Open letter to the LED industry

[conix67]

There are many issues..............hot spots, spectrum, lack of coverage, shadowing to much or too little par.

A lot depends on the design also........discs, banks of diodes, strips of diodes, ect.

That's all outside of getting the spectrum right which is challenging because they are working with single color diodes.

For me, I think single color diodes will never work for our application. Mixed color matrix type chips are when they may finally get a clue.

I'm still using T5's as you are because nothing out there has come close to meeting my expectations.

If people are already knee deep in spending on LEDs, supplementing is a good option.

Here's a thread I started a while back..............there's a lot of info that has been added and a ton of links---

http://reefcentral.com/forums/showthread.php?t=2198810

Great link. I guess based on the known "existing" problems, the only one that is unique to "LED" is, again, the spectrum coverage. All other issues also exist on MH fixtures.

Considering how "UV" is perceived as harmful to corals, the lack of spectrum coverage is what's covering upper range of 400nm to 500nm?

 

[conix67]

For me when I'm looking at a light fixture I'm basically looking for spectrum and intensity above all else. Then I'll look at long range cost over roughly 5 years including electrical usage. The extra bells and whistles like multiple channels, storm effect, and fade in fade out, remote control I view as added possibilities of something to go wrong and count them as undesirables.

Spectrum and intensity is crucial, but I would argue the fact that you can control intensity is also big benefit which is not possible with MH fixtures. All other bells and whistles are derived from the dimming capability, and costs simple solid state electronics plus microcontroller which should outlast all other components so I wouldn't worry too much about its reliability.

 

[TropTrea]

It is possible to build a light with 420, 430, 455, 475 and 495 leds using two suppliers. The lumia you went with is close to what you consider optimum. I'm not sure where you would find 410 violets.

Yes there are 410's as well as 380's out there. if you go through the Chip manufacturers data you will find them. However most DIY'ers use the pre made up stars and I have not found many people that sell tem as they are low demand item. There are also 430's out there but I yet to find a quality chip in 430 range.

Often they list the chips as 410-420 and in most cases the chip manufacturer has various bins numbers in this range and they will buy which ever is lower cost at the time.

As hint you can look through medical grade chips and find a much larger selection. But I doubt anyone would want to pay for the quality of medical grade LED's for an aquarium. Basically you would add another digit to the cost of the system.

Are you saying that the warm whites do not offer enough light in the green/yellow range for good colour balance? I'm still trying to get a handle on what is really needed in a diy build.

No the Yellow Green is not the issue with warm whites. It is that they produce a lot of red light. If not used sparingly they could very easily produce so much red red light that your geting a negative return from them. Neutral whites also produce the same red light but at a much lower level which makes them easier to balance in with the Blue end LED's.

All white however show a dip in there spectrum in the blue green area of the spectrum, near 500nm. This is why I recommend a little supplementation with blue/green chips int the 485 to 495 nm area,

Right now each manufacturer has their own idea of the ideal mix. Over time those differences will converge based on user experience. That doesn't really help for someone who is equipping a system today though

I do not think it is that easy. When you have complete control of the lighting you can have four separate areas of concern.
1. Coral Growth
2. Florescent Color
3. Reflective Color.
4. Usurers color preference

The middle two can be very subjective as when you increase one you decrease the other. What is the balance between the two you want to maintain. Some individuals complain because the reflective colors are washed out, but there florescence look fantastic, Other complain their reflective colors are fantastic but they do not have the spark as the other guy has from the florescence.

This is why I recommend two channels. One to vary the LED's that boast the florescence and the other for the LED's that boast the reflective colors.

Oh and then for the low cost manufacturers there is the cost difference. They are looking to keep down cost and to do so having the least number of the lowest cost components is important. There are still systems out there only using Royal Blue and Cool White LED's because they can get these at the lowest cost and still create a fixture that shows most corals acceptably even though it is not at there peak for any of the four above variable.

 

[Fredfish]

...
So, if the new gen stuff doesn't work as well, will everyone just say it was "a bad example" to evaluate LEDs by?

Well, that's my fallback position... :crazy1:

If manufacturers don't bother to apply what we have learned about corals and light, I suppose it is inevitable that the next gen will fail again. That would surprise me though.

Halides went through a similar evolution on their way to becoming a gold standard. As I recall, the early adopters in the '80s were using 4000k halides over their tanks nobody would think of using now.

