Serious question: Are LEDs actually more efficient than T5 or MH?

[GroktheCube]

The title says it all. We often see efficiency touted as an advantage for LEDs, but I've never actually seen lumen/watt comparisons of common aquarium lights.

Doing some reading lately, I was surprised to learn that in the "real world" (not labs), LEDs are at best on par with T5s and gas discharge lamps (metal halides) in terms of luminous efficiency. The data I read suggested that off-the-shelf LEDs are usually in the 50-90 lumen watt range (using more generous numbers), with T5s and MHs both falling into the 60-110 lm/w range (including ballasts). Virtually every direct comparison I've seen shows T5s being more efficient, usually substantially so, than LEDs. I have not found as many direct comparisons for MHs, so I'm going by the "Standard" quoted figures.

I've done a few google searches, but can't find any data comparing commonly used aquarium T5, MH, and LED bulbs/lighting units.

It seems to me that if this same trend holds true for aquarium lighting, many LED users may be dramatically under-lighting their tanks relative to T5 and MH users. If that is the case, it goes a long way towards helping explain why some LED users have had results that didn't quite live up to their expectations. One might actually need just as many, if not more, watts of LED light to replace T5s or MHs.

Can any of the resident experts chime in?

 

[alton]

I am no expert but they are about 30% more efficient. The lie they don't create much heat is BS. Place a couple in a canopy and see how long before they overheat and shut down. The biggest difference is LEDs create heat on the back side which creates heat in the room but not directly into the tank. Getting an LED to have only a 20% difference between foot candles and PAR reading is key. Many of your new fixtures accomplish this, and why you do not hear the horror stories of burning corals as much anymore. I fried a green birdsnest with a marineland led once. Only 200 PAR but 4000 foot candles directly under one of the leds.

 

[ca1ore]

Certainly not an expert, but my general view is that if something that's purpose is to not generate heat is generating heat then there is energy being lost. Seems to me that MH generates far more heat than LED so the latter must be more efficient if light generation is the goal.

As to why some reported poor results with LED, I'd be inclined to assign blame to either over-lighting or poor spectrum before under-lighting. LED is simply experiencing the same kinds of growing pains that MH did 25 years ago. Most have not been around long enough to have direct experience, but when I bought my first MH light (a DuplaSun in about 1988) the only bulbs available were Osram 5000K. I was some years before we saw broader, higher kelvin, choices appear.

 

[jda]

Do you mean in light output or total electrical consumption/cost?

I can tell you that here in Colorado where even right now in August where the lows are 50-60s, heat is my biggest friend and the MH give the heaters a break when they are on. I welcome the heat from lighting and would otherwise have to run the heaters to make up the difference. There are no electrical savings for me.

Like everybody else, I doubt that we have any idea about light efficiency... but shouldn't effectiveness also play into this? Who cares if a horrible spectrum (either for coral, eyeballs or both) is more/less efficient.

 

[FlyPenFly]

Do you mean in light output or total electrical consumption/cost?

I can tell you that here in Colorado where even right now in August where the lows are 50-60s, heat is my biggest friend and the MH give the heaters a break when they are on. I welcome the heat from lighting and would otherwise have to run the heaters to make up the difference. There are no electrical savings for me.

Like everybody else, I doubt that we have any idea about light efficiency... but shouldn't effectiveness also play into this? Who cares if a horrible spectrum (either for coral, eyeballs or both) is more/less efficient.

Heat from the MH is not even close to energy efficient in terms of heating to your tank like a properly sized heater would. We are talking orders of magnitude.

 

[Wazzel]

In a watts per lumen I believe so.

 

[jda]

Heat from the MH is not even close to energy efficient in terms of heating to your tank like a properly sized heater would. We are talking orders of magnitude.

I have a ton of experience in this and have watched and measured it closely. If you have done such things, I would love to hear how this worked out in NY.

I am not saying that MH > heaters. I am talking about combo of lights and heaters: Heater + LED > Heaters + MH. Nobody is saying that bulbs are the best way to heat your tank... only that the heat from something that you are using anyway is very beneficial in my case and that orders of magnitude don't apply when you are using something anyway.

