My take on the cluster

[jedimasterben]

As soon as you can fit at least ten Luxeon TX into the same 10mm die area of the Vero 10, let me know.

[farkwar]

Excuse me for interjecting,

 

 

Exactly how does it matter that a white LED is rendering 4800 Kelvin instead of its rated color of 5000 Kelvin?

 

Especially when you are already altering the total color with other color LEDs.

 

 

All of those 70 and 80 CRI LEDs are rendering warmer than rated anyways. That's why its easier to find 90CRI in the 4000 Kelvin and lower ranges.

 

Ben, youve been recommending NW and WW over CW for as long as Ive read your stuff. So how does it really mattter. It doesnt matter to the corals. It does'nt matter to us, I think we've all given up trying to photograph our tanks accurately out of Camera without post color balance.

 

It doesnt matter to us that have channels and controllers. Dont like how the light looks, tweak up this or tweak down that. It might have mattered with a MH bulb, because once you plugged it in, it had that color. If you wanted to tweak it, you needed to change ballasts.

[jedimasterben]

If it doesn't matter to you whether or not colors show up, then that's up to you. There are people lighting their tanks with only royal blue LEDs because they think that looks great.

[farkwar]

CRI matters if you photograph for NG or AD. Even those, youre still allowed to color balance (do it in camera for the purists).

 

Or were you a cinematographer making stuff on film still. Digital? Just run the whole file through a filter and fix it in post.

 

Can you explain or post a link as to why 90 CRI white LEDs are important to us, as hobbyist reefers, please?

 

 

I was writing this post while you wrote the above

 

If it doesn't matter to you whether or not colors show up, then that's up to you. There are people lighting their tanks with only royal blue LEDs because they think that looks great.

What colors are missing that matter to you?

 

A 5000 Kelvin 70 CRI is going to be more red and orange than stated. Its not going to be bluer.

 

You recommend a warmer color LED anyway. And twice the blue(as white) on the blue channel anyway. Which is a great recommendation, dont get me wrong. But now your CRI is so blown with mixed in colors, you need $100,000-1,000,000 equipment to see the difference.

[farkwar]

There is NO way, if you put 100% power to each of DNKs 12 LED array that you could look at the light and know, and say

 

"The white emitters on this array are 70CRI, I can see the difference".

[Horerczy]

I notice an interesting efficiency those hadn't Ben brought up.

 

Output per square mm. Yes the rebels are more output than the Z but they're also ~6 times larger. The rebels certainly don't get 6 times output though.

 

Anyway we can debate this till the end of time but frankly it doesn't matter since bpaston already sent the files for his mcpcb out.

 

Also we don't need to use the most efficient leds or sinkpad for our lights. Leds have survived just fine without sinkpad up till now and we were lighting or tanks with lower efficiency leds a while ago and still managing to fry corals.

 

Just my $0.02

[DNK]

As soon as you can fit at least ten Luxeon TX into the same 10mm die area of the Vero 10, let me know.

I see no need to do this. But if you want - no problem. Please check - Vero 10 has a area around 19x19=361 square millimeters. Luxeon TX has a 2.04x1.64=3.35 square millimeters. So, on area of one Vero 10 I can place approximately 100pcs of Luxeon TX. To be honest, exact quantity will be less, but I sure, 50pcs - without any problem. Have you any more question?

 

Also, please be informed - evenly distributed white LED on area of LED assembly is better that one white LED. Is this need to be explained?

[DNK]

Output per square mm. Yes the rebels are more output than the Z but they're also ~6 times larger. The rebels certainly don't get 6 times output though.