 

[rtparty]

Well, that's my fallback position... :crazy1:

If manufacturers don't bother to apply what we have learned about corals and light, I suppose it is inevitable that the next gen will fail again. That would surprise me though.

Halides went through a similar evolution on their way to becoming a gold standard. As I recall, the early adopters in the '80s were using 4000k halides over their tanks nobody would think of using now.

The first halides used were crazy yellow compared to today. I personally never used them because I wasn't even born in the early 80s....lol

But that's all they had. As a hobby we've always had to adapt everyday items to our likes and uses. It takes a while for someone to invest the needed capital to create a hobby specific item. Even today, that isn't being done. Kessil is the only company I know of that is even close to it. Everybody else is just using everyday items and making them work for us.

IMO the big problem is that everyday LEDs will never accomplish what we need and want, at least for the masses. It will take someone creating a specialized multi chip for our specific usage. That will take millions of dollars however and is a giant risk.

JMO of course.

 

[rgulrich]

....wasn't even born in the 80's...sheesh, now I really feel old.
At any rate, I fondly remember outfitting my reef back then with some great 5500K bulbs until I got my hands on some (if I remember correctly) Iwasaki Eye 6500K. Supplemented with VHO Actinics, of course.

 

[TropTrea]

Well, that's my fallback position... :crazy1:



Halides went through a similar evolution on their way to becoming a gold standard. As I recall, the early adopters in the '80s were using 4000k halides over their tanks nobody would think of using now.

If you l;ook at some of Sanjay's tanks he is still using 6,500K Metal Hides. And I would not call any of his tanks faulty in any way. There are very few people that can take a small frag and have it outgrow a 200 plus gallon tank in under two years.

 

[TropTrea]

IMO the big problem is that everyday LEDs will never accomplish what we need and want, at least for the masses. It will take someone creating a specialized multi chip for our specific usage. That will take millions of dollars however and is a giant risk.

JMO of course.

I do not believe the multi chip will be the way to go. Sure a perfectly balanced base chip would be nice but a powerful multichip would go back to some of issues others mentioned like extreme shadowing. Then you know yourself that the same light balance that makes one person happy does not make a lot of other people happy. This is especially true of the balance between blue and full spectrum.

And yes if want to shop around you can find the chips to put a perfect LED system together. The issue is the DIYer cannot get himself the best chips unless he is willing to pay through the nose for them. Look at the Spectrum's on some of Pacific Suns Lights. Can you find a fault in them?

 

[rtparty]

I do not believe the multi chip will be the way to go. Sure a perfectly balanced base chip would be nice but a powerful multichip would go back to some of issues others mentioned like extreme shadowing. Then you know yourself that the same light balance that makes one person happy does not make a lot of other people happy. This is especially true of the balance between blue and full spectrum.

And yes if want to shop around you can find the chips to put a perfect LED system together. The issue is the DIYer cannot get himself the best chips unless he is willing to pay through the nose for them. Look at the Spectrum's on some of Pacific Suns Lights. Can you find a fault in them?

Just how smoothed out they are. I've seen true LED spectral plots and they are much more spiky than any of these companies show.

 

[TropTrea]

Just how smoothed out they are. I've seen true LED spectral plots and they are much more spiky than any of these companies show.

Yes if you look at some of the individual plots for specific wave length LED's they show up very spikey. However the spikes do have wave patterns with usually 50% of there peaks 5 to 10 nm from there spike. This varies by the manufacture and bin numbers. Now if you put another colored LED with it that is 10 to 15 nm you will often get the overlapping point at 50% or more and the additive effect does remove a good section of that spike.

Now if you look at true high resolution plot of T-5's or MH's you will also see those spikes in the pattern. The only true light source with minimum spikes is the sun and that even has spikes in it only they are less extreme.

 

[Fredfish]

The first halides used were crazy yellow compared to today. I personally never used them because I wasn't even born in the early 80s....lol

But that's all they had...

No, that's all they knew. The 6500k, 10000K... lights existed and were used in other applications. Its no different than today where the aquarium industry uses off he shelf diodes in custom mixes to create what they believe to be the perfect light.

If you look at some of Sanjay's tanks he is still using 6,500K Metal Hides. And I would not call any of his tanks faulty in any way.

Most people today would fault it because it does not give the T5 pop. Fault can be found with anything.

I do not believe the multi chip will be the way to go.

That's an interesting comment given the route you chose for your current builds. I wonder what the smallest practical multichip is. It would be easy to build a chip that has 1/2 of what the current Lumia has. Now you have 2 chips to cover that ~24" area and better light spread. What happens if you can break the lumia down into 4 mini-chips? Do you get Halide like spread then? You could put them together like lego into whatever configuration you want...

