LED Tank Lighting Experiment with 1st look at results...

I found a source that explains the difference between Lux and Lumen with relations to LED lighting...

http://www.brillianz...ments/Lumen.pdf

Quoting some of the text -

"To clarify the relationship of watts to light output and LED bulbs, we talk about the number of 'lumens per watt' (lm/w) that the device produces. Incandescent bulbs are very inefficient and only generate around 12 lm/w, whereas LEDs and halogen sources now have almost equivalent efficiencies of 40 lm/w. Some fluorescent tubes generate 60 lm/w (which is where LEDs arepredicted to be next year!)

It is always better to talk about lumens (and lux required)than watts when it comes to lighting decisions.

Another point to bear in mind is that unlike incandescent light bulbs that illuminate in a 360-degree spherical pattern regardless of the shape of the bulb, LEDs depend on the shape and composition of their lenses and package structure to direct the light where it is needed.

When we illuminate an area with a 'spherical bulb' source,even with a reflector, only a small proportion (perhaps 50% at best) of the lumens are delivered as lux to the surface we are lighting – the rest is lost

With LEDs, properly applied, over 80% of the lumens can be directed to provide lux at the surface. That is why LEDs can seem such bright pinpoints when you look at them. So LEDs are as bright per watt as halogen and nearly as bright as fluorescent – and when properly applied they can deliver more lux per watt than either of these onto a surface such as a display case floor."

Looks like Lumen gives us a better comparison taking into consideration LED characteristics but angle, area and distance also seem important. Not so clear. I think that the important thing is that we establish a standard to use and stick to it in all of our calculation so we can standarize our results.

So - for Lumen calculation we can take the Lux value and stick it into this calculator:

http://www.ledrise.c...ent.php?coID=19

This adds a little more to the understanding.

http://www.intl-ligh...r/luxlumenscalc

And this one.

http://online.unitco...lumination.html

I will look at all these and create a table for my light once I deal with the issue of putting the light meter on the gravel inside my tank. I need to do these measurements at night as the sun comes through the window during the day and would skew the data during the day.

Usil

Edited by Usil, 24 September 2011 - 12:13 PM.

Usil, although I have experience in photography, to my shame, I have to admit that I've never studied lighting in this context before, so for my own clarification I need to ask; Am I correct in gathering from this that we can say that Lux is the measurable amount of Lumens actually reaching our target? The target being the aquarium water.

Thanks for the work.

Steve.

The T8 spectrum posted above is just one spectrum from one T8 bulb. Each bulb design can have an extremely different spectrum, with some designed to be office white, some full spectrum, and some specifically designed with peaks in the red and blue to grow plants.

Since we're sharing what lights we use, I'll tell you about my current setup. I don't have a way to split the light into its spectrum, but I do know from over a year of using these bulbs that aquarium plants love them. Every one of the dozen or so species I've put under them has grown very well. It would be inexpensive for other NANFA members to use this setup and be able to grow aquarium plants if they so desired.

The bulbs I use are $8 for two bulbs at Home Depot. They are Lithonia 32 watt, four foot long T8 fluorescent bulbs, which fit into a 4 foot shop light that costs $20. It is a very convenient light for aquariums, being exactly the four feet in length that a 55 gallon fish tank is. Here is a photo of the light on my tank: http://img.photobuck...imiru/039-1.jpg If you buy two, they cover the 55 gallon tank's top so no fish can jump out.

Link to light fixture: http://www.homedepot...catalogId=10053
"Lithonia Lighting All Weather 4 Ft. 2 Light T8 Fluorescent Unit Shop light"

Link to bulbs:
These bulbs cost about $6 to $8 for two (the price on the link is for a 10 pack). I replace mine about once or twice a year.
1. http://www.homedepot...catalogId=10053 "Philips 32-Watt 4 ft. T8 Daylight Deluxe Linear ALTO Fluorescent Light Bulbs" 2,750 lumens, 6500 K, 24,000 hours of life
2. http://www.homedepot...=1&ddkey=Search "Philips 32-Watt Neutral Full Spectrum Linear Fluorescent Light Bulb" 2850 lumens, 5000 K, 20,000 hours of life.

