Minimalistic multichip DIY LED build

http://www.flickr.com/photos/45627954@N02/8044080537/in/photostream

Updated my Multichip 50w hybrid with 3 20w royal blues on a dimmer. Should I put lens on the RB or should I just leave it as is?

Lasse - Could I ask your expertise on this? I have a
20w RB ---50w Hybrid---20RB---50w Hybrid--and 20 watt Rb
Over a 48x18x18 hung 2 ft from the water. Would you recommend I add any other kinds of chips and should I put lens on my RB chips? I have 60 degree lens on the 50 watt hybrid chips.

I do not own a par meter so should I raise or lower my chips?
 
@DetroitReefer : I have talk to my "electronic guru" and he has not a quick fix to this problem either. He agree with me that it should be wise not to exceed 60 % for the 10 watts. Some PWM dims has a construction which in practice gives a current adjustment. It is difficult to know if your drives have this construction but if you can measure that the current is varying with dim level it should probably work.

In the long run, both He and me thinks that it is better to replace your Sure drives to those that provide up to 1000 mA.

I fully understand why you prefer this solution - it provides full control of your lighting. At some of my solutions, I have max 2 multichip on one drive - just to be in full control. I also let the "sun" rise at one end of the aquarium.

Sincerely Lasse

I was looking at the datasheets for the drivers and the difference in the 10/20 watt drivers and the 5w drivers is 500MaH. I took my desoldering iron and separated 9 of my 15 drivers into their lesser-current 5W counterparts. I am now running the 10W at 1023MaH. Heat is almost non-existent and they are just as bright to my naked eye. It would have cost me $60 for 9 new 5W drivers, so I've managed to save some cash and I am driving the LEDs at their proper V and I. The power supply is nearing its limit, but they are working perfectly. The Power supply is 24V at 14.6A. I suspect I am somewhere near 11A draw (I purchased a back-up power supply just in case).

One power supply, 15 drivers, completely controllable led arrays. This thread, and you, Lasse, have been an inspiration. Thank you!
 
Only one additional thing to keep in mind. And that is that they now have 10 Watt single chip LED's. It would not surprise me if within a year or two we will be seing single chips up to 25 Watt.

My plans for my next build will be using the 10 Watt single chip LED's. Hopefully by then they will be available in all colors.

The other thing to keep in mind is that LEDs are more efficient at low wattage, so 10pcs 1w is more efficient than 1pc 10w.

Multichips allow you to mix different colours in a small footprint for better colour mixing and single light source shimmer, similar to MH.

RWB (red, white, blue) mutichips give you colour shifting flexibility that higher wattage single chips do not. While this is not particularly useful for an end user, it is beneficial for manufacturers who need to cater to varying end user taste. In other words, if you like a blue tank, order a higher ratio of blue chips, rather than wasting resources on chip rows you will inevitably dim.
 
I my opinion, multichip LEDs offer a tailor made spectrum in the smallest possible footprint. As a result, colour mixing is optimized and natural single light source shadows are created (no disco effect). Metal halide is limited by phosphor mixed in a single tube, and larger LED arrays give you a spotty "Lite Brite" mix of colours.

The large glass dome lenses used with multichips provide more uniform coverage than a collection of smaller lenses used with an LED array.

The dream chip below illustrates how colours would mix in a large array (12" x 12") array of 3w chips, vs. a 2" x 2" multichip with single lens.

file-22.png
 
LOL!!!!!!!!

Where can I order this one?

I'm still ironing out the details and have a bunch of beta testers lined up. I'm working on software to have the eyes blink for lightning simulation and have the mouth slowly smile as it gets brighter at noon peak sun levels. Sanjay's lighting talk at MACNA edified my belief that the jaunty red cap will not bleach corals :)
 
mr wilson i have a few questions

1 where can i get that panel ?
2 are all five colors on a separate channel ?
3 would i need to buy five drivers could i not get away with three ?
4 would all the LEDs go out if one burns out ?
5 do you think it is wise to put all your eggs (LEDs) in one basket ?
6 my wife said she thinks she has seen that some ware before is it available ? now

Marc
 
mr wilson i have a few questions

1 where can i get that panel ?
2 are all five colors on a separate channel ?
3 would i need to buy five drivers could i not get away with three ?
4 would all the LEDs go out if one burns out ?
5 do you think it is wise to put all your eggs (LEDs) in one basket ?
6 my wife said she thinks she has seen that some ware before is it available ? now

Marc

Your wife recognizes one of the most iconic toys of the 60s & 70s, Lite Brite. It is a pegboard panel that looks remarkably like the LED arrays we use.

