DIY LEDs - The write-up

[HeneryH]

Can someone help me understand the Cutter pricing? If I add the XPG to my cart and select the 3-up serial packaging it gets confusing.

Am I entering the quantity of LEDs or 3-up Packages? The adder of $20 for the 3-up packaging seems crazy if it is $20x3 for each package but on the other hand the pricing for the LEDs seems low if the quantity is for the complete package.

Thanks

 

[stugray]

an someone help me understand the Cutter pricing?

YEAH ME TOO!!

I have been buggin them for almost two weeks to JUST TELL ME HOW MUCH the The "7 up.. 4 x XPG Cool White and 3 x XPE Royal Blue on a 40mm Dia Metal Core PCB" IS.

It shows ~$41 ea, but I cant tell if you have to buy the LEDs for them to populate the thing separately or are they included in the price.

Also, the 7up is an oddball are the 4 XPGs on a separate circuit from the blues or are they all in a string???

If they are all in series, it will be harder to drive them with meanwells since you cant put two in series or that is 14 LEDs.

Stu

 

[HeneryH]

Sorry for all the questions lately.

I'm hearing a lot of comments about how many LEDs per Mean Well.

Since the LEDs are serial, how are you calculating the number of LEDs in a string? I'm an EE from long ago and understand the basic equations but not sure with the LEDs which parameters are the constant and which are the variables.

Do you just take the Mean Well wattage rating and divide by 3.3 W (when driving the string at 1A)?

That says a 60W Mean Well can drive 15 XPGs at 1A?

The spreadsheet is my attempt at figuring this out rather than just taking comments here at x LEDs per driver.

Thanks

 

[der_wille_zur_macht]

Sorry for all the questions lately.

I'm hearing a lot of comments about how many LEDs per Mean Well.

Since the LEDs are serial, how are you calculating the number of LEDs in a string?

It's all about the voltage:

1) Determine your model of LED and the target drive current
2) Find the datasheet for your LED and look up the forward voltage of the LED at your chosen current.
3) Divide the driver's maximum output voltage by the Vf for your LED.

For example, the Meanwell ELN-60-48 has a max of 48v output. Imagine we have an LED that drops 3.7v at some target current. 48/3.7 = 12.9, so we round down, and we can safely drive 12 of these LEDs on this driver.

In practice, if the math comes out REALLY close, you should probably round down another whole LED. So, if our LEDs drop 3.6v, the equation would tell us 13.3, but I would probably do 12 to have a safety margin. This will account for variations in Vf from the published spec for your LED. Otherwise, you might end up underdriving your string.

 

[kgk]

We have (I think) just taken 2 inch spacing as a good idea. This may give a better idea for spacing. I will just throw it out there and you can laugh at me :lolspin:

I am not laughing, great post! This is the model I have thought about as well. Figure out how deep your tank is, and how high you want to hang your fixture. Then you can figure out which optics & spacing you want based on the method you elaborated on. Amount of overlap would be completely up to the builder.

 

[TheFishMan65]

I am glad some likes it. I found out that changing the spacing does not work. And I should have plotted the grpah with dimmed light based on distance (1 / radius squared thing). Of course water quality will have a big effect especially at wide angles.

 

[HeneryH]

Has anyone considered how the refraction at the boundary of the air/water change the calculation. I'm thinking that we could spend a lot of calculations and come out way off.

p.s. slow day here at work with the snow on the east coast so while I should be working I am designing my LED rig Shhh

 

[TheFishMan65]

I think that is the index of refraction. Straight on no reflection. As the angle increases you get more at a certain angle all is reflected. Of course if we have a well agitated surface for gas exchange, it get way past simple math.

[EDIT]
I have that backwards. The light will bend sharper which is good since it will bend away from the sides.
Index of Refraction for Water 1.33
Index of Refraction for Air 1
Here is the angle in Air and what it will be in water using snells law
00 0
10 7.502100527
20 14.90149465
30 22.08241319
40 28.90108454
50 35.1678191
60 40.62813065
70 44.95375334
80 47.77035587
90 48.75346663
For the record refraction of water varies by temperature and light wave size (nm or color)

 

[stugray]

Posted back on page 51:

"Another thing to keep in mind when trying to figure out light distribution:

Snell's law ( tyrin to be as nerdy a kcress above ).

This means lets say we are using 90 degree optics (or a bare XR-E).
Light comes out of the emitter ( essentially a point source ) in a 90 degree cone and the edges hit the water at a 45 degree angle ( the angle of incidence ).

Well, the air/water interface bends the light, so the cone UNDER the water is now a 64 degree cone.

And as far as the fish are concerned, it looks like the LEDs are farther away from the surface than they really are.

As you use tighter & tighter optics the effect lessens.



And kcress - you brought back memories of nyquist plots
My Brain ..... it hurts

Stu"


Stu

 

[der_wille_zur_macht]

And I should have plotted the grpah with dimmed light based on distance (1 / radius squared thing). Of course water quality will have a big effect especially at wide angles.

Has anyone considered how the refraction at the boundary of the air/water change the calculation. I'm thinking that we could spend a lot of calculations and come out way off.

Also, it's fair to note that the LEDs aren't all going to emit light with the same distribution, especially with optics - you can't count on a single formula to accurately describe the intensity across some plane at some given distance from the LED based purely on the FWHM or viewing angle, unless you're taking into account the spatial distributions of each LED and optic used - and I don't think we have that info for all of the components we're playing with, do we? Cree plots the spatial distribution of their LEDs in their datasheets but it's just a plot, not a formula, so you'd have to fit something yourself. And with optics, it's anyone's guess - some cut off VERY sharply around their FWHM angle; others have a narrow, strong peak in the middle but taperoff much more gently, and appear WAY WAY wider than their FWHM angle would imply.

