CNCreef Asis Pro 824 LED/T5HO Evil Cluster Hybrid

[jedimasterben]

That's it, I'll be sending you irritating PM's until I can figure out how to construct something along these lines for my new build. Like it.

Buy LEDs. Buy power shit. Solder. Plug in. Need corrective surgery after.

[bpaston]

Jedimasterben I have been following this thread closely and received my 950b's in today. On the boards one has a number 04 and the other a 15, what does this indicate the bin they are from? If so will these have a different look when compared to each other.

[jedimasterben]

blasterman would be a better person to direct that question to, but he is subscribed to this thread and will hopefully be able to answer that. I've seen a lot of people say I'm the resident guru, but I'm pretty sure it's a tie between him and evil.

[dacdaddy]

Four clusters for SPS. You can get away with three for LPS and softies, but the fourth array will be more seamless

Ben led this thread with 1 bridgelux : 3 Luxeon M's. Is this a good ratio or what are your experiences?

[Psychosis]

Buy LEDs. Buy power shit. Solder. Plug in. Need corrective surgery after.

On that note, how many clusters would I need for a 36x30x12? I don't imagine penetration will be an issue, and I intend to have the fixture relatively high off of the water (1.5-2') since I built the stand for easy top down viewing. Would four give adequate dispersal and blending?

 

For that matter, would four clusters built as pendants be an option? I'm still 50/50 on the process.

[jedimasterben]

On that note, how many clusters would I need for a 36x30x12? I don't imagine penetration will be an issue, and I intend to have the fixture relatively high off of the water (1.5-2') since I built the stand for easy top down viewing. Would four give adequate dispersal and blending?

 

For that matter, would four clusters built as pendants be an option? I'm still 50/50 on the process.

blaster was going to post some coverage measurements using the Ledil reflectors in the meantime until I can get mine up and start measuring everything, so at this point, I know how much intensity you'd need approximately, but at what height for coverage is up in the air.

 

Remember that with four clusters, evil's original design, using less efficient and lower-output LEDs hit 1500 PAR at 18".

 

Ben led this thread with 1 bridgelux : 3 Luxeon M's. Is this a good ratio or what are your experiences?

I did a 1:3 at the same power to keep a higher color temperature.

[evilc66]

Considering the tank is fairly shallow, and almost square, 4 clusters in a square(ish) pattern would do the job. You would be amazed at how well reflectors spread light evenly compared to the TIR lenses most of us are used to. I'm trying desperately to use only reflectors from this point on in all my builds, and Ledil is the company to go to. With so many options from them, you can mount low with wide spread, or mount high with tighter spread.

[dacdaddy]

Considering the tank is fairly shallow, and almost square, 4 clusters in a square(ish) pattern would do the job. You would be amazed at how well reflectors spread light evenly compared to the TIR lenses most of us are used to. I'm trying desperately to use only reflectors from this point on in all my builds, and Ledil is the company to go to. With so many options from them, you can mount low with wide spread, or mount high with tighter spread.

I have a 36x18x20, I presume 4 clusters more in a straight line would be better for me?

 

edit: also w the Molex LED holders, with or without the covers? Not sure about this since optics are being put on this. I want to get going on some of these pieces

[evilc66]

Three in a row would be sufficient for that setup.

[blasterman]

I'm just going to shoot some pictures of the light spots against a wall. Let you guys figure out what to use.

[jedimasterben]

I'm just going to shoot some pictures of the light spots against a wall. Let you guys figure out what to use.

sweet.

[dacdaddy]

Question http://shop.stevesleds.com/SPS-Grade-LED-Retrofit-Kit-Package-34-44-tanks-SPS-Retrofit-Kit-34-48.htm

 

Steve's LED's has an SPS kit, that has 21 NW's at 3w apiece, for 63w. Those Bridgelux 950bs are only 10.5w each. Whats your opinion of the 21 smaller chips vs. the 3-4 Bridgelux cannons as far as output, and anything else considered?

[Paleoreef103]

Question http://shop.stevesleds.com/SPS-Grade-LED-Retrofit-Kit-Package-34-44-tanks-SPS-Retrofit-Kit-34-48.htm

 

Steve's LED's has an SPS kit, that has 21 NW's at 3w apiece, for 63w. Those Bridgelux 950bs are only 10.5w each. Whats your opinion of the 21 smaller chips vs. the 3-4 Bridgelux cannons as far as output, and anything else considered?

