Needing some help with designing my led light.

taproot

Well-Known Member
I read a lot last year from supra and others on leds and understand the basic concepts like running the cobs in series and looking at the voltage of the cobs and the amount you want to run while adding their voltages all up into one sum to make sure the driver has enough forward voltage for them all and enough amps to meet their minimum requirements or push them harder if the driver can do that. I've been using hps but it's just too fucking hot and I need to scale things down anyways and will be growing in a custom cab or a grow tent. The space I'll be using for now will be a 4ft wide x 2ft deep x 5 ft high shelf so I need enough for that area but if possible I'd like to design a system that could scale up a little to say a 4x4. From what I've read if you size the driver right you can run the cobs dimmed and then turn them up if needed should the driver support that like some of the mean wells. I'm not sure if the same old 50 watts per square foot applies to leds but I'd figure a 4x2 would need around 300 watts at least and more likely a 400 watt light would work well as my 400 hps did. But later down the line if I went to a 4x4 I'd like to just turn up the driver and get the amount of light I need. Last year when I read a lot on this here it seemed the vero 29's were the way to go and the color temp of 3500 were the ones to get. I've found with hps the plants vegg and flower just fine with the same bulb so would 3500's work for both? The vero's are in gen 7 so that's what I figured I'd buy once digi key carries the 3500 if those are still recommended. As for the heat sink I remember a lot of folks using long pieces of extruded aluminum from heat sinks usa but now I've see it recommended on some youtube vids that the pin heat sinks are better so I'm not sure what to use here. So, to sum things up I was planning on running vero29/s in series with some sort of passive heat sink and hopefully a single driver that can dim and then possibly handle a few more cobs latter for a 4x4 area. I've never used leds so I'm not sure how many watts or cobs I might need for now in my 4x2 and latter a 4x4 that can compete with a 400 or 600 watt hps. I understand leds are more efficient in their operation so I'd assume comparing watt to watt between the two technologies might not be a good comparison. I've also read that having more cobs but running them about half of their total potential is more efficient. So for example if I needed 200 watts and each cob can run at 100 watts at their full potential it would be better to get four of them and run them at half power is this correct.

1. How many cobs are recommended for my 4x2 space and how many total watts?
2. Are those reflectors recommended ..have they been to be worth their price as compared to running the cobs without them?
3. Heat sinks, large ones from heat sink usa or the pin style ones?
4. Drivers....one that will work for my 4x2 and latter my 4x4 if possible.

I found these drivers..somebody was saying they are pretty good.

https://www.cutter.com.au/proddetail.php?prod=cut2999
https://www.cutter.com.au/proddetail.php?prod=cut2893

Thanks in advance!
 

GrowLightResearch

Well-Known Member
A good 400W HPS will generate 50,000 lumens with a CCT of 2100K. Divide 50,000 by the CoB's lumen output and that's how many you will need. It's all about the lumens per watt.
The more CoBs you have the better the uniformity and the closer to the canopy you can get. If you reduce the height of the fixture above the canopy from HPS height of 1 meter down to 18", you will get 4x increase in flux per m². You have to compare lumens per $ and balance it with uniformity. If you have a 3000 lm CoB vs three 1000lm CoBs at the same price I would consider the 3 CoBs due to the improved uniformity and simplified thermal management.

If you want to match the HPS you will need to lower the CCT, supplement with Deep Red (650nm) LEDs, and or raise the CRI.

LEDs are usually most efficient at the rated current. If you push the power up then you will run hotter, reducing the efficiency.

The power supply is fair deal. You have to calculate the cost per watt and compare it with the alternatives. I like the 7 year warranty of the Mean Well HLG series. The PF-P400 supply has a better MTBF than the Mean Well. Generally you get a the best cost per watt with supplies in the 200 Watt range.
 