There are like a dozen different 32 watt T8 bulbs to choose from, but both of those above work. Avoid any with less than 5000 K; they're yellow and they look icky in my opinion. For a white-white color, you want 5,000 K and above. Here's an image from wikipedia that demonstrates what I'm talking about: http://en.wikipedia....-comparison.png

Bulbs usually use lumens, not lux, as a measurement of their brightness. I'm not sure what these bulbs' light output is in lux. Lux seems an odd way to measure things since it is lumens/square meter.

This conversation is great and got my curiosity elevated. I went looking for other waterproof LEDs and came cross this page. I have racks of tanks for breeding and as such just want to light them as cheap as possible. Plus with LEDs you have a long life span.

http://www.buyonme.c...2v-p-62420.html


Bill F.

That looks interesting if you need lots of lights. One of the 'problems' I think we will run into is that these LED lights are actually too efficient. It will provide too much light in the tank and surpass what the normal plants will need while enhancing the growth of algae which we want to avoid. How do you know this without testing light setups?

I am considering building a PAR meter using an RGB photodiode so that I can measure PAR values inside the tank. This is ambicious and have no idea if I can do this but am already investigating. I understand electronics so will work on this for a winter project.

Usil

I have spent some time looking at how to measure PAR from light sources and keep coming back to thinking that it might just be easiest to buy a PAR meter. Still, I am not there yet because of the cost. Then I found this artical. It is an inexpensive spectrophotometer that would allow you to view and compare the spectrum of your aquarium lights which should be very useful information for us as we make light comparisons. You have to have the correct spectrum first for growing plants before doing anything else and this looks like it might allow us to see if cheap LED lighting is providing the light in the correct spectra. This is a good start. It would also allow you to measure and record the differences over time as your lights become less useful with age. Here is the artical.

http://www.advanceda...006/7/aafeature

Here is a link to the Spectrometer.

http://www.scientifi...ectrometer.html

I have ordered one to experiment with. While this does provide for a means to review the spectrum of light provided by a light source (I will use this instead of the diffraction grating), it does not measure the intensity of that light recieved by the plants. Still need a PAR meter to do that. Still, this will provide useful information.


Usil

Edited by Usil, 25 September 2011 - 08:53 AM.

I found a nice source explaining LED light and thier spectra on the Olympus Microscope website.

http://www.olympusmi...tion/index.html

http://www.olympusmi...urcesintro.html

In addition, they have a neat software download for subrtracting the background of a photographed image. Very easy to use and interactive. Might be useful for those that photograph fish.

http://www.olympusmi...nddownload.html


Usil

Edited by Usil, 25 September 2011 - 09:41 PM.

Yes, I just noticed that too. It looks like the links were truncated for some unknown reason and that broke the links. And, to make matters worse, for some reason - maybe the admins can tell me - I can't edit my post. I remember editing it in the past. Now I can't fix the links. I have never seen the edit button go away on my posts in other forums I am a member. Is there some reason for this? It would be much better to have the edit button not go away. Any way to get it back?

Usil

...to make matters worse, for some reason - maybe the admins can tell me - I can't edit my post. I remember editing it in the past. Now I can't fix the links. I have never seen the edit button go away on my posts in other forums I am a member. Is there some reason for this? It would be much better to have the edit button not go away. Any way to get it back?

There seems to be a maximum amount of time after the post was last posted in which you can edit it. I don't like it either; sometimes I come back later and realize I spelled something horribly wrong and then can't change it. But maybe it makes the forum take up less space or there is some reason for it that I don't know about. Drew would be the person to ask.

Actually, it makes you take up more space since to make modifications or corrections you then have to post entirely new posts. This software that the forum is using is the same used by all the others I am on. I am sure it is just a quick option change to bring the Edit button back. Hope that he can do this.


Usil

Ok - Back to the DATA and this project.

I have spent hours pouring over hundreds of links with lighting information. To say there is a lot of information on fresh water and salt water tank lighting needs is an understatement. There is also a lot of old information that is no longer useful with respect to tank needs i.e. watts/gallon rule. LED tank lighting is rewriting the book based on their effectiveness and output.

