4 x CLU058-1825 @ 3.2A over 4'x4'

[mrNLK]

Hello fellow cannabis enthusiasts,

This is my first post at rollitup, but I've been lurking on the various forums, studying and cultivating cannabis for about 8 years now. I'd like to start my post by thanking everyone who has contributed to the wealth of knowledge and data that is now available in our community and industry, I certainly wouldn't be where I am today if it wasn't for all of you and your generosity of knowledge.

For my first build, I am trying to get the most power out of the lowest cost set up possible in a 4'x4' area. I've spent a lot of time reading about various builds and combinations, and I understand that in most cases running more cobs at a lower amperage results in the highest efficiency. That said, I would really appreciate some feedback/criticism/tips from the community on the following set up:

4 x CLU058-1825 3500K 80-90CRI = $160
4 x HLG-240H-54A = $220

200 total volts @ 3.2A = 640W
640W / 16 square feet = 40W / Square foot

I'm not certain of the efficiency, but I'm under the impression that the 1825 performs best in the 150-160w range.

Can anyone speak to the efficiency and PAR @ 12"-36" that a 1825 produces running at 3.2A?

 

[CobKits]

if youre only gonna run 3.2A per cob youre leaving a lot of driver on the table

you can get there with 2x 320-48A (or 54B) drivers cheaper, and it would be more efficient

 

[mrNLK]

@CobKits

Thank you very much for your response and input. I'm excited to see a cheaper/more efficient driver option available, but I'm trying to avoid parallel wiring. With the proposed drivers, that would be a necessary, no? Do you have any input on the ability/performance of these cobs at the stated amperage? Thanks again!

X

 

[CobKits]

Thank you very much for your response and input. I'm excited to see a cheaper/more efficient driver option available, but I'm trying to avoid parallel wiring.

with the voltage limiting on the A drivers there is really no disadvantage whatsoever to parallel wiring

if youre married to series, hlg-480h-c3500B is your only path into less # of drivers. again youre leaving a lot of driver capacity on the table

Do you have any input on the ability/performance of these cobs at the stated amperage?

they would be my choice

 

[mrNLK]

Thanks again for the feedback! I'll have to do some more reading to increase my comfort level and understanding of parallel wiring. In the mean time, would you recommend I push the cobs to a higher amperage, do you know at which point we see significant harm to efficiency(sub-HID SE/DE)?

 

[CobKits]

do you know at which point we see significant harm to efficiency(sub-HID SE/DE)?

just about any current cob run to its limit will still match HID (and beat it i suppose when considering lumen maintenance over time). there are exceptions but most people arent running the 500W cobs to their limits

 

[E2DX]

What about 16x Vero18's in a 4x4 with 2 HGL-240H-C1050 at 500w? Or step it up to 2X HVGC-320-1050A and get 640w?

 

[CobKits]

What about 16x Vero18's in a 4x4 with 2 HGL-240H-C1050 at 500w? Or step it up to 2X HVGC-320-1050A and get 640w?

16 of any chip (one per square foot) will rock your world. and efficiency of all chips in the 30-40W range is not all that different so you might as well go cheap, citi 1212, vero 18/22, etc.

 

[mrNLK]

Having 16 cobs improves efficiency, allows one to run the fixture closer to the crop canopy, and provides improved spread of PAR through the increased cob density.

I understand those benefits and advantages, but shouldn't my fixture @ 640w across four cobs be relatively comparable(within 10% of performance) to a setup @ 500w-640w across 16 cobs?

Thanks for the discussion and information!

 

[CobKits]

I understand those benefits and advantages, but shouldn't my fixture @ 640w across four cobs be relatively comparable(within 10% of performance) to a setup @ 500w-640w across 16 cobs?

a good rule of thumb ive found for leds is with the same heat dissipation, if you cut current in half efficiency goes up about 10%

so 8 cobs are 10% more efficient than 4 cobs
16 cobs are 10% more efficient than 8 cobs

so 16 cobs can be 20% more efficient than 4 cobs. lets take a look at actual measurements of a citi 1825 on the bench

X axis = watts
Y axis = ppfd/watts measured with a par meter at 12" on axis in free space.