21 NWs to 21 Royal Blues? Not a great combination there. Honestly, I'd 21 neutral whites on a 36 or a 48" tank is a ton. You'd be much better off taking 7 of those neutrals and swapping them to cool blues on that combo. Compare 14 NWs at 3W for 42Ws for a better comparison between the two. The 14 -3W is more efficient with raw emitters, but with the ledil reflectors on the Bridgelux emitters it should be close to wash (plus 4k 950s are rumored to produce better color than any 3W emitter).

[jedimasterben]

Never pay attention to wattage when comparing two LEDs unless all you want is less power consumption. At 1000ma, the 950B puts out about 2000lm, at the same current it takes a bit over six Rebel ES in Steve's bins to match it. About 22w for the Bridgelux, 18w for the Rebel ES. If you're comparing 'test' drive currents, 500ma for the BL and 700ma for the Rebels, that's 1150lm and 10.5w and 235lm at 2w each, respectively, so five Rebels to match the BL.

[blasterman]

Another issue is that Steve calls 'neutral LEDs' 5000k. If you want the Neutrals, he calls them 'warm whites'.

 

You can use any color temp you want, but lets please compared horses to horses. At 5000k a 1:1 ratio of royals to whites is more common, but with neutrals you need a higher ratio or royals. Like, 1.5 royals to 1 4000k white. So, because steve is referring to much cooler LED's that's why his ratios are weird.

 

That means a single Bridgelux is equivelant to at least four, if not a lot more Rebels. No, this doesn't mean the Rebels are inferior. What it does means is a sh_t load less soldering. The greater number of rebels will allow you spread them out if that's the look you want. If you're going to cluster them the Bridgelux route is the better option.

 

Sorry guys, but it's going to be a day or two before I can shoot those reflectors. Just super busy at work and finishing up a new frag tank.

[jedimasterben]

Another issue is that Steve calls 'neutral LEDs' 5000k. If you want the Neutrals, he calls them 'warm whites'.

 

You can use any color temp you want, but lets please compared horses to horses. At 5000k a 1:1 ratio of royals to whites is more common, but with neutrals you need a higher ratio or royals. Like, 1.5 royals to 1 4000k white. So, because steve is referring to much cooler LED's that's why his ratios are weird.

In reality, they don't look any cooler. Just with my 12x 5K whites and 32x royals, all at 1000ma, the color temp is still somewhere around 12K. At first I thought something was wrong and that maybe a driver wasn't ramping up, but verified it all with my multimeter.

 

Sorry guys, but it's going to be a day or two before I can shoot those reflectors. Just super busy at work and finishing up a new frag tank.

No worries, just post em when you get em.

[dacdaddy]

Never pay attention to wattage when comparing two LEDs unless all you want is less power consumption. At 1000ma, the 950B puts out about 2000lm, at the same current it takes a bit over six Rebel ES in Steve's bins to match it. About 22w for the Bridgelux, 18w for the Rebel ES. If you're comparing 'test' drive currents, 500ma for the BL and 700ma for the Rebels, that's 1150lm and 10.5w and 235lm at 2w each, respectively, so five Rebels to match the BL.

Thanks for the info. Another noob question, I notice you're running the current higher through them, and what I had been talking about was the test current. I suppose this is what the driver is for, and the dimming capability is essentially adjusting the current flow?

 

I think i remember you mentioning some sort of sheet with build info. I am only familiar with the basic circuit formulas ie P=VI. I am not sure with the drivers what's toggling what, and its effect on the lumen ouput. Is that sheet around or still in the works? Thanks this is a great thread, I'm learning a lot!

[dacdaddy]

Closer and closer. The LDD got pushed back again though.

 

You think only 2 Cool Blue's is enough against those big stack RB's? Again, a steve's LED reference, but he mentioned something like a 5:2 RB:CB ratio on the normal 3w stars. I know this is a lot of test phase here but what do you think?

 

btw, this looks bada$$

[jedimasterben]

Thanks for the info. Another noob question, I notice you're running the current higher through them, and what I had been talking about was the test current. I suppose this is what the driver is for, and the dimming capability is essentially adjusting the current flow?

