Minimalistic multichip DIY LED build

[bhazard451]

Led color preferences aside, I think its really cool that we will be able to wire up one chip with multiple drivers within a few minutes, and create a fixture that would normally take hours

 

[BeanAnimal]

I am not sure we are all on the same page and it may be that we can't agree on a chip combo as a result.

To be very claer: I could give a rats-*** about algal growth if it means compromising the color appeal of the tank. If martian green grew coral and no algae, geat, I don't want it because I don't want to look at it

 

[bhazard451]

You know what else grows super poofy green tufts of algae? Royal blue Crees, I have a "junky softy" attachment to my main tank, I just tossed on a few groups of LEDs, and on one side there's nothing but royal blues, and that side gets the thickest algae growth, the side with significant whites (4:5 CW:RB, with CW's running at 1.05A to 700mA of blue) not a single lick of algae...

You know, you do have a solid point. Blue can be just as bad.

This is my Freshwater tank that is being converted to SW soon. I'm waiting on a new mag-float from Amazon to arrive... hasn't been cleaned in over a week. This is the side of the tank with 120w of 3:2 Royal blue to 10-14k Epistars, 2 reds 2 greens, 2 uv (uv is useless, I prefer 420nm). (test fixture)

This is the side with warmer smds and very few rb, no more than 20 watts or so total for this entire side.

Its not really a fair comparison, but yes lots of blue can grow algae and diatoms pretty well.

 

[ronreef]

Change of pace. This is on the horizon from ac-rc, or some variant thereof.

https://picasaweb.google.com/103048...&authkey=Gv1sRgCMvxy6Wgj--EVw&feat=directlink

 

[jerpa]

I am not sure we are all on the same page and it may be that we can't agree on a chip combo as a result.

To be very claer: I could give a rats-*** about algal growth if it means compromising the color appeal of the tank. If martian green grew coral and no algae, geat, I don't want it because I don't want to look at it

Yup

To that end, it may be prudent to order these in 10W single color versions so that mix and match can be done before a final chip is settled upon.

You build the drivers I'll buy the LEDs.

 

[BeanAnimal]

For testing purposes, the drivers can be as simple as my bench supply set to constant current mode...

 

[bhazard451]

I think we're all on the right track into designing something special. I think the end result will be something that catches the attention of a lot of reefers and maybe even some bigger companies.

With that said, one custom chip could never satisfy everyone because of different preferences, but its a good start. The options to choose from for this custom chip already beat out anything I would order otherwise. There is huge appeal in one powerful chip that can be driven over my nano in the color channels and intensities I desire, with the added benefit of a single point source for optimal shimmer and less shadows.

 

[tomservo]

Cyan looks good in very low numbers, like 4 leds tops per 24" of length, and mixed with a red and blue next to it. I have 4 of them in my 48 led fixture on a separate string of colored leds only. Any more than that (applies to greens and reds too) are extremely overpowering.

To be clear, those 12 OCW were spread out over 78x26" of tank, half of what you're recommending and they were awful at any but the lowest possible dimming level, and at that point on vs off made no difference.

 

[gtfcniles]

That was my point. Almost every commercial fixture has too many blue for my taste as do almost all DIY builds. If I could deal with the disco effect I would build an array to suit my tastes. I like the color blending of multichip arrays but I can't find a commercial fixture that satisfies me aesthetically. My ideal multichip would be dominated by lower kelvin whites, UV, and at least two blues. IME I would need a 1:1 ratio of whites to blues to run it anywhere near full power. In my situation your ratios would look awful to me no matter what white I used. I could deal with one low Kelvin white but would prefer two whites. Aesthetics are a big part of that and to each his own but i fail to see how any of these options will allow me to get a satisfactory look. Perhaps these LEDs are able to accurately render reds and pinks. In that case I would still want it dominated by 10000K whites. The only ones I have seen that do warm colors well use lower Kelvin whites and based only on my experience i would need an emitter with a large number of them in relation to blues.

Why not just use a 10k chip if thats the look you want mate?

 

[gtfcniles]

ac-rc has informed me that if we can manage a total of 100 pcs for this group buy we can do two different color configurations. I will get more details.

So far, like it or not, what it really comes down to are these two options for the first color configuration... so far.

Configuration A:
1) 20 x 10000K
2) 20 x 455nm
3) 20 x 445nm
4) 10 x 420nm, 10 x 430nm
5) 20 x 15000K

Configuration B:
1) 20 x 10000K
2) 20 x 455nm
3) 20 x 445nm
4) 10 x 420nm, 10 x 430nm
5) 20 x 6500K

Tomorrow, ac-rc will inform me of what additional LED's are available <6500K.
Then we can take another stab at a whiter (less blue dominant) color configuration.

If its down to these two I vote for configuration A, i'm bowing to lasse's superior knowledge.

