Energy Assessment: How long does it take for LED luminaries to pay for themselves?

31 to 51 of 51

- CommentAuthorJSHarris
- CommentTimeJan 27th 2012 edited
<blockquote><cite>Posted By: wookey</cite>This mostly confirms my opinion that GU10 mains LEDs are crummy engineering. 62C with the lid off doesn't sound good.

I've made a nice current-driven XPG-based LED lamp to test my ideas on how it should be done. Busy right now but will post some pics soon.</blockquote>

Exactly my views. The engineering challenges of trying to fit a very efficient and reliable 240 V AC to ~3.5V, constant current DC, converter into the base of a standard GU10 lamp almost certainly exceed the capabilities of many (most?) of the Far Eastern manufacturers.

I decided a while ago to opt for a low voltage DC lighting bus in the new build, partly for efficiency and partly to allow the use of MR16 LEDs. I've been using these for a while in my off-grid workshop lighting system and they seem to be excellent in terms of reliability, even the really cheap multi LED ones I bought direct from China.

Partly that seems to be down to them not having any electronics in the base, just a bridge rectifier and a few resistors. These multi-LED MR16s don't use true constant current drive, but just hook the LEDs up in strings of three in series and feed each string via a small resistor. The losses are tolerable, the 2.4 W ones seem to waste around 160 to 180 mW in the resistors and rectifier (they are designed to run on 12V AC as well as DC), which is probably close to, or perhaps better than, the efficiency of the cheap switched mode converters.
- CommentAuthorDamonHD
- CommentTimeJan 27th 2012
You could probably up the efficiency nearly a third if space and polarity wasn't an issue (either something keyed unlike MR16) by dumping the bridge rectifier (saving 1.2V) and replacing the resistor (another ~2V?) with a switching current regulator. Wookey may well be doing that already of course.

Seems a shame to have super-efficient LEDs and waste some in the bridge at the last moment for example.

Rgds

Damon
- CommentAuthorWillInAberdeen
- CommentTimeJan 27th 2012
I'm pleased with a GU10 240V Aurora led (only had it since reading CW's LED trial thread so no reliability issue yet) but...
... that finned case with vent holes in the face, makes an air flow path through the kitchen ceiling into the loft. For the 23.5 hrs each day when the lamp is not on, I can feel a cold draught coming down through the vents into the room.
This seems to be a big drawback compared to the Gu10 CFLs which completely fill the light fitting.
Moral of the story is dont fit recessed downlighters that bridge the air tightness envelope.
The 'adjustable angle' downlighters are even worse, they have a clear air gap to allow the lamp holder angle to be adjusted.
- CommentAuthorDamonHD
- CommentTimeJan 27th 2012
My GU10s are in fittings below the ceiling, thus getting cooling and not breaching the plasterboard.

Rgds

Damon
- CommentAuthorpmagowan
- CommentTimeJan 27th 2012
You could make a box in the loft to prevent this air flow. Most light fittings only require a specific clearance as fas as I am aware. Use plasterboard and that is furthur fire protection for if they go pop.
- CommentAuthorjoe90
- CommentTimeJan 27th 2012
I created plasterboard boxes in the loft to stop draughts and create enough vent room when installing down lighters for a customer, I just cut 5 off 150mm squares (for each light) and glued them together and to the ceiling with plasterboard adhesive. proper job!
- CommentAuthorWillInAberdeen
- CommentTimeJan 27th 2012
Did consider that, but as others said, I think that restricting the airflow will spoil the reliability. Same as just blocking the vent holes on the lamp face....!
Better to have lamp holders hung underneath the ceiling as Damon said.
- CommentAuthorJSHarris
- CommentTimeJan 27th 2012
<blockquote><cite>Posted By: DamonHD</cite>You could probably up the efficiency nearly a third if space and polarity wasn't an issue (either something keyed unlike MR16) by dumping the bridge rectifier (saving 1.2V) and replacing the resistor (another ~2V?) with a switching current regulator. Wookey may well be doing that already of course.

Seems a shame to have super-efficient LEDs and waste some in the bridge at the last moment for example.

Rgds

Damon</blockquote>

In practice it doesn't seem that the losses from the crude resistor/rectifier drive are that bad, though. The switched mode current regulators I've used on some of the LED lamps I've made seem to be around 85 to 90% efficient, which is very close to the efficiency of the simple rectifier and resistor units.

The 12 V MR16s I have barely get warm anywhere except on the front face. The rear part of the housing seems to stay pretty cool. The voltage drop across the rectifier/resistor combination seems to be around 1.7 V when running on my 12.5 V supply (which is a set of usually fully solar charged lead acid batteries of uncertain parentage), which gives an efficiency of around 87%, probably very close to that for a switched mode converter and without the potential reliability issue of having more components in a potentially hot zone.
- CommentAuthorChris P Bacon
- CommentTimeJan 27th 2012
http://www.downlightcoversdirect.ie/
- CommentAuthorCWatters
- CommentTimeJan 29th 2012
I've decided to torture test one of the GU10 LEDs I've been using. Have fitted it into a sealed IP rated downlight housing in my bathroom. The fitting is a metal can with rubber seal at the bottom and a thick glass bezel. No airflow at all that I can see. Because it's a bathroom it won't get the same usage as say a kitchen but we'll see what happens.
- CommentAuthorWillInAberdeen
- CommentTimeJan 29th 2012 edited
That finned metal case is presumably not earthed (there is only 2 wire connection to the lamp). What is the risk of a failure mode in the power electronics that makes the case live?
(I'm still very pleased with the LED much better than the halogen it replaced)
- CommentAuthorJSHarris
- CommentTimeJan 29th 2012
<blockquote><cite>Posted By: WillInAberdeen</cite>That finned metal case is presumably not earthed (there is only 2 wire connection to the lamp). What is the risk of a failure mode in the power electronics that makes the case live?
(I'm still very pleased with the LED much better than the halogen it replaced)</blockquote>

