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

@CobKits

With my current design:

680 CoB W/16SQ = 42.5 CoBW/SQ

42.5 CoB W/SQ @ 45.65% = 19.15 PAR W/SQ

20 PAR W/SQ = 900 PPFD

If 20 = 1, then 19.15 = 0.9575

900 PPFD multiplied by 0.9575 = 861.75 PPFD

Is PPFD in this instance our estimate for real world, or is this PPFD 'units'?

Thanks!

------------------------------------------------------------------------------------------------------------

EDIT: This is not accounting for driver efficiency. These numbers are probably more realistic at a 10% performance reduction:

680 x .90 = 612 CoB W/16SQ = 38.25 CoB W/SQ
38.25 @ 45.65% = 17.24 PAR W/SQ
17.24 PAR W/SQ = 86%(0.86) relative to 20 PAR W/SQ.
900 PPFD multiplied by 0.86 = 775.6 PPFD

When guesstimating 20 PAR W/SQ = 900 PPFD, are you accounting for driver efficiency?

ANC said:

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

Click to expand...

I don't want to spoil the cost of setup by introducing massive passive cooling. Initially will use some variety of CPU cooler or fan attached to alternative heat-sink medium. Later I will likely move to passive or water cooled systems.

Thanks for your input!

mrNLK said:

20 PAR W/SQ = 900 PPFD

If 20 = 1, then 19.15 = 0.9575

900 PPFD multiplied by 0.9575 = 861.75 PPFD

Is PPFD in this instance our estimate for real world, or is this PPFD 'units'?

Click to expand...

20 parW/SF is in fact 900 PPFD on the plants. the loose variable is understanding efficiency to apply 20 PAR W

mrNLK said:

EDIT: This is not accounting for driver efficiency. These numbers are probably more realistic at a 10% performance reduction:

Click to expand...

not if you run a 95%+ efficient driver

mrNLK said:

When guesstimating 20 PAR W/SQ = 900 PPFD, are you accounting for driver efficiency?

Click to expand...

20 PAR watts is completely independent of driver or lights used to get there (assuming similar spectrum)

COB DC watts are not reduced by poorly efficient driver, you just use more AC to get the same DC output

CobKits said:

not if you run a 95%+ efficient driver

Click to expand...

You were implying that two 320-54B in parallel would be more efficient than my current driver selection. Since you suggested the option, do you know if they would run more efficiently than my selection of the 240H-54A?

I'd love to maintain that 95%+ driver efficiency if it's not going to cost me. I feel more comfortable with parallel driving now.

I'm getting the feeling that my design may, in theory, function to my needs. The only limitation I'm concerned about is proximity to canopy and relative coverage. Hoping that my plants will be comfortable and I'll obtain a decent canopy coverage while hitting the 800+ PPFD range. I figure 4 points of light has to be better than 1(HID).

Thank you!

mrNLK said:

You were implying that two 320-54B in parallel would be more efficient than my current driver selection. Since you suggested the option, do you know if they would run more efficiently than my selection of the 240H-54A?

Click to expand...

efficiencies are pretty similar in any class of driver (all 240s are similar, all 320s are similar etc)

320:

mrNLK said:

I'm getting the feeling that my design may, in theory, function to my needs. The only limitation I'm concerned about is proximity to canopy and relative coverage. Hoping that my plants will be comfortable and I'll obtain a decent canopy coverage while hitting the 800+ PPFD range. I figure 4 points of light has to be better than 1(HID).

Click to expand...

definitely more uniform and less stress if your girls overgrow close to a light

I see, my mistake, the efficiency is affected by the use of 110v-240v.

Thank you so much for your shared info.

I should be ordering parts in the coming days, as the build progresses I will share progress and data.

mrNLK said:

I don't want to spoil the cost of setup by introducing massive passive cooling. Initially will use some variety of CPU cooler or fan attached to alternative heat-sink medium. Later I will likely move to passive or water cooled systems.

Thanks for your input!

Click to expand...

the 133 heatsinks i use have mounts for 50mm fans (which i have, but not on site)

you can run those to 150ishW each easily with a $5 fan on top

Thanks for that info! Is there a product data sheet that expresses the amount of surface area the 133mm unit boasts?

I'm ordering my first 1825 CoB, heatsink and driver today. Once it arrives, I'll hook them up and test the performance at 3.2A. Figure I'll test one out over a 2x2 before trying it on a 4x4.

Thanks for your input @CobKits !

I got nearly same setup instead the 54a its the 54b on one clu058 but will definitely upgrade this to a second one.

war74u said:

I got nearly same setup instead the 54a its the 54b on one clu058 but will definitely upgrade this to a second one.

Click to expand...

Hey War74u, thanks for your contribution to the thread!

To clarify, you are running a CLU58 1825 @ ~3.2A? Can you speak on it's distance from canopy, foot print, and any other data(Lumens, PAR, etc)?

No t

mrNLK said:

Hey War74u, thanks for your contribution to the thread!

To clarify, you are running a CLU58 1825 @ ~3.2A? Can you speak on it's distance from canopy, foot print, and any other data(Lumens, PAR, etc)?

Click to expand...

No problem nice to see others experiencing with similae equipment, but no so far i only got hands on a Multimeter and i just checked the voltage to see if the poti is running properly. The B version of the meanwell driver is different than the a version i did get that only after buying, i will check tomorrow the current output at maximum given voltage(which is less than with the 54a) and see if the 3.2a at least can be reached. Before buying i thought the b version just the more comfortable one.

war74u said:

No t


No problem nice to see others experiencing with similae equipment, but no so far i only got hands on a Multimeter and i just checked the voltage to see if the poti is running properly. The B version of the meanwell driver is different than the a version i did get that only after buying, i will check tomorrow the current output at maximum given voltage(which is less than with the 54a) and see if the 3.2a at least can be reached. Before buying i thought the b version just the more comfortable one.

