DIY panel design. Compatibility/reality check? Resident experts please help!

bicit

Well-Known Member
Hi folks, I was hoping a few of the resident experts on DIY lighting would pop in and offer some advice. Inspired by picrogravs bridgelux vero 29 build I though that I would build something similar. I'm looking to grow flowers in a 2'x3', ideally I'd like to match the performance of an Area-51 XGS-190.

I'm thinking about using 3, vero 29 COB led's, wired in series using the Molex pico EZ mate connectors.
www.bridgelux.com/wp-content/uploads/2013/10/DS33-Bridgelux-Vero-29-Datasheet-2013.07.18.pdf
http://www.molex.com/pdm_docs/sd/688014225_sd.pdf

I want to mount them too an 18"(L)x4.6"(W) heat sink and mount two 80mm fans and one 120mm fan to it for cooling. One fan over each emitter, USB powered, won't be adjustable.
http://www.molex.com/pdm_docs/sd/688014225_sd.pdf

Powered by a Meanwell-HLG-185H-42. Which means each emitter should run with 38 volts at 1.7 amps or just over 55 watts of power per emitter. I could also use the HLG-240H which would allow me to run at about 75 watts per emitter. A question about power supplies now, do they just draw what's being used by the emitters? Or does a 185watt PSU draw 185 watts of power all the time?
http://www.meanwell.com/search/hlg-185h/default.htm
http://www.meanwell.com/search/hlg-240h/default.htm

I'm not going to bother with dimming, seems like an unnecessary complication. I am however looking for some suggestions on what color of emitters to get? I've been debating either getting three, 3500k emitters or two, 2700k and one 4000k. But I'm open to suggestions. I plan on experimenting with 660 and 730nm emitters as well as UV-B down the road, if that impacts anything.

What do you guy's think? Suggestions? Am I completely off base here? As it stand I'm at roughly $420 per panel after shipping from 4 suppliers. There is a negligible cost difference of like $10 for the 240watt supply. Suggestions and constructive criticism would be awesome.

ETA: I'm operating in sealed room conditions and will be supplementing CO2 at elevated levels. Should I increase the amount of light? Theses emitters can be ran quite hard, though their efficiency starts to diminish.
 

Doer

Well-Known Member
Kick me if I am wrong, but LED still seems like a scam to me. It is much., much cheaper, easier, effective to just use CFL. They are not hot. And you can get the warmer colors, 2700K, you use for bloom.

I don't get it. But I will ask you. Doesn't that seem like very dim light, compared to sunlight? And why will they not give the foot-candles? That's all that matters.

So, if you show me an LED rig that can produce Sun Power, ie, 10,000 foot candles at any distance, in any frequency of light, I will show you how you can get the same thing for much less, and more better.

Somehow these LED guys want us to believe the good light for the plant is not visible to us. BS!! These lights are not intense enough for any cost, to me.
 

caretak3r

Well-Known Member
not an expert, but I've got some 3000K vero13s and some 3500K vero18s. Between the two, i like the light from the 3500K better -- 3000K vero is more yellow than you'd expect if you're used to looking at Crees. I don't have any grows on the 3500K so i can't speak authoritatively by any means.
 

bicit

Well-Known Member
Kick me if I am wrong, but LED still seems like a scam to me. It is much., much cheaper, easier, effective to just use CFL. They are not hot. And you can get the warmer colors, 2700K, you use for bloom.
Thank you for your impute, CFL's are boring though. I don't even have any CFL's in my home, all LED or halogen.
 

Observe & Report

Well-Known Member
It won't work. When you wire up LEDs in series the forward voltages are added together. You'll either need a single high voltage driver or three separate ones each sized for a single COB. The power rating of the driver is the maximum it is capable of. Actual power consumption estimated by looking at the graph in the COB spec sheet to find out the voltage at the driver output current and LED junction temperature (which depends on your cooling solution.) Volts * Amps * driver-efficiency = power consumption. Actual tests of drives indicate that the efficiency specs are usually optimistic.
 

vostok

Well-Known Member
Wall of text...that looks like spam but it ain't spam is it ....? No its another LED builder who is even skimming on a dimmer ....lol
 

SupraSPL

Well-Known Member
Kick me if I am wrong, but LED still seems like a scam to me. It is much., much cheaper, easier, effective to just use CFL. They are not hot. And you can get the warmer colors, 2700K, you use for bloom.

