Minimalistic multichip DIY LED build

[tomservo]

will the wattage ber equal for each of these channels? Sound to me like you will need five seperate drivers with the driver selection being dependent upon how many of each of these multichips you will use in the final build.

I do like the idea of using a 420-430 nm LED rather than the 380 to 420 so many use that is marketed as a near UV led. However rather than the 10,00K LED I personaly would select a 5,000K neutral white considering you have 3 LED's in the blue range plus the 16,000K led that is also strong in the blue range. To me this is roughly a 1 to 3.5 white to blue ratio. fortunatly with the seperate adjustable drivers it sounds like it is tunable to ones personal color taste. But if your turning some channels down to 50% or less to get a balance you like then you will need more total wattage potential than if it balanced so you ideal is with everything at full power.

Actually I have finally setup one of my DC's over my tank, and I find it a bit on the white side for me. I have 5 ELN60-48P drivers to run it, and have dialed the current limit down to 500ma @ 100%.. Currently the drivers are set to 30% blue and 20% white. I prefer a look that is a little on the blue side but still crispy white. I've noticed that the white phosphors do get excited by the blue/violet emitters when the white channels are off. I only have the basic reef angel so I just have two light channels to play with but I do want to get the expansion at some point.

I'd have a picture but I've been wanting to scrape the algea from the glass first and it's been a lazy week for me.

Best part - NO DISCO LIGHTS.

 

[Lassef]

Hi everyone. Got a question about this chip -

Ebay link removed~dc

which contact is *+* and whih contact is *-* ?

http://ac-rc.com/images/original_size/LED/EPISTAR_60W_10KK+445nm.jpg

:worried2:

The lower metal clips is minus - the upper is plus.

Between the plastic and the LED matrix you see a thin metal string. Over the minus sign you see a plate and it has contact with this metal string. This metal string to the left is going down to the bottom through a rounded corner and is connected to the metal clips - therefore minus . To the right - over the + sign you see another metal plate - it is connected to the metal string going up wards to the upper metal clips , therefore upper metal clip is positive

Sincerely Lasse

 

[Lassef]

guys a simple question,I dont understand the whole meanwell driver issue and how it works, as i noticed that its better to use meanwells and not the cheap hong kong drivers.
I want to use a mixture of 10W and 20W chips on the nano that have and there is a reefer here selling Meanwell ELN 60-48 P drivers, Can someone please tell me how do to figure out how many of the multichip 10W or 20W chips i can cannot to this driver?

Also how good are the drivers below?
are they worth getting?
anyone use them?

Ebay link removed~dc

thanks in advance

This answer is only valid if you use the AC-RC type of 10 and 20 watts - there is a lot of different chips out there

You can use 3 or 4 pcs of 10 watt in a daisy chain to one ELN-60-48 P- no more - no less (P means it is dim able with PWM - if you have 1-10 V you should use the D version) You have to set down the current to 1000 mA. There is an internal pot to use. You can use the same amount of 20 watts to the driver but you will under drive them (1300 mA instead of 2000 mA)

To run 3 or 4 pcs of 20 watt in a daisy chain at 1880 ma (max 2000 mA) you can use LPF-90D - 48. That driver is dim able both with a pot, PWM and 1-10 V

The drivers you link to is only dim able manually but they work. I have one of this types (but for 3000 mA) and it works rather good.

Sincerely Lasse

 

[ziyaadb]

excellent thanks lassef

 

[reefer859]

Ziyaadb. I use those ac-rc drivers and it works fine for me. My next project will be with meanwells so I can use a controller though

 

[ziyaadb]

thanks reefer

 

[megadeth72]

Hi everyone. Got a question about this chip -

Ebay link removed~dc

which contact is *+* and whih contact is *-* ?

http://ac-rc.com/images/original_size/LED/EPISTAR_60W_10KK+445nm.jpg

:worried2:

please post how you like that combo, I was considering it, or the 50w version that is 20x20k and 30x450-453

they have another combo that's 24x 10000-14000K + 24 chips 455-465nm Blue

What I really want is one that covers both 450-453 and 455-465 something like a "20x 10k + 20 blue + 10 royal blue" but I have not seen that combo yet

 

[hillscp]

I have that 60W combo. I run it at 80W. It is comprable to the 3 150W DE Phoenix they replaced but I would call it more purpleish pinkish than blue. A lot of white. I have three over my 125 long. Wish I had gone with the 100W and supplemented with some 130NM and 10K Whites. I'll be adding those soon.