Last edited:

Stone_Free

Well-Known Member
@taproot I don't know if you've considered LED strips at all? The Samsung F Series use the same diodes as the Quantum Boards - LM561 and are very efficient, especially if you use a ScrOG as your growing style.
Also I came across a really good PAR/PPF/DLI tool which looks to be excellent as it is backed by proper science.
I can't post links atm so have attached a .txt file with the link for the tool and also a link to a Reddit post which helps with DLI info.
The beam angle for the LM561 diodes is 120 degrees which you'll need to enter into the tool.
Hope this helps dude!
:peace:
 

Attachments

GrowLightResearch

Well-Known Member
Quantum Boards - LM561 and are very efficient,
What I do not like about the LM561 is they are medium power ( more real estate and manufacturing cost. I would never use them because they do not have an electrically isolated thermal pad.

I do like the Cree XP-3G which are the most efficient high power LED available today.

For an easy DIY Opulent Americas sells a 6" x 2" metal core PCB with 12 XP-3Gs for about $17.

SEE: https://www.rollitup.org/t/diy-inexpensive-cree-xp-3g-led-hps-equivalent.951740/#post-13854619
 

Stone_Free

Well-Known Member
What I do not like about the LM561 is they are medium power ( more real estate and manufacturing cost. I would never use them because they do not have an electrically isolated thermal pad.

I do like the Cree XP-3G which are the most efficient high power LED available today.

For an easy DIY Opulent Americas sells a 6" x 2" metal core PCB with 12 XP-3Gs for about $17.

SEE: https://www.rollitup.org/t/diy-inexpensive-cree-xp-3g-led-hps-equivalent.951740/#post-13854619
Fair enough dude, your grow, your rules. I have to point out though, the LM561 diodes are killing it when it comes to growing weed. Samsung FB24B strips are the nutz!
:peace:
 

taproot

Well-Known Member
Thanks to both of you for your input. I've never even heard of these strips so I'll have to look into them, what are the pros to them vs the cobs? Can you get the same power output and efficiency with them...what I see off hand is they spread the light out more than a cob but I'm concerned about there ability to have enough penetration power...but again I'll need to look into them. Are you using them, pm me some links to some grows that are killing it with them so I can read about it. Thank again for both of you taking the time to provide feedback.
 

Stone_Free

Well-Known Member
The strips are very efficient. You have Bridgelux and Samsung to choose from. Both are very good. Samsungs are more $'s but also a bit more efficient. In terms of penetration you can mitigate this completely by growing in a ScrOG/LST style which gives a very even canopy that isn't too deep and therefore penetration is not an issue and you will use less energy to grow in this way; it's more efficient than a few tall plants where the light doesn't penetrate unless you turn the power up. If you aim for a PPFD of 800 you can get that at the top but not 2 feet down into the plant unless you wack up the power!
People often talk about efficiency of the lights but not the grow style, which is important!
If you do a search for led strips or bridgelux or samsung strips you'll see the threads. Hop on over to @Rider509 or @graying.geek threads - those are just the 2 off the top of my head, there are a few more guys killing it with these strips!
If you have a 4x2 ft. space, you use 2 x Samsung FB24B strips with a ScrOG you'll absolutely kill it. You can get those strips 2 inches from the canopy and your plants will fucking love it!!
My set up which I'm buying in a week or so will cost me £130/$180 for 215W of 3000K and you'll be mega happy with the results, I'm sure!
:peace: :bigjoint:
 

Attachments

nfhiggs

Well-Known Member
A good 400W HPS will generate 50,000 lumens with a CCT of 2100K. Divide 50,000 by the CoB's lumen output and that's how many you will need. It's all about the lumens per watt
Actually its all about uMole/s per joule. You can't directly compare very different light sources by their lumen output, especially differing CCTs and CRI's, it just doesn't work that way. You need to first convert the lux to PAR, and the conversion factor is different for different light sources and spectrums.

If you reduce the height of the fixture above the canopy from HPS height of 1 meter down to 18", you will get 4x increase in flux per m²
This assumes a fall off rate corresponding to the inverse square law, which only works for point sources radiating 360 degrees. Once you have multiple emitters spread over a larger area, with the light directed in less than 180 degrees and with reflective walls, the inverse square law simply does not apply. Its impossible to say what the increase will be (too many unknown variables) when you halve the distance with a multiple cob fixture, but it will most certainly be less than 4x.

