What to Do? (Help My Build!) 4x CXB3590 72v and 3x 1400mA Driver Ordered

TLDR How should I space my lights for a 3ft x 6ft or 6ft x 6ft area, is 3000K alright to start with veg, how should I use my extra driver... help me build something cool!

A ton of research has been done, to the point where it was just going in circles trying to decide between overall light/energy output, upfront cost, efficiency, color warmth/spectra etc... So just to take action and get started, 4 CXB3590-0000-000R0HCB30G (3000K 30G CB 72V 3-step) and 3x 1400mA HLG-185H-C1400B have been procured; so I have an extra driver to use up, sell, keep as spare, etc.

I am working with just over 6' x 6' total area and made my original decisions based on that, but am thinking I will design this to eventually be used in 3' x 6' of that and having two chambers in the future. There is a pole in the middle of the area So I will probably be making 3 separate fixtures for modularity.

I was planning to run the lights at the full 1400 mA, slightly overdriving them, giving me some real intense output per light, which would give me a much higher light output for seemingly lower initial cost. Or I could back them down if I wanted. I was aware of the hit on efficiency but caved in; I know the efficiency pays for it over time of course, but the lights aren't the only thing I needed and I am not on a relatively high budget at all.

Trying to cut to the chase, I need help figuring this all out. For one, I want to use the spare driver and get some 6500K, 5000K, or similar to start with, then down the line at the right time also get some 2700k and hooking up some switches to make a super versatile light! These need to be lower cost LEDs but I don't want them to be totally useless.

The problem is I am having trouble finding them since most places don't have stock of 6500k, anyone know good sources in the USA? Any other brands to look for other than Cree and Bridglux?

They also need to make proper use of the driver's 71v-143v 700-1400 mA output, or I would have to get cheap drivers (scared of fire risk). In the future I can also increase the efficiency by getting other drivers / lights and make a new combination.

However, could I get by with just the 4x 3000K lights for now, putting out about 11,000 lm or more each, to fill either 3ft x 6ft or 6ft x 6ft? Is there enough lower wavelength light for that stage?

I am going crazy trying to figure out the spacing, actually using trigonometry for once. I may design it as more of a bar style, all 4 LEDs in a row, and then supplementing that with what I create with the other driver to help with the width. At 18 inches and a 60 degree angle of light I calculated I should place them 14.8 inches apart to spread 6 ft. Does that sound right? Is 60 degrees too small of an angle? Is this a good idea for 3ft x 6 ft? What angle of light do you use? I plan to use reflectors.

This is too long winded, so please just give your input on what would you do with the other driver, light spacing, light wavelengths, etc.

Are you aware of the fact that 2 CXB3590 at 1400mA need 148-149V while HLG-185H-C1400 provides only 143V?

4 x 3000k for now is enough. I was using 4 x 5000k cxa3590 for veg in a 2ft x 5ft x 5ft tent. Only difference for you, you're going to get a little lower temperature than me. I was getting 70%-75% humidity light on and about 80% at light off. I didn't use a intake, exhaust nor did I have any fans inside blowing at the plants. The only fans were the PC fans cooling down the heatsinks. My temperature inside was probably at 25c-30c. Maybe 28-29c tops, only saying 30c because my temperature sensor was at the bottom of the tent, so my temp could've been a tad higher.

To answer your question. Use 50-100w on a heatsink covering a sqft each. I'm currently at 100w sq/ft, with the light @ 5 - 8" from the plant. The only time they bleached my plant was, when I left for a week. Came back and they were touching the cree chips. I, also don't have any fan blowing at the cree chips. I use its heat to warm up my plants.

alesh said:

Are you aware of the fact that 2 CXB3590 at 1400mA need 148-149V while HLG-185H-C1400 provides only 143V?

Click to expand...

And no efficiency bonus over Vero at that current. 3 Vero29 @ 1.4 amps on a HLG-185 is a pretty sweet combo.

alesh said:

Are you aware of the fact that 2 CXB3590 at 1400mA need 148-149V while HLG-185H-C1400 provides only 143V?

Click to expand...