 

[megadeth72]

I have that 60W combo. I run it at 80W. It is comprable to the 3 150W DE Phoenix they replaced but I would call it more purpleish pinkish than blue. A lot of white. I have three over my 125 long. Wish I had gone with the 100W and supplemented with some 130NM and 10K Whites. I'll be adding those soon.

I'm using a pendant so I'm stuck with whatever single chip I choose

 

[ShipWreck2]

So spent almost two days reading and still have the last 20 pages of post to go through but had to finally ask. The thread started out with an idea, then gave lots of information, had lots of users purchase and build lights, and even plans to design the next led or take over the world ( i got a little confused in the middle there) but what it is lacking is any real feedback on the success or failure of any systems.

So does anyone have some real world experiences with their setups? What size tanks, what size and type of LEDs, and what type of corals?

Now I ask this as I have a 6'x3'x3' (probably will be 32" of water to sand when done) and trying to decide what I should do. I had planned on putting 2 400w MH in lumenbright reflectors but would like to do the multichip LED to reduce the electricity consumption as well as the maintenance. On the other hand I don't have thousands to invest so want to do it as inexpensive as possible and not waste money on any more LEDs then necessary. I plan the tank to be primarily SPS/clams. If you have some specific LEDs you would recommend please PM me the source so the links don't cause issues.

 

[HeavyChicken]

Again I ask, would these multi-chip leds allow me to grow coral in a tank with eight foot of water depth?

 

[bhazard451]

So spent almost two days reading and still have the last 20 pages of post to go through but had to finally ask. The thread started out with an idea, then gave lots of information, had lots of users purchase and build lights, and even plans to design the next led or take over the world ( i got a little confused in the middle there) but what it is lacking is any real feedback on the success or failure of any systems.

So does anyone have some real world experiences with their setups? What size tanks, what size and type of LEDs, and what type of corals?

Now I ask this as I have a 6'x3'x3' (probably will be 32" of water to sand when done) and trying to decide what I should do. I had planned on putting 2 400w MH in lumenbright reflectors but would like to do the multichip LED to reduce the electricity consumption as well as the maintenance. On the other hand I don't have thousands to invest so want to do it as inexpensive as possible and not waste money on any more LEDs then necessary. I plan the tank to be primarily SPS/clams. If you have some specific LEDs you would recommend please PM me the source so the links don't cause issues.

The issue that is apparent with multi chips, is that the best implementation of them is not fully ready, nor readily available.

The "dream chip" which is a 5 channel chip that contains multiple wavelengths and a wide spectrum in one chip, is not readily available to purchase, and is expensive to produce. It also has some kinks to be worked out. Once a process is available to create these chips with selectable color channels, and at a good price, they will be great. I myself am not 100% satisfied with the current channels being used to use them in my builds. They also use epistar LEDs, which do not match up to Cree LEDs in efficiency.

The other issues are, that smaller 10w multichips are not as efficient as a 3-up star or 3 Cree LEDs placed tightly together. 3 Cree XT-E chips in that same space running 5w each will produce a brighter light. You can also create tight clusters with the led colors of your choosing that will be brighter and have higher par than 50-100w multichips, and provide a wider range of spectrum. You would have to use several different multi chips as of now to do something similar, and in my eyes that defeats the purpose of a single point source multi chip anyway.

You are also right, in that despite months of posts, it is mostly a concept, with very few tanks actually using them. The best implementation is Mr. Wilson's, and it is an unfair comparison as he has custom chips provided by Orphek, as he is an Orphek distributor. Similar chips cost $800 from Orphek if you buy their fixture that uses them.

I think the PAR reading in the sandbed for one of the dream chips at 24" down a few pages back was 150-200? My apogee meter reading for my 12" 180w fixture using Crees and Luxeons w 90 degree optics reaches an average of 500 with minimal color shadows at 24" at full blast. It has also been running for months with excellent results using neutral and warm whites that many are against in this thread. Neutral and warm whites are also not an available choice in a dream chip implementation.

