Cheap and Cheerful DIY using Citizen cobs

[Fastslappy]

I use themostat wire from homedepot (tho most of my builds use 2 conductor as im wiring in parallel)

it is CMR/CL2R rated and printed right on the jacket. if you look up that wire it is 300V rated

it's that garden green shit with no labeling at all & it is in the hardware section as doorbell wire with tan &/or white covering
that's the solid wire i'm talking about , i've seen pix of guys using it on builds

 

[CobKits]

how long have u been growing w/ the citizen's? are they 4000k ?

Jorge hasnt been around for a little while ill attempt to answer your questions tho

whats the difference other than the kelvin values of the cree cxb 3070,cxb3590, and the citizen chips?

different mfrs, different models, same concept/technology

cxb3070 and 3590 are both the latest generation from cree and are almost 2 years old. they are two different sized chips, so the same thing, but are designed for slightly different size applications.

citizen has a bunch of diferent sized chips. their latest generation (gen 5) is about 10 mos old and are among the best chips out there in terms of efficiency and cost

vero has a new gen that jsut came out (gen 7), which is on par with the citizens for efficiency and cost effectiveness

ive been told u get what u pay for so the higher the price the better the led??

not necessarily. good stuff isnt cheap but just because something is expensive doesnt make it better. a $25 citizen or vero will perform as well as a $35-$40 cxb3590.

imo the led that runs colored isnt as good as the white light produced by the latest led's like the 3070,3590, and citizen

both grow well if you have good stuff. there are good purples and shitty purples, and same for white light.

not to mention its easier to see in the grow room,,not orange or red tinted light???is this correcvt?

a lot of people like cobs for this reason. is a lot easier to see deficiencies

 

[wietefras]

Cree doesn't just have two sizes. They have a whole array of them:
http://www.cree.com/LED-Components-and-Modules/Landing-pages/CXA2

It's not 2 year old tech either. Over that time their COBs have moved up a bin and perhaps more. We won't know exactly how much unless we would keep testing new COBs.

Kingbrite has also started selling Citizen COBs which gives us a good comparison. The price difference between the CXB3590 (or CXB3070) and the Citizen 1825 and Citizen 1212 seems to be about 30% cheaper in PAR/$ in favor of the Citizen (when compared to Cree at the same efficiency). So you'd get the same amount of light at the same efficiency for 30% less in cost. For the 1812 and 1818 the price difference seems a lot less (around 15%).

 

[CobKits]

Cree doesn't just have two sizes. They have a whole array of them:
http://www.cree.com/LED-Components-and-Modules/Landing-pages/CXA2

yes, he specifically asked about two of the models tho.

It's not 2 year old tech either. Over that time their COBs have moved up a bin and perhaps more. We won't know exactly how much unless we would keep testing new COBs.

pretty sure 3500k has been in CD bin since late spring/early summer 2015. not quite 2 years but getting there

Kingbrite has also started selling Citizen COBs which gives us a good comparison. The price difference between the CXB3590 (or CXB3070) and the Citizen 1825 and Citizen 1212 seems to be about 30% cheaper in PAR/$ in favor of the Citizen (when compared to Cree at the same efficiency). So you'd get the same amount of light at the same efficiency for 30% less in cost. For the 1812 and 1818 the price difference seems a lot less (around 15%).

datasheets are data sheets but ive never seen an 1818 that doesnt run close to a cxb3590 in the real world, and they are about 33% cheaper

its difficult to compare the 300+W 1825 to the CXB, its an entirely different class of cob size-wise, but yes i think the 30% you quote is close, maybe a little less from my experience

 

[wietefras]

If you compare them at the same efficiency it should be quite easy. For instance the CXB3590 @1400mA vs the CLU058-1825 @ 1400mA. That's both around 2.45PAR/W. In that case you'd need 30% less of the Citizen COBs to produce a similar amount of light using a similar number of watts. Price is about the same for the COBs.

At higher efficiencies the advantage for the Citizen's drops off and at lower efficiencies it goes up. For instance the CXB3590 @1750mA and 1825 @1900mA should both produce around 2.33PAR/W. Then you need 37% less Citizen COBS for the same output.