4 cobs at 160W/chip = 8.4 PPFD/W
8 cobs at 80W/chip = 9.5 PPFD/W (+13% efficiency)
16 cobs at 40W/chip = 10.2 PPFD W(+7.4% over 8 cob, 21.4% over 4 cob)

you can even drop your overall wattage as you add cobs to maintain the same amount of light

16 cobs @ 32W/chip = 10.5 PPFD/W (+25% over 4 cob)

even though "PPFD" in this measurement is arbitrary and unrelated to actual PPFD in a garden we can use it for scaling

4 cobs at 160W/chip = 8.4 PPFD/W X 640W = 5376 "units" (i dont want to call them "PPFD" so as to confuse newbs with the actual PPFD we use to calc a garden. these are relative units and only useful when compared to other numbers in this example)

8 cobs at 80W/chip = 9.5 PPFD/W X 640W = 6080 units
16 cobs at 40W/chip = 10.2 PPFD W X 640W = 6528 units
16 cobs at 32W/chip = 10.5 PPFD/W X 512W = 5376 units (25% more efficient, we reduce input power by 25%, light output is the same). So 16 cobs at 512W is as efficient as 4 cobs at 640W (and give better coverage which we will disregard here)

electrical savings = 0.138 kW X 12 hrs x 365 days x $0.15/kWh = $91/year

so youre adding 12 cobs at maybe $50 each ($+600) with heatsinks so it will take you a while to get your money back

but notice how the improvement from 16 to 8 cobs (+13%) was more dramatic than 8 cobs to 4 cobs (+7.4%)

iteratively lets try running the 8 cobs at 70W/chip (560W)
8 cobs at 70W/chip = 9.65 PPFD/W X 560W = 5404 units. pretty close to our 4 cob rig

electrical savings = 0.08 kW X 12 hrs x 365 days x $0.15/kWh = $53/year but only a $200 addl investment up front so it covers the cost in 3 years then keeps on giving you money back after that.

buying LEDs is like starting a savings account you forget about that sits and builds interest year over year ("Compound" interest if electrical rates are rising)

 

[mrNLK]

@CobKits

I appreciate your feedback tremendously, thank you so much for sharing this data and information.

Based on what is provided, assuming I'm reading the graph correctly, a single cob @ 160w provides 1320 PPFD 'units' (At 12" from light source, at an unknown angle). Calculated at 8.25 PPFD 'units'/W.

I like the efficiency aspect, and my later fixtures will be designed to have higher CoB density and efficiency. At the moment I am restricted to <$500 and trying to squeeze something comparable to an HID system in power.

On the topic of longer term efficiency and cost return. My first series of fixtures are meant to be comparable to an HID system in cost and performance with a ~25%-35% reduction in wall draw. I figure with the legalization of recreational cannabis, 280-E tax law reform, and the amount of cash this industry and crop is about to receive and generate, in the coming year or two there will be significant and rapid improvements in LED driver and lighting technology. At that time, I would have more fluid cash to design a significantly better system with better hardware.

Thanks again Cobkits, I sincerely appreciate your contributions in data and hardware sourcing for the community. You rock!

 

[CobKits]

so lets look at a different example. why run expensive 1825 chips at 32W when smaller cheaper chips do the same. bigger chips are great at higher wattage. their advantage at lower wattages is slim to none. in the graph below you will see that "bigger chips" like vero29C and citi 1825 gen5 are 9% more efficient than cxm22@160W (technically well out of cxm22s 125W range), but only about 3% more efficient at 80W and not any more efficient at all below 50W



lets look closeup at teh 32W range. red would be efficiency of 1825 or cxm22 (equal), green would be wattage of 1212 at that same efficiency



lets say a cxm22 chip heatsink holder is around $35. 16 cob cxm22 rig (like the 1825 rig) is as efficient at 512W as the 1825 4-cob rig is at 640W

but adding chips is cheaper.

4 1825 assemblies = $200
16 1825 assemblies = $800
16 cxm22 assemblies = $560

so your payback period is significantly shortened

you can do the same math with the 1212s
you can see why im bullish on the cxm22s tho. if youre really squeezing your cobs for efficiency at low currents there is no reason to spend twice the price on a chip, as the cxm22s are phenomenal up to 70W or so.

 

[mrNLK]

@CobKits

I think my ideal build for a 4'x4' at low CoB wattage would be 20 citi 1212's, but for a first build I don't want to go that wild or work that hard. If I knew there was a setup including 8-16 cobs providing similar PPFD 'units' at a similar build cost, then I would be highly interested in exploring those options. I love your feedback, and thank you for keeping the discussion going. I'm hoping someone can provide some real-world PPFD data on the clu058-1825 @ 3.2A.

Thanks again!

 

[CobKits]

Based on what is provided, assuming I'm reading the graph correctly, a single cob @ 160w provides 1320 PPFD (At an unknown distance and angel from the light). Calculated at 8.25 PPFD/W.

yeah its all relative, point being its static. all the measurements in the graphs were taken at 12" in free space under similar circumstances. ive since taken to measuring inside of a sphere which gives better low current data (no stray light). PPFD is just what my meter, in that location, output and is independent (though correlatable) to actual ppfd of light in the garden. as a wild guess in the graph above 10 ppfd/W is prob right about 2 umol/J at the chip level

I like the efficiency aspect, and my later fixtures will be designed to have higher CoB density and efficiency. At the moment I am restricted to <$500 and trying to squeeze something comparable to an HID system in power.

