I think there is also a large difference between Chinese eBay leds in terms of quality. I've been running a 20W multichip on my 7g nano from a well known Chinese seller on eBay and I've had excellent SPS growth. I switched then to 3up Cree (8x RB + 4x CW) and the growth was nowhere near that. So, I was actually better served by a Chinese led than by "high-end" leds from Cree.
Minimalistic multichip DIY LED build
- Thread starter maglofster
- Start date
Anyway, I think 400w might not be enough for a 200g depending on its dimensions. Figure 1 watt of multichip for every 5in^2 for tanks up to 24" deep, and one watt per 4in^2 for 30 inch deep tanks. For deeper than that, 3.5in^2 per watt should be good. 90* optics for the first, 60* for the second, 40* for the last. 400w will be okay for SPS in the mid-top, but for SPS on the sandbed go with the rules posted above.[/QUOTE]
Hi, I don't understand the above formula. On a 24" deep tank for example, how to do the math? it looks like 1 watt @ every 25sq. inches of tank, but I know I'm not getting it??---Rick
Hi, I don't understand the above formula. On a 24" deep tank for example, how to do the math? it looks like 1 watt @ every 25sq. inches of tank, but I know I'm not getting it??---Rick
Based on my experience and research, I find this difficult to believe. 20 "watts" of multi-chip was more powerful than 36+ "watts" of Cree emitters?
Besides the chips what else did you change?
optics?
drive current?
placement above tank?
photo period?
I'm not saying it was more powerful, I'm saying it better grew coral for me. Both lights were set in the hood, so exact same placement and exact same distance from water surface. No lenses on both. I drove the Cree leds on max power and the only thing I can suppose is that these were simply too strong for such a small tank.
Probably strength or light shock. If you acclimate corals to the power of Cree (start at 50% or less and go up 5% per week) then they will end up doing better.
Ok, that would make sense then, at full power the Cree chips were likely delivering nearly twice the light, which may indeed have simply been too much, or as asid61 mentioned the coral did not have time to acclimate to the new light levels.
Lassef
Member
@ TropTrea: You maybe right if you are standing on dry land. But if you going down a couple of meters below the surface - your not right. During the sunrise when the angle to the sun is below 45 degree - the water surface will reflect more than 50 % of the radiation. After 45 degrees more goes down in the water and at 90 degree all going down. When the sun set below 45 degree again - most radiation is reflected by the water surface. You most also take in shadowing effects from the coastline, cliffs and vegetation. Real measurements at actual coral reefs show that max radiation is only achieved during one to two hours a day.
An other thing: Corals do not care if the brand is Cree, Bridgelux, Epistar, Epileds or no names. They do not care if it 10, 20, or xx watt. They do not care if the efficiency is 100, 120 or 80 lumen's/watt. They only care of to get the right wavelengths for their zooxanthelles. And for most corals - its certain wavelengths in the range of 410 - 470 nm
I also think that acabgd´s experience depends more of wavelengths than the intensity. Corals that get to much of radiation not only stop growing - they will be damaged too
Sincerely Lasse
An other thing: Corals do not care if the brand is Cree, Bridgelux, Epistar, Epileds or no names. They do not care if it 10, 20, or xx watt. They do not care if the efficiency is 100, 120 or 80 lumen's/watt. They only care of to get the right wavelengths for their zooxanthelles. And for most corals - its certain wavelengths in the range of 410 - 470 nm
I also think that acabgd´s experience depends more of wavelengths than the intensity. Corals that get to much of radiation not only stop growing - they will be damaged too
Sincerely Lasse
Last edited:
Intensity is greater than wavelengths in most cases. While it's bad for corals to be exposed to strange wavelengths of light for extended periods, intensity is extremely important to the coral. Light shock is really apparent in many cases, such as switching from Chinese leds to Crees. So while corals don't care about brands, you still have to take into account efficiency.
Chris Abbott
New member
Is there anything negative about not having all led bunched up on one heatsink other then the increased shimmer? I have a second hand fixture that has LEDs arranged in three rows over a 40 breeder. They are about 8 inches over water in a cabinet. 28 royal blue, 9 cool white. no optics.( 3watt Cree)Was thinking of adding 12 regular blue to help with a little better blue spectrum and to increase par a bit. I'm getting about 200 par 4 inches down and 75 on the sand. I can add these for a little over 50 bucks, think it will make any difference in par reading?