Alternatively when you pick 2.6PAR/W the CXB3590 @1050mA and 1825 @700mA should be around that efficiency, but then you'd actually need 3% more Citizen COBs than Cree's.

But indeed that's all based on datasheet figures.

 

[CobKits]

But indeed that's all based on datasheet figures.

they really all converge around 30-50W and thats where the smaller clu048s really shine as an incredible value

 

[wietefras]

I wouldn't compare them at wattage levels though. The 1212 does 2.1PAR/W at 50W, but then the 1825 does 2.5PAR/W. Completey incomparable and this is what leads to the nonsensical notion of "this COB is more efficient than the other".

Efficiency is simply a parameter which detemines the overall cost of a fixture. Wanting a higher efficiency/efficacy means you need more COBs. In the end really the only relevant metric is price. You can run all these COBs at 2.3 PAR/W efficacy up to say 2.6PAR/W, but the question is which gives you the cheapest fixture at your preferred efficacy. Or perhaps which gives you the nicest uniformity if the price difference is small.

Trouble with the smaller ones is the cost of cooling them though. Indeed the CLU048-1212 is nice and cheap at less than a third of the price of a CLU058-1825. They are also a third of the "size" of a 1825 so you need three times as much of them a lot of them to fill a room too. If you put each COB on it's own CPU cooler you'd pay about as much for the coolers as for the COBs. Which would completely negate any price advantage on the COBs.

On the other hand, I kinda like the 1825 being able to fill a square meter with only 4 COBs (90W each) while still maintaining good efficacy at around 2.36PAR/W. At $35 per COB that's 6 PAR/$. Which should make it possible to build fixtures costing around a dollar or euro per Watt.

 

[TheChemist77]

I wouldn't compare them at wattage levels though. The 1212 does 2.1PAR/W at 50W, but then the 1825 does 2.5PAR/W. Completey incomparable and this is what leads to the nonsensical notion of "this COB is more efficient than the other".

Efficiency is simply a parameter which detemines the overall cost of a fixture. Wanting a higher efficiency/efficacy means you need more COBs. In the end really the only relevant metric is price. You can run all these COBs at 2.3 PAR/W efficacy up to say 2.6PAR/W, but the question is which gives you the cheapest fixture at your preferred efficacy. Or perhaps which gives you the nicest uniformity if the price difference is small.

Trouble with the smaller ones is the cost of cooling them though. Indeed the CLU048-1212 is nice and cheap at less than a third of the price of a CLU058-1825. They are also a third of the "size" of a 1825 so you need three times as much of them a lot of them to fill a room too. If you put each COB on it's own CPU cooler you'd pay about as much for the coolers as for the COBs. Which would completely negate any price advantage on the COBs.

On the other hand, I kinda like the 1825 being able to fill a square meter with only 4 COBs (90W each) while still maintaining good efficacy at around 2.36PAR/W. At $35 per COB that's 6 PAR/$. Which should make it possible to build fixtures costing around a dollar or euro per Watt.

so ur saying a larger cob is better than running multiple smaller ones?? i thought just as with hps, running 3 400 watt units able to get closer to canopy and light spread is better than running 2 600 watt hoods.. in my expierience running more smaller is better than 1 larger,,i got better gpw running 2 400 watt hps so 800 watt total than running a 1k hps 1,000watts over a 4ftx6ft table.. then 2 400 watt hps pluss 1 400 watt mh got better gpw than 2 600 watt hps...
so wouldnt running multiple 100-200 watt cobs produce a more even distribution of light and better gram per watt average than running a single 400 watt cob? thanks for the information guys!!

 

[wietefras]

Not so much that it's better, but that it could be cheaper to run bigger ones. Or you need to be more clever with the cooling.

Indeed the spread of the light is better when you use more points of light. So in that regard running 8 48W COBs is better than 4 90W COBs. Although I doubt it matters that much. It's a less work building a frame for 4 COBs instead of for 8.

It's always a compromise though. You want to run the COBs as soft as possible for maximum efficiency, but then the costs go up. You can calculate which is the most economical efficacy though. For me (23 cents per kWh) it's around the 2.3PAR/W mark. So I go with that and then the other parameters follow.

 

[CobKits]

discrete heatsinks will almost always favor larger cobs

bar heatsinks will almost always favor smaller cobs

 

[sleepless_canuck]

I run 4 x the clu 1825 3500k and love them in my 3x3 on 16" centers.