Another example: heat from a chiller used is thought to be a bad thing. ...but in the fall when I lived in the midwest when the temp drops quickly, the heat from the chiller gets my fish room warm about 6:00 right before the temp drop and keep the heaters from having to work as much during the evening and early night. The timing is perfect. Of course, heating a tank with air from a chiller is a horrible idea, but it works out great when you are using it anyway.

 

[GroktheCube]

I am no expert but they are about 30% more efficient. The lie they don't create much heat is BS. Place a couple in a canopy and see how long before they overheat and shut down. The biggest difference is LEDs create heat on the back side which creates heat in the room but not directly into the tank. Getting an LED to have only a 20% difference between foot candles and PAR reading is key. Many of your new fixtures accomplish this, and why you do not hear the horror stories of burning corals as much anymore. I fried a green birdsnest with a marineland led once. Only 200 PAR but 4000 foot candles directly under one of the leds.

Interesting, thanks for the info. For which fixtures/bulbs did you find that 30% figure?

Certainly not an expert, but my general view is that if something that's purpose is to not generate heat is generating heat then there is energy being lost. Seems to me that MH generates far more heat than LED so the latter must be more efficient if light generation is the goal.

As to why some reported poor results with LED, I'd be inclined to assign blame to either over-lighting or poor spectrum before under-lighting. LED is simply experiencing the same kinds of growing pains that MH did 25 years ago. Most have not been around long enough to have direct experience, but when I bought my first MH light (a DuplaSun in about 1988) the only bulbs available were Osram 5000K. I was some years before we saw broader, higher kelvin, choices appear.

My understanding is that the heat from MHs and T5s isn't just waste heat, but also IR.

Look at it this way:
An MH consuming 150 watts at 15% efficiency produces 127.5 watts of waste heat.
An LED consuming 150 watts at 20% efficiency produces 120 watts of waste heat. Alternatively, and LED producing the same amount of light at 33% greater efficiency still produces 90 watts of waste heat.

In all 3 cases, not exactly a huge difference.

It think the bigger difference is in the relative lack of IR from LEDs, and the efficient heat sinks most LEDs are attached to that keep the fixture from feeling as hot.

Obviously, if one is using 300 watts of LEDs instead of 700 watts of MH, less heat will be produced. My question is whether or not LEDs using so much less energy can actually produce comparable amounts of light in the real world.

All the actual data I found suggested otherwise, but I came up blank when I tried to find that information for aquarium lighting.

 

[GroktheCube]

Do you mean in light output or total electrical consumption/cost?

I can tell you that here in Colorado where even right now in August where the lows are 50-60s, heat is my biggest friend and the MH give the heaters a break when they are on. I welcome the heat from lighting and would otherwise have to run the heaters to make up the difference. There are no electrical savings for me.

Like everybody else, I doubt that we have any idea about light efficiency... but shouldn't effectiveness also play into this? Who cares if a horrible spectrum (either for coral, eyeballs or both) is more/less efficient.

I'm talking luminous efficiency. Lights alone (to eliminate other variables). Lumens/watt. While the big picture matters for an individual making a decision about how to light their tank, what I'm driving at is that people may be using far fewer LEDs than they actually need to in order to provide the same amount of light as T5s or MHs.

The most often quoted figure for T5s is about 70-105, which translates to approximately 10 to 15% efficiency. 65-115 lm/w is what I keep seeing quoted for MH, which translates to roughly 9.5 to 17% efficiency. The most generous numbers I can find for LEDs are right around 100, or 14-15% efficiency. The vast majority of quoted luminous efficiency dated for LEDs I've seen generally pegs them in the ~55-85 range, which is roughly 8 to 13% efficient.

Maybe aquarium LEDs are substantially more efficient? Some sources seem to suggest that blue LEDs might be more efficient than white LEDs (which makes sense, given white LEDs are usually blue LEDs + phosphor coating), but I haven't seen that quantified.