When you bunch them up, it stop the disco effect, which is where the colors appear separated from each other on the sandbed and rocks. It's not a nice effect.
Lassef
Member
And why should I use LEDs with high intensity bad wavelengths for my corals? (read: high lumen output 3500 "“ 6000 K LEDs)
I rather use high K LEDs with a higher peak in the 450-460 nm range and much lesser of photons in the "œlumen window" (500 "“ 600 nm) These LEDs in the 14 "“ 20 000 K range will not be found in the human lighting industry.
I prefer to put my money in LEDs of the 410 "“ 470 nm range for the corals pleasure and high Kelvin LEDs for my viewing pleasure. There is RB in the range 450 "“ 465 nm from Cree, Philips, Osram and Bridgelux but to found LED in the 410 "“ 450 nm range "“ the Taiwan based companies Epistar and Epileds is best.
I also prefer the high K LEds because they give a good colour rendition and most of all "“ they are not strong in the "œLumen window" and therefore not blank out the weak radiation from corals fluorescence.
Most of the beautifully colours in coral is caused by fluorescence and to much of reflecting wavelengths in the range of 500 "“ 600 nm just fool our eyes and we do not see the weak colours of fluorescence (compared with the high intensity reflecting wavelengths in the Lumen window)
One of the commercial actors (Pacific Sun) has gone all the way and have no white LEDs in some of their fixtures "“ just in order not to blank out the colours from the fluorescence. They trick instead our eyes to see white through using the RGB trick
Sincerely Lasse
Chris Abbott
New member
So in my case would I be better of adding uv leds or more RB's
than the blues that are 465-485?
than the blues that are 465-485?
Lassef
Member
IMO - yes
But 410 is not UV. UV is below 400 nm and at this moment I do not want to recommend LEDs with peak lower than 410. By my self - - the lowest I use is 420 nm. In the original Dream Chip - the peak wavelengths are 420,430,445 and 455. In your case (as I understand no multi chip) its maybe not wrong to have some in the range 475-485 nm but I think that most of your new LED should cover 410 - 470 nm. The normal "bandwidth" for a monochrome LED is plus/minus 20 nm.
Sincerely Lasse
TropTrea
New member
First of I strongly agree that anything under 420 is either a waste or could possibly have negative effects.
Now you list that your dream chip had LED's that peaked at 430 and 445 can you give me a hint where I can fnd chips in this range. Since I have found some 420 nm chips I would lovew to get some at roughly 427 to help with the gap between the 420 and 455.
I think a little caution neds to kept in mind when you ay they have a band width of 20 nm + and -. Yes a 455 nm Chip probably produces zero or near light at 434nm and at 476 nm. But do to the nature of leds chances are it will be oriducing 50% of the light at or less at 450 and 460 nm. Then you go to 445 and 465 nm the light is probably under 25% of what you have at 455 nm. I will say this does vary by manufacturer and even by run numbers from a single manufacturer. But chances are the more powerfum they are at the specified wave lenght the more narrow the band with is.
A while back I took a spectrum photograph of a combination of equal wattage of 455,470, and 505 nm chips. There was a definatly obvious banding of the frequencies where the gap between the 470 and 505 showed a band of zero light, And a band between to 455 and 470 showed less than 30% of what I had at 470 and 455.
I don't see how that is related to my post. My post had nothing to do with colors of leds; it only stated that wavelength are not everything.
Plenty of people have used high lumen output 3000k-6000k leds for years upon years, and even now most fixtures use them. They are not harmful to your coral, unless you run them very high, which is true of other leds as well.
Violet leds are extremely useful. But in order to make the tank seem brighter, you have to use white leds and royal blues leds or a RGB style fixture, which is risky if you don't know much about leds.
Steve's LED Royal Blues are the 445 bin if your not looking for a multi-chip. I've seen plenty of multi-chips on ebay labeled 445nm as well.
Just google '430nm hyper violet' and you'll find a couple different sources for those.
TropTrea
New member
Well I just had my first "multi Chip" attempt fail. This was a 30 Watt total unit where I basicy used 4 three up chips on a air cooled heat sink. I orignaly started running them at 250 ma and monitored the temp of the heat sink driectly between the chips. First 24 hours running a 250 ma showed a temp of 28C bairly above room temp. second day running at 350 ma temp went up to 29.5C, Day three at 500 ma only went up to 30.4 C. Day 4 I was running at 650 ma after 12 hours I was up to 31.2C but when I checked it at the 24 hour point the ligts were all dark.