I would like to remind one more time. Efficacy of Luxeon Z is worst than Luxeon Rebel, Rebel Es, Luxeon T and Luxeon TX is more that 40%. Only this is enough to forget of Luxeon Z in anyway. But, if you wish, let's count of luminous density of Luxeon Z comparative to Luxeon T, regardless of efficacy (sic!). As I showed before, Luxeon T has a surface area 3.35 square millimeters. It has at this surface for Luxeon T up to 1040*1.6=1664mWt, i.e. 1664/3.35=497mWt per square millimeters. Lets take, for example, Royal Blue LED Luxeon Z. For luxeon Z it will be approximately 600*1.75=1050mWt per 1.7*1.3=2.21square millimeters, i.e. 1050/2.21=475mWt per square millimeters!

 

As you can see, Luxeon T has a more radiation density than a Luxeon Z. Please don't forget about this simple conclusion. It will be near the same for all other types of Luxeon Z.

 

I ask to all people to check my calculations and make sure - Lexeon Z is approximately 40-50% worst that Luxeon T (or TX, Rebel ES, usual Rebel) and has LOWER density of radiation! I can't see any reason to use Luxeon Z in fixture for marine aquaria. How about you?

and still managing to fry corals.

Excuse me, but for what you talk about "fry corals"? We will construct our home sea not to fry corals, but for outstanding visual appearance, isn't it? Corals will have enough food in anyway. But make for it best possible visual appearance - our primary target, isn't it?

 

Second target - best efficacy. It second reason is why we use LEDs. Why we should use LEDs with worst efficacy, with low radiation density, unconvenient to use with lens, unconvenient to use with usual driver. Who is could explain me?

 

To be honest, I think - hobby - as sex. Process does matter, result (in many cases) - no. But it's hard for me to see a non-optimal design of DIY LED fixtures. Why did you have to make exactly the same mistakes that we already have made along the way to understand what I'm saying?

 

If my thoughts is not comfortable for people in this forum, I'll be silent. If our experience does matter for you, please try to understand what about I talked. If I should make any explanation to avoid any unclear points please let me know.

 

In any case, I have a big request to all. If you are going to oppose to my opinion, please bother to look at the data that provided by the manufacturer of LEDs. I would not like to waste nor your time nor my, by banal calculations as shown above.

 

PS Jedi, I respect your educational work. But I have a great request. Please understand that best possible CRI for wide of CCTs may be achieved only in one way - by tuning spectra of carefully selected color LEDs with base of high-efficacy cool white LEDs. No reason to use high-CRI white LED with bad efficacy. Manufacturers of LEDs is forced to promote high-CRI LEDs because a constructing a good LED assembly with good controller for usual application is pricey. Extremely pricey for usual light. In other words, it is impossible pricey for most of usual applications. But we can to afford for us the best possible solution, isn't it? So, please try to understand what I try to explain.

 

I'm sorry again for my bad English. If my English is bad for understanding, let me know, I'll try to be more clear.

[bpaston]

Unfortunately, using of such big LED will cause serious problem. We can't use lens (at least for all LEDs), it means we will forced to hang this fixture very low. Or, if we would like to use lens in anyway, we forced to design LED assembly very spaciously, it will cause strong disco.

I sincerely hope you have read my posts prior to commenting which it seems you have not. So without a full understanding of what I am trying to accomplish this statement is of no value. In regards to the leds and use of lenses, you can see that there will not be individual lenses per led. This project will make use of a lens collimator. So the goal was to actually use leds with no dome optics installed. The only reason luxeon m's are being used is there are no royal blue luxeon mz's. The use of the 74mm lens collimator also dictated the overall footprint of the leds, which led to the decision to use the luxeon z's to retain footprint and also the fact of no dome optic installed. Also the plan is for this fixture to have an overall distance of somewhere between 4' - 5' between optic and sandbed.

 

OK. Let's check any other LEDs type. For example, cyan.