I like the idea of mini-clusters, but I don't know how much impact that has on manufacturing cost. I would guess it to be significant as you would have to house each mini-cluster separately.

 

[Danny_B]

I just read the whole thread. I have gone back to 4 x 39w t5 on my 29g. Id like to play with my diy still though. So what are we thinking instead of white leds, how about OCW ocean coral white 3 on star led?

 

[Pacific Sun RD]

Just how smoothed out they are. I've seen true LED spectral plots and they are much more spiky than any of these companies show.

Hello Ryan,
They are not smoothed.
This readings was taken on following config(on Ocean Optics spectrometer):
- integration time: 35ms
- scans to average: 2
- boxcar widh: 2
- electric dark correction: enabled
I think that our spectrum is most credible in compare to other manufacturers charts - we have measured some of competitors lamps and few of them have "little" other looking spectrum charts(as I remember I posted that kind of "sample" even in this thread).
We dont have to worry about our charts - we are taken full responsibility for them. I'm curious to know that all other manufacturers also..

 

[rtparty]

Hello Ryan,
They are not smoothed.
This readings was taken on following config(on Ocean Optics spectrometer):
- integration time: 35ms
- scans to average: 2
- boxcar widh: 2
- electric dark correction: enabled
I think that our spectrum is most credible in compare to other manufacturers charts - we have measured some of competitors lamps and few of them have "little" other looking spectrum charts(as I remember I posted that kind of "sample" even in this thread).
We dont have to worry about our charts - we are taken full responsibility for them. I'm curious to know that all other manufacturers also..

It's all companies and all types of lighting.

 

[Pacific Sun RD]

Are you sure that all spectrum charts advertised by some manufacturers are "real"?

 

[rtparty]

Are you sure that all spectrum charts advertised by some manufacturers are "real"?

Certainly not. I don't fully trust any of them and have said that numerous times.

 

[nynick]

I have a couple of questions about the "spiky" spectrum.

I did read that light blends, especially when passing through a surface (such as the top of the water) or being reflected. IF this is the case, how would a multi chip that is 100% reflected light work? Instead of pointing the light down, point it up onto a mirror/reflector and then down into the tank.

I have no clue how correct/possible any of this would be but would love to find out.

 

[conix67]

I have a couple of questions about the "spiky" spectrum.

I did read that light blends, especially when passing through a surface (such as the top of the water) or being reflected. IF this is the case, how would a multi chip that is 100% reflected light work? Instead of pointing the light down, point it up onto a mirror/reflector and then down into the tank.

I have no clue how correct/possible any of this would be but would love to find out.

Using the reflectors, such as those used for MH fixtures, as an indirect source of LED lighting could benefit one thing - spreading the light across larger area better than lenses alone.

However, this will not change anything on the spectrum side of things. No matter how much you blend different spectrum of lights, in the end you'll get what is there.

 

[TropTrea]

I have a couple of questions about the "spiky" spectrum.

I did read that light blends, especially when passing through a surface (such as the top of the water) or being reflected. IF this is the case, how would a multi chip that is 100% reflected light work? Instead of pointing the light down, point it up onto a mirror/reflector and then down into the tank.

I have no clue how correct/possible any of this would be but would love to find out.

I think what you refering to is the prism effect. Where if took pure white light (all wave lenght) and run it through a prism you can separate all the colors and create a rainbow.

Yes you can get some of this effect with angling LED's over the surface of the water but since they are not a narrow beam of light the separation is not very drastic at all. However it does create more of a shimmer effect if there is some movement on the water surface.

As far as the spikiness on all spectrum plots this has a lot to do with the resolution of the instrument. Usually if your running a 3 nm resolution most of that spikiness is removed from the graph. On most of the vendor charts I suspect they are displaying plots with about 5nm resolution. Another thing is the number of samples they use to determine the plots. With most power supplies there is some AC ripple and this will show up in a single sample but taking multiple samples which average out this ripple.

 

[nynick]

Maybe I didn't explain my question well enough.

What I was wanting to know was how much of the color blend you get from different lights are created in our heads and how much is real. For example, if you take a blue lamp and a red lamp and project them onto the same area of a white surface you get purple light. Waves do affect each other but to what degree did the properties of the light actually change? Is it only purple to us or is it actually some new wave lengths of light in there and to what extent.