Thank you for pointing that out, I guess you are the first person to save my ass on this I did do a lot of research and calculations etc so I feel dumb now, my brain is pretty scrambled trying to calculate the possibilities of so many configurations and I'm just anxious to get things moving. It was another model I was going to overdrive. Going back and looking at the chart on the data sheet, I was also misreading the graph (attached). It will require more voltage at a lower temperature, so I wasn't taking that into account either. It's looking like I would have to run it at about 1000-1100mA max? What should I run it at for an optimum blend of efficiency and power in a 3x6 space? Time to go back to the drawing books...

BecauseIgotHigh said:

4 x 3000k for now is enough. I was using 4 x 5000k cxa3590 for veg in a 2ft x 5ft x 5ft tent. Only difference for you, you're going to get a little lower temperature than me. I was getting 70%-75% humidity light on and about 80% at light off. I didn't use a intake, exhaust nor did I have any fans inside blowing at the plants. The only fans were the PC fans cooling down the heatsinks. My temperature inside was probably at 25c-30c. Maybe 28-29c tops, only saying 30c because my temperature sensor was at the bottom of the tent, so my temp could've been a tad higher.

To answer your question. Use 50-100w on a heatsink covering a sqft each. I'm currently at 100w sq/ft, with the light @ 5 - 8" from the plant. The only time they bleached my plant was, when I left for a week. Came back and they were touching the cree chips. I, also don't have any fan blowing at the cree chips. I use its heat to warm up my plants.

Click to expand...

Hmm, it seems like you aren't using reflectors or lenses? I assume you had them in a row rather than side by side? I was planning on trying to get some intense light, now things are working backwards. Lower mA and wider angle. On the plus side, my efficiency numbers are increasing with every step back I take on what mA I'll be driving them at.

I was looking at Vero29s and may end up using them with these drivers in the end, this isn't going to be the optimal setup right off the bat, but at least I will have something going. Then in a couple months I will reconfigure when I build another setup for the other chamber.

helloseahorse! said:

Thank you for pointing that out, I guess you are the first person to save my ass on this I did do a lot of research and calculations etc so I feel dumb now, my brain is pretty scrambled trying to calculate the possibilities of so many configurations and I'm just anxious to get things moving. It was another model I was going to overdrive. Going back and looking at the chart on the data sheet, I was also misreading the graph (attached). It will require more voltage at a lower temperature, so I wasn't taking that into account either. It's looking like I would have to run it at about 1000-1100mA max? What should I run it at for an optimum blend of efficiency and power in a 3x6 space? Time to go back to the drawing books...



Hmm, it seems like you aren't using reflectors or lenses? I assume you had them in a row rather than side by side? I was planning on trying to get some intense light, now things are working backwards. Lower mA and wider angle. On the plus side, my efficiency numbers are increasing with every step back I take on what mA I'll be driving them at.

I was looking at Vero29s and may end up using them with these drivers in the end, this isn't going to be the optimal setup right off the bat, but at least I will have something going. Then in a couple months I will reconfigure when I build another setup for the other chamber.

Click to expand...

You can run 4 CXB3590 on HLG-185H-C700 approximately 50 watts per COB @ 51.67% efficiency VS the Vero 29 @ 1.4A 51 watt 43%
HLG-185H-C700 is good for 286V and 4 CXB3590 should draw about 280V @ 50c

4 CXB3590 CB @ 700ma = 103 par watts / 97 heat watts
4 Vero 29 @ 1.4A = 87.72 par watts / 116.28 heat watts

If you run a meanwell cen100-36, you can run from 2 to 6 Vero 29s in parallel. I am running 4 Vero 29 on 2 of my lights with one driver per light (25 watts each) and 2 Vero 29s on another driver (50 watts each). I like the options I have with these drivers, plus I already had them. They are also cheaper (under $50)

nevergoodenuf said:

If you run a meanwell cen100-36, you can run from 2 to 6 Vero 29s in parallel. I am running 4 Vero 29 on 2 of my lights with one driver per light (25 watts each) and 2 Vero 29s on another driver (50 watts each). I like the options I have with these drivers, plus I already had them. They are also cheaper (under $50)

Click to expand...

I thought it wasn't recommended to run COB's in parallel?