see luminus rec above. way more bang for the buck than citi

I figure with the legalization of recreational cannabis, 280-E tax law reform, and the amount of cash this industry and crop is about to receive and generate, in the coming year or two there will be significant and rapid improvements in LED driver and lighting technology. At that time, I would have more fluid cash to design a significantly better system with better hardware.

not a bad angle. drivers are at top of their curve as tht type of power supply is old hat. hard to improve on 96% efficiency

**OH and that was a significant part we missed. your OG 240W driver setup is up to 93.5% efficient (less than 90 if youre running your drivers at 120V. Running at 240V is the free lunch that everybody misses)



lets look at HLG 600:



a single HLG 600 driver is more efficient at 115W than your HLG240 is at 120V. its a full +3% efficient at 240V. thats like knocking your 512W rig under 500W at the wall

in fact since were around 500W now youre into the territory of a single HLG480


95.5% at 240V and 93% at 120V not too shabby. that driver has 10.75A at 51.7V and would run those 16 luminus perfectly at 32W each (512 output watts).

you could start out with as few as 8,10, or 12 cobs on that single driver and add more later. Parallel FTW!

 

[CobKits]

I'm hoping someone can provide some real-world PPFD data on the clu058-1825 @ 3.2A.

i dont know anybody running them that hard. your best bet is to estimate using the relationships above (which are similar on datasheet and citi calculator

as a guess at 50W the cxm22 and 1825 are 55% efficient and you need 20 PAR W/SF to get 900 ppfd

so at 55% efficient we need 36W/SF (DC side of driver) to get ~900 PPFD

so at 3.2A/170W its 17% less efficient than at 50W. you would need 36W/SF x 1.17 = 42W/SF to get the same ~900 PPFD

its all relative

 

[PilouPilou]

Hi @CobKits what is your opinion in term of final weight of a grow in the same same space: per example 4 x 1825/or vero29C at around 100W each can do better or same or less than 9 x 36V/(or more) COB driven with less intensity but with the same total wattage? thx for your reply

 

[mrNLK]

You are a wealth of information, it is awesome. I'll look at the luminous CoB build cost, I'm impressed with individual CoB performance @ low watt.

When you say that 3.2A/170W is 17% less efficient are you saying that the CoB would be running in the range of 38% efficiency, or 45.65% efficiency(55 multiplied by 0.83)?

If it is 170W @ 38% efficiency, then I would be pulling 16 PAR W/SQ on a good day. That isn't enough.
If it is 170W @ 45.65% efficiency, then I would be nearing 19.50 PAR W/SQ on a good day. That is sufficient.

I'm not certain I'm understanding how you are getting 42W/SF to meet 20 PAR W/SQ.

42W/SQ. @ 38% = 15.96 PAR W/SQ.
42W/SQ. @ 45% = 18.9 PAR W/SQ.
42W/SQ. @ 48% = 20.16 PAR W/SQ.

Thanks!

 

[CobKits]

Hi @CobKits what is your opinion in term of final weight of a grow in the same same space: per example 4 x 1825/or vero29C at around 100W each can do better or same or less than 9 x 36V/(or more) COB driven with less intensity but with the same total wattage? thx for your reply

well all the info to calc that is given above so give it a shot and ask here if you need help

steps:
lookup the relative efficiency of your chips at 100W each (9.2)

look at what chip wattage of 1212s is that efficiency (50W is right about 9.2)

you need twice as many smaller chips to be as efficient. if youre trying to stay passive 100W heatsinks cost twice as much as 50W heatsinks so thats a wash

the chips are less than 1/2 the cost do thats a savings

driver cost the same

coverage = better

 

[ANC]

Start with active CPU coolers, upgrade to passive later on.

 

[CobKits]

When you say that 3.2A/170W is 17% less efficient are you saying that the CoB would be running in the range of 38% efficiency, or 45.65% efficiency(55 multiplied by 0.83)?

the latter, its relative. at 50W chip is 10 ppfd/W, at 170 its 8.3 ppfd/W

I'm not certain I'm understanding how you are getting 42W/SF to meet 20 PAR W/SQ.

42W/SQ. @ 38% = 15.96 PAR W/SQ.
42W/SQ. @ 45% = 18.9 PAR W/SQ.
42W/SQ. @ 48% = 20.16 PAR W/SQ.

oops i messed up there, 17% less efficient isnt the same as 1.17 on the other side

1/0.83 = 1.204

we really need 20.4% more watts so 36*1.204 = 43.34 W/SF

at 55% efficiency
36W/SF @ 55% = 19.8 PAR W/SF

at 17% less relative efficiency, 45.65% efficiency
43.34W/SF @ 45.65% = 19.8 PAR W/SF

43.34 is 20.4% more than 36
36 is 17% less than 43.34

depends if you use 36 or 43.34 as the denominator