Luxeon Z for 700mA current has a radiant efficacy 38LPW, Luxeon Rebel has a 56LPW. In other words, Z worse than Rebel on (56-38)/38*100=47%! And it is not all. You can use Rebel on 1000mA also, and it will be dissipate less heat than Z on 700mA! In other words, you have more 30% of light in case of Rebel absolutely free of any obligation, including excessive heat

You data is actually incorrect, you are using an old datasheet that does not include the cyan led I plan on using and have purchased. You are quoting data from part number LXZ1-PE01-0072, I am using LXZ1-PE01-0080. Which has a 52lm/W efficiency at 500mA and 44lm/W at 700mA, please take a little more time familiarizing yourself with my project before comparing apples to oranges to make your dramtic points. So using your formula (56-44)/44*100=27% thats a 20% difference due to use of wrong data.

 

But if we only run the z's at 500mA we are alotted a 52lm/W efficiency...so running these at 700mA may not be the best option for my design but I will figure that out during testing. It would definitely allow for a cooler junction temp.

 

Now when designing PCB the idea of using luxeon rebels for color was considered and mocked up, but only 3 leds would fit in between the luxeon m border due to design constraint of collimator optic. So lets say the 3 luxeon rebels were the same cyans we used the formula on earlier I would be looking at 133 lumens per rebel at 700mA for a total of 399 lumens. But with the use of luxeon z's I was able to fit 5 or 6 per side. So lets low ball it and use the side with 5, at 100 lumens per z I have a total of 500 lumens as opposed to 399 lumens using the rebels. Or if I decide to run the z's at 500mA then I would be looking at a total of 405 lumens.

 

So overall yes these are less efficient, but not by as much as you would lead us to believe. High luminous flux bins were chosen for this project with that in mind. But when considering the overall footprint of each the z takes up only 2.21 sq. mm and the rebel takes up 14.48 sq. mm. That being said running these at 700mA the luxeon z projects 45.25 lm/sq. mm and the rebel projects 9.18lm/sq. mm.

[DNK]

The use of the 74mm lens collimator also dictated the overall footprint of the leds

OK, if this type of lens is suit you, I have no claim. Sorry for inconvenience.

You data is actually incorrect, you are using an old datasheet that does not include the cyan led I plan on using and have purchased. You are quoting data from part number LXZ1-PE01-0072, I am using LXZ1-PE01-0080.

I just checked official page of Luxeon Z http://www.philipslumileds.com/products/luxeon-z and download latest version of datasheet. Please check screenshot of page from this datasheet:

Can you please find LXZ1-PE01-0080? I'm not.

 

Please be informed - most of sellers is indicate in description of LEDs is completely bullshit. They do this to mislead the consumer. Only one example. Please check page of this seller: http://www.newark.com/lumileds/lxz1-pe01-0080/led-luxeon-z-cyan/dp/45W6291?ost=2148789

Please open datasheet for this LED from that page: http://www.farnell.com/datasheets/1641576.pdf

It is preliminary datasheet! Can we trust such datasheet?

 

Do you like to deluding oneself? LXZ1-PE01-0080 does not exist.

 

Also please keep in mind. About bullshit from sellers. This example is not the most outstanding deception. LXZ1-PE01-0080 at least has been in very early preliminary datasheet. But part of sellers contrive to sell the LEDs, that not exist in any datasheets!

[bpaston]

Well if you are correct that what newark is selling is not LXZ1-PE01-0080 then your math was correct. But at the end of the day whether I purchased the 0072 or the 0080, the efficiency is still less than a rebel. But as footprint size and optic choice were determining factors behind choice in leds, the efficiency of led was not the final determining factor of design. I understand it was in yours, but my overall goal is to obtain a solid reef light that could be hung 5' from the substrate. The point I was trying to make was that decisions were made off of information given to obtain the most efficient led available, in a non domed optic led. Which at the end of the day is the luxeon z in either part number 0072 or 0080 for this cyan.

 

 

Thank you for pointing out the descrepancies between datasheets, I had noticed but figured the prelim was accurate also, I dont see any reason not to trust it phillips did release it. Maybe a run of LXZ1-PE01-0080 were actually produced and newark is selling, hard to say without confirmation. What I do know is that newark is the only one to use the minimum luminous flux with the part number as outlined in datasheet. All other sellers just show the part number as LXZ1-PE01 and good luck knowing what luminous flux you are getting. But I will actually be emailing newark and phillips regarding this to get a straight answer, regarding that part number. I will post my findings.