 

[bigguyuk]

Sorry Guys been away for a few weeks are we now looking at custom chips from ac-rc
and can these be dimmable as a hole chip or each color
Sorry For the question

 

[gtfcniles]

Sorry Guys been away for a few weeks are we now looking at custom chips from ac-rc
and can these be dimmable as a hole chip or each color
Sorry For the question

They are dimmable if you use seperate drivers for each channel that have a dimming function. You could also wire 2, 3 or 4 channels with the same driver if you only want one channel dimming or dimming together.

 

[bigguyuk]

Cool What Cost will they Be sound very Intresting

 

[zone]

Option A for me, mark me down for 3 chips for starters, I actually need about 9 in total.

 

[BeanAnimal]

Why not just use a 10k chip if thats the look you want mate?

I think we keep going in circles here...

Some of us (after looking at numerous fixtures) have come to the conclusion that the light they emmit looks like crap! We have found that the only fixtures that use multiple (5-6) colors of LEDs have a true visual appeal.

To that end, most of the fixtures are far too blue at any reasonable level of output. The blue is nice to be able to dial up for a short period, but we don't want it all the time. That is, we would rather have the look of a CRISP WHITE halide with an actinic suppliment like the Iwasaki with a UVL actinic. You don't see blue water, just pop.

Furthermore, each company uses different phosphors and therefore terms like 10000k or 65000k really mean little.

So when askef if we want a "custom" chip, we want one that can be tuned to any specific color. This can be done with individual LEDs but then you get disco. With the multi-chip you get a point source.

In addition, if there are "so many" colors on the chip that it is not powerful enough (bright enough) at certain drive ratios, then we will simply add more chips to thje build. They are point sources, so we won't get disco.

I think a few of you are having trouble understanding this and frankly trying to replicate the same old crappy 2:1 RB:CW ratio that most of us hate to begin with.

 

[jerpa]

Why not just use a 10k chip if thats the look you want mate?

Part of the draw with LEDs is the ability to dial in the exact color I want without changing chips. I for one would like the ability to go from say 8K-14K. Ideally the chip would let us go from 8K-22K but I think a chip with that ability would be inefficient for those looking for something towards either end of the range. If the 10K accurately renders all colors then I would use that as the base white and supplement with 2-3 channels of blue. I just doubt this LEDs ability, or any single high Kelvin chip, to accurately render colors across the board.

If we can get to 100 chips I think we could do a white dominated version with a range of 8-15K and a blue dominant diode that would cover 14-22K. I think this would give all prospective buyers a chip they can dial in as they see fit without needing an absurd number of chips for coverage and intensity due to running only one channel at full power. To be honest the only option I'm confident I could get a pleasing look out of at any drive current on any of the channels is the option with 6500K.

I'm waiting to here back as to whether they use the same phosphor across the various wattage of LEDs. If they do I will be ordering some LEDs in the lower wattages to see what ratio I would actually need to get a color I like.


Or you can just read Bean's post above. He obviously types faster..

 

[gtfcniles]

lol

I'm just going into the background now, but i'll say that I won't be purchasing any chips that has 6500k leds in it. If I want to add these I would run 20w or 50w either side the multichip so you wouldn't get the disco effect. I don't want 6500k in my display for algae reasons.

 

[jerpa]

lol

I'm just going into the background now, but i'll say that I won't be purchasing any chips that has 6500k leds in it. If I want to add these I would run 20w or 50w either side the multichip so you wouldn't get the disco effect. I don't want 6500k in my display for algae reasons.

In that situation you will still get color separation though. You will have two point sources with different colors. IMO this negates my reasoning for going with a multichip. You could just not drive the 6500K channel if algae is your concern. Run the higher Kelvin white channel with the blues and lose some intensity. I could choose to vastly underpower the blues and with the 6500K dial in the color I like while also losing intensity. We are both sacrificing intensity to get the color we like, but if you want a one-size-fits-all emitter I think this is necessary. I believe option 2A, or whatever the h@!! number we are at, gives the widest range of colors possible with one chip

 

[gtfcniles]

In that situation you will still get color separation though. You will have two point sources with different colors. IMO this negates my reasoning for going with a multichip. You could just not drive the 6500K channel if algae is your concern. Run the higher Kelvin white channel with the blues and lose some intensity. I could choose to vastly underpower the blues and with the 6500K dial in the color I like while also losing intensity.

I'm not going to pay out $120-200 and only use 80% of it. I think the colour spectrums are right in A and so does lasse, so thats good enough for me.

 

[jerpa]

I'm not going to pay out $120-200 and only use 80% of it. I think the colour spectrums are right in A and so does lasse, so thats good enough for me.

This is why I think two emitters is more feasible in order to please all. I understand I am in the minority and as that goes the multichip that is eventually decided on may be unsuitable for my needs. I would be willing, and would actually prefer, to use multiple emitters at lower power for better coverage and to be able to get the color I want. If it doesn't work out I won't lose any sleep over it.