I probably sound like a broken record, and I mean this with all kindness - I really applaud the effort you guys are putting in to modeling this stuff, but again, IMHO there's no substitute for just ordering a small number of LEDs and optics and trying them out on an actual tank. I think we're luckily at a point where there has been enough of this experimentation that "rules of thumb" will get people close enough, and a little experimentation will seal the deal.

 

[TheFishMan65]

Man go to edit and everyone chimes in

 

[TheFishMan65]

Lenses

Lenses

Evil over on NR has not been brought up in a while. I read that thread last week (yes, the whole week). It is interesting that he uses lens angle to represent MH bulb type. Don't quote, but 80 degree is 150 w, 60 is 250 and 20 is 450 or some such numbers. I think this is a fair way to think about. The number of LEDs just determine how good the coverages.

But he does not really take height above the water into account. Anybody have any thoughts on this. Goes back to the age old question for this much PAR what do I need. Answer with the right lens and ONE led you can get what ever PAR you want - but only in one spot.

So I guess what I am asking is does anyone have thought on how initial output, lens, and height can be translated to PAR.

 

[Skeptic_07]

I just ordered some of these optics. Not sure yet if i'll use them, but they're cheap enough to experiment with at ~$8 for a 4' piece. Previously I could only find these in the Uk which driving the price way up (like $28 a piece plus shipping).

http://www.wustlich.de/images/onecms/site/Produkte/PDF/cree/Continuous Strip Optic.pdf

http://www.futureelectronics.com/en.../multiple-led-optics/Pages/5858329-10397.aspx

we'll see what happens.

 

[ReefEnabler]

I habe several different xre optics. Ive been aiming them at the ceiling and taking photos with the same iso/aperature/shutter to somewhat compare them just to get a basic idea. Havent Backed it up with actual par readings yet tho.

Would people be interested to see those even without numbers to go along?

Still waiting on the rest of my xres to arrive from dealextreme

 

[der_wille_zur_macht]

Different topic, but it looks like some guy named Nate Enders is going to be writing a series of articles on DIY LEDs for Reefkeeping Magazine:

http://www.reefkeeping.com/joomla/index.php/reefkeeping-blog

 

[ReefEnabler]

I habe several different xre optics. Ive been aiming them at the ceiling and taking photos with the same iso/aperature/shutter to somewhat compare them just to get a basic idea. Havent Backed it up with actual par readings yet tho.

Would people be interested to see those even without numbers to go along?

Still waiting on the rest of my xres to arrive from dealextreme

 

[der_wille_zur_macht]

Evil over on NR has not been brought up in a while. I read that thread last week (yes, the whole week). It is interesting that he uses lens angle to represent MH bulb type. Don't quote, but 80 degree is 150 w, 60 is 250 and 20 is 450 or some such numbers. I think this is a fair way to think about. The number of LEDs just determine how good the coverages.

But he does not really take height above the water into account. Anybody have any thoughts on this. Goes back to the age old question for this much PAR what do I need. Answer with the right lens and ONE led you can get what ever PAR you want - but only in one spot.

So I guess what I am asking is does anyone have thought on how initial output, lens, and height can be translated to PAR.

It's a complicated subject, for sure. I appreciate his approach, but I agree that it doesn't fully take spread and height above the water into account. As we all know, optics don't create more light or inherently make the light more intense, they just change where that light goes.

That said, my approach is this:

1) Pick drive current and LED count to reach your overall desired intensity. I start with this because, well, it's a simple fact - the overall amount of light generated is generally controlled by these two variables.

2) Pick optics to match tank height and fixture height above the water. Again, simple fact - optic choice alters where the light goes, so I like to tie it to where the LEDs are in relation to where you want the light.

3) Adjust as required for extreme situations (i.e. if you're using a VERY small LED count for a fish only tank, you'd probably forgo optics even if the first two steps told you that you needed them.)

I habe several different xre optics. Ive been aiming them at the ceiling and taking photos with the same iso/aperature/shutter to somewhat compare them just to get a basic idea. Havent Backed it up with actual par readings yet tho.

Would people be interested to see those even without numbers to go along?

Still waiting on the rest of my xres to arrive from dealextreme

Definitely interested to see the images. Measurements will help too, but at least the images will give a clear picture of the shape of the distribution.

 

[Skeptic_07]

Would people be interested to see those even without numbers to go along?

for sure! definitely post those. it would be good to post your camera settings, distance from emitter to the illuminated area, and some kind of size comparison, if possible.

 

[Skeptic_07]

Different topic, but it looks like some guy named Nate Enders is going to be writing a series of articles on DIY LEDs for Reefkeeping Magazine:

http://www.reefkeeping.com/joomla/index.php/reefkeeping-blog

wonder who that guy is? probably some kook who doesn't know what he's talkin about anyway.. jk

 

[bstohrer]

Posted back on page 51:

This means lets say we are using 90 degree optics (or a bare XR-E).
Light comes out of the emitter ( essentially a point source ) in a 90 degree cone and the edges hit the water at a 45 degree angle ( the angle of incidence ).

Well, the air/water interface bends the light, so the cone UNDER the water is now a 64 degree cone.

Stu

Since I'll have the LEDs only a few inches off the surface (24 inches to the bottom), I may not need optics after all. I will have optics to try in a week or two, so it will be trial and error.