Neptune Aquatics

Improved Color LED thread

I'm all finished.

Wired, enclosure done, knobs on, hood on tank, and LEDs doing their thing. Went smoothly... have tons of pictures and will post them up soon. Do you think I should post a lengthier thing here, RC, or on my blog? I can't decide where to put the bulk of it.


Edit... Initial thoughts:

Royal Blue is BLUE... seems almost beyond what I've seen in a 20K setup but my experience is limited. Cameras don't do it justice. With blues off, I get a nice 10K look with only whites. With the two arrays even, it's 12 to 14K.

Both arrays dialed to max, at 900mA, is extremely bright. I think I could satisfy any light demands in this tank and I may end up having to part with some of my LPS if it won't acclimate.

Blues on at 3pm, Whites on at 4pm, Whites off at 1am, Blues off at 2am. Mostly because I spend time around my tank during the evening and at night only.
 
Tamazula said:
I'm all finished.

Wired, enclosure done, knobs on, hood on tank, and LEDs doing their thing. Went smoothly... have tons of pictures and will post them up soon. Do you think I should post a lengthier thing here, RC, or on my blog? I can't decide where to put the bulk of it.

Up to you.
I consider this an "open" thread. Feel free to post anything.
Especially directly related to LED color.

But if you have a nice build blog, you may want to follow through on that.
If you put it somewhere else, please make sure to add a link in bold here though.
 
Tamazula said:
Both arrays dialed to max, at 900mA, is extremely bright. I think I could satisfy any light demands in this tank and I may end up having to part with some of my LPS if it won't acclimate.

Yes, be careful with that. Note that PAR is likely higher than normal for perceived brightness as well.
Meaning, if it looks brighter, PAR may be WAY higher.
Depends of course on what you were comparing it with.

You might want to ask around about a PAR meter as well. I think someone had one, not sure it is
LED compatible though. You can borrow my not-as-waterproof-as-advertised lux meter,
but that is borderline useless for what we want.
 
yardartist said:
The reds leave color bands in my tank. I like that you have groups of three together close as that would seem to lessen the band effect.

Does anyone have suggested high temps to run these off the top of your head?


Depends on brand but crees and most of the other common types are good to 130 - 140 degrees measured at the heat sink, right next to the led. You really need to measure the emitter, but the really isn't possible without special equipment.
 
Qwiv said:
yardartist said:
The reds leave color bands in my tank. I like that you have groups of three together close as that would seem to lessen the band effect.

Does anyone have suggested high temps to run these off the top of your head?


Depends on brand but crees and most of the other common types are good to 130 - 140 degrees measured at the heat sink, right next to the led. You really need to measure the emitter, but the really isn't possible without special equipment.

It really depends on your objective.
Qwiv is correct that 130-140 C at the heat sink is about max.
The Cree is rated to 150 deg C at the junction, and a good estimate is 10 deg TJ to sink.

However:
Increased temperature has a sizable effect on both efficiency and longevity.
From 25C to 150C, you lose 30% of the light output.
From old semiconductor days, longevity is is roughly proportional to the SQUARE of the junction temperature.
Assuming LEDs are similar, a 25C LED would last 30 times longer than a 150C LED.
Although I don't think it is similar, since the reason it is the square is because
temp reduces resistance, increasing current. But LEDs are constant current.
Whatever. Heat = Bad.

My opinion : The LED star package itself should just be a bit hot to the touch on a warm day.
 
This is totally subjective, but my LEDs seem to be running very very cool. I feel no warmth at all coming out of the exhaust duct or on the plastic portions of the hood. It's only 20 LEDs on a fairly massive aluminum heatsink so maybe I was overestimating how much heat they would generate.
 
Tamazula said:
This is totally subjective, but my LEDs seem to be running very very cool. I feel no warmth at all coming out of the exhaust duct or on the plastic portions of the hood. It's only 20 LEDs on a fairly massive aluminum heatsink so maybe I was overestimating how much heat they would generate.

If memory serves, you are running 450 mA (900 at half power)
So 3.4 * 20 * 0.45 = 30W.
Now lets guess that 60% of that is heat.
So that is only about 20W of heat to deal with. Not a big number.
 
rygh is correct on running them cooler extending life. Data is tough to find on this and we only have life studies on LEDs not applicable to this at my office, but the principle would be the same
I was only stating MAX temp. Cooler the better, and the limit on that is your design.
 
rygh said:
Tamazula said:
Both arrays dialed to max, at 900mA, is extremely bright. I think I could satisfy any light demands in this tank and I may end up having to part with some of my LPS if it won't acclimate.

Yes, be careful with that. Note that PAR is likely higher than normal for perceived brightness as well.
Meaning, if it looks brighter, PAR may be WAY higher.
Depends of course on what you were comparing it with.

You might want to ask around about a PAR meter as well. I think someone had one, not sure it is
LED compatible though. You can borrow my not-as-waterproof-as-advertised lux meter,
but that is borderline useless for what we want.