 

[alton]

Okay the questions is, where did I come up with a 30% savings of LED versus everything else?
Listening to lighting reps when they come in to my office and tell me about the savings from led over fluorescent and metal halide. They always go back to replacing old technology when giving their cost savings. Example one rep might come in and say you can save up to 50% if you go from 3 lamp 32 watt 2x4 lay in fixture to LED. Sounds great except in our industry most 2x4 lay ins are now 2 lamp 28 watt fixtures, so you can see the 50% is actually a lot less. On the metal halide side same thing except for 400 watt comparison to LED but today most locations that in the past used 400 watts, now can get by with 320 watt lamps with pulse start ballast. Oh yea you hear that you never have to change lamps with led! Yes that is true you get a five year warranty and if the fixture fails you get a new one. After five years and one goes bad you have to change the whole fixture not just a lamp or ballast. Remember the 50,000 hours for LED, new Fluorescent lamps are claiming 30,000 hours and 20,000 hours for metal halide with the correct ballast. As the rare earth phosphor market continues to be squeezed by the Chinese, fluorescent companies know they have to do something to compete. Looking at one companies literature on LED, a 2 x 4 – 4800 lumen fixture draws 48 watts. On a 2 lamp fluorescent fixture using 28 watt lamps should be about the same lumens at 56 watts.

 

[GroktheCube]

Okay the questions is, where did I come up with a 30% savings of LED versus everything else?
Listening to lighting reps when they come in to my office and tell me about the savings from led over fluorescent and metal halide. They always go back to replacing old technology when giving their cost savings. Example one rep might come in and say you can save up to 50% if you go from 3 lamp 32 watt 2x4 lay in fixture to LED. Sounds great except in our industry most 2x4 lay ins are now 2 lamp 28 watt fixtures, so you can see the 50% is actually a lot less. On the metal halide side same thing except for 400 watt comparison to LED but today most locations that in the past used 400 watts, now can get by with 320 watt lamps with pulse start ballast. Oh yea you hear that you never have to change lamps with led! Yes that is true you get a five year warranty and if the fixture fails you get a new one. After five years and one goes bad you have to change the whole fixture not just a lamp or ballast. Remember the 50,000 hours for LED, new Fluorescent lamps are claiming 30,000 hours and 20,000 hours for metal halide with the correct ballast. As the rare earth phosphor market continues to be squeezed by the Chinese, fluorescent companies know they have to do something to compete. Looking at one companies literature on LED, a 2 x 4 "“ 4800 lumen fixture draws 48 watts. On a 2 lamp fluorescent fixture using 28 watt lamps should be about the same lumens at 56 watts.

Thanks for the information.

I'd definitely be curious to see someone take some measurements for aquarium lighting.

 

[ca1ore]

LMy understanding is that the heat from MHs and T5s isn't just waste heat, but also IR.

Agreed, however, I'd consider that simply another component of 'waste heat'. Certainly excess heat can help to reduce the use of tank heating, but seems to me unlikely to do so as efficiently as a purpose built heater. The reality is that figuring out how much cheaper LED may, or may not, be is far more complex than just looking at the light generation efficiency. The kicker for me was not needing to run a chiller anymore; even if my Winter heating costs are slightly higher.

 

[CuzzA]

Im not an expert and dont have experience with metal halides, but I think it's safe to assume that a 100 watt LED probably running at an average of 50-70% is consuming a lot less energy than a 400 watt medal halide. Not to mention a majority of that 400 watts isn't even being transferred to actual light output. Yes an LED in an air tight canopy is going to produce heat, but a lot less than a metal halide would. Most canopies should have a vent. Also, just because one LED bulb fails doesn't mean you have to replace the whole unit. Simply replace the individual bulb. So in my opinion LED's are much more efficient in terms of energy consumption and longevity. There's a reason the US government banned so many light bulbs in the past years... The technology is old and not efficient.

 

[GroktheCube]

Agreed, however, I'd consider that simply another component of 'waste heat'. Certainly excess heat can help to reduce the use of tank heating, but seems to me unlikely to do so as efficiently as a purpose built heater. The reality is that figuring out how much cheaper LED may, or may not, be is far more complex than just looking at the light generation efficiency. The kicker for me was not needing to run a chiller anymore; even if my Winter heating costs are slightly higher.

