DIY CXB2530 3000K 80cri U2 bin (Vero Killers!)

Great build... I just ordered some CXB3590's today, I'll definitely be going with solid core at the ideal connectors, I had the same issue with the BJB connectors on my reef LED setup, they needed to either be tinned or solid core wire to make a really good connection.

As for hanging these hunks of aluminum, I'm using a hanging kit from the company that I bought my DIY LED stuff for on my aquarium. A little expensive, but really worth it IMO. I can easily raise the lights up or down, and then if they're not sitting flat I can move them front to back as well and lock it in at that setting.

https://reefledlights.com/shop/hanging-kit/

Here's how they look on my heatsink...




Used the same hangers over my old reef tank with a ridiculously heavy T5 fixture...

FESTER665 said:

Great build... I just ordered some CXB3590's today, I'll definitely be going with solid core at the ideal connectors, I had the same issue with the BJB connectors on my reef LED setup, they needed to either be tinned or solid core wire to make a really good connection.

As for hanging these hunks of aluminum, I'm using a hanging kit from the company that I bought my DIY LED stuff for on my aquarium. A little expensive, but really worth it IMO. I can easily raise the lights up or down, and then if they're not sitting flat I can move them front to back as well and lock it in at that setting.

https://reefledlights.com/shop/hanging-kit/



Used the same hangers over my old reef tank with a ridiculously heavy T5 fixture...

Click to expand...

That's a really nice aquarium. It's great seeing reefers just as interested in cxb as weed growers. The hangers I'm making are a lot cheaper to make, but take about 10-15 minutes to make each using a swagging tool. I cut the 1/16" stainless cable 14 inches long and the loops take up 2 inches of cable each, leaving 4 inch cable segments. I've also been using ratcheting pulleys to make raising lights a lot easier.

Show us what you setup when you get the cxbs! I'm interested in seeing reef DIY builds as well!

It's funny because I took the same progression I think many on here took... The old tank had huge MH ballasts and reflectors:


But you can only take so much heat issues and ridiculous electric bills... So I switched to T5's with crazy ballasts and reflectors, and then switched to LED's...

FESTER665 said:

It's funny because I took the same progression I think many on here took... The old tank had huge MH ballasts and reflectors:

But you can only take so much heat issues and ridiculous electric bills... So I switched to T5's with crazy ballasts and reflectors, and then switched to LED's...

Click to expand...

How much power are you saving, do you think?

ttystikk said:

How much power are you saving, do you think?

Click to expand...

Hard to tell because I downsized everything at the same time.... Just got crazy when I live alone in a small 915 sq./ft house and was running up $250 a month electric bills... LOL

My average bill lately is less than $100, so we'll see how much it changes with the new LED setup I'm building...

Sorry to hijack your thread church! I'll be keeping and eye on the thread though!

FESTER665 said:

Hard to tell because I downsized everything at the same time.... Just got crazy when I live alone in a small 915 sq./ft house and was running up $250 a month electric bills... LOL

My average bill lately is less than $100, so we'll see how much it changes with the new LED setup I'm building...

Sorry to hijack your thread church! I'll be keeping and eye on the thread though!

Click to expand...

LOL, my bills are never UNDER $1000 a month...

tty, how much /kWh there? using hps with AC?

we pay 12-15 cents/kWh for supply with delivery. $150/mnth for elec, ~$200 when there is humidity. I guess that would be a lot lower if we had gas dryer and water heater, hard to compare with those kinds of variables.

SupraSPL said:

tty, how much /kWh there? using hps with AC?

we pay 12-15 cents/kWh for supply with delivery. $150/mnth for elec, ~$200 when there is humidity. I guess that would be a lot lower if we had gas dryer and water heater, hard to compare with those kinds of variables.

Click to expand...

I have a gas dryer, a gas range (electric oven tho, dammit!), and gas hot water heat.

I use water chilling for environmental control and a compressorless chiller when needed. In the winter, my chiller sits inside where the rejected heat warms my home all winter- in Colorado. I haven't run my gas forced air furnace in going on four winters now. My gas bill never exceeds $31 a month.

My system is highly productive and efficient. I do use HPS thouies in bloom. I'd like to run a custom fixture made from this latest generation of COBs head to head at the same wattage in the same setup to see what productivity differences there might be.

10¢/kWh, 13.3 for the summer peak AC season.

It might help to read my thread, link in my signature line.

ttystikk said:

I have a gas dryer, a gas range (electric oven tho, dammit!), and gas hot water heat.

I use water chilling for environmental control and a compressorless chiller when needed. In the winter, my chiller sits inside where the rejected heat warms my home all winter- in Colorado. I haven't run my gas forced air furnace in going on four winters now. My gas bill never exceeds $31 a month.

My system is highly productive and efficient. I do use HPS thouies in bloom. I'd like to run a custom fixture made from this latest generation of COBs head to head at the same wattage in the same setup to see what productivity differences there might be.

10¢/kWh, 13.3 for the summer peak AC season.

It might help to read my thread, link in my signature line.

Click to expand...

Jealous. Just gas delivery is $36/month in Ohio. Having a gas water heater, I pay like $40, the 36 plus 4 for actual usage. Electricity is about $.12/kWh after delivery.

So your telling me the same amount t of cxa3070s driven at 1.4 amps this light you built beats it out in lumens?

95'ZR1 said:

So your telling me the same amount t of cxa3070s driven at 1.4 amps this light you built beats it out in lumens?

Click to expand...

Yes. about 155-160lm/W from what I understand if Tc is 35C and current is 500mA.

Going by CXA 3000k 80cri phosphor's official LER of 325W/lm, that would make efficiency somewhere between 47.7% (155lm/W) and 49.2% (160lm/W).

churchhaze said:

Not only will this kill martians and bad dogs, it should also be more than capable of killing veros and even cxa3070 3000k AB!