 

 

I did also read your other post regarding binning, trust me we all understand binning. What you need to understand again is that without purchasing a minimum quantity of 250+ we are not able to request certain bins at all. So all we can source is what is being sold on newark, mouser, etc. and take the sellers word that what they are describing the led to be is true. Just like if we purchase the semilied n5 from your site, as a hobbyist I have no way to prove your claims that the led is what you desceibe it to be. I just have to trust that you are honest...simple as that. I think the big disconnect between us here and you is the fact that you have designed and sell a product for profit, where you can obtain specific bins, and place minimum orders for sinkpads with ease compared to us. We are at the mercy of what sellers decide to carry and sell, and define product as.

[DNK]

Because I sell LED, which is not described in datasheet of manufacturer, I have gave for customers the official letter from manufacturer. This letter has confirmed data for this LED. What your opinion - can you receive such letter for LXZ1-PE01-0080? Definitely no

 

But, as I understand your goal for designing fixture, worst efficacy of Luxeon Z is "unpreventable evil" in your case. I'll not bother you by other idea, because that may cause disappointment because your project is already half ready and other decision may require painful costs.

[bpaston]

Because I sell LED, which is not described in datasheet of manufacturer, I have gave for customers the official letter from manufacturer. This letter has confirmed data for this LED. What your opinion - can you receive such letter for LXZ1-PE01-0080? Definitely no

 

But, as I understand your goal for designing fixture, worst efficacy of Luxeon Z is "unpreventable evil" in your case. I'll not bother you by other idea, because that may cause disappointment because your project is already half ready and other decision may require painful costs.

 

Feel free I am open to all ideas and am thoroughly enjoying this convo, as I am learning as we go back and forth! I look forward to hearing your idea on alternative...

 

In regards to led can I obtain letter, I dont know never thought of obtaining a letter for a hobby project, why would I? But I am looking into this, as newark should not be advertising these leds with that part number if it is not truely that product. Like I said I will post findings of inquiry to both phillips and newark.

[DNK]

About improvement of your project. You have a very small area for color LEDs because you have plan to use ugly big Vero. If you will use the white LED with smaller footprint - for example Luxeon M or N5TL-W you will get enough space to use Rebel LEDs with high efficacy. Please understand - efficacy is extremely important for you because you have plan to use usual MCPCB with poor thermal conductivity.

[bpaston]

Well just a quick update...boards are complete and in hand where the leds will be reflowed. Hoping to get some before and after pics of the board during the reflo process, when I do I will post them up. I am ready to have these up over my tank already to see what they look like...but in the meantime I have been hard at work on a suitable cooling method some will think im crazy but I think it just maybe worth a shot.

[Horerczy]

exciting times

[bpaston]

And after a little over two months of silence......the finished product is finally in hand! Gonna see if I can upload some pics, and then get started on a heatsink setup for these.

[bpaston]

[jedimasterben]

Holy damn.

[Horerczy]

Epic

[Giga]

can I have one? lol

[gus6464]

That is one of the sexiest things I've ever seen.

[farkwar]

Very nice work.

[Moolelo]

These are wicked awesome! Like a mix between jedi's setups and a borealis array. Not to be ridiculous in this moment of success, but how are you wiring the vero molex connector? Looks like it would have to pass over the luxeon zs?

[bpaston]

Not over just around, and a little dab of hi temp silicon will hold the wires in place so it doesn't interfere with the luxeon z's The Vero will make us of the pico ezmate harness and will be soldered to two pads in front of the molex connecter. At this point on Mk. 1 of the board it was the best I could come up with.... I am sure once I build these I will come up with a better way to go about it in Mk. 2.