Preliminaries:
I did my tank experiment tonight and measured the EV values underwater. I took measurements using my Sekonic light meter in a clear plastic zip lock bag. Measurements were taken at 4" under water, 8" under water and on the bottom at 16" under water. I also took readings at 1" from the fluorescent light (air reading) and the same for the LED, however the LED presented a problem in that it has 30 emitters. I held the meter 1" from one LED emitter for this test.

The light meter in the bag produced a EV reading that was 0.4 less than if the meter were not in the bag so using this as a constant, I adjusted all readings taken in the water by this value.

After much searching, I found a value for the lumens for the Perfect-A-Lamp F15T8 bulb I have. It is rated new at 700 lumens by the manufacturer.

The Table below contains my data:


Note that the LED lighting has double the efficiency of the fluorescent bulb in penetrating the water. This is expected as the emitters on the LED light are completely focussed downward providing nearly 100% of the light output downward. The fluorescent, however, distributes it's light in a 360 degree angle so only about 30% gets directed downward plus what might be directed from the white reflector of the lamp housing. This illustrates one of the benefits for using LED lighting. Teaming this up with the LED low wattage, it appears that we gain several benefits.

The air readings I took of the fluorescent and LED gave these results:
Fluorescent EV = 11.5; Lux = 7140; Lumen = 672.6 (Interesting to note and encouraging that the Manufacturers rated Lumen value was 700).
LED EV = 12.0 (I used only 1 of the 30 emitters for this test); Lux = 10,240; Lumen = 951.3

Using this link: http://books.google....epage&q&f=false
I found a statement that says: "Plants that grow on the bottom need about 300 lux. Plants that grow upright in a tank need about 1500 lux. Most other plants fall in between these values. " This might be useful for reviewing the data.

The current lighting scheme I am using is pictured below:


I am going to order a second light strip and do the same thing for the second tank light. Currently, the vendor is out of stock for this light so I am waiting.

I am tired so I leave this for your review. If we look at the manufacturer's Lumen reading of 700 for the bulb I use and compare it to the calculated Lumen value of the 'air test' I did (672.6 lumen) this appears to indicate that the data in the table may be close enough to start making some preliminary 'total' light measurements in a tank. Understanding that this is only a first step and does not relate to the PAR or PUR values we ultimately need, it does give us a start for overal visual comparisons. I have ordered a low cost spectrophotometer from Edmond Scientific so that eventually we can move to the second step of measuring light spectra.


Usil

Edited by Usil, 26 September 2011 - 11:29 PM.

It's really interesting that only a fifth of the fluorescent light reached the bottom. That's such a dramatic difference. I understand now why foreground plants that grow close to the ground have such high light requirements; if only a fifth of the light reaches the bottom, then even if their light requirements are the same as a plant in the middle of the water column, the initial light would have to be much more bright to penetrate that far into the water.

That LED has quite an advantage over the fluorescent by having 36% instead of 18% of its initial light value all the way on the bottom of the tank. It's enough to make me want my next light to be LED.

What's your fluorescent light like? Does it have reflectors?

The tank light is pretty standard. Here are two shots of it. The 'reflector' part is just the standard white casing shell for the light.






Usil

I just ordered the 3w version.
http://www.dealextre...c-12v-900116211
I will post when I receive them. the were $6.13 ea when I bought 5 of them. My tanks have jave fern, java moss,najas grass, anubias and duckweed. I think I will have enough light to keep those growing. I am going to use these for a rack I am going to build. I like the tiny size. I may mount them in a length of aluminum U channel. This will give me more roon between shelves to work.

The tank light is pretty standard. Here are two shots of it. The 'reflector' part is just the standard white casing shell for the light.

Ick, yeah, the lights that come built for aquariums are never (in my opinion) built very well. A single bulb, with no reflector, and so much of the space for the light fixture taken up by plastic housing and not bulbs....well, that's about the worst setup for a light that you can find. I prefer lights that have a reflector all the way around and two bulbs that go the entire length of the fixture. Picture: http://fp.vendaria.c...itle=&pf=t&fr=t

Maybe you could add some aluminum foil to your light housing to make it reflect more? But really, yeah, I can see why you bought those LEDs. *nods*

Edited by EricaWieser, 27 September 2011 - 02:29 PM.