I can push each up to about 113 watts are 80%.

I keep them around 90 watts and everything is happy.

I went this route as it saved on heat sinks and cranked up I can blow the socks off my 1m square.

 

[phpatrick]

I run 4 x the clu 1825 3500k and love them in my 3x3 on 16" centers.

I can push each up to about 113 watts are 80%.

I keep them around 90 watts and everything is happy.

I went this route as it saved on heat sinks and cranked up I can blow the socks off my 1m square.

If ya don't mind me asking sleepless, what drivers are you using on those 4 clu 1825?

Thanks
phpatrick

 

[sleepless_canuck]

If ya don't mind me asking sleepless, what drivers are you using on those 4 clu 1825?

Thanks
phpatrick

2 x HLG 240h 2100 b

 

[pop22]

Bet you missed the specs:

Thermostat Wire, 60°C, 150 Volt, UL 13, Type CL2

It says not to exceed 30 volts

 

[Pulses]

where the hell is whore hey

I run 4 x the clu 1825 3500k and love them in my 3x3 on 16" centers.

I can push each up to about 113 watts are 80%.

I keep them around 90 watts and everything is happy.

I went this route as it saved on heat sinks and cranked up I can blow the socks off my 1m square.

what heatsinks are you using. I have same setup with passive sst-x sinks and was hoping would cool enough turned up

 

[sleepless_canuck]

I am using the sstx

 

[Tomula]

Hello there, I'm thinking about building a 450W flowering light with 6x CLU048 1825 80CRI- 3000K or 3500k?. I have my eyes on MW HLG-150H-54 A as a power sources (3 of them). Some old Pentium 4 coolers (tower), put them all in ducting, get 3x Arctic 12 continuous fans (great fans with high CFM). Add a 20W UVA tube. What do you think, I should work just fine right? Thank you and have a nice day.

 

[Bustos]

I am real glad to see the info of enlightenment shining upon us...

a few weeks ago i placed an order with Mr. Cobkits for 6 each of the 1825 cob engines at 3500K, and drivers for the lights and fans... figuring on running them in pairs of 2, I cannot wait till they arrive and get those babies set up
living in alaska in the great white North, i am curious also aside from the benefits of growing, about the possible benefits of spectrum light that will be beneficial to those affected by long hours of darkness.... I know, wrong forum, but ... Just saying.
It is comforting to see though that my choice of the 1825 is turning up to look like a good choice...

soo thanks all of you for contributing to the base of Knowledge here and cant wait to see the effect of these awesome lights on my vegetative matter, for sure.

best regards

 

[CobKits]

just playing around with the numbers for the new cobs, looks like the gen 6 citis are just a hair under 3% more effcient at nominal currents. part of this is due to better thermal resistance, so they may do better at higher currents, like maybe 4%+ efficacy bump at high currents. when i apply the 2.8% average nominal efficacy (37W) across the entire curve, the new 1212 gets dangerously close to a cxb, esp if as i suspect it would do better at higher currents, it would look better than this data here at 50-70W



the ~$24 1818s should be a few percent up on the vero D's as well
we'll see how the 1825s do, i expect the same. should be a real winner in the 80-150W zone

per datasheet/citi calc the gen 6 1825 is 152 lm/W at 115W, the gen 5 3618 is 149 lm/W at same wattage....

 

[Tomula]

just playing around with the numbers for the new cobs, looks like the gen 6 citis are just a hair under 3% more effcient at nominal currents. part of this is due to better thermal resistance, so they may do better at higher currents, like maybe 4%+ efficacy bump at high currents. when i apply the 2.8% average nominal efficacy (37W) across the entire curve, the new 1212 gets dangerously close to a cxb, esp if as i suspect it would do better at higher currents, it would look better than this data here at 50-70W

the ~$24 1818s should be a few percent up on the vero D's as well
we'll see how the 1825s do, i expect the same. should be a real winner in the 80-150W zone

per datasheet/citi calc the gen 6 1825 is 152 lm/W at 115W, the gen 5 3618 is 149 lm/W at same wattage....

There is a new gen. of citis? That's great. Thank you for the info man, appreciated!