Minimalistic multichip DIY LED build

[Lassef]

With commercially made lights the cooling standard has to be higher, as the end user may live in hot Texas or chilly Sweden (or worse Canada). They may have a well vented open top or a tightly sealed heat trap. As a general rule, DIY hobbyists never finish anything so the DIY LEDs will not have any contributing heat or venting issues, and ambient heat will be manageable.

It's easy to take temp readings when the light is in the bare bones stage, but it's the completed project with waterproof seals, lenses affixed and housings installed where the real heat test comes into play.

Yes and I´m living in Sweden - Noplay 180 in Florida - its a long way for him to trawel to me just in order to cut my head of

Sincerely Lasse

 

[mr.wilson]

Yes and I´m living in Sweden - Noplay 180 in Florida - its a long way for him to trawel to me just in order to cut my head of

Sincerely Lasse

Ah yes, the great distance makes it easy to give advice

 

[TropTrea]

Look Good Lasse and 12mm will be a lot better than 1/4 way 2 thin
Best to be over cool than not enough

Keep in mind that surface are is almost a direct ratio to cooling capacity, while thickness is almost a direct constant. So the big issue is how moch actual contact you have between the mounting plate and the cooling system for transfering heat from one to the other.

 

[tntneon]

Exactly! The only thing I would add is that metal halide radiates heat over greater distances, thus heating your water, while LED is very hot....

Hi mr. wilson ,

-If MH radiate over a bigger distance does it not means that there is more energy converted into heat (IR) than with leds ?
Also the mass / heat of the MH bulb is higher than a led , leaving me to think again that the led is more efficient than MH in converting energy into light?

-I know for sure that you guys know more about lighting then i do , but it still it leaves me thinking ....


greetingzz tntneon

 

[Lassef]

Keep in mind that surface are is almost a direct ratio to cooling capacity, while thickness is almost a direct constant. So the big issue is how moch actual contact you have between the mounting plate and the cooling system for transfering heat from one to the other.

The contact area is the transfer area but whats matter mostly is the temperature differences between the two systems. The heatpipe constructions move the the heat out to cooling fins in a great way and therfore keep the temperature low at the base. I imagine that a larger mass of this particular design is a moderating effect of rapid increases in power and thus more heat. With the VGA coolers You also get an air stream down over the plate which helps a bit with the cooling

Heatpipes works like a heat pump - and are very effective to move heat from one point to another. The small footprint of the VGA cooler is able to take care of app 250 watt heat. You can compare with water cooling.

The only problem I can see with this construction is that instead for one transfer area we get 2. Therefore it is very important to making the joints properly.

Sincerely Lasse

 

[TropTrea]

But if you do not belive me -and thats up to you - please, stop trying to indicate that I have opinions that I do not have. Once again - I do not like the blue look!

The "dream chip" adventure starts with a question -which would be the optimum configuration of a 100 watt chip to provide both a good appearance and that the light would not be limiting for growth. SIncerely Lasse

I hpe this was not directed at me. I believe everyone has a different idea of what the perfect coloration is. I have repeatedly thought a tank looks too blue and someone else commented that it lacks blue light. I have also had the opposite where I thought it was too white and sopmeone else said it looked to blue.

The most amaizing thing is that when I used some 420 nm LED's sparingly on my system everyone commented they made my tank look very pink. Some of my friends are frequent divers in the ocean and all commented that the light in the ocean gets bluer but not pink. the removal of 4 420 nm chips and replacing them with 2 505 nm Chips on a 24 chip system completly changed everyone thoughts on the color balance.

However going back to some of my optics and light classes in college and my knowledge of the ocean today I wounder if it even poosible to get that perfect white light and still have a good balance for coral growth. Te true perfect white would have to excite the three colors the human eye is sensative to in an equal ratio. However if this were done we woul not have enough blue light, or we would have a dangeriously high level for corals at the other wave lenghts.

My experience is that a blance between the Neutral White Chips 4,000K to 5,000K is the most effective way of getting the balance against the Blue Chips we use photosynthesis. To my eye a 3 Blue to 1 Neutral White looks much crisper in all the colors than some systems that had a 1 to 1 ratio using so called 20,000K white chips.

But the combination I thrive for is.
1. Producing enough Light in the right frequencies to suport the corals needs not only in photosynthesis but also in there florescense.
2. Producing a color balance that is pleasing to the observers eye. The second being the hardest since everyones color taste is different. Andnone are more wrong or right than someone elses since there tastes naturaly vary.

 

[Lassef]

Hi mr. wilson ,

-If MH radiate over a bigger distance does it not means that there is more energy converted into heat (IR) than with leds ?
Also the mass / heat of the MH bulb is higher than a led , leaving me to think again that the led is more efficient than MH in converting energy into light?

-I know for sure that you guys know more about lighting then i do , but it still it leaves me thinking ....


greetingzz tntneon

The important thing for us according to a light source is the first row - Visible light- This row contains the energy we are interested in. First, the energy needed for photosynthesis and secondly the energy that allows us to see our surroundings.

Whatever name we give to the waste (the part we can not use for our purposes) is not interesting. Looking at the table in this way, it turns out that the three main light sources provide about the same amount of the wavelengths we are interested in. Thus - for our purposes - producing visible light - they are about as effective in relation to the energy input.

Sincerely Lasse

 

[Lassef]

@ Trop Trea

I´m sorry if I missunderstand you - I got the impression that you thought that I advocate the blue look. Enough of this.

The most amaizing thing is that when I used some 420 nm LED's sparingly on my system everyone commented they made my tank look very pink.