 

I think i remember you mentioning some sort of sheet with build info. I am only familiar with the basic circuit formulas ie P=VI. I am not sure with the drivers what's toggling what, and its effect on the lumen ouput. Is that sheet around or still in the works? Thanks this is a great thread, I'm learning a lot!

Yes. The LEDs can handle a maximum amount of current (that they're rated, anyway, they can usually be pushed beyond their maximum without damage as long as they're cooled properly), and the driver is what sets that current. When the driver is dimmed, it lowers the current.

 

And that's assuming you use a 'normal' constant current driver. You can use a normal power supply that delivers constant voltage, but small changes in voltage can equal massive amounts of current, and I don't even fully understand that, so I always just say use a CC driver.

 

You think only 2 Cool Blue's is enough against those big stack RB's? Again, a steve's LED reference, but he mentioned something like a 5:2 RB:CB ratio on the normal 3w stars. I know this is a lot of test phase here but what do you think?

 

btw, this looks bada$$

I normally do a 2:4:1 ratio of NW to RB to CB. I don't want too much on my tank, but we'll see if I want to add more once it is all together.

[Paleoreef103]

I normally do a 2:4:1 ratio of NW to RB to CB. I don't want too much on my tank, but we'll see if I want to add more once it is all together.

I have to imagine you'll need less cool blue with those Ms because they peak at 450 instead of the 440 of Rebel ES. Not a huge difference, but I bet that just NW and RBs are less purple than normal. Anyway, looking forward to it actually firing up!

[blasterman]

Rebel ES don't peak at 440. Dominant wavelength is about 447-448 with royal rebels, unless you have data sheets to the contrary.

 

I have some 10watt 440's...totally different than rebels.

[jedimasterben]

Steve said his measured a bit over 445nm peak (the Rebel ES, not the M), if I remember right.

[MeepNand]

Steve said his measured a bit over 445nm peak (the Rebel ES, not the M), if I remember right.

Did he mention something in the area of 447-448nm?

[blasterman]

In reality, they don't look any cooler

 

I'm a bit concerned about this because 4000k and 5000k are pretty dramatic differences over a reef tank along with visually. So, I'm wondering if Steve is not getting the bins right before he reflows. What do you think? CCT is otherwise tough to measure, even with a spectrometer unless you have additional software that plugs into your calibration tools. I have some Fortimo arrays that were calibrated at nearly a spot on 5000k, and I use them as a subjective reference for quickly judging white LEDs. Neutral Bridgelux are significantly different than the Fortimos. Retail bins of cool-white XPGs and XMLs stand out terribly next to the Fortimo, even if CCT is similar.

 

Peak and dominant are often confused, and because it's easier to read a peak wavelength it's usually marketed that way. If it's a metal halide application, peak is usually more important due to the strong spikes. For LEDs dominant is more important. I should have some royal M's arriving today or tomorrow, and if I can get my spectrometer to work with Win8 on my laptop I'll post the graphs.

[jedimasterben]

I'm a bit concerned about this because 4000k and 5000k are pretty dramatic differences over a reef tank along with visually. So, I'm wondering if Steve is not getting the bins right before he reflows. What do you think? CCT is otherwise tough to measure, even with a spectrometer unless you have additional software that plugs into your calibration tools. I have some Fortimo arrays that were calibrated at nearly a spot on 5000k, and I use them as a subjective reference for quickly judging white LEDs. Neutral Bridgelux are significantly different than the Fortimos. Retail bins of cool-white XPGs and XMLs stand out terribly next to the Fortimo, even if CCT is similar.

 

Peak and dominant are often confused, and because it's easier to read a peak wavelength it's usually marketed that way. If it's a metal halide application, peak is usually more important due to the strong spikes. For LEDs dominant is more important. I should have some royal M's arriving today or tomorrow, and if I can get my spectrometer to work with Win8 on my laptop I'll post the graphs.

I think the same thing, but I think now they're going through more LEDs and may not be keeping up 100% with the bins. I like how the 5k look, and actually they're very similar in look to the NW XM-L I've got (they're binned at 4500k, a couple look slightly cooler than the rest))

 

 

That would be awesome to get the spectral readouts! If you want I can send you some of the 5K Rebels I've got so you can take a look at them, I should have some on single stars in my spare parts box. Steve also sent me a 5K M with my shipment of RB M.