The AC-DC converter module looks to be transformer isolated, plus the circuit board has pretty good isolation between the AC input and DC output side. All told I'd guess that these are reasonably effectively double insulated, although the sample from qeipl doesn't have the double insulated approval mark.
- CommentAuthorPaul_B
- CommentTimeJan 29th 2012
Interesting thread, I've been using LEDs both MR16 and GU10 for around 4 years now. Quality varies hugely from one design and manufacturer to the next My most expensive purchases have all lasted but some of the cheaper units failed within a year and these tend to pop indicating the ancillary electronics have failed rather than the LED. I'm glad in an earlier post someone mentioned lumens rather than watts as it is the former that is relevant. I'd also add that the angle of the light is also another important consideration, many LEDs are advertised with high output but they are really spot lights with the light coming in a very tight beam. Interesting that people are considering running LV circuitry alongside mains, many electronic items want 12v DC but the downside would be the losses in thin cable at higher currant.

Paul
- CommentAuthorCWatters
- CommentTimeJan 30th 2012
The latest trend seems to be to use multiple 5050 form factor LEDs and these all seem to have a very wide beam angle.

http://www.google.co.uk/search?hl=en&q=5050+led+gu10&gs_sm=e&gs_upl=2870l4509l2l4797l5l5l0l0l0l0l147l698l0.5l5l0&bav=on.2,or.r_gc.r_pw.,cf.osb&biw=1074&bih=615&wrapid=tlif132788162659810&um=1&ie=UTF-8&tbm=isch&source=og&sa=N&tab=wi&ei=pN0lT9uzDI200QXD8ujQCg
- CommentAuthorDamonHD
- CommentTimeJan 30th 2012
With your prompting I've just ordered one such SMD-based GU10 to review it and write it up. Next week I'll be able to whisper the result!

If as described, it matches the efficiency of the best LEDs that I currently have at over 100lm/W.

Rgds

Damon
- CommentAuthorCWatters
- CommentTimeJan 30th 2012
When I tested one on it's own I wasn't sure the were going to work out due to the wide beam angle but now I've done a whole room full I'm pleased I took the gamble.

Looking for dimmable versions now.
- CommentAuthorborpin
- CommentTimeJan 31st 2012
Posted By: CWattersWhen I tested one on it's own I wasn't sure the were going to work out due to the wide beam angle but now I've done a whole room full I'm pleased I took the gamble.
How did you space them to get an even light? In a 3Mx3M room I was thinking of 4 equally spaced from the corners but I wondered about a 5th in the middle.
- CommentAuthorCWatters
- CommentTimeFeb 1st 2012
Spacing to get an even light should be less critical with these wide angle 5050 SMD LEDs than it is with halogen. I was replacing a 3 x 5 grid of halogen in a kitchen about 4m x 5.5m so the spacing was already determined. In the narrow direction it's about 950mm and in the longer dimension about 800mm. The gap around the edge is larger at around 1m - probably to avoid lighting the tops of the wall units and creating shadow beneath them.

More important to have enough to get tbe brightness you want. I have 15 in an area about 22sqm which works out at around one per 1.5sqm. The result is plenty bright enough for a kitchen and too bright for other rooms like a living room.

On that basis 3 x 3 = 9 and 9/1.5 = 6. Perhaps string up 4 or 5 like christmass lights and see what the end result looks like before you cut holes.
- CommentAuthorSteamyTea
- CommentTimeFeb 1st 2012
Posted By: CWattersMore important to have enough to get tbe brightness you want. I have 15 in an area about 22sqm which works out at around one per 1.5sqm. The result is plenty bright enough for a kitchen and too bright for other rooms like a living room.

If there is an assumption that there is enough light, what is the W/m^2.
I use about 1W/m^2 with my CFLs
- CommentAuthorCWatters
- CommentTimeFeb 1st 2012
I'm using a mix of GU10 LEDS from two sources both using multiple 5050 LED die. One is rated 5W at 400L and the other 3.2W at 350L. Both appear equally bright but I haven't tried to measure the power consumption.

400L for 5W = 80L/W which I suppose is plausible
350 for 3.2W = 109L/W which might be too good to be true?

So if we assume 400L at 5W then I have 75W in 22sqm = 3.4 W/m^2 compared to the 34 W/M^ I had with halogen.

On an earlier thread I reported measurements using a light meter on two LEDs and two 9W CFL...

Make, On axis, 45 deg, 60 deg (in Lux)
Interlux 33 13 6
Leder 35 13 6
Megaman CFL 18 8 3
Ring CFL 20 8 4

Now my light meter cost all of Â£12 so might not be very accurate. Other have pointed out that light meters are colour sensitive and the LEDs although called warm white are colder than the CFL.

I won't be buying CFL again unless these LEDs prove unreliable in the longer term.
- CommentAuthorborpin
- CommentTimeFeb 2nd 2012
Posted By: CWattersOn that basis 3 x 3 = 9 and 9/1.5 = 6. Perhaps string up 4 or 5 like christmass lights and see what the end result looks like before you cut holes.
Thanks for that. Gives me a good starting point at least.

31 to 51 of 51

Add your comments



	Site Map \| Home \| View Cart \| Pressroom \| Business \| Links	Contact Us
Logout
© Green Building Press

Green Building Forum

Categories

Feeds

Energy Assessment: How long does it take for LED luminaries to pay for themselves?

Username	Password