Given that IR isn't always useful, I agree entirely.

Also, just to be clear, those IR emissions are not included as lumens (non-waste) when measuring luminous efficiency. A 150w MH and a 150w LED, if both are producing ~70 lm/w, will both produce the same amount of heat. It's just that in the MH's case, rather than all of it being traditional waste heat, a substantial portion of it will also be IR.

I wasn't suggesting that the IR was "extra" heat beyond the loss measured when determining luminous efficiency, but rather that it means more heat being transmitted to the tank.

Im not an expert and dont have experience with metal halides, but I think it's safe to assume that a 100 watt LED probably running at an average of 50-70% is consuming a lot less energy than a 400 watt medal halide. Not to mention a majority of that 400 watts isn't even being transferred to actual light output. Yes an LED in an air tight canopy is going to produce heat, but a lot less than a metal halide would. Most canopies should have a vent. Also, just because one LED bulb fails doesn't mean you have to replace the whole unit. Simply replace the individual bulb. So in my opinion LED's are much more efficient in terms of energy consumption and longevity. There's a reason the US government banned so many light bulbs in the past years... The technology is old and not efficient.

Obviously 50-70 watts of LED lighting is consuming less energy. The question is how much light is it producing, relatively speaking? The answer seems to be that in the best case scenario for the LED, it's producing about a third as much light as the 400w MH and operating as 50% greater efficiency. In the best case scenario for the MH, the MH is producing about 13 times as much energy, and operating at more than double the LED's efficiency.

Most likely, the truth is somewhere in between. Unless the LEDs used in aquarium lights are dramatically more efficient than average, it seems more likely that LEDs are at best 10-20% more efficient in relative terms. Meaning that you should actually be comparing 330-360 watts of LED lighting to 400w of MH or T5.

The bulbs being banned by the US government are not gas discharge lamps like MHs or T5s. It's older types of bulbs like incandescents, which are many fold less efficient than all the above types of lights. All the evidence I've seen pretty clearly shows that is not the case with LEDs vs T5s or MHs. In fact, the only evidence I've found to suggest that LEDs are meaningfully more efficient than T5s comes either in the form of marketing from companies selling LEDs, or tests performed on single diodes under laboratory conditions that are not reflective of real world performance.

 

[Hzuiel]

Most bulb makers have shy'd away from posting the specs of their bulbs, to the point that their "data sheets" basically say nothing but the part number, size, and wattage of the bulbs. It reminds me of disposable batteries, they all fight over who is the best, and advertise being like some gifts from god, but only the off brand batteries tend to post their full data sheets easily accessible to view.

You can get a full data sheet for any cree led that is like 12 pages long. http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Discrete-Directional/XLamp-XPG2

By comparison most t5 makers i looked at don't provide any details worth reading about their products. I found this article that is just about general commercial lighting, which sometimes uses high output t5 bulbs. http://www.lightingtaxdeduction.org/technologies/t5.html

About 5000 lumens for a 54 watt t5, according to this. I do not know how much I believe this, but if it is true, that makes a T5 ho, about 92.59 lumens per watt.

In any case, an xml2-u2 produces 728 lumens at 2amp, and should have a forward voltage of 3.1 volts, which makes for 6.2 watts. That is 117.419 lumens per watt. At 1 amp, your forward voltage is about 2.9-3 volts. 3 volts at 1 amp is about 3 watts. At that power it is putting out 412 lumens, or about 137.3 lumens per watt. You can drive an xml2-u2, to 3000ma at 3.3v, max, and it should put out around 1000 lumens, at about 9.89 watts, but as you can tell this gets less efficient, down to about 101 lumens per watt. Driving them that hard will probably sacrifice longevity as well.

According to this http://www.nofs.navy.mil/about_NOFS/staff/cbl/lumentab.html
A 400 watt metal halide light puts out about 25,000 lumens. That is 62.5 lumens per watt.


