1% or less if you wanted to without complex, costly controls. Another advantage is very long life. Even at 700 mA if the LEDs are properly heatsinked you'll be looking at 50,000 hours or more until they reach 70% of initial brightness. That's about 23 years if used 6 hours a day. Since I assume that since you want dimmability the LEDs will be operating at far less than 700 mA most of the time then they will probably outlast you. Disadvantages? Cost for one. The LEDs alone, even discounted, will probably run you $1200. After that you'll need circuit boards and heat sinks. In the end you would probably be looking at upwards of $3000 or more. A second disadvantage is that color rendering will be good, but not as good as with fluorescent. A third disadvantage will be tint variation, even among LEDs from the same tint bin. Like I said, LED isn't quite there yet. Check back in about 5 years and my guess is you would be able to do this for under $500, perhaps even less, and the LEDs would have better color properties. On an even more interesting note Cree and Seoul Semiconductor plan to ramp up efficiencies of their commercial products to 145 lm/W within 15 months so an LED solution then would use less power and fewer LEDs than the current one. Current off-the-shelf solutions mostly use cheap 5mm LEDs which will lose a lot of their brightness after a few thousand hours.

This takes us to the last solution-incandescent. $1400 is not exactly cheap. In fact, you can probably do the setup with dimming ballasts for the same or less. The only hard part is finding them. Of course, with incandescent you'll have infinite dimmability. However, the disadvantages are many. First off, those so-called long-life incandescents get their long life by underdriving the filament. This makes them even less efficient than regular incandescent which is already nothing to write home about. Second, dimming incandescents doesn't save as much power as you might think since efficiency drops further upon dimming. For example, incandescents dimmed to 10% of their maximum output are still using about 30% of their maximum power. Basically as you dim them they turn more and more into miniature heat lamps. Third, as you well know there is by definition a color temperature shift when dimming incandescent. They get more red as you dim them, and their color temperature to start with isn't exactly optimal for lighting. Can anyone say major eyestrain? LEDs and fluorescents don't have this color temperature shift when you dim them. Well, they do, but to a much smaller degree which is barely noticeable. Fourth and already touched upon in other threads, incandescents are basically available only in one color temperature-a sickly yellow one. Either you like it or you don't. My feelings on this are already pretty well known. They do have blue filtered incandescents but I don't know if any are available for the system you mentioned.

I think of all these choices I'd choose for now between the two flourescent options I mentioned. In five years time, maybe sooner, I would certainly go LED. Either way, once you finish the project please post some pictures. I'd love to do a similar setup, perhaps in the basement.
 

ddrueding

Fixture
Joined
Feb 4, 2002
Messages
19,729
Location
Horsens, Denmark
Thanks for the lengthly response JTR:

I was considering a stepped Flourescent solution, but one of the things that really appealed to me about the Xenon system I linked to is the very low profile. The ceilings here are of the standard 8-foot variety, and the GF and I are tall (5'10" and 6'2"), so there is little room for the lighting install anyway. One of the things this low-profile solution would allow me to do is install it in a very small railing (perhaps 2" tall, protruding another 2"). If I were to do it with Flourescent, I suspect I would need something similar in dimensions to a rain gutter (4" tall and 4" deep).

This larger solution could work, if I only did it along the main long wall (20'). If I did 4 parralel banks of 4 lights, Than light levels could be 25%, 50%, and 100% without uneven light distribution.

How large a fixture would I need to fit this kind of installation? Could I use paintable vinyl rain gutters?

Also; this is an apartment, so special controls will need to be surface mounted, and master control/power will be from a switched outlet on the wall.
 

jtr1962

Storage? I am Storage!
Joined
Jan 25, 2002
Messages
4,375
Location
Flushing, New York
This larger solution could work, if I only did it along the main long wall (20'). If I did 4 parralel banks of 4 lights, Than light levels could be 25%, 50%, and 100% without uneven light distribution.