So, if you show me an LED rig that can produce Sun Power, ie, 10,000 foot candles at any distance, in any frequency of light, I will show you how you can get the same thing for much less, and more better.
I am not good at detecting sarcasm, this is a joke? If not, and you are happy with CFL results, you would be floored by diy LED results. New 3000K COBs are generating 140 lumens/watt. I run my 4200K XML2 at 161 lumens/watt. Fluoro tubes can reach 93 lumens/watt creating much more heat and they suffer lumen depreciation and shine in every direction relying heavily on reflectors.

6800 lumens minimum, 46W
IMG_0041a.jpg IMG_0042a.jpg DSC06692b.jpg

IMG_9906a.jpg DSC06698a CnK LED.jpg DSC06744a 16.jpg
 

SupraSPL

Well-Known Member
Op your best best bet is a simple constant current driver. You can choose from 650mA for $6, 1000mA for $11 or 1600mA for $11 free shipping. The softer you run it the more efficient it will be but even at 1600mA there is nothing like it.
 

bicit

Well-Known Member
It won't work. When you wire up LEDs in series the forward voltages are added together. You'll either need a single high voltage driver or three separate ones each sized for a single COB. The power rating of the driver is the maximum it is capable of. Actual power consumption estimated by looking at the graph in the COB spec sheet to find out the voltage at the driver output current and LED junction temperature (which depends on your cooling solution.) Volts * Amps * driver-efficiency = power consumption. Actual tests of drives indicate that the efficiency specs are usually optimistic.
Indeed I seem to have gotten the idea backwards in my head somehow. Back to the drawing board.
 

Scotch089

Well-Known Member
Boring is not the word I would use for CFLs,

Outdated. Old. Eletrictly savvy over 23w, no.. the larger the bulb the less and less efficient you get..

Like bicit stated some people don't even run them in their household fixtures anymore..

If my are going to hop onto a DIY thread, and state footcandles matters... There are some links out there you should really read first.. good luck OP
 

bicit

Well-Known Member
Well it looks like the Meanwell HLG-185H-C would be a good fit, if I'm doing my math right it should be more than adequate to power a string of 3 vero 29's at 1.4 amps. Only problem is that I cannot find any one who stocks them. I wanted to avoid running individual drivers for cost/simplicity reasons. May not be able too.

ETA: I am able to find the HVGC series in stock. link The HVGC-150-1400B seems like it would work for a string of 3 vero 29's. What do you guy's think?
 

Gaius

Active Member
Well it looks like the Meanwell HLG-185H-C would be a good fit, if I'm doing my math right it should be more than adequate to power a string of 3 vero 29's at 1.4 amps. Only problem is that I cannot find any one who stocks them. I wanted to avoid running individual drivers for cost/simplicity reasons. May not be able too.

Any suggestions on a power supply(s)?
It's far cheaper right now to buy 1 driver per chip for these high-wattage COBs. I just compared a wide variety of options, and all the single-driver solutions were tremendously more expensive, and usually not even stocked.
 

SupraSPL

Well-Known Member
Cheaper, more efficient and safer, long strings of LEDs in series create high vF which can overcome the resistance of your skin. This is just one example I am sure there are other similar drivers. $11 1.6A 88% efficient. Personally, I would not run them that hard though, $11 1A 88%.
 

Gaius

Active Member
The 1.4A ones I got a decent compromise? Supra, you previously mentioned the 1.4A drivers I bought were 36% efficient with the 3070s, so where does that stand up against the 1A for example?
 

bicit

Well-Known Member
Cheaper, more efficient and safer, long strings of LEDs in series create high vF which can overcome the resistance of your skin. This is just one example I am sure there are other similar drivers. $11 1.6A 88% efficient. Personally, I would not run them that hard though, $11 1A 88%.
Those look very nice indeed! Just out of curiosity, why do you recommend running them at a lower current?
 

SupraSPL

Well-Known Member
The lower you run them the more efficient they will run (current droop) but the trade off is a higher cost up front for the COBs and a lower cost for drivers, heatsinks, electricity.