It should be mentioned that there is power lost when converting electrical currents, the ballast in a t5 or halide fixture, or the driver for led's loses some power to heat. The only true way to calculate efficiency is to use some sort of monitor to see how much power is being drawn from the wall outlet, then use light meters of various sorts to to see how bright the output of the light is, and what the par readings are underwater, etc.


In general though, now that LED's are becoming a more mature technology, i don't know how there is even a contest. Metal halides should be replaced every 6-12 months. A bulb is between 80-160 dollars for a 400 watt. You can buy an equivalent amount of LED's for 240, that should last at least 5 years(and should be saving you money on electricity that whole time), if not longer(when properly cooled). If one burns out, it's a 7 dollar replacement for the xml's and 3-5 dollars for some of the others. The lighting fixtures for led are simpler(and in some cases cheaper), just a heatsink, fan, and power drivers. They allow for open top. Some people that run halides are running chillers. Chillers are not free, nor do they consume no power. The amount of flexibility that is possible with LED's is incredible too, being able to add LED's of different spectrums and put them on multiple channels so you can dim or brighten them for fine tuning. Ease of directly controlling multiple channels of LEDs to an anal retentive degree if so desired, through a controller.

 

[sirreal63]

Metal halides should be replaced every 6-12 months.

Only if the bulb is being over driven, which few people do. Many of us go well over a year between bulb changes and the majority of people do not use 400 watt bulbs, we use 250's. I am going to try going two years on my current bulb, as some people I know have done this, measured the output and have noticed no major difference. This is also in line with this article, written many years ago, and is still very relevant today.
http://www.reefkeeping.com/issues/2007-04/ac/index.php

Over the years I have noticed that vast majority of LED pushers take a worst case scenario for MH, the 400 watt bulb being over driven, and use it as a basis to show how horrible MH is. Yet that isn't the real world most reefers live. They also have used such baseless comparisons as this LED is equal to a 250 watt bulb, but they never say which bulb or ballast or give any specifics. The variability in output between the worst performing 250 watt MH and the best is a vast gulf, not a small creek.

Some things against LED's, many people are discovering they did not save any money by switching to LED's, they in fact lost money on replacing an older yet excellent performing light, and gained nothing except poor results and a smaller bank account.

Speaking of efficiency, the LED pushers seem to be mostly new reefers who have no understanding of bulb, ballast and reflector combinations and how to use that to their advantage, they only know LED, and on paper LED shines bright. In reality a properly driven MH bulb in a properly designed reflector covers a lot more area than the LED fixtures available. It is using this experience that allows me to completely cover a 40x40 area with a single 250 watt bulb. Not just cover with light, but a good even spread of PAR.

Very few LED fixtures even cover a 24x24 area with adequate PAR, let alone the 36x36 area some tout, yet when you look at their distribution graphs, the area beyond 30" is getting zero PAR, corals dislike zero PAR, and the PAR inside the 24x24 area is mostly a hotspot and falls off rapidly beyond a 12x12 area. There is nothing equivalent in that, just hype and marketing. The saddest thing is some of the cheapest fixtures out there do a better job at spreading out the light than the expensive ones with all of the bells and whistles and bling.

Though I am very much in favor of the LED tech growing for the hobby, sorting out the fluff, BS and marketing is hard for most people, they fall for it like a mouse to a trap with cheese.

 

[Hzuiel]

Only if the bulb is being over driven, which few people do. Many of us go well over a year between bulb changes and the majority of people do not use 400 watt bulbs, we use 250's. I am going to try going two years on my current bulb, as some people I know have done this, measured the output and have noticed no major difference. This is also in line with this article, written many years ago, and is still very relevant today.
http://www.reefkeeping.com/issues/2007-04/ac/index.php

Over the years I have noticed that vast majority of LED pushers take a worst case scenario for MH, the 400 watt bulb being over driven, and use it as a basis to show how horrible MH is. Yet that isn't the real world most reefers live. They also have used such baseless comparisons as this LED is equal to a 250 watt bulb, but they never say which bulb or ballast or give any specifics. The variability in output between the worst performing 250 watt MH and the best is a vast gulf, not a small creek.