I looked at the Cree spec sheet and TBH, I'm not really impressed. A 6W diode with the efficiency spec'd at a mere 15% of its max current rating? If they spec'd them at a more realistic 30-40% of max current, they're not nearly as impressive.

After flowering my last plant under strip lights, High power LED's and COB's are not what I'm looking for anymore. Its maximum light distribution I'm after.
 

graying.geek

Well-Known Member
If you have a 2' x 4' grow space, check page 4 of my signature thread; you'll find a parts list and build instructions for my EB-series build. 400W+ total and 750 umol/sqft at 4", all for $250 US. Seriously, if you're comfortable using a screwdriver and cordless drill, this is the way to go.
 

nfhiggs

Well-Known Member
Thanks to both of you for your input. I've never even heard of these strips so I'll have to look into them, what are the pros to them vs the cobs? Can you get the same power output and efficiency with them...what I see off hand is they spread the light out more than a cob but I'm concerned about there ability to have enough penetration power...but again I'll need to look into them. Are you using them, pm me some links to some grows that are killing it with them so I can read about it. Thank again for both of you taking the time to provide feedback.
The "penetration power" of brighter emitters is a myth. Here is an example of light intensity measurements taken under a 2x4 ft Bridgelux strip light fixture at various distances:

Inches LUX
===== ===
4 51000
8 48200
12 45000
16 43000
20 39900
24 30200

As you can see, from 4 inches to 24 inches there is less than 50% fall off. Not only that, but the spread was incredibly uniform, varying only about 20% across the 2x4 grid.
 

nfhiggs

Well-Known Member
If you have a 2' x 4' grow space, check page 4 of my signature thread; you'll find a parts list and build instructions for my EB-series build. 400W+ total and 750 umol/sqft at 4", all for $250 US. Seriously, if you're comfortable using a screwdriver and cordless drill, this is the way to go.
Hope you don't mind me reposting your intensity numbers.
 

GrowLightResearch

Well-Known Member
, the LM561 diodes are killing it when it comes to growing weed
Bottom line I believe you know what you are doing more than most growers.

There are many LEDs that are "killing it". It's about getting the most photons to the leaves per kWatt. That requires significant thermal management.

I look at this a low volume manufacturer rather than buying off the shelf strips. The XP-3G works better for me than LM561s. I'm doing between 48 and 192 high power LEDs (Cree XP or OSRAM Olson) per 48" strip.

I was not familiar with the F-Series Gen3 boards. After taking a closer look, I may have to go along with you and forget the Opulent Americas boards. The Samsung F-Series Gen3 appears to be a better value than the Opulent Americas board. I like the four foot length. That is rare and can only be pulled off by someone with extensive manufacturing capabilities. Just buying PCB panels of that size is not easy. I have to go with 3 x 16" strips to get four foot fixtures as the most economical PCB panel size is 16" x 20". No low volume pick and place robot can assemble 48" boards.

2 inches from the canopy
That I like. At 2" you do not need that much radiant flux. You do need exceptional uniformity to get that close.

£130/$180 for 215W of 3000K
215 watts means nothing unless it is specified as radiant watts per area which can be mathematically converted µmol/s/m². Furthermore you need to specify BOTH the wall watts AND radiant watts to say anything meaningful.

If you are referring to 215 W of the FB24B strips, 215w makes even less sense as they are 43V with no specific wattage.

I'd go with the F562B with the single row of LEDs. The uniformity will be much better with the rows spread evenly and the difference in cost between one FB24B and two FB22Bs is minimal. Thermal management will be simplified with less heat per sq. inch.

If you aim for a PPFD of 800
All photons travel at the speed of light no matter their source. 800 µmol/s/m² at the canopy is the same no matter the height of the source, it's wattage, or method the photons were generated. Penetration is the same. What does matter is the uniformity across the entire canopy.

For off the shelf strips I will agree with you, the F-Seies Gen3 will be difficult to beat. I prefer strips to CoBs as well. CoBs uniformity is inferior to strips.
 

GrowLightResearch

Well-Known Member
Actually its all about uMole/s per joule.
I totally agree with you but the needed information, to do it accurately, is not available nor practical for a forum post. For a guesstimate where the comparison HPS output range is 400-600 watts, it will suffice. I do not expect the OP has the background necessary to calculate an accurate comparison.

