Introducing CobKits.com - specializing in DIY and Citizen COBs

stupidcracker00

Well-Known Member
just getting in, saw your email

if you noticed when you ordered the heatsink/color/holder combo you specified werent stock, those are "backorder" (aka custom builds), theres one more in front of you but im doing bench work today and tomorrow so hope to get them out soon, those passives go pretty quick
Perfect! Thanks!
 

pop22

Well-Known Member
The wattage doesn't change. Volts x amperage = watts so 36V x 1000ma = 36 watts. As a measurement of the potential of the chip, it is totally meaningless.

Also
"At 50%, the new chips don't carry the brightness as the 1212s. At 100%, they're brighter then the 1212s."
how did you measure that? If the newer gen chips are more efficient, it means they put out more light at the same wattage as the previous chip.




Mr Cob,

Everything is installed. The chips are in the mail. Super Thanks by the way.

Question. Any idea what wattage the previous 1212s were putting out? I'm trying to figure out the compared wattage with the new chips.

I'm kinda impressed. Overdriven by 10volts for over 4 months. The temperature of the sinks, chips and the driver have fallen by 20°f. Now it has a full range of voltage. Going from 42v up to about 51v.

At 50%, the new chips don't carry the brightness as the 1212s. At 100%, they're brighter then the 1212s.

I need to add two more chips to the array. According to the driver specs, 6-7 chips are the best performance and the highest wattage yield on the curve. With wattage dropping off after an 8th chip is added.

Am I on track with the number of chips?
 

CobKits

Well-Known Member
sorry southcross i didnt see your post yesterday.

theres a couple things there

most notably the fact that with 36V chips on that driver, dimming via Vo probably does nothing as in all cases the voltage the driver is providing is far above the chips operating voltage. the chips take their share of current from the driver and voltage follows along, about 36V per chip

with the 50V chips, when you reduce Vo you are actually restricing voltage to the chip and pushing it down the voltage/current curve. as far as Io with those drivers, it generally adjusts from about 40% to 110% of driver current. but by using Vo you can dim it below 40% as again youre restricting the voltage with the 50V chips which limits the current the driver would pull.

hard to say exactly where you were operating it with the 36V chips, but with driver turned all the way up on both screws it should be 40% brighter

current is the same, but voltage is 40% higher, so wattage is 40% higher

as for how many chips fit on that driver its essentially unlimited. after reading this
I need to add two more chips to the array. According to the driver specs, 6-7 chips are the best performance and the highest wattage yield on the curve. With wattage dropping off after an 8th chip is added.

Am I on track with the number of chips?
it would help if maybe posted a screenshot of the driver datasheet that got you to that conclusion, so were talking apples to apples

with constant voltage drivers assuming you can fill it with X number of chips and operate them at full power (i.e for the current you are giving the chip you are not capping its voltage), youre using the same amount of driver power by adding more chips.

in other words say the max voltage on your driver was 53.5V and this corresponded to 2.1A for a 50V chip (just making up hypothetical current all chips are a little different)

if the driver is 7.6 A you hook up one cob and that driver is gonna put out 2.1A@ 53.5V
2 cobs: 4.2A @ 53.5V
3 cobs 6.3A @53.5V
4 cobs 7.6A @ maybe 53.3V
5 cobs 7.6A @ maybe 53.1V
6 cobs 7.6A @ maybe 53.0V

so i see what youre saying, as voltage falls you are starting to get less output out of your driver
its not clear whether the 7.6A limit is 100% fixed or if that will bump a little as voltage drops

my 2 cents is the increased efficiency of the cob chips makes up for the driver not operating at its absolute peak output. driver efficiency curves are pretty flat above 50% load

that would be an excellent study though! at some point, the overall light output (which grows plants) may decrease with heavy stacking on constant voltage drivers, or at best level off to where adding more cobs in fact adds no light at all because the incremental efficacy gain by dropping current per cob is exactly equal to the reduction in array voltage

certainly adding an infinite number of cobs makes the array more effcient per watt, BUT we are looking at overall light output as what grows plants so at some point youre adding cobs for nothing

the efficiency difference between 4 cobs vs 5 cobs and then 5 cobs vs6 cobs is somewhat pronounced

between 35 and 36 its tiny.

so we need to do some more homework to find the sweet spot i guess
 

mcnasty_nug

Well-Known Member
didnt notice a difference between the two honestly. if im gonna run one per sq/ft and drive them soft i would go with the 1212. the cxm22 are more efficient at higher wattage. my light with 666W and 9 COB is made from cxm22s air cooled in an old xxxl magnum hood but i havent ran it yet.
DesertHydro, Do you have a thread on the build? I'm working on switching over to LED lighting and I have a ton of magnum hoods I'd love to re-use. What did you use to build the frame into the hood?
 