In time these could be great, but they aren't the best or most efficient implementation to me as of now, and don't provide the flexibility and color choice of single emitters.

Sent from my SGH-T999 using Tapatalk 2

 

[TropTrea]

The issue that is apparent with multi chips, is that the best implementation of them is not fully ready, nor readily available.

The "dream chip" which is a 5 channel chip that contains multiple wavelengths and a wide spectrum in one chip, is not readily available to purchase, and is expensive to produce. It also has some kinks to be worked out. Once a process is available to create these chips with selectable color channels, and at a good price, they will be great. I myself am not 100% satisfied with the current channels being used to use them in my builds. They also use epistar LEDs, which do not match up to Cree LEDs in efficiency.

The other issues are, that smaller 10w multichips are not as efficient as a 3-up star or 3 Cree LEDs placed tightly together. 3 Cree XT-E chips in that same space running 5w each will produce a brighter light. You can also create tight clusters with the led colors of your choosing that will be brighter and have higher par than 50-100w multichips, and provide a wider range of spectrum. You would have to use several different multi chips as of now to do something similar, and in my eyes that defeats the purpose of a single point source multi chip anyway.

You are also right, in that despite months of posts, it is mostly a concept, with very few tanks actually using them. The best implementation is Mr. Wilson's, and it is an unfair comparison as he has custom chips provided by Orphek, as he is an Orphek distributor. Similar chips cost $800 from Orphek if you buy their fixture that uses them.

I think the PAR reading in the sandbed for one of the dream chips at 24" down a few pages back was 150-200? My apogee meter reading for my 12" 180w fixture using Crees and Luxeons w 90 degree optics reaches an average of 500 with minimal color shadows at 24" at full blast. It has also been running for months with excellent results using neutral and warm whites that many are against in this thread. Neutral and warm whites are also not an available choice in a dream chip implementation.

In time these could be great, but they aren't the best or most efficient implementation to me as of now, and don't provide the flexibility and color choice of single emitters.

Sent from my SGH-T999 using Tapatalk 2

Wow I think you summerized this up fantasticly. I do believe though that the multichips out there today do have there place in some of the very tall tanks. But pricing and there effeciency are a big factor. Personaly I think some of Cree's newer single chips rated up to 10 watts will eventualy knock out most of the multi chips when they become available in more colors.

 

[Lassef]

I think the PAR reading in the sandbed for one of the dream chips at 24" down a few pages back was 150-200?

Sent from my SGH-T999 using Tapatalk 2

Source for that statement, thank you.

Tomservo has done some test with the Dreamchip and 100 degree lens and his results says another thing. He has done it in air (in water you will read a slight higher PAR because the water surface act as a second lens. This I have seen many times) Post 2935 show his results.

He get a PAR reading of 400 with 100 degree lens at 24" depth.

Sincerely Lasse

 

[Lassef]

Again I ask, would these multi-chip leds allow me to grow coral in a tank with eight foot of water depth?

8 ft is around 250 cm if I know my math. I do not think so - at least not sps . The highest wattage with anymulti chipp that suits reef aquariums is around 250 and the intensity is probably not enough for that deep. However if yourmulti chipp has a lot of thewavelengthss around 420 - 460 probably mushrooms and softies will survive. Just now - the most narrow lens is 60 degree and I think it is nor narrow enough. I have test 100 wattsmulti chipp (60 degree lens) at 4 metersdepthh - At this depth it gives a light which for your eyes is similar with 400 MH in the center but to grow corals there is an other issue.

Sincerely Lasse

 

[Lassef]

So spent almost two days reading and still have the last 20 pages of post to go through but had to finally ask. The thread started out with an idea, then gave lots of information, had lots of users purchase and build lights, and even plans to design the next led or take over the world ( i got a little confused in the middle there) but what it is lacking is any real feedback on the success or failure of any systems.

So does anyone have some real world experiences with their setups? What size tanks, what size and type of LEDs, and what type of corals?