Click to expand...

gen 6 should get there. gen5 will prob top out at around 3000 mA depending on temp

To make the whole thing a bit more compliceted...
Why not using six 4ft. DoubleRow Samsung F-Series strips? 48$/per strip

https://www.digikey.de/product-detail/en/samsung-semiconductor-inc/SI-B8VZ91B20WW/1510-2229-ND/6676693?cur=USD&lang=en

They use the same diode as QuantumBoards(Samsung LM561c) and distribute the light in a way that COB's can not!
Driven at nominal current of 2,24A six strips together is ~620w net., 168lm/w(52,25% eff.) and 104.040lm.
That's 2,54μMol/J or above 1000μMol/s/1,44m²(4x 4')!

You could still use two HLG-320H-48A and would probably end with up to 700w max. at the wall.!
https://www.mouser.com/Search/m_ProductDetail.aspx?Mean-Well/HLG-320H-48/&qs=sGAEpiMZZMvV8Y9YugmIgk5X04kZkB/a/zQv/yOJEtA=

For cooling you only need six 2x 2x 2" aluminum c-channels and can go strictly the passive route too...
From the scrap yard at kilo price base...20bucks!

And what is the best.. you have probably a few bucks left in your pocket because all thats costs you less than 500$...


6 strips á 48$ = 288$
2 HLG-320H-48A á 93$ = 186$
6 4ft. C-channels =20$
__________________________
Total costs: 494$


Have a look at here if you want to see these strips in action:

https://www.rollitup.org/t/riders-first-foray-into-led-strip-lighting-samsung-h-series-gen3.950197/

Randomblame said:

They use the same diode as QuantumBoards(Samsung LM561c) and distribute the light in a way that COB's can not!

Click to expand...

well its different but at a certain height (typically around that of cob spacing, 18" or so in this example) either will be equally as uniform.

those are great strips though - its just the difference between cobs and strips/boards isnt really as night and day as some claim. If youve ever measured cob lights in a garden they can be extremely uniform at proper distance

Deffinately not a big difference but a lot less wiring effort. I like the idea to use more COB's running low and own a 150w veg lamp with 2 COB's per st² @350-700mA.

Slightly better # vs. 8x CXB3590
https://www.rollitup.org/t/riders-first-foray-into-led-strip-lighting-samsung-h-series-gen3.950197/page-3#post-13861433

I like them mainly because of the simplicity and the fact that it is cheaper to work with it because one need only a c-channel for cooling instead of expensive pin heatsinks.

@Randomblame

I like the idea a lot! Thank you for sharing this option with us. I've yet to discover anyone driving the CoBs as hard as I would want to. Where as @Rider509 four strip fixture performed impressively well at 500W. The 800-1000PPFD range is what I am looking to be in. You seem to be a big fan of this application and hardware, are you running it yourself?

My Arduino arrived yesterday and I want to make time this afternoon to play with it, trying to make my first fixture "smart," but I'll spend a little time researching the strip lighting approach before I put in my final hardware order.

In theory, six strips would be more wiring for me than the 4 cobs and two drivers.

Thanks again for input!

I can run my four 3590 four-COB groups to almost 320W each for over 1200W in a 4x4, but I like the penetration I'm getting with the strips better. Less shadowing of the lower leaves by upper leaves.

mrNLK said:

@Randomblame

I like the idea a lot! Thank you for sharing this option with us. I've yet to discover anyone driving the CoBs as hard as I would want to. Where as @Rider509 four strip fixture performed impressively well at 500W. The 800-1000PPFD range is what I am looking to be in. You seem to be a big fan of this application and hardware, are you running it yourself?

My Arduino arrived yesterday and I want to make time this afternoon to play with it, trying to make my first fixture "smart," but I'll spend a little time researching the strip lighting approach before I put in my final hardware order.

In theory, six strips would be more wiring for me than the 4 cobs and two drivers.

Thanks again for input!

Click to expand...

Thanks, mate!
Yeah, I like them a lot! It is the diffused but intense light that makes these LEDs so well suited for our purposes.

I'm currently building two new 250w "hood-lights" using 864 3k/LM561c diodes, four V18/CRI90 and two 1750k COB's together with an 24w UVB-tube. Still waiting on some misc parts and xmax is also near so it will take some time before I can start. Maybe in january, maybe end of december...

And to your question about running high powered COB's.
@nevergoodenuf is using COB's with up to 300w in his high power garden, Citi3618 for example.
Here is the link:
https://www.rollitup.org/p/13419954/

My concerns about hard driven COB's are so much because of bleaching. You will need at least 24-30" distance if you drive the COB's so hard! With stripes you would not have this problem, because the light and heat is much better distributed. Especially for LED beginners, I would therefore always recommend strips.
Mounting and wiring are also easier as you do not have to solder. Thermal grease is also not necessary! Only a few selftapping screws or double sided thermal tape to mount the strips to the c-channels.

CobKits said:

gen 6 should get there. gen5 will prob top out at around 3000 mA depending on temp

Click to expand...

thx good thing to know when it comes to discernment between gen‘s. i had it at 53v with around 2000ma when turned up to 54,8v( which was max but maybe cause of that low cost poti) it was around 2900ma so probably gen5.
What i recognized it that extrem temperature increase form 2000ma to 2900ma i now wouldn’t dare to run it with that glass lense in front even with active cooling.