Although the differences we are splitting hairs over is insignificant. 40 years ago I was growing 5 plants under four 4' Grow-Lux florescent tubes. They did just fine.

the inverse square law simply does not apply.
Yes and No. I agree with you here too. What I said was very true, just not all that applicable to many situations. While the flux increases the coverage area is reduced. But is is true reducing the height from 39" to 18 inch will give more than a 4x increase in flux for a given area (e.g. m²). That's why it's called a Law. How it is applied can get very complex with multiple sources. It fairly simple to apply to a single HPS.

I looked at the Cree spec sheet and TBH, I'm not really impressed
An LED will always be spec'd at the most advantageous current. That works well to fool most engineers.

Bottom line it is, today it is THE diode with the highest efficacy at about 80% for ideal conditions. You cannot just look at lumens and lm/watt. It is actually easier to look at the blue LED under the phosphor. When comparing LEDs the forward voltage, and flux output at a specific current and temperature of both LEDs is required. A cursory peek at the datasheet is most often misleading.

Medium vs. High power is mostly a manufacturing preference. Less parts is less assembly cost. I am populating my strips with between 48 and 192 high power LEDs per 4' strip.

Here is an example of light intensity measurements taken under a 2x4 ft Bridgelux strip light fixture at various distances:
There is only one precise ideal height for any given fixture. That is the height with the best uniformity. No other height matters. Your measurements are very misleading. Those measurements apply only to that particular fixture and are meaningless. Also the method used to do the measurements is a huge factor. Measuring at a single point with multiple light sources is not very meaningful. This is about as misleading as my statement on the 4x increase per m².

Its maximum light distribution I'm after.
That is the one of the most intelligent ever said on this site. I call it maximum uniformity. And I want the fixture close to the canopy with maximum uniformity. I also want the best PPFD per Joule ratio as well.
 

Randomblame

Well-Known Member
Yeah, LM561c S6 bin is hard to beat currently.
3000°k diodes at 80mA is 180lm/w according to samsung mid-power LED calculator.
http://www.samsung.com/global/business/led/support/tools/calculator-pkg

LER is 321,6lm/w this corresponds to an efficiency of over 58% at 80mA.
QER is 4,86μMol/j and this means they deliver up to 2,81μMol/w ppf at 80mA.
Don't know how low a XP-G3 needs to be driven to reach that point and what it costs compared to these diodes.
I've read that Cree have concluded an alliance with another company and they are also planning to develop an efficient mid-power diode. But so far nothing concrete yet from this direction.

*LER and QER was cherry-picked from HLG website, thanks HLG-Team!
 

Chip Green

Well-Known Member
If you have a 2' x 4' grow space, check page 4 of my signature thread; you'll find a parts list and build instructions for my EB-series build. 400W+ total and 750 umol/sqft at 4", all for $250 US. Seriously, if you're comfortable using a screwdriver and cordless drill, this is the way to go.
Read this thread ^^^
I elected to "cheap out" and go Bridgelux EB and have never regretted my choice. Not once. Not one time.....excellent option for "entry level" builders.
Maybe I'm missing out on greatness, but I care not.
I was a BRAND new LED user, let alone DIYer. In 6 months I went from not knowing what a COB was to building quality, photon sufficient lighting for bargain basement cost.
 
Last edited:

taproot

Well-Known Member
I've read a few of these threads by folks like rider509 and graying.geek regarding the led strips and personally I think that it's better for me and my application over the traditional cobs. I like the fact that the light is more easily spread out by default vs cobs, you don't need reflectors, you don't need traditional expensive heatsinks, they run cooler which is a huge plus and what heat is generated is spread out instead of hot spots. I also like the fact I can get the lights closer to the canopy since they run cooler which also increases efficiency for what we're using them for. Thanks stone_free for the introduction to these strips!! I've looked at both the Samsung F series and the Bridgelux EB series. I can see the Sammy F series are more efficient but If I'm looking at the spec charts right the lumens per watt between them and the EB series doesn't seem that much at not least to justify a double price tag. I've used a few calculators to play around with the volts/amps that are in the spec charts for these two lights and what I can see is that the Samsungs do offer the ability to produce more at a higher amp than the EB series can at there max amps...the sammy's can be driven a little harder. But comparing the different loads they can handle and imputing the volts/amps into the calculator I doesn't seem to be such a huge factor to pay double. Are the F series just that much better as I see lots of people really like them, am I missing something about them over the EB? I want a good light and don't mind paying for them but they just seem to not be worth double which seems to be the conclusions of a few others. With the EB series I seem to be able to get close to what I wanted, here's what I've been looking at...I'm just posting everything that I was considering for conversation. What I'm considering is 8 of the below EB series @3000K for my 4Wx2D space. I already have one HLG-185H-C1050A driver that never got used and it can drive four strips all the way to 1A. With another driver and the other four strips, according to the calcs, I should be able to get close to the 400 watts I wanted. Each strip running at 1040mA / 45.3V gives=47.112 watts. That's 377 watt potential which as pointed out to me with the proper growing techniques should be enough. If I go this route it looks pretty close to greying geeks which seems to be doing awesome with his build. I really liked the " LT-FB24B" dual row leds but dam they are so expensive and I figured I'd get better spread with more strips so once again the EB series seems to shine on price point. Does this sound on point or are there better options I missing?

1. https://www.digikey.com/product-detail/en/bridgelux/BXEB-L1120Z-30E4000-C-A3/976-1502-ND/6236288

2. https://www.digikey.com/product-detail/en/samsung-semiconductor-inc/SI-B8V521B20WW/1510-2225-ND/6624009?WT.srch=1&gclid=CjwKCAjwssvPBRBBEiwASFoVd7ow9fYnB1gKVyjaAupjFtFBVn_vaiyhhlbVFP0Mu9eJ1Jb0KUyiqRoCAYgQAvD_BwE

3. https://www.digikey.com/product-detail/en/samsung-semiconductor-inc/SI-B8VZ91B20WW/1510-2229-ND/6676693?WT.srch=1&gclid=CjwKCAjwssvPBRBBEiwASFoVdzMTK4JZ0ks5jWXU2sEN0LbKRlrziDlne2MeYVAfRNPNA2T-5ANYNxoCw0UQAvD_BwE

Here are the tech sheets I found on them.

http://www.bridgelux.com/sites/default/files/resource_media/DS130 EB Series Datasheet Rev A_0.pdf

http://www.samsung.com/global/business/business-images/led/file/product/products/201704/Data_Sheet_F_Series_G3_Rev.1.0.pdf

Here are the calculators I used for the basics.

https://www.google.com/search?q=convert+milliamps+to+amps&oq=convert+mill&aqs=chrome.1.69i57j0l5.5663j0j7&sourceid=chrome&ie=UTF-8

https://www.supercircuits.com/resources/tools/volts-watts-amps-converter
 

graying.geek

Well-Known Member
<snip>I've looked at both the Samsung F series and the Bridgelux EB series.
Folks are attracted to the F-series due primarily to their enthusiasm for higher efficiency, and I've seen cocktail napkin calculations that suggest the higher efficiency will pay off the difference in cost in something like 2 years (I'm sure someone will correct me if I'm misremembering), hypothetically. There are also more subtle differences in efficiencies at lower power, etc. Still, you won't be surprised to hear me say that I believe you're making a wise choice. 8-)

: said:
here's what I've been looking at...I'm just posting everything that I was considering for conversation. What I'm considering is 8 of the below EB series @3000K for my 4Wx2D space. <snip>
Yup, other than using the 3000K (I went with the 4000's for veg and flower) you're ending up with the same general setup as mine. Trust me, you won't be disappointed.
 

CobKits

Well-Known Member
so I need enough for that area but if possible I'd like to design a system that could scale up a little to say a 4x4. From what I've read if you size the driver right you can run the cobs dimmed and then turn them up if needed should the driver support that like some of the mean wells.
constant voltage drivers in parallel might be better suited for you

do some reading and ask some questions
 
Top