- 11x Cree CXB2530 3000k 80cri U2 bin
- 11x IDEAL 50-2102CR chip-lok cob holders
- 2x 10" profile cut to 15"
- 1x 10" profile cut to 5"
- M3 0.5 pitch 6mm length screws (preferably socket head, with flat bottom like panhead).
- A 94% efficient constant current driver that will go unnamed for safety reasons. (if you can't figure it out on your own, you aren't allowed to know which it is)

So far I have one of the panels done.

View attachment 3361612

View attachment 3361613

Almost ready to kill vero 18s. This one will orbit earth.

Click to expand...

This sir, is seriously badass!

Cant do this with a spreadsheet!

I got 8 Cree 3070 6 running at 1.5amps 2 running at .9 amps in a 2x3 box and thought I was insane. It growing some amazing bud. Can't wait I'll I clean out my basement more and get either a 4x4 or 6x4 or even better 8x8.

What about a HIGH POWER array, using as much as 4000W? This would be fed by a 30A@240V circuit, just like current HID systems.

I'm envisioning scaling this technology up a bit... something I've seen absolutely no information on, anywhere.

ttystikk said:

What about a HIGH POWER array, using as much as 4000W? This would be fed by a 30A@240V circuit, just like current HID systems.

I'm envisioning scaling this technology up a bit... something I've seen absolutely no information on, anywhere.

Click to expand...

I've been scaling up little by little in fear of significant advances making my investments obsolete and the lost opportunity cost.

It's possible cree will release a cxb2530 3000k 80cri U4 bin in a few months, blowing the U2 bin out of the water. It's also likely that in a few weeks, the price of the U2 bin will significantly drop in price as other major electronics retailers get them (like mouser, newark, jameco, arrow, etc).

churchhaze said:

I've been scaling up little by little in fear of significant advances making my investments obsolete and the lost opportunity cost.

It's possible cree will release a cxb2530 3000k 80cri U4 bin in a few months, blowing the U2 bin out of the water. It's also likely that in a few weeks, the price of the U2 bin will significantly drop in price as other major electronics retailers get them (like mouser, newark, jameco, arrow, etc).

Click to expand...

LOL at fear of obsolescence- we are ALL on our way to the dustbin of history, so don't wait!

There are also so many ways you could setup the heat sinks, whether active or passive.. how long of fin channels for active setups, etc. Fin/channel lengths should be shorter for passive setups, but still can't be too long for active setups. I'm not sure what the optimal fin/channel length is for a standard 120mm coolermaster DC fan at 5V and 12V. Supra has done a lot of testing on different heat sinks with fans at different speeds on the heat sinks thread.

If you're in any way OCD, you will probably have a new heat sink design every few days.. and thus have constant dreams to make new panels...

visual considerations for passive designs:

churchhaze said:

There are also so many ways you could setup the heat sinks, whether active or passive.. how long of fin channels for active setups, etc. Fin/channel lengths should be shorter for passive setups, but still can't be too long for active setups. I'm not sure what the optimal fin/channel length is for a standard 120mm coolermaster DC fan at 5V and 12V. Supra has done a lot of testing on different heat sinks with fans at different speeds on the heat sinks thread.

If you're in any way OCD, you will probably have a new heat sink design every few days.. and thus have constant dreams to make new panels...

visual considerations for passive designs:

View attachment 3365922
View attachment 3365923

Click to expand...

I thought after i had Tesla's panels built i would stop dreaming about amps and volts and drivers and heatsinks...maybe cuz they are still sitting here or maybe cuz i am now addicted to DIY LED. I love it though...each build brings a sense of pride and accomplishment.

ttystikk said:

What about a HIGH POWER array, using as much as 4000W? This would be fed by a 30A@240V circuit, just like current HID systems.

I'm envisioning scaling this technology up a bit... something I've seen absolutely no information on, anywhere.

Click to expand...

Sure you can scale it up as much as you like. A low cost example, if you run Vero 29s at 2.45A they dissipate about 95W ea at 35% efficiency, $0.85/PAR W. You can put each COB on a CPU cooler with a 50mm lens and 6 of those would match a 1000W HPS. Scaling up a 49% efficient build tho, would cost a pretty penny and parts availability could become an issue in the short term.

As far as being slow to pull the trigger, there may be some merit to that. Just a few years ago the best DIY LEDs were 26% efficient, labor intensive to build and cost $7+/PAR W just for the LEDs alone. So when I needed to scale up, I bought 600W HPS setups instead. Now the best flowering COBs are pushing 50% and cost $2.50/PAR W. The best vegging LEDs can be up 65% efficient and cost $6/PAR W. As you can probably guess, we cannot exceed the theoretical 100% efficiency and in practice we may not exceed 80% at consumer level (my conjecture). Once we are pushing those levels we can only try to continue decreasing the cost. We still a lot of room for gains, but you can see we will be approaching the physical limits at some point.

The best Cree test in the lab so far for a white LED was 303lm/W (My conjecture, these are about 85% efficient assuming a greenish tint with LER of 350 and the blue LED it was made from may have been 90%+ efficient).

From March 2014:
"Cree reports that the LED efficacy was measured at 303 lumens per watt, at a correlated color temperature of 5150 K and 350 mA. Standard room temperature was used to achieve the results."

@SupraSPL Rather than attempt to interpolate what the equivalent lumen output might be relative to HPS, my plan to to pull an actual thousand watts... or two or three.

The differences in output really, at the end of the day, don't mean shit- after all, it's about plant active response.

If I make something spectacularly bright/hot, I can always alter my setup.

I just don't want to be in the boat of wishing I had more output...