Aluminum foil provides only minimal increases as reported by others. Actually, the LED project has many reasons I am pursuing but ultimately it will be a more efficient system for me.


Usil

I just ordered the 3w version.
http://www.dealextre...c-12v-900116211
I will post when I receive them.

What SKU did you order as I do not see one unless it is the blue. Please describe how you will use them when you get them and evaluate your needs. Also, take pictures of before and after if you can. Do you have a light meter to take EV values in the tank before and after? White LEDs are generally rich in blue and I am hoping to confirm this when I get my spectrophtometer. An unknown is how much red is delivered.

It appears that many mount the LED lights on channels hanging them above the tank by several feet. This is for several reasons. First, the commercial Aquarium LED lights are powerful emiters made by Cree who combine InGaN with proprietary G-Sic sibstrates to deliver high intensity LEDs. These LEDs burn hot and require a substantial heat sink which is why DIY users mount them on Aluminum channel. Most needed materials for DIY can be bought at a local hardware store and CREE LED modules can be purchased from Deal Extreame for high powered flash lights. Not too difficult to do but these high powered LED units go far beyond what I am trying to do using a much lower powered and lower cost LED strips.

Here is a ghigh powered LED DIY project describing what I am talking about. He will add white and maybe even a few red to optimize his setup before he is done.



My thoughts on my project were "What can we obtain from low powered LED strips" as they seem to now be available for very little money. These do not heat up like the high powered Cree Modules and do not need heat sinks. In addition, they require only minimal power requirements. The Cree modules will require LED driver circuitry, substantial heat sinks and larger power supplies. The upside of all this is that as powerful LED lights are developed at the high end, the low end get the benefits of the design and manufacturing requirements. Hence, why not test the low end LED strips and see what is produced which is the thrust of this project.

Keep us posted on how your project works out.

Usil

I have not recieved the spectrophotometer yet but it is on it's way. I had forgotten I had already ordered the holographic diffraction grating and I recieved it today.

I just did a preliminary peek at the white LED lighting and I see a full spectrum from beyond blue to beyond red with everything in between. I will try to be more quantitative tomorrow but it will still be by visual observation. I will see if I can photograph it and post it.

Usil

Edited by Usil, 29 September 2011 - 11:54 PM.

Well, it is dark so I thought I would do a quick look at the spectra of the lights used in my aquarium. As noted before my aquarium originally came with two florescent bulbs marked Perfect-A-Lamp F15T8 (18 inches long and 110 volts/15 watts each) rated at 700 lumen by the manufacturer. I tested this bulb (in air) in a previous post and measured and calculated a lumen value of 672.6. As the bulb is not new this is not an unreasonable value to obtain. It is supposed to be similar to a Gro-lux bulb providing critical action spectra for growing plants.

I have added to that on one side a strip light of 30 white LEDs rated at 12 volts/4.5 watts. I tested and calculated the lumen value (in air) for the LED light at 951.3 lumen.

Also previously noted, the LED has an inherent benefit of intensity as they are designed to have nearly 100% of it's light focused downward which is advantageous in an aquarium where you want as much of the light as possible directed downward into the water. This is not the case for any florescent lights as they direct their light in a 360 degree angle and only have about 30% of it's light pointed downward. So, right off the bat, they have a design issue. Even with better reflectors above florescent lights, users report that only minimally they get slightly better results.

A direct benefit of the LED design is that they are able to measurable penetrate the water inside the aquarium with much more efficiency. My data table, previously posted shows that in my experimental case, the amount of light reaching the bottom for the LED light (18 inches to the top of the gravel) was 44% of the total output in air. The florescent light had only 29% of it's light reaching the bottom. A value of 300 lumen (minimally) is reported needed for plants at the bottom of a freshwater aquarium. The florescent light alone provides only 196.9 lumen while the LED alone provides for 422.2 lumen. If the reference is acurate it looks like these low power LED's do provide sufficient light for bottom plants. (Note - we are only talking about fresh water aquariums here. Salt water aquarium have different needs and standards).