Some things against LED's, many people are discovering they did not save any money by switching to LED's, they in fact lost money on replacing an older yet excellent performing light, and gained nothing except poor results and a smaller bank account.

Speaking of efficiency, the LED pushers seem to be mostly new reefers who have no understanding of bulb, ballast and reflector combinations and how to use that to their advantage, they only know LED, and on paper LED shines bright. In reality a properly driven MH bulb in a properly designed reflector covers a lot more area than the LED fixtures available. It is using this experience that allows me to completely cover a 40x40 area with a single 250 watt bulb. Not just cover with light, but a good even spread of PAR.

Very few LED fixtures even cover a 24x24 area with adequate PAR, let alone the 36x36 area some tout, yet when you look at their distribution graphs, the area beyond 30" is getting zero PAR, corals dislike zero PAR, and the PAR inside the 24x24 area is mostly a hotspot and falls off rapidly beyond a 12x12 area. There is nothing equivalent in that, just hype and marketing. The saddest thing is some of the cheapest fixtures out there do a better job at spreading out the light than the expensive ones with all of the bells and whistles and bling.

Though I am very much in favor of the LED tech growing for the hobby, sorting out the fluff, BS and marketing is hard for most people, they fall for it like a mouse to a trap with cheese.

I was going by marine depot's replacement guide which actually i realize now says 9-12 months on metal halide, and 9-18 with t5ho. Anyway, if it's 1.5-2 years, it's still less than 5ish.

According to that chart a 250 watt halide is less efficient, at 14333 lumens for 250 watts, which is 52.32 lumens per watt.

"but they never say which bulb or ballast or give any specifics." If the manufacturers own data sheet doesn't have any specifics, how am i supposed to have them? Compared to some companies, ushio actually has specs for their bulbs. Here is an example. http://www.ushio.com/files/specs/Aqualite.pdf

10600 or 11000 lumens for 250 watt. That is 42.4 or 44 lumens per watt. 6000 hour recommended operating life, which is about 2 years at 8 hours per day. 23-42% lumen loss at 6000 hours. At only 2000 hours 11-24%. Kelvin loss of around the low end of the lumen loss figures, per hours of operation.

That data doesn't look good at all, but like i already said, you can only gauge actual power consumption at the outlet, and actual performance output with light sensing instruments inside the tank.


As far as coverage goes, that is entirely subjective to what you buy. I did my fixture diy, and that is only limited by what lengths of wire you use and how big of a heatsink you buy. Obviously a puck or pendant is going to have a harder time spreading out, and the height required to achieve a certain spread is usually listed by the manufacturer. There are LED lenses and diffusers as well. LED's are easy to dim via controller, can be split into multiple channels for different levels of light in different areas or at different times of day. If you are anal retentive enough, you could make a separate channel of LEDs with narrow lenses to spotlight one coral with exact spectrums and intensity.

If people just take a huge wad of cash and buy the most expensive LED fixtures without researching and planning, and the result is incinerating corals in one spot and withering ones in a dark spot, is that the fault of the technology? Or is it the fault of the person? If the corals don't grow well, again poor planning is the culprit. If they spend too much money because they didn't shop around, or they assumed more $ = better, again that is not anybody's fault but theirs. Suckered in by marketing gimmicks? It's their fault.

The bad experiences of people who did not research their purchases and utilize the full potential of the technology are not the bar by which to judge LEDs

 

[sirreal63]

I was going by marine depot's replacement guide which actually i realize now says 9-12 months on metal halide, and 9-18 with t5ho. Anyway, if it's 1.5-2 years, it's still less than 5ish.

They sell bulbs, not exactly unbiased because they have a financial stake in the advice they give. How many people have actually kept their LED for 5 years? The Solaris came out in 2006, how many of those are still in use? How many first gen AI's? Why?

According to that chart a 250 watt halide is less efficient, at 14333 lumens for 250 watts, which is 52.32 lumens per watt.

Efficiency is more than numbers on paper, if you can light a larger area with less electrical efficiency, then you are being more efficient in the use of the product, this has advantages over more electrical efficient but less capable LED's.