Further incestigation showed that one of the chips was no longer on the heat sink . The thermo epoxy for that chip was black both on what was left on the chip as well as on the heat sink, Then the solder connections melted off so I'm assuming the Sar must have gotten extremly hot.
Interestingly I checked the LED's on that multi chip and they are all lighting up individualy. It looks to me like there was a problem with the thermo transfer epoxy that caused to problem. The other 3 chips sho no sign of simular issues though.
Note the Heat sink I used was found at American Scientific. It is about 15 cm diameter and about 7.5 cm tall including a built in fan. No labeling or ratings on it and they only had two in stock which I picked at $12.00 each.
Further incestigation showed that one of the chips was no longer on the heat sink . The thermo epoxy for that chip was black both on what was left on the chip as well as on the heat sink, Then the solder connections melted off so I'm assuming the Sar must have gotten extremly hot.
Interestingly I checked the LED's on that multi chip and they are all lighting up individualy. It looks to me like there was a problem with the thermo transfer epoxy that caused to problem. The other 3 chips sho no sign of simular issues though.
Note the Heat sink I used was found at American Scientific. It is about 15 cm diameter and about 7.5 cm tall including a built in fan. No labeling or ratings on it and they only had two in stock which I picked at $12.00 each.
Call all Wise Multi Chip advisers out there! I need some wisdom for a system i am going to build. Guide Me all wise Multichippers!
Im building a 92 gallon corner tank. It will mainly be sps. I would like to do the least amount of chips over the tank. The tank is a semi circle. Its 38 inches from the back corner to any point in the front of the tank.
What is best solution with the best controlability. Its looking like DreamChip or Lumia 5.1. Maybe there are other options I dont know about.
Could I get away with one dreamchip or would I need two?
Same with the Lumia could I get away with one or will i need two?
I have plenty of room over the tank for optimal height.
Please Advise me!
Im building a 92 gallon corner tank. It will mainly be sps. I would like to do the least amount of chips over the tank. The tank is a semi circle. Its 38 inches from the back corner to any point in the front of the tank.
What is best solution with the best controlability. Its looking like DreamChip or Lumia 5.1. Maybe there are other options I dont know about.
Could I get away with one dreamchip or would I need two?
Same with the Lumia could I get away with one or will i need two?
I have plenty of room over the tank for optimal height.
Please Advise me!
TropTrea
New member
The shape and dimensions of the tank would make a single chip very difficult. Since the sides are 38" long it means you have a thanl that is 51 inches from corner to corner. plus about 38" from the back corner to the front glass. With simply math using 45 degree lenses would mean you lights should be about 40" above the water line to cover this entire tank. You culd use 90 degree lenses to bring the lights down closer to 25 inches above. But my big worry with a single light source that shadows would be extrem plus the higher you go the more light is lost. Therefor I'd recommend three identical lights minimum.
For a 93 Gallon tank I would recommend a minimium of 184 Watts if you going to be keeping the higher light demanding SPS corals. Dividing that into the 3 light sources would give you 62 Watts plus per source.
Ideal situation would be having the colors adjustable on at least two channels 1 being the whites and the other beng the Blues. With the adjustability you can go brighter on your wattage and run them at only a fraction of there total power capability.
For Color I would look at putting into each Source a minimium of 16 Watts of Neutral Whites and minimium of 50 Watts of Blues. Note it you use cool whites you will want to increase it to a minimium of 24 Watts.
The seletion of the Blue chips is were you have ideal and what is available. For the 50 Watts I use a minimium of 8 Watts near UV (420 nm) 30 Watts of Royal Blues (455nm) and 15 Watts of True Blues (470 nm).
So the problem now comes to finding a mutli chip that gives you the closest combination to the ideal. Very often you can find a combination that will meet 3 of your needs but will be excessive on one of them ore insufficient. The excesses can be handleled by putting that color on a seperate chanel and adjustin it down. But the channel that is short leads to a different issue and you need to consider the addition of additional ligh sources to bring that lighting level up. Note many individual use a combination of Multichips and single chips.