I talked with Bridgelux about this and there is no safety issue. There is just a chance that one COB draws a little more power than the other. Since we are not running them at max and are driver is well under the max power rating of the COB, if one COB fails everything else will be fine. If you had 2 COBs on a 200-36 and one has thermal runaway, then you will eventually loss both. This is what I was told by Bridgelux.

helloseahorse! said:

Thank you for pointing that out, I guess you are the first person to save my ass on this I did do a lot of research and calculations etc so I feel dumb now, my brain is pretty scrambled trying to calculate the possibilities of so many configurations and I'm just anxious to get things moving. It was another model I was going to overdrive. Going back and looking at the chart on the data sheet, I was also misreading the graph (attached). It will require more voltage at a lower temperature, so I wasn't taking that into account either. It's looking like I would have to run it at about 1000-1100mA max? What should I run it at for an optimum blend of efficiency and power in a 3x6 space? Time to go back to the drawing books...



Hmm, it seems like you aren't using reflectors or lenses? I assume you had them in a row rather than side by side? I was planning on trying to get some intense light, now things are working backwards. Lower mA and wider angle. On the plus side, my efficiency numbers are increasing with every step back I take on what mA I'll be driving them at.

I was looking at Vero29s and may end up using them with these drivers in the end, this isn't going to be the optimal setup right off the bat, but at least I will have something going. Then in a couple months I will reconfigure when I build another setup for the other chamber.

Click to expand...

Go buy 1.4A first then, after your harvest, go buy better CXBs

BecauseIgotHigh said:

4 x 3000k for now is enough. I was using 4 x 5000k cxa3590 for veg in a 2ft x 5ft x 5ft tent. Only difference for you, you're going to get a little lower temperature than me. I was getting 70%-75% humidity light on and about 80% at light off. I didn't use a intake, exhaust nor did I have any fans inside blowing at the plants. The only fans were the PC fans cooling down the heatsinks. My temperature inside was probably at 25c-30c. Maybe 28-29c tops, only saying 30c because my temperature sensor was at the bottom of the tent, so my temp could've been a tad higher.

To answer your question. Use 50-100w on a heatsink covering a sqft each. I'm currently at 100w sq/ft, with the light @ 5 - 8" from the plant. The only time they bleached my plant was, when I left for a week. Came back and they were touching the cree chips. I, also don't have any fan blowing at the cree chips. I use its heat to warm up my plants.

Click to expand...

Da fuk did I just read? You're running 100W of COB LED lighting in your space?!

Somewhere something is seriously amiss with the math...

ttystikk said:

Da fuk did I just read? You're running 100W of COB LED lighting in your space?!

Somewhere something is seriously amiss with the math...

Click to expand...

My flowering setup it's 2 Cree cxb3590 @ 50w each. 4 heatsinks. Each sink with a pair of cxb3590 @ 700ma.

helloseahorse! said:

It will require more voltage at a lower temperature, so I wasn't taking that into account either. It's looking like I would have to run it at about 1000-1100mA max?

Click to expand...

Although it is true the driver will not be able to output the full 1.4A, it should be able to put out near 1.05A and since the voltage would be maxed you should get the full driver efficiency. As the COBs warm up and the Vf drops, the driver will automatically increase its output higher. So if you already bought the drivers and/or COBs no worries. Assuming 1.05A output, each CXB will dissipate ~75W. A good value in terms of driver cost and significantly outperforms the Vero29 in terms of cost/efficiency

CXB3590 3500K CD 72V class @ 1.05A = 75.5W @ 50% cost 1.26/PAR W
Vero 29 V2.0 3500K---------------@ 1.4A = 51W @ 43% cost $1.29/PAR W

That said Vero 29 X 4 @ ~1.3A is a better value in terms of driver cost and can give you a more uniform spread.

SupraSPL said:

Although it is true the driver will not be able to output the full 1.4A, it should be able to put out near 1.05A and since the voltage would be maxed you should get the full driver efficiency. As the COBs warm up and the Vf drops, the driver will automatically increase its output higher. So if you already bought the drivers and/or COBs no worries. Assuming 1.05A output, each CXB will dissipate ~75W. A good value in terms of driver cost and significantly outperforms the Vero29 in terms of cost/efficiency

CXB3590 3500K CD 72V class @ 1.05A = 75.5W @ 50% cost 1.26/PAR W
Vero 29 V2.0 3500K---------------@ 1.4A = 51W @ 43% cost $1.29/PAR W

That said Vero 29 X 4 @ ~1.3A is a better value in terms of driver cost and can give you a more uniform spread.

