Top bin COB comparison 2

Freezy_

Active Member
Whats up guys, Just joined RIU.

Forgive me for being a complet noob when it comes to leds. I was all set to pull the trigger on building out a light for my 4 x 6 canopy and was going to use 24 3500k cxb 3590's with individual heat sinks. If I use a long shared heat sink would I be better off cost wise with a bunch of 1212's? like 2 or even 3 times the amount of led's? If someone was to recommenced me the most efficient way to knock it out of the park for a 4 x 6 area what would you recommend? From what I understand here the 3590's are great chips they are just over priced?
 

churchhaze

Well-Known Member
Whats up guys, Just joined RIU.

Forgive me for being a complet noob when it comes to leds. I was all set to pull the trigger on building out a light for my 4 x 6 canopy and was going to use 24 3500k cxb 3590's with individual heat sinks. If I use a long shared heat sink would I be better off cost wise with a bunch of 1212's? like 2 or even 3 times the amount of led's? If someone was to recommenced me the most efficient way to knock it out of the park for a 4 x 6 area what would you recommend? From what I understand here the 3590's are great chips they are just over priced?
I think that's somewhat of a matter of preference. Obviously the long sinks have a structural element to them while a fixture with individual sinks requires a more complicated structure.

For active cooling, the long sinks allow you to have 1 fan in the middle to cool multiple cobs.

This is where OCD can really start to hold you back! Which is the best when I want both?
 

BOBBY_G

Well-Known Member
amps = the current fed to a given chip which can be any number from zero up to maximum amperage of chip (typically between 2000 and 5000 mA). brightness is a function of current. higher currents push the chips harder and are less efficient. so if you double the current you only get like 190% more light

voltage = the design voltage of a chip. can be 36 or 72V like a cree, or 36/50/100 like a citizen or vero

driver cost is based on wattage. regardless of the voltage chip you select and what current you push your chips at, drivers are almost always 0.20-0.40/watt
 

optzulu

Well-Known Member
We need two values not just amps. It's amps * volts... For measuring device output we always use watts even with little dc electric motors we measure output in watts.
P(W) = I(A) × V(V) (WATTS EQUALS, AMPS TIMES VOLTS)
I(A) = P(W) / V(V) (AMPS EQUALS, WATTS DIVIDED BY VOLTS)

Just because we are using DC current exclusively now it doesn't change this.
Yes I thought that you and jorgegonzales and bobby_g are smart ppl an you can figure it out but NO so lets explain it one more time.

So this thing started when I said :
''The 1212 will win from the 1818 from 1W to 36W then the 1818 will get ahead, but the 1212 costs also 50-60% less then a 1818 and only have 1/3 less power.''

So I wanted to compare the 2 chips and because they run both on a different Voltage 1212= 34V and 1818=50V Its going to be hard compare them head to head and besides of that in real world we using this chips to cover a space right ? in this case I said 1m2 I think thats a bit smaller then what you guys call 3x3 square feet and to cover this area for blooming cannabis you need around 400W of power ( heyyy there is your first value !!!)

Good because now we can compare and see what we need to get this power. The first and most important thing to do Is to chose your Amperes !! Because they will determine your heat,efficiency,and power ! and what to buy also very important !

In the next post I showed them how much cobs you need to cover this area ore get this power It Is what you prefer
'you need either sixteen 1212 cobs for 170 euros ( 700ma x 34.6v= 24w x 16 = 384W )
Ore eleven 1818 chips for 240 euro. ( 700ma x 50v= 35Ww x 11= 385W)

And I think bobby_g charts confirmed what I said about the 1212 killing the 1818 at lower currents
 

BOBBY_G

Well-Known Member
Yes I thought that you and jorgegonzales and bobby_g are smart ppl an you can figure it out but NO so lets explain it one more time.
:: pops popcorn::

So this thing started when I said :
''The 1212 will win from the 1818 from 1W to 36W
maybe we have different definitions of "win". while the 1212 did better than expected it certainly wasn't brighter than the 1818 below 25W and 1818 pulled away significantly above 25W. in any case, data below 25W is highly suspect if you look at the graphs and i wouldnt be drawing any definitive conclusions


So I wanted to compare the 2 chips and because they run both on a different Voltage 1212= 34V and 1818=50V Its going to be hard compare them head to head
not at all. a Watt is a Watt. voltage of chip is 100% irrelevant to the discussion. which is why 36 and 72V cxbs act identically at a given wattage


and besides of that in real world we using this chips to cover a space right ? in this case I said 1m2 I think thats a bit smaller then what you guys call 3x3 square feet and to cover this area for blooming cannabis you need around 400W of power ( heyyy there is your first value !!!)
power required is dependent on light efficiency. 250W of cobs will give the same light as 400W of HPS, 600W of T5, or 1000W of incandescents. i guess i dont get where youre going with this point