Now I ask this as I have a 6'x3'x3' (probably will be 32" of water to sand when done) and trying to decide what I should do. I had planned on putting 2 400w MH in lumenbright reflectors but would like to do the multichip LED to reduce the electricity consumption as well as the maintenance. On the other hand I don't have thousands to invest so want to do it as inexpensive as possible and not waste money on any more LEDs then necessary. I plan the tank to be primarily SPS/clams. If you have some specific LEDs you would recommend please PM me the source so the links don't cause issues.

007Bond has a similar tank and he is going to do a build with 4 pcs of dreamchip.

This is a new thing (with multichip) and there is not so much of experiences with tanks as deep as yours. In air (100 degre lens) and with a Dream Chip tomservo measured the following PAR

PAR measured in air / 100 degree optic (inches)
Under light
18 - 675
24 - 400
36 - 185
38 - 165

12" offset (24" spread)
18 - 340
24 - 265
36 - 165
38 - 148

24" offset (4 foot spread)
18 - out of light cone
24 - 100
36 - 108
38 - 100


He run it around 220 W (max 250 Watt)

PAR readings in water will normally be a little bit higher compared with reading in air at the same distance. The water surface act as a week lens.

Sincerely Lasse

 

[BeanAnimal]

in air (in water you will read a slight higher PAR because the water surface act as a second lens. This I have seen many times)

No...

Par readings can sometimes be higher in a small tank compared to a large tank simply due to the glass reflecting some of the light back to the sensor. The water absorbs far more light than the air, per distance traveled and any collimating effect of the water (surface refraction) is negated by the scattering and reflectance of the (not smooth) water surface. That is, even if the water surface was a collimating lens (breaking the inverse square law), the losses through the body of water are greater than that of the collimating effect, save the reflection from the glass.

Lets not make LEDs or the "dream chip" into some magical device that is not bound by the laws of physics

 

[mcgyvr]

Are the "dream chips" ready to be purchased yet by those that weren't in on the group buy?
Or is there a thread specifically for those chips? maybe on another board or something?

I'm gonna be ready to build the lighting for my new tank VERY soon and would like to go in the multichip 50W+ direction versus star PCB leds.

 

[TropTrea]

I do believe that the single dream multicip will have its place however I do see some big draw backs. The biggest is that it a single light source. With a single light source it is impossible to get around the inverse square rule of physics. Sure lenses creating a more beamed effect will decrease the effect of this however but it is still something that needs to strongly considered.

It is the inverse square law that has put the HT-5 bulbls at an aedvantage over the MH's. Basicly you have a point source with the MH and a long thin source of light with the florescent bulbs. This puts the florescent bulbs at a slight advantage when your looking at penetration. Simularly going with multi LED chips allows a for a grid work of chips that work together spreading out the sourcing of light compared to one a single point of light.

the biggest advantages I see with multi LED's rather than multi chips is the eveneness of the light distribution in the aquarium. If you have 30 individual LED sources of light each shinning on one particular point on the substrate it becomes indistinguishable as to which sorce is creating which baem of light. When an object is put in the way of a single beam of light it prduces a hard shadow while with multi sources multi shadows are produced at different angles producing a much more even blend.

A good example can be seen in many pictures of tanks. Do you sometimes notice the coral on the top being brightly light up yet the area onder the corals in there shadows are appearing almost black with almost no light reaching them. Some individuals may like this type of effect however I personaly do not.

To me the goal should be to get as even of a light distribution as possible. It should not create hard shadows even though some shadowing is unavoidable, unless you had light sources on all sides.

As the single Chip LED's are becomming more powerful I believe they will make the multichips obsolete before they we know it. Cree is making 10 Watt chips at this point which are more effecient than most multi chips. These higher Wattage LED's are ideal for taller tanks and as single chip move to even higher wattages they will be even more effecient for the tallet tanks.

Another big factor is cost for many people. Cree 10 Watt chips are now selling in the 6 to 7 dollar range. Can you get a 50 watt Multi chip as effecient as 5 10 watt chips for under $35.00 or a 250 Watt multi chip for under $175.00.

Finaly you have cooling. Whith the multi chip cooling can veret easily be a bigger expense than anything else. Yet with multiple individual chips the ability to distribute the heat throughout the fixture becomes considerably less expensive.

 

[jason12345]

Anyone using these to supplement their MH? I have a 175 watt in a lumenbright, how strong of a Chip would i need to make an effect on the color spectrum?