How large a fixture would I need to fit this kind of installation? Could I use paintable vinyl rain gutters?
If you go with T5 tubes instead of T8 then I think 4 sets in parallel could be made to fit into a rain gutter. T5 tubes are a different length that T8s, and use a different ballast. They cost more also, but not prohibitively more. Anyway, since the tube is 5/8" in diameter instead of 1" then four should comfortably fit in a rain gutter. You may need to offset the ends slightly since the sockets are larger than the tube. Actually, with four banks of lights you could have four light levels- 25%, 50%, 75%, and 100%.

In the meantime I'll keep a lookout for dimmable ballasts in case you may want to go that route instead.

The main drawbacks I personally see with that Xenon system would be cost and power consumption. To get decent light levels in a 15'x20' room using indrect lighting would mean a good 1000 watts or more power consumption. Even then, your light level would still be less than 25% of the fluorescent solution.
 

udaman

Wannabe Storage Freak
Joined
Sep 20, 2006
Messages
1,209
If you go with T5 tubes instead of T8 then I think 4 sets in parallel could be made to fit into a rain gutter. T5 tubes are a different length that T8s, and use a different ballast. They cost more also, but not prohibitively more. Anyway, since the tube is 5/8" in diameter instead of 1" then four should comfortably fit in a rain gutter. You may need to offset the ends slightly since the sockets are larger than the tube. Actually, with four banks of lights you could have four light levels- 25%, 50%, 75%, and 100%.

In the meantime I'll keep a lookout for dimmable ballasts in case you may want to go that route instead.

The main drawbacks I personally see with that Xenon system would be cost and power consumption. To get decent light levels in a 15'x20' room using indrect lighting would mean a good 1000 watts or more power consumption. Even then, your light level would still be less than 25% of the fluorescent solution.

Ok, what about T6 with an dimmable ballast from Icecap? T6 are easily obtained in 'daylight' (and I think full-spectrum versions). I'd have to check with the acquarium forums where most of my knowledge comes from (haven't visited in about a year). Icecap 660 ballasts (I think they came out with a less expensive model to drive 2 t5's, but still more expensive than you can probably fine else where) are easily obtained used on e-bay, but the older ones don't have a port to connect to the dimming control module, which it's self is expensive, IIRC. Used IC660's tend to sell for around $100-130 on e-bay, occasionally you can find them on sale new for ~140-150. I think the IC660 (outputs 440w?) will drive 4 T5's or 4 T6's @4ft lenght, 3@5ft lenghts. T6 tubes are typically made of thicker walls than T5, are less likely to break, and also because of the thicker walls seem to dissipate the heat more uniformly, feel a little less hot to the touch, so I've read. There are cheap T6 made with end connectors to fit standard T12 end sockets, but IIRC, other T6 are made with smaller end connectors....it's been quite a while since I researched these, as they were brand new on the market here in the USA, but are supposedly commonly used in Japan for retail business lighting fixtures, even more so than T5's..at least that's what the Chinese distributor here locally, for one of the major producers in China, told me?

you may see 135lm/w commercially, but in what size (what rated current draw- 100ma or less, or are we talking 700ma?), and for what cost (kind of like SSD's are still very expensive, but dropping quickly, though I doubt we'll see the same preciptious price drops for LED's...who knows, I'm not quite as optimistic about progress or price drops on LED's as jtr is :( )???

As jtr knows, at the higher ma ratings the efficiency drops quite a bit, unless the heatsink can keep the LED at manufacturers 'idealized' temperature rating...which is usually quite lower than most operating conditions). will we see 3watt LED's at that 135l/m efficiency, 5watters ??? A 3w LED that can be heat sinked properly would make one nice 400lm mini flashlight, like a Mag AA using 2 of the newer low loss Sanyo 2000ma AA NiMH rechargeable's. http://www.stevesforums.com/forums/view_topic.php?id=96500&forum_id=51

Of course jtr already knows this is an impossibility as 1watt LED's are already as hot as you can put into a smaller flashlight, 3watters put into Maglite-D cell conversions use massive heatsinks and they get really HOT.