This is strange but interesting. Which type of 420 nm LED did you use. I use one 20 watt 420 nm chip in one of mine tanks (epiled 45 mil). I run it very low. It does not give an overall pink color but my white xenia just below this chip looks a little bit rose/pink. The 420 nm is know as the wavelenght that probably promote fluorocense best in corals. There is other wavelenghts also involved in this but 420 seems to be the most common. I´ll see if i can get a good picture of this tomorrow.

My experience is that a blance between the Neutral White Chips 4,000K to 5,000K is the most effective way of getting the balance against the Blue Chips we use photosynthesis. To my eye a 3 Blue to 1 Neutral White looks much crisper in all the colors than some systems that had a 1 to 1 ratio using so called 20,000K white chips.

For my eyes (and I´m 62 and do not see blue wavelenghts very good) a ratio of 1:2 between 455 nm blue and 10 - 16 000 K white looks best but I agree totally in the sentense

And none are more wrong or right than someone elses since there tastes naturaly vary

Sincerely Lasse

 

[tomservo]

Not sure why the 420 would have made people see a pink tint.. I currently have 18w of 420nm over my 12g, and I expect about 18w of 445 and 10-15w of 4100k white. Definitely not a pink tint. The 420 by itself is a very dim illumination, but the corals just glow like crazy.

 

[TropTrea]

@ Trop Trea


< I mentined the pink effect from 420 nm LED's>

This is strange but interesting. Which type of 420 nm LED did you use. I use one 20 watt 420 nm chip in one of mine tanks (epiled 45 mil). I run it very low. It does not give an overall pink color but my white xenia just below this chip looks a little bit rose/pink. The 420 nm is know as the wavelenght that probably promote fluorocense best in corals. There is other wavelenghts also involved in this but 420 seems to be the most common. I´ll see if i can get a good picture of this tomorrow.

Sincerely Lasse

The "420 nm" LED's I picked up the distrinutor did not list the brand name. They physicialy look very simular to the design Phillips uses rather than Cree, and the curent to voltage range seems to be much closer to the Philips LED's than the Cree's as well.

My thoughts on the pink effect is that since 420 nm is on the edge of the visual spectrum what is actualy happening is that violet pigments that give us the night vision are being activated by these led's. IT is also being slightly activated by the Royal Blues as opposed to the Blues which is why the Royal Blues look more purple to some of us than the True Blues. Keep in mind that this is my personal theory only and could be completly wrong or right on?

For my eyes (and I´m 62 and do not see blue wavelenghts very good) a ratio of 1:2 between 455 nm blue and 10 - 16 000 K white looks best but I agree totally in the sentense

Sincerely Lasse

Well I do have a few years on you. So perhaps this is qhy I like that 1 to 3 ratio with 4,000K LED's compared to smeone else liking a 1 to 1 Ratio. But my thoughts are make the corals happy first then add in any other colors to comprimise there preference to your personal likes.

Yes I have done a lot of experementing with single emeter LED's. But very little with multi emiter LED's. going back years ago I also had excess to spectrum anylizers and did a lot of reaserch on T-5's. However I do not have a excess to a spectrum anylizer today and it ourt of my budget range. From the research I did I will say I have little trust in what any any manufacturer claims there lights actualy put out.

I did a stint with a lighting manufacturer and know how there specs were created and maintained. If they rate a light source at 20,000K in reality it could be anywhgere from 16,000K to 28,000K. Simularly if it is rated at 10,000K to could be anywhere from 8,000K to 14,000K. They simply look at three specific wave lenghts and look at the average balance between them. Ten manufacturing specs are to keep each of these within 10 % of there ideal. But if red is 9% high and blue is 9% low it is still in there spec but the true K rating may be considerably lower than advertised. In the event this is reversed the opposite can occur but the K rating jumps even more than in the prior case.

 

[Lassef]

@ Trop Trea

Is this something for you? One is on the way to Sweden just now

Picture in this post

Sincerely Lasse

 

[Screener]

I reelly hate it when native English speakers spell so pourly

There is also this lasse
Here

 

[bhazard451]

How's the progress of the dream chip coming along? I look forward to seeing setups with it, and that custom driver sounds promising.

 

[jimmy_beaner]

I reelly hate it when native English speakers spell so pourly

Oh the irony.

In other news, is there a preferred manufacturer for these chips?

 

[Lassef]

I know that I write with a slight swedish accident but I hope you understand me nevertheless (I´m not sure on this sentense but Google says so)

More than 60 chip's are manufactured right now and they will be shipped to people in the United States, Great Britain, South Africa, Canada, Sweden and Norway. Everyone knows that it is an experiment and that it is a project where you do not know how it will work fully. It will take some months before the first constructions will see the light of day. Hopefully there will be a collective documentation of the various constructs.

Sincerely Lasse

 

[Screener]

Oh the irony.

In other news, is there a preferred manufacturer for these chips?

Did you miss the sarcasm? We all know lasse is not a native English speaker.

Spectrometer

 

[tntneon]

@ Trop Trea

Is this something for you? One is on the way to Sweden just now

Picture in this post

Sincerely Lasse

-Thx for sharing this link with us Lasse .
I was going almost nuts when i saw the DIY spectrometer , this wil/ must be my next DIY project !!! :thumbsup:
-What hits me was that the warm white xm-l almost has the same wavelenght signature than sunlight w/o the sub 400 wavelenghts.

greetingzz tntneon

 

[kuya_cesar]

@ Trop Trea

Is this something for you? One is on the way to Sweden just now

Picture in this post

Sincerely Lasse

This is so cool, and I am currently building one for my self, but decided to do the mobile phone version.

 

[Lassef]

It looks like they change something so you can´t buy directly. When I visit - I got a link here and buy from them.

Sincerely Lasse

 

[BeanAnimal]

there was a unit for sale, but now it appears the instructions are using a webcam...