Click to expand...

ELG-150-C1050 or HLG-120H-C1050 would be a great fit for 2 72V CXB3590. Or ELG-150-C2100 for 2 36V CXB3590. This actually seems like an awesome combo -
(2) CXB3590 - $100
(2) holder - $6
(1) ELG-150-C - $40
HS, wire, fan, etc. - $56
A nice 150W light for about $200/$250 tops.

BecauseIgotHigh said:

My flowering setup it's 2 Cree cxb3590 @ 50w each. 4 heatsinks. Each sink with a pair of cxb3590 @ 700ma.

Click to expand...

This reply doesn't say how big your grow space is, nor how many watts you're lighting it with in total.

SupraSPL said:

Although it is true the driver will not be able to output the full 1.4A, it should be able to put out near 1.05A and since the voltage would be maxed you should get the full driver efficiency. As the COBs warm up and the Vf drops, the driver will automatically increase its output higher. So if you already bought the drivers and/or COBs no worries. Assuming 1.05A output, each CXB will dissipate ~75W. A good value in terms of driver cost and significantly outperforms the Vero29 in terms of cost/efficiency

CXB3590 3500K CD 72V class @ 1.05A = 75.5W @ 50% cost 1.26/PAR W
Vero 29 V2.0 3500K---------------@ 1.4A = 51W @ 43% cost $1.29/PAR W

That said Vero 29 X 4 @ ~1.3A is a better value in terms of driver cost and can give you a more uniform spread.

Click to expand...

Thank you for your input! You are the head guru around here so I was hoping you'd chime in... I did the math and plan to run it at about the 1.05 (or actually just 1) amp. Having the driver perfectly maxed out for full efficiency is actually one reason I decided on this setup... I think I had things right in my head when I made the order, just not when I wrote the post... a lot of sleep deprivation going manic over the whole project (not just the light) combined with Metal Gear Solid V.

The math actually adds up perfect up to where I could actually get the exact resistors (71.5k/n drivers) I would need to achieve exactly 1000 mA, which also matches up with the voltage limit of the of the driver exactly even with the COBs at room temperature. I decided to go with resistors because a potentiometer wouldn't be as accurate, and a PWM signal from an Arduino, RPi or other device doesn't sound great since it's essentially just switching the power on and off at an extremely fast rate if I'm not mistaken... although I'm into Arduino and RPi so I will inevitably play with some dimming control in the future especially if I can design a circuit that switches the current through a resistor after the PWM based dimming stage (slowly turning lights on/off to reduce shock then switching to the more stable method when a constant output is needed).

alesh said:

ELG-150-C1050 or HLG-120H-C1050 would be a great fit for 2 72V CXB3590. Or ELG-150-C2100 for 2 36V CXB3590. This actually seems like an awesome combo -
(2) CXB3590 - $100
(2) holder - $6
(1) ELG-150-C - $40
HS, wire, fan, etc. - $56
A nice 150W light for about $200/$250 tops.

Click to expand...

Is there an advantage to running the driver at full current? I actually think I made the right choice here if I don't want to run much lower than 1A. Shouldn't it be an advantage to not have to run it at the maximum amperage; less heat / wear and tear on the components? Of course if I wanted to run it at 700 mA I'd want a 1050. Is there any efficiency advantage or disadvantage to running the driver at it's amperage limit?

ttystikk said:

This reply doesn't say how big your grow space is, nor how many watts you're lighting it with in total.

Click to expand...

I'm pretty sure he's saying he has 4 heatsinks with two COBs on each, each COB producing 50W for a total of 400W. Claiming 100W / sq ft he must be working with 4 ft x 4 ft.

BecauseIgotHigh said:

Go buy 1.4A first then, after your harvest, go buy better CXBs

Click to expand...

What do you mean by, "better"? You mean the different wavelengths I want to supplement with?

helloseahorse! said:

Thank you for your input! You are the head guru around here so I was hoping you'd chime in... I did the math and plan to run it at about the 1.05 (or actually just 1) amp. Having the driver perfectly maxed out for full efficiency is actually one reason I decided on this setup... I think I had things right in my head when I made the order, just not when I wrote the post... a lot of sleep deprivation going manic over the whole project (not just the light) combined with Metal Gear Solid V.