Good because now we can compare and see what we need to get this power. The first and most important thing to do Is to chose your Amperes !! Because they will determine your heat,efficiency,and power ! and what to buy also very important !
i am sorry im trying to be polite but you are making zero sense

In the next post I showed them how much cobs you need to cover this area ore get this power It Is what you prefer
'you need either sixteen 1212 cobs for 170 euros ( 700ma x 34.6v= 24w x 16 = 384W )
Ore eleven 1818 chips for 240 euro. ( 700ma x 50v= 35Ww x 11= 385W)
thats fine. as long as you dont need holders, or lenses, or reflectors, and are married to large extruded heatsinks as opposed to discreet pin fin/cpu cooler/etc type designs.

And I think bobby_g charts confirmed what I said about the 1212 killing the 1818 at lower currents
no, it did not. in a very fuzzy area of the data it appeared to match it in a limited area. i would suggest further testing before drawing any conclusions
 

JorgeGonzales

Well-Known Member
Patience of a saint.

I think I mentioned this in PM, but it would be interesting to see a couple samples of 1212s or 1818s run against each other, to see some real world min/max/typ spreads. Might explain some supposed anomalies.
 

SaltyNuts

Well-Known Member
Man, this all jibes pretty good with my head, (I'm using stoner-speak, lol).... the efficiency graphs (thanks Bobby-G!) look like the 1212 is gonna be roughly 10% less efficient than a 3590 at about 35W... all very rough but that is close enough for me! $12.50? I'll take sixteen! Those $10 Mechatronic heatsinks too, yo, "cheap and cheerful" as the man JG says. I'm thinking that sixteen 1212's at 1050mA would be pretty effin' bright. 36W per foot! Looks like two HLG-320H-C1050A would do the trick to power all sixteen cobs? I'd mount the drivers remote and use the internal pots for the seasonal dimming if need be, should drop to 25W/per cob at that spread. Now I just need an extra $800 for the build... shit my life sucks... who else here has high monthly expenses and lives on beans and rice...
 

optzulu

Well-Known Member
power required is dependent on light efficiency. 250W of cobs will give the same light as 400W of HPS, 600W of T5, or 1000W of incandescents. i guess i dont get where youre going with this point
Hmm let me see we are on a sub forum specially for LED and we are talking and comparing 2 Citizen COB leds of the same generation and family so 400W of what do you think that I am talking about ?
I am out because you don't want to understand it and we are on a different page.

Ooh and btw the charts you make where CREE is getting slapped make zero sense since we are after a high g/w ratio put a gavita de in your charts since the are all about $/wat ratio maybe that fair ?
And since the CXB3590 CD price dropped the hype is over and you guys can move on.
 

PurpleBuz

Well-Known Member
any chance we can see a citizen 1212 against a cxb3070 or some other cobs with about the same LES size ?

don't get me wrong the diffs between the different citizens and a cxb3590 are interesting. But there is a definite correlation between LES size and efficiency. LES size seems to make a big difference in performance. I presume because of heat management.

But I deploy base on the size that fits my application the best first. For example how does citizen compare to the nichia for the 14.6mm size ? gen7 vero15s, cxb1830's ?
 

BOBBY_G

Well-Known Member
any chance we can see a citizen 1212 against a cxb3070 or some other cobs with about the same LES size ?
i dont have any 1212s in 80 cri 3500k to compare with the 3500k 80 cri 3070s i have, sorry. i can do 1818 vs 3070 but that wont be a fair fight
 

PurpleBuz

Well-Known Member
i dont have any 1212s in 80 cri 3500k to compare with the 3500k 80 cri 3070s i have, sorry. i can do 1818 vs 3070 but that wont be a fair fight
oh well sigh ... do you have anything to compare to a 14.6mm nichia ? I can send you one of those.
 

JorgeGonzales

Well-Known Member
i dont have any 1212s in 80 cri 3500k to compare with the 3500k 80 cri 3070s i have, sorry. i can do 1818 vs 3070 but that wont be a fair fight
Isn't that a perfectly even fight, dollar for dollar? I think cutter sells 3070s for $28, and that's a great price.

I'd love to see that shootout.

Editing to add...I still wonder if the 3070 would have been a better choice for me at 35W.

But I deploy base on the size that fits my application the best first. For example how does citizen compare to the nichia for the 14.6mm size ? gen7 vero15s, cxb1830's ?
Same, what about price? Those Nichias are dark horses in this race. 108B vs 1212, look out.
 
Last edited:
Top