Only ones local I ever hear about are low end Apogees (like I have) which tend to not make good LED PAR meters. Most hobbyists don't' want to pony up 1K+ for a PAR meter to work with their LEDs :(
 
One fixture done-ish.
The wiring and LEDs are done.
In the process of gluing the wires, cleaning up, sealing, etc.
A bit hard to tell so far, but I think I even like the color as-is.
Still plan PWM controls though.

 
Ah nah I'm running both arrays at 900mA. Still not too hot though. My enclosure is a little warm, actually, but not too bad and that's after it has been running for several hours.
 
LED fan recommendation.

I received these last week, and I am really happy with them.
Perfect for a big heat sink.
Key - dead silent. Rated at 7 dB, and it really is.
Really. You can put your ear inches away, and it is just a slight whir.
And 33 CFM is about right for a heat sink fan. Not like you are cooling computer.

http://www.sidewindercomputers.com/no12ulnonf33.html

Standard 12V, 120 mm, even comes with little rubber feet.
 
One light fixture is in and installed.


The funny thing is - both sides use about the same amount of power.
Maybe 15% more on the LED side is all.

I am having no luck at all getting the camera to respond properly to show color. Argh!

-----

Initial thoughts:

1) Wow, bright.
Most likely needs to be toned down a bit, which is good, since it allows color tweaking.
Way brighter than my T5+LED setup.

2) I like the color. About in the 10,000K range as hoped.
(I am not a big fan of the 20,000K blue look)
The ratio is 1 CW to 2 RB, but with the XP-G performance, and extra drive, the white is still dominant.

3) The red really helps.
It is obvious, even with just white coral that it is making a difference.
Definite red-pop on that spectrum end.
But really a bit too much. Needs to be turned down a bit.

4) The neutral whites and standard blue were probably a waste of time.
Unless you put in a LOT of them, the difference is basically undetectable.
But perhaps once XP-Gs are tuned down, it may show.

5) Seems like a very good CRI.
You can get close, and it is the clarity is great.

5) I have a very odd shimmer.
The shimmer is crazy. Lots of it, and moving very rapidly.
I think due to the sheer number of point sources.
PLUS: There is a distinct RED shimmer.
There are only 4 reds. I did not get banding, but I do get a strange shimmer effect.
Basically, the shimmer sparkles red at times.
I thought I might like that much shimmer, but I think I don't, and it may have to be fixed.
First easy fix is less turbulence at the surface. Kindof overdone anyway.
But the fix for the reds might be a bit more complex. Need to think about that.

---

So pretty happy so far. But still need a LOT more testing to do.
All I am looking at is white dead live rock and sand.
 
Yea, I noticed that (5, the second one) too. The multiple point sources, especially of different colors, make for some strange patterns as the surface agitation refracts blue here or white there. Far more chaotic than 1-2 MH bulbs, but I got used to it pretty quickly.

To diffuse the reds, maybe a small translucent baffle placed roughly an inch below the LED would work so that it could both reduce the amount of red light going into the tank (you mentioned you might want to turn it down) and also turn it from a sharp point source to a larger area of emission. Maybe get a lens and pop out the actual lens portion, and replace it with something translucent to your liking?
 
Interesting idea.
I tried something really simple - I put in that micro-diamond-diffusion grid.
Standard home depot lighting stuff I had leftover.
Definitely made an improvement, but not a huge difference.
Made it dimmer as well though, so some loss in light.
Needs some more experimenting.

They actually sell lenses with diffused gratings.
 
So as per main tank thread, Lighting PWM controller is done.
I wrote a simple program to cycle through all the values.
Took a movie.
But unfortunately, the camera had a really tough time of it.

Here it is anyway.
First ramp is reds
Second ramp is royal blues
Third ramp is cool whites. (looks yellow in movie, but is not)
Fourth ramp is neutral white + standard blue.
Fifth ramp is everything equal.

 
sfsuphysics said:
Yikes that red one makes you look like you a massive cyano outbreak :D
Pretty freaky looking huh.
I do have a big diatom outbreak, but that was expected.

Just the red, on very slightly, are fun at night. Fish don't seem to notice.
I think I will assign that to the one of the keypad buttons.
 
The plexi shield is less than an inch from the lamps. I added two inch white paper disks under the red LED units and the color is much better. There is red and blue banding on the wall. The room is purple from the street. Yellow fish look deep orange. Knocking the red back makes the blue greens come back to life.

Good thing about personal taste, I like it so far. Color is improving on the acros. A green one is sporting a blue sheen on the tips. Cyphastrea might have to be lowered or tucked under a ledge. One spot is fading. The blue sponge does not have the same bright color it did under MH. More steel blue and much less purple overtones. Odd given there is more red light to reflect back.

I changed the daylight to 55% and the actinic to 45%. Both started at 50%. Adjust again in about three weeks.

The algae is growing fast as nitrates had risen. Back to near zero with fuzzy rocks again. Chaeto ball is now fair size so sea hare to munch the other back and see if I can get better balance away from the hair.


Great video Mark. At this rate we will be holding a LED tank tour soon.
 
Back
Top