The math actually adds up perfect up to where I could actually get the exact resistors (71.5k/n drivers) I would need to achieve exactly 1000 mA, which also matches up with the voltage limit of the of the driver exactly even with the COBs at room temperature. I decided to go with resistors because a potentiometer wouldn't be as accurate, and a PWM signal from an Arduino, RPi or other device doesn't sound great since it's essentially just switching the power on and off at an extremely fast rate if I'm not mistaken... although I'm into Arduino and RPi so I will inevitably play with some dimming control in the future especially if I can design a circuit that switches the current through a resistor after the PWM based dimming stage (slowly turning lights on/off to reduce shock then switching to the more stable method when a constant output is needed).



Is there an advantage to running the driver at full current? I actually think I made the right choice here if I don't want to run much lower than 1A. Shouldn't it be an advantage to not have to run it at the maximum amperage; less heat / wear and tear on the components? Of course if I wanted to run it at 700 mA I'd want a 1050. Is there any efficiency advantage or disadvantage to running the driver at it's amperage limit?



I'm pretty sure he's saying he has 4 heatsinks with two COBs on each, each COB producing 50W for a total of 400W. Claiming 100W / sq ft he must be working with 4 ft x 4 ft.



What do you mean by, "better"? You mean the different wavelengths I want to supplement with?

Click to expand...

4x4=16 ft²

ttystikk said:

4x4=16 ft²

Click to expand...

Ugh sorry, as you can tell my math is pretty rusty and I'm a bit scatter brained. I meant 4 square feet, 2'x2' which seems small for so many lights but who knows!

helloseahorse! said:

Ugh sorry, as you can tell my math is pretty rusty and I'm a bit scatter brained. I meant 4 square feet, 2'x2' which seems small for so many lights but who knows!

Click to expand...

I'm trying to see what others run for lighting.

I've been running 20-30W/ft² of HID lighting for years now in my vertical setup. It's good, but I'm not getting eye popping results and I'm thinking I just may need more light.

4 x CXB3590 3500K COB chips running at fifty watts each will produce 824 PPfD over 6 ft². Four of these = 800W over 24 ft², or 33.3W/ft²...

I can't imagine multiplying that by three?

helloseahorse! said:

Thank you for your input! You are the head guru around here so I was hoping you'd chime in... I did the math and plan to run it at about the 1.05 (or actually just 1) amp. Having the driver perfectly maxed out for full efficiency is actually one reason I decided on this setup... I think I had things right in my head when I made the order, just not when I wrote the post... a lot of sleep deprivation going manic over the whole project (not just the light) combined with Metal Gear Solid V.

The math actually adds up perfect up to where I could actually get the exact resistors (71.5k/n drivers) I would need to achieve exactly 1000 mA, which also matches up with the voltage limit of the of the driver exactly even with the COBs at room temperature. I decided to go with resistors because a potentiometer wouldn't be as accurate, and a PWM signal from an Arduino, RPi or other device doesn't sound great since it's essentially just switching the power on and off at an extremely fast rate if I'm not mistaken... although I'm into Arduino and RPi so I will inevitably play with some dimming control in the future especially if I can design a circuit that switches the current through a resistor after the PWM based dimming stage (slowly turning lights on/off to reduce shock then switching to the more stable method when a constant output is needed).



Is there an advantage to running the driver at full current? I actually think I made the right choice here if I don't want to run much lower than 1A. Shouldn't it be an advantage to not have to run it at the maximum amperage; less heat / wear and tear on the components? Of course if I wanted to run it at 700 mA I'd want a 1050. Is there any efficiency advantage or disadvantage to running the driver at it's amperage limit?



I'm pretty sure he's saying he has 4 heatsinks with two COBs on each, each COB producing 50W for a total of 400W. Claiming 100W / sq ft he must be working with 4 ft x 4 ft.



What do you mean by, "better"? You mean the different wavelengths I want to supplement with?

Click to expand...

There's nothing wrong with a driver running at 100% load. But there's also nothing wrong with a driver running at smaller load as long as the voltage stays close to max voltage of the driver (as in your case).

The main point is that ELG-150-C is only $40 while HLG-185H is $60 and - in this case - offers nothing more.