"but they never say which bulb or ballast or give any specifics." If the manufacturers own data sheet doesn't have any specifics, how am i supposed to have them? Compared to some companies, ushio actually has specs for their bulbs. Here is an example. http://www.ushio.com/files/specs/Aqualite.pdf

You are not understanding, many mfg's of LED fixtures would claim it was equal to a 250 watt MH, but would never say what they tested against, some 250's have pathetically low PAR, like under 20 and some have over 100. They would never say because they used it as a marketing ploy, not a reality. I prefer to deal in reality.

10600 or 11000 lumens for 250 watt. That is 42.4 or 44 lumens per watt. 6000 hour recommended operating life, which is about 2 years at 8 hours per day. 23-42% lumen loss at 6000 hours. At only 2000 hours 11-24%. Kelvin loss of around the low end of the lumen loss figures, per hours of operation.
Irrelevant, that is why we change bulbs and it isn't $160 bucks a piece, I pay $80 for a Radium, if it makes two years, that is $40 a year. I frink more coffee than that in a few weeks. Comparatively, it is really cheap.

That data doesn't look good at all, but like i already said, you can only gauge actual power consumption at the outlet, and actual performance output with light sensing instruments inside the tank.

Sanjay has tested the most popular ballasts and bulbs showing the wattage draw at the outlet. I am getting the feeling you have zero experience with MH. Is that a reality?


As far as coverage goes, that is entirely subjective to what you buy. I did my fixture diy, and that is only limited by what lengths of wire you use and how big of a heatsink you buy. Obviously a puck or pendant is going to have a harder time spreading out, and the height required to achieve a certain spread is usually listed by the manufacturer. There are LED lenses and diffusers as well. LED's are easy to dim via controller, can be split into multiple channels for different levels of light in different areas or at different times of day. If you are anal retentive enough, you could make a separate channel of LEDs with narrow lenses to spotlight one coral with exact spectrums and intensity.

Most people buy a fixture for lack of ability, time or desire to DIY, this is just a reality.

If people just take a huge wad of cash and buy the most expensive LED fixtures without researching and planning, and the result is incinerating corals in one spot and withering ones in a dark spot, is that the fault of the technology? Or is it the fault of the person? If the corals don't grow well, again poor planning is the culprit. If they spend too much money because they didn't shop around, or they assumed more $ = better, again that is not anybody's fault but theirs.

The bad experiences of people who did not research their purchases and utilize the full potential of the technology are not the bar by which to judge LEDs

Perhaps you can create a guide from your research, and take that and go back in time to around 2006 so that no one from then until now can make a poor decision. You do realize that LED's have been evolving, and evolution takes time, failures and successes. Most hobbyists don't have time to pour over information like a few others do, they trust that what they are buying is good from the start. For you to say they made a bad decision and it is their own fault is silly, unless you are willing to build everyone who wants to replace the units you feel are a bad choice with one of your DIY's, for free of course. All of us need to thank those who were the early adopters of LED's, their dollars, failures and successes are what led us where we are today. For a long time in this hobby, the DIY crowd has done the most research and development, and that is usually adopted by mfg.'s in the next great LED of this year, which will be replaced by next years latest and greatest. Meanwhile many of us are still running our old and antiquated equipment that works as flawlessly today as it did when new, many years earlier.

I am still guessing you have virtually no experience with MH.

 

[sirreal63]

Grokthecube, the answer would have to be maybe, depending on the setup and how you define efficiency. On paper LEDs are the clear winner, yet over your tank you may get better results with T5 or MH, as far as spread and correct spectrum are concerned. If the goal is purely electrical efficiency, go with LEDs, if your goals are proper time tested spectrum, great coverage and removing light as a variable to the success of your tank, then perhaps traditional light sources are what you want.

 

[GroktheCube]

Most bulb makers have shy'd away from posting the specs of their bulbs, to the point that their "data sheets" basically say nothing but the part number, size, and wattage of the bulbs. It reminds me of disposable batteries, they all fight over who is the best, and advertise being like some gifts from god, but only the off brand batteries tend to post their full data sheets easily accessible to view.