I won't singe out any one brand because they all share the same weakness, but 3w arrays offer a surplus of one spectrum in one area while a deficit in another. The net result is red, white and blue hot spots throughout the reef. While some people like the look of a red spotlight on a particular coral, it certainly isn't natural, and it's arguably detrimental.

When we look at spectrographs of various LED fixtures on the market, they don't take into account the directional nature of the individual LEDs. The graph may show lots of UV light, but it may only be available in a couple of small locations in your tank. The all too familiar disco effect of purple and yellow shadows can be avoided with multichips and the lenses that come with them.

Multichips are able to mix the colours by simply locating them close together and through larger dome lenses which are harder to find in 3w array fixtures. I saw at MACNA that Acan Lighting has a fixture that uses a 3w array and clusters them together under a larger dome-like lens. Blueline and Orphek also had single dome lenses in their new 3w arrays, but colour separation is still an issue.

In the end, lens technology may prove to be more important than chip technology. Some manufacturers are offering lenses with phosphor coatings so one can alter an existing lens. In the future the end user may be able to manipulate these phosphors to alter the spectrum of the light, This could be a simple colour wheel application as seen with RGB lighting or a more sophisticated chemical method of achieving the same goal.

The idea of a fully adjustable LED fixture may appeal to many, but if your corals had a say in the matter, they would request a steady photoperiod at 100% illumination with a spectrum that represents that of the sun in their natural environment. The corals we buy are adapted to a depth of 6-10' and will be exposed to full spectrum light, including red.

There are so many factors that contribute to the colour and growth rates of coral, it is silly to think that one can change the programming of an LED fixture and expect to see a measurable change. Im not saying you wont see varying results, but keep in mind we (the hobby & scientific community) have been experimenting with reef lighting for 30 years so you won't make any earth shattering discoveries overnight.

Pick a colour you like that hopefully benefits your tank and build a light that will achieve that goal. I don't think we will be seeing cloud and storm simulations offered in the coming years, and a slow ramp-up of light will also be a passing trend. Dawn and dusk simulation has benefits for the viewer and inhabitants of the tank so it is here to stay. The single most important lighting feature in my opinion is a lunar simulation. It has been proven to induce spawning in fish and coral, yet many of the most complex fixtures on the market overlook this feature???
 
forgive me i have read the thread but i can not remember is there a way to compare the chips to MH
like the
50w & 60w 45mil with optic is = to a 250w MH
100w 45mil with optic is = to a 400w MH
 
From my observations...

50w multichip = 150w mh
100w multichip = 250w mh

The 50w chips are suitable for a tank 20" tall or shorter. The 100w chips are suitable for up to 30" deep. Of course this is with appropriate lenses. 90 & 120 lenses are best for most applications.
 
Mr. Wilson, do you think just All On / All Off on time switches is fine?

There doesn't appear to be a benefit to varying the light intensity throughout the day. In general, the ambient light from windows and home lighting tend to create their own dawn and dusk sequences, but exclusive blue light in the morning and at night for an hour looks cool.

Most multichips don't allow you to independently control blue and white channels, so it's either all on or all off. You can slowly increase intensity over an hour instead. When white chips are dimmed they look bluer, do you get a similar effect.

Moonlight is full spectrum as it is the sun's light reflected in the moon, not blue, and sunrise and sunset appears red/orange as the light is transmitted through the earth's atmosphere, and not blue as often seen in our tanks.
 
I think Mr Wilson's point is that once you get past 20" tall, you're already talking about a big-ish tank, so that 50w LED will likely be asked to cover an 18x18x20 area or more. I think this is enough for some corals but not all.. Unless we're talking about a 45mil "50 watt".

That said, I have made a bunch of 38 watt bridgelux 5600k fixtures for planted tanks, and they cover 18x18x24 with 115-125 PAR on the gravel, with 100 degree lenses and a spread of at least 75% at the glass lid. But this is a strictly white light emitter, so the PAR numbers aren't really comparable to what we need for a reef tank.
 
Thanks for the help but sill I want to clear some doubts

Here is a diagram I made to show how I was thinking to connect everything




Everything would be in parallel

Please guide me if I am on the right track.

Thanks

Please anyone
 
Back
Top