SouthCross

Well-Known Member
sorry southcross i didnt see your post yesterday.

theres a couple things there

most notably the fact that with 36V chips on that driver, dimming via Vo probably does nothing as in all cases the voltage the driver is providing is far above the chips operating voltage. the chips take their share of current from the driver and voltage follows along, about 36V per chip

with the 50V chips, when you reduce Vo you are actually restricing voltage to the chip and pushing it down the voltage/current curve. as far as Io with those drivers, it generally adjusts from about 40% to 110% of driver current. but by using Vo you can dim it below 40% as again youre restricting the voltage with the 50V chips which limits the current the driver would pull.

hard to say exactly where you were operating it with the 36V chips, but with driver turned all the way up on both screws it should be 40% brighter

current is the same, but voltage is 40% higher, so wattage is 40% higher

as for how many chips fit on that driver its essentially unlimited. after reading this


it would help if maybe posted a screenshot of the driver datasheet that got you to that conclusion, so were talking apples to apples

with constant voltage drivers assuming you can fill it with X number of chips and operate them at full power (i.e for the current you are giving the chip you are not capping its voltage), youre using the same amount of driver power by adding more chips.

in other words say the max voltage on your driver was 53.5V and this corresponded to 2.1A for a 50V chip (just making up hypothetical current all chips are a little different)

if the driver is 7.6 A you hook up one cob and that driver is gonna put out 2.1A@ 53.5V
2 cobs: 4.2A @ 53.5V
3 cobs 6.3A @53.5V
4 cobs 7.6A @ maybe 53.3V
5 cobs 7.6A @ maybe 53.1V
6 cobs 7.6A @ maybe 53.0V

so i see what youre saying, as voltage falls you are starting to get less output out of your driver
its not clear whether the 7.6A limit is 100% fixed or if that will bump a little as voltage drops

my 2 cents is the increased efficiency of the cob chips makes up for the driver not operating at its absolute peak output. driver efficiency curves are pretty flat above 50% load

that would be an excellent study though! at some point, the overall light output (which grows plants) may decrease with heavy stacking on constant voltage drivers, or at best level off to where adding more cobs in fact adds no light at all because the incremental efficacy gain by dropping current per cob is exactly equal to the reduction in array voltage

certainly adding an infinite number of cobs makes the array more effcient per watt, BUT we are looking at overall light output as what grows plants so at some point youre adding cobs for nothing

the efficiency difference between 4 cobs vs 5 cobs and then 5 cobs vs6 cobs is somewhat pronounced

between 35 and 36 its tiny.

so we need to do some more homework to find the sweet spot i guess
Here's the driver.

https://cobkits.com/product/meanwell-hlg-320h-48a/


Going by these figures.

Max current: 7.6A can power 48V COBs at the following approximate wattage (actual wattage varies slightly by particular chip):

4@1.9A (~85W ea), 5@1.52A (~76W ea)
6@1.27A (~64W ea) , 7@1.09A (~55W ea)
8@0.95A (~40W ea), 9@0.84A (~42W ea)
10@0.76A (~38W ea)


The wattage drops off after the 7th cob is added. 10 COBs are in the area of 380. 6 COBs are 384. 7 COBs, 385. 8 COBs are 320 and 9 COBs are 378.

Seems according to the figures. There's a wattage curve with how many COBs are on the array per one driver. With 7 COBs rating the highest.
 

CobKits

Well-Known Member
those are all round numbers and not really indicative of a trend. cob voltage is a factor and vary not only from mfr to mfr but model to model (1818, 1825 are both 50V cobs but 1825 is lower voltage)
 

SouthCross

Well-Known Member
I downloaded the datasheet for the CMX22 Gen 3 80cri. I'm tracking a lot more on what you're saying. Relative output flux vs forward current. Forward current vs voltage. Time to pull the multimeter back out.

The maker says the output flux is a typical 8285Im. Comparison charts I've seen say 3000Im equals 200watts of incandescent bulb. 50w of florescent/LED.

So if I'm following the figures right. 4 at 100% will blow away a 600watt HID.
 
Last edited:

CobKits

Well-Known Member
theres a few assumptions there. fluorescent and LED shouldnt be in the same group as LEDs are 50-100% more efficiewnt depending on how hard you run them

4 cxm22 at 100W each would probably equal a 600W HPS
 

BobCajun

Well-Known Member
I found a good lens for COBs. It's an Alibaba page but could be useful anyway to some people who buy stuff from China. It's 60 degrees, which seems about right to me, 90 is actually too wide.
 

BobCajun

Well-Known Member
polycarbonate typically has pretty low transmittance. most people who use lenses opt for ground glass or silicone. the manufacturer gives no transmittance specs on those so its anybodys guess how efficient they are

probably best to post it in robin's thread about lenses as it will just get buried here

https://www.rollitup.org/t/lens-and-reflector-optics-for-cob.893660/
It's actually PMMA, acrylic, the plastic of choice for optical devices. The page says >90% transmission. But yeah I might post about it on some other thread.
 

hour

Well-Known Member
Aside from having to solder and not being able to use chip holders, is there anything special I should do in wiring up my new 9 cob 1212 setup, driven by a meanwell hlg-480h-1400?

My 640w setup has 3 separate drivers so this 9 cob/1 driver setup will be higher voltage than I've worked with. All wiring I purchased is rated for 600v fwiw. Thanks-can't seem to find many people running this in DIY setups, or my searching sucks
 

hour

Well-Known Member
wire is rated by current. 1400 mA isnt much

18 ga solid will be fine and is easy to solder
Sorry, yes, my wire is 18 gauge. The 600v is stamped on the jacketing, and if I'm not mistaken I wouldn't want to be running 300v through 18 gauge wire wrapped in insulation that I can scrape off with my fingernail, like doorbell(er..thermostat) wire, right? That stuff tends to not even be marked with a voltage rating that I've seen in my big box store perusing.
 

CobKits

Well-Known Member
depends.

cable assemblies are usually rated higher than primary wire, you can find 600V rated 18 solid on the internet at the wire suppliers
 
Hey cobkits, I was wondering what would be a nice setup for a 4x4 tent was looking at your clu048s that you've got on sale for $9, what would be better for a flowering light 80 or 90 cri, was looking at 3000k or 2700k ? Or a mix of both :-) or is there something else I should consider for a 4x4 space? Thanks :-)
 
Top