You can get a full data sheet for any cree led that is like 12 pages long. http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Discrete-Directional/XLamp-XPG2

By comparison most t5 makers i looked at don't provide any details worth reading about their products. I found this article that is just about general commercial lighting, which sometimes uses high output t5 bulbs. http://www.lightingtaxdeduction.org/technologies/t5.html

About 5000 lumens for a 54 watt t5, according to this. I do not know how much I believe this, but if it is true, that makes a T5 ho, about 92.59 lumens per watt.

In any case, an xml2-u2 produces 728 lumens at 2amp, and should have a forward voltage of 3.1 volts, which makes for 6.2 watts. That is 117.419 lumens per watt. At 1 amp, your forward voltage is about 2.9-3 volts. 3 volts at 1 amp is about 3 watts. At that power it is putting out 412 lumens, or about 137.3 lumens per watt. You can drive an xml2-u2, to 3000ma at 3.3v, max, and it should put out around 1000 lumens, at about 9.89 watts, but as you can tell this gets less efficient, down to about 101 lumens per watt. Driving them that hard will probably sacrifice longevity as well.

According to this http://www.nofs.navy.mil/about_NOFS/staff/cbl/lumentab.html
A 400 watt metal halide light puts out about 25,000 lumens. That is 62.5 lumens per watt.


It should be mentioned that there is power lost when converting electrical currents, the ballast in a t5 or halide fixture, or the driver for led's loses some power to heat. The only true way to calculate efficiency is to use some sort of monitor to see how much power is being drawn from the wall outlet, then use light meters of various sorts to to see how bright the output of the light is, and what the par readings are underwater, etc.

The issue with manufacturer data sheets, and the reason for my OP, was that the relevant factors beyond the bulb/diode are not accounted for. Unless I'm very mistaken, Cree's numbers are from "perfect" lab conditions, without accounting for losses from wiring, drivers, temperature, etc. I've seen those data sheets before which is why I was very surprised when I did some reading and found that LEDs usually don't approach that level of efficiency in the real world.

In general though, now that LED's are becoming a more mature technology, i don't know how there is even a contest. Metal halides should be replaced every 6-12 months. A bulb is between 80-160 dollars for a 400 watt. You can buy an equivalent amount of LED's for 240, that should last at least 5 years(and should be saving you money on electricity that whole time), if not longer(when properly cooled). If one burns out, it's a 7 dollar replacement for the xml's and 3-5 dollars for some of the others. The lighting fixtures for led are simpler(and in some cases cheaper), just a heatsink, fan, and power drivers. They allow for open top. Some people that run halides are running chillers. Chillers are not free, nor do they consume no power. The amount of flexibility that is possible with LED's is incredible too, being able to add LED's of different spectrums and put them on multiple channels so you can dim or brighten them for fine tuning. Ease of directly controlling multiple channels of LEDs to an anal retentive degree if so desired, through a controller.

I agree that LEDs have improved substantially in recent years. I have little doubt that in fairly short order they will be consistently and substantially more efficient than old-school lighting sources in the real world.

I'm not an LED hater. I've seen far too many absolutely world class tanks lit primarily or solely by LEDs to say that they can't grow corals. My little Maxi-Mini nano is lit exclusively by a PAR 38 LED bulb, and it's been absolutely perfect on that tank. I use two BML strips as supplemental lighting on my 120 mixed reef. I believe that LEDs are the way of the future.

I will say that a substantial portion of people seem to have problems with them, and for whatever reason, have better results with "old school" T5 or MH lighting. I think the biggest reasons for this are coverage/spread (LEDs are extremely directional) and spectrum (lots of narrow peaks, often weak below 450nm). Both of these can be corrected, and I think they're a big part of why many of the best LED lit tanks I've seen have been lit with custom/DIY LED systems that used a far greater number and variety of diodes than pre-fab lighting manufacturers usually recommend.

I'm now wondering if pre-fab LED manufacturers have also been recommending too few watts of lighting to their customers, in addition to often suggesting too few diodes, too little spread, and not broad enough spectrum.

I do agree with your later post that the issue with LED use is typically implementation, rather than being the technology itself.