Here is how they perform at each current. The first column is the drive current and the last gray column is the resulting efficiency (assuming you can maintain a junction temp of 50C). The efficiency is referring to the conversion of electricity into heat and light. For comparison a brand new HPS bulb is about 36% efficient.
CXA3070.png

Gaius you are correct 36.25% at 1.4A. I am not saying there is anything wrong with that at all, badass performance and a good compromise. I run softer just because of my personal preference as a stickler for max efficiency and somewhat experimental.
 

bicit

Well-Known Member
The lower you run them the more efficient they will run (current droop) but the trade off is a higher cost up front for the COBs and a lower cost for drivers, heatsinks, electricity.

Here is how they perform at each current. The first column is the drive current and the last gray column is the resulting efficiency (assuming you can maintain a junction temp of 50C). The efficiency is referring to the conversion of electricity into heat and light. For comparison a brand new HPS bulb is about 36% efficient.
View attachment 2993712

Gaius you are correct 36.25% at 1.4A. I am not saying there is anything wrong with that at all, badass performance and a good compromise. I run softer just because of my personal preference as a stickler for max efficiency and somewhat experimental.
So far I like what I see. I need to spend sometime on excel and play with some numbers. What do you think about running 5 vero 29's with the 1 amp drivers that you linked? Those are AC-DC correct?
 

SupraSPL

Well-Known Member
Here is how the 3000K Veros stack up. The CXA3070 slightly outperforms them in terms of cost and efficiency but the Vero might currently be available in more color temps. Yes those drivers would work well with Vero29.

Vero 18 Vero 29.png
 

Doer

Well-Known Member
I am not good at detecting sarcasm, this is a joke? If not, and you are happy with CFL results, you would be floored by diy LED results. New 3000K COBs are generating 140 lumens/watt. I run my 4200K XML2 at 161 lumens/watt. Fluoro tubes can reach 93 lumens/watt creating much more heat and they suffer lumen depreciation and shine in every direction relying heavily on reflectors.

6800 lumens minimum, 46W
View attachment 2993461 View attachment 2993466 View attachment 2993489

View attachment 2993453 View attachment 2993487 View attachment 2993488
Lumens don't matter to me. That is the measure for comparing outputs. What counts is foot candles. Photon density. As I understand this, they are similar but different for for the surface delivery of energy for photosynthesis.

Foot-Candles vs. Lumens -

A "foot-candle" is a unit of measure for quantifying the intensity of light falling on an object. A "lumen" is a unit of measure for quantifying the amount of light energy emitted by a light source. In other words, foot-candles measure the brightness of the light at the illuminated object, while lumens measure the power of the light radiated by the light source.

Both foot-candles and lumens are units measurements of electromagnetic radiation detectable by the human eye. Foot-candles are used to measure "luminance," the density of light energy reaching a reference surface at a given distance from one or more light sources. Lumens are used to measure "luminous flux," the amount and rate of light energy radiated by a particular light source.

A foot-candle is based on the English system of measurements. A foot-candle equals 1 lumen per square foot. The international standard (SI) counterpart of the foot-candle is the "lux." A lux equals 1 lumen per square meter. The equation used to convert foot-candles to lux is: 1 foot-candle = 10.76 lux.

http://www.dlulighting.com/tech_data_step2.php?pcat_id=18
----------------
So, one measures the energy as leave the emitter, lumen and lux, foot-candles the measurement of the 2D density at the receiver surface.

I'm not interested in the emitter, (agnostic on that,) I'm interested only in the energy density at the receiver end, for photosynthesis. That is why watt is yeild.

We know that Sun Power max in the summer is 10,000 FC at sea level. At the landrace levels, say 3000 meters, it is 16,000 FC.

The sun is about 93 lux.

It takes 10.76 Lux to make one foot candle. (1 lumen per square meter.) So, this mean you need 10,000 x 10.76 = 107,600 lumen = Sun Power.

You are dissipating 46 watts to produce 46 x 150(avg) = 6900 lumen. You are 2 orders of magnitude below Sun Power. So, you need in that set up 2 orders of magnitude more watts or 4600 watts.

I can get Sun Power from my 1000w HPS at 26". And it doesn't cost me 4.6kWh to get it.

I am just trying help. The math and the logic is quite simple.

Why would you grow with LED? Because you don't have the space or want less heat. But, no one can tell me it grows more pot per sq foot.

Not enough foot candles. So, LEDs are efficient, but to me, not as effective for the money. BTW, I don't feel the need to indulge in sarcasm. Count on it. :)

 
Top