Samsung just released a LED backlight LCD 20in monitor @$1,999 MSRP , but it's supposed to be a professional tool, with 114% of NTSC color gamut (ie much better than CCFL backlight LCD's). http://www.samsung.com/PressCenter/PressRelease/PressRelease.asp?seq=20061104_0000298342

Sorry for the OT commentary, nasty habit I picked up at SR :) .
 

jtr1962

Storage? I am Storage!
Joined
Jan 25, 2002
Messages
4,375
Location
Flushing, New York
Thanks for reminding me about the IceCap ballasts and T6s, udaman. I think I've even seen full-spectrum T6s somewhere a while back. This might be another viable solution to Dave's lighting project. Also, I haven't gotten around to answering your latest emails but I'll try to find time this weekend.

On the new Crees, the projected 145 lm/W efficiency in 15 months would be at 350 mA drive current. At 700 mA based on my tests you would get about 80% of that or around 108 lm/W. Figure power level at 700 mA is roughly 2.5 watts so that would be about 270 lumens output. Although 1 watt might be about the power dissipation limit in a smaller flashlight some interesting things are happening now that efficiency is up. In the Luxeon days heat more or less equaled the power input since the LED was only about 10% or less efficient. Now we're around 25% so you can drive a Cree at maybe 1.3 watts in the same body that limited a Luxeon to 1 watt. At the 145 lm/W level you could probably go to about 1.6 watts since 0.7 watts would come out the front end as light. Even more impressive are the lumen figures. The Luxeon was maybe 30 bulb lumens. The current crop of Crees will get you to about 90 bulb lumens, maybe 100 with the better bins. The Crees in 15 months should be good for over 200 lumens in that same body with the same heat production. As we tend towards 100% efficiency heat will become less and less an issue with LEDs.

As for cost, the XR-E unbinned is presently a little over $3 in moderate quantities (1000s) and about $6 to $7 in 100s. I don't know what future price drops are forecast but the increased efficiency will ultimately mean fewer emitters to do the same task even if they cost the same. In 15 months I could probably do Dave's lighting project with only 100 to 120 emitters instead of 200. If I had other uses for the Cree so I could buy in thousands the LEDs for this would cost under $400.

I'd love to do a side-by-side comparison of Samsung's LED backlit model with the same one backlit by CCFT. Although they're claiming superior color gamut I wonder how well they bother calibrating these from the factory? Superior gamut is all well and good but not if you need to spend another few hundred on equipment to calibrate a brand new monitor. I heard Samsung at least is a little better than most out of the box with regards to calibration.
 

ddrueding

Fixture
Joined
Feb 4, 2002
Messages
19,729
Location
Horsens, Denmark
Great info guys...thanks.

WRT power consumption: In this apartment, utilities are included ;) further, we have the heater cycling on about 10% of the time anyway; and it's noisier than the light fixture. But if we were to move, than this would be of greater concern.

So far I like 2 options:

1. One wall with 4 banks of full-spectrum bulbs
2. Xenon all around

I'm tempted to wait for future LED lighting technology if it's so close. Further, I'm tempted to choose the cheaper solution so I can change to LED when it becomes more practical.
 

ddrueding

Fixture
Joined
Feb 4, 2002
Messages
19,729
Location
Horsens, Denmark
OK, let's get the ball rolling on a flouescent solution.

Here is a sample gutter. 4.5" wide at the top, and about 3" deep. What should I get to put in there? If you could include links to tubes, ballasts, etc. that would be great. Don't worry about dimmers, this should provide enough light control for me.

~TIA
 

jtr1962

Storage? I am Storage!
Joined
Jan 25, 2002
Messages
4,375
Location
Flushing, New York
Go with the aluminum gutter. If something overheats there's no chance of igniting it like the PVC.

Here's some full-spectrum T5 tubes. I'll see if I can find T6s later.

You'll probably need to custom rig something to hold the sockets. Overall if you do this with switches it'll cost only a couple of hundred dollars, and in a few years when LED becomes viable you can make a nice dimmable LED setup around the perimeter of the entire room. In fact, I could probably make the LED assemblies for you. I'd just wait until the 145 lm/W LEDs are out though. Once that happens there probably won't be any great efficiency improvements, just slow incremental ones, so no point in waiting much beyond that.
 
Top