The action spectra needed by plants is summed up in the next two images.





The spectra utilized by the plants incorporate regions of the blue and red spectra. Plants are green because they reflect green light.



Utilizing a holographic diffraction grating I just bought and a camera I took pictures of the spectral output of the florescent and LED bulbs I am using. The first one is the florescent bulb.


You will note that there are indeed spectral lines in the red and blue zones and this is supposedly by design by the manufacturer who makes claims that their bulbs support plant growth. The intensity is not very bright and the red and blue regions are not very wide but they do exist. The next image is for my LED light setup.


You can see right away that these white LED lights are a full spectrum bulb that incorporate much of the blue and red zones. As long as these red and blue zones are supplied they will provide for the needs of plant growth. The benefit of being a full spectrum bulb is that you get a normal appearance when viewing your tank. If the bulbs were specifically designed to only provide for plant growth they would look like this.


Your plants would grow well but I am not sure I would want to look at the aquarium in this kind of light. The plants need the red and blue spectra but for us humans we prefer to see light as it comes from the sun in a full spectra.

I have a LED flashlight that uses a CREE high power module in it. These high power LED CREE modules are used in the expensive commercial LED light systems now proliferating in the market. I did a spectral image for it as well.


Note that it is very bright compared to the low powered LED unit but it too has a full spectrum supporting the red and blue regions very well.

I think that this part of the experiment is coming to a conclusion. The low powered LED lighting unit I have added to my aquarium does indeed support the active red and blue zones and it does it in a far more efficient way than the florescent bulbs, in terms of cost effectiveness, power requirements, output of light in the active zone regions, light penetration to the bottom of the aquarium and longevity of the bulbs in use. To me, it is not even a close call. In fact, the florescent bulbs have a puny output compared to the LED set-up. So much so, that I am going to remove entirely the florescent fixtures in their housings and totally replace them with an LED light system. I will duplicate the system for each light housing.

I hope that this provides useful information for others with a little DIY experience and the desire to make significant improvement in growing plants in their aquaria. I may be able to add a bit of extra information to this when I get a spectrophotometer I ordered from Edmond Scientific. It may be able to directly tell me the exact spectral frequencies supplied by these lights. As the spectrum appears to cover most of the red and blue regions I think that we will see that the action spectra are well represented. I will post the data when I make the final tests.

If you have a camera light meter capable of providing EV readings or you purchase a low cost Lux meter like this one: http://www.amazon.co...17436719&sr=8-1 plus the low cost holographic diffraction grating I used, you should be able to duplicate this experiment and provide for your own measurements as you design or adapt a LED lighting system for you tanks.

Usil

Edited by Usil, 30 September 2011 - 09:59 PM.

I've been following your experiments closely and eagerly awaiting the information you get at this stage. The whole time I've been thinking that the main hangup could be in whether or not the LED's have the appropriate frequencies to grow plants. It appears that they certainly do. This is some impressive work and quite worthy of some serious consideration. Thank you for your time and work on this.

Do you have any live plants in your aquarium? If not, do you think you'll get some to see how well they do?

Thanks again for some really good info. we all can use.

Steve.

Frogwhacker - If you go to my Fish Cam link this weekend during the day you can see my 55 gallon aqaurium in real time and the plants (live).

http://71.170.193.21...ndex.html?cam=0

There are live plants in the aquarium. Notice the ones on the right compared to the one on the left. All of these were about 3 to 4 inches tall and planted at the same time. The center and right side of the tank (with the lighter colored wood) has the LED lighting and the left side (darker drift wood) only the florescent light. The one plant on the far left is rather sickly compared to all the new growth you see on the plants in the center (mid section and back wall) which are exposed to the LED light. This is only for a two week period when I first planted them. (The plants next to the slate and Java Fern on the dark driftwood are newer and have not had the exposure the others have.)

So, experimentally, there is already evidence that the LED light is doing it's job very well while the florescent provides anemic light that seems to place the plants in a stressed condition. I do not add CO2 to the tank.



Usil

Edited by Usil, 30 September 2011 - 11:15 PM.