Fanleaf's Huge 42 COB Array Build Plus Other Builds

jimmyclone said:
There is a well known member on riu that did tests on the cxb3590's and found that there kelvin rating is actually higher than what is stated.. There is a huge thread on it.. He tests multiple chips and the consensus was that a cree cxb3590 3000k/80 tested out to 3400k almost every time.. Yes there were variances between chips but that was the average over a lot of chips.. And he tested 4000&5000's as well.. All seem to be higher in the blues.. His results led him to believe that this was how cree was getting higher lumen per watt ratios.. Please dont shoot the messanger... Now im not saying cree dosent make good products, i have cree light bars all over my truck.. But its a very interesting read, i think its over a 1000 pages.. There are some cob and qb guru's on that site that make my head spin.... And im an electrician by trade...
Cheers
Click to expand...

Do you know how he tested that theory? There's really only two ways to test luminosity over colour temperature. First you need the spectrographic data. This is the "white" light being divided into it coloured spectrum (with a prism) and each individual wavelength being measured on a photometer. So split it up and then measure each colour essentially. I'm an engineer (BioTech not electrical) and that sounds daunting and expensive even for me. So I'm curious how a guy on ROI has access to this kind of tech? Not that I dispute your claim.
But that being said I know for a fact that somewhere in china right now is a guy sitting in front of a machine that measures each SMD for Luminous output as it comes of the line (This is before the SMD's are integrated into COBs.). That guy is "binning" the chips. There's obviously a human error factor not to mention SMD technology isn't exactly 100% acculturate to begin with...otherwise there would be no need to "Bin" them. Correct? But each tiny little SMD is tested by hand at the Cree factory. (Which probably accounts for the higher cost of Cree COBs)
So though I am somewhat dubious of this claim I believe it is possible that every COB off the line is not marginally the same. But a colour temperature of that wide a variance makes me want to know the science involved and see this experiment performed by an unbiased 3rd party.
 
To be quite honest I hesitated to even come back to this thread " my thread" because of so much of the " professional" advice given the last 15-20 pages. To those who know no better some of this advice seems genius. To others they just shake their head and move on.
I'm in no way a know it all. But what I do know I share. Like 2 months ago when you didn't know how to calculate what Meanwell driver you needed for a bunch of different setups Jimmy. I must have calculated 15 different drivers out for you right? Every time you asked me "what driver would I need for this"? I did it. Then 1 month later you argue me when all the facts are on my side and I prove it by showing it.
Then after multiple failures of wiring a simple switch to run 2 cobs/ 4 cobs at the simple flick of a switch I draw it up for you and once again it works flawlessly.
Now, 2 more months and you have "ParHorta Horticultural Lighting Company"? Are you shitting me bro?
I see almost nothing but inefficient advice given. That is the exact opposite reason for my whole thread.
If people want to use as much or MORE power than hps then maybe that should be its own thread not this one.
Im not slamming Bridgelux or anything of the sort. I like them and they have their place. What I am slamming is companies now selling cob lights that advertise how much better their lights are and how efficient they are. Thats a bunch of bullshit. If a company wants to really sell "efficient" lights it wont be 2,4,6 chip units or tell someone they only need 10 cobs in a 5x5 area to smoke a hps in efficiency. Nope, because they want to sell their lights and the way they explain their lights to be soooo efficient a customer would need to put 3 times the chips in there to see the efficiency advertised.
Its all a bunch of lies from EVERYONE who sells cob arrays and I wont be a part of it. Instead, I build my own, put the extra money up to do it the way it really beats hps and offer the advice.
90% cant afford my advice and end up skimping. Thats fine, I dont care but I am in this to save $$ in the long run where the savings will be far greater in the end.
You get what you pay for plain and simple.
 
fanleaf said:
Lets not forget Jimmy that I'm the one who found and sent you a link to the 1750k vero. After that we went over it together and found it interesting.
The whole thing is photons guys. The purpose of most cob builds is to save power to end up with the same result. If you think you can do that by using chips that provide less photons per watt than hps you have some learning to do.

Will the 1750k chip grow plants? Hell yes it will. Will it do it efficiently? Hell no it wont. A plant uses photons to grow and when a light source puts out half the photons at the same power as another source that makes it inefficient.
Cannabis is proven to adapt to whatever spectrum is provided to it to grow within reason. Thats why some growers go start to finish under a 6500k and some with great results.
It always goes back to this.....we spend a bunch more money going cob for a reason. If not for efficiency then why not just go hps or blurple LED's?
A chip producing 180+ lumens per watt inside the par spectrum with a decent spectrum will outgrow a chip that produces 90 lumens per watt every time.
Click to expand...

I got 1200 micromoles at 8" on a Vero 18 (half the power of the 29). So the photons are coming from somewhere. Let's not forget that lumens are for humans and red is lowest energy photon in the bunch. But that being said 83 lumens per watt is a disgusting number. Lol
So how does a chip that only has 83 Lm/watt get over 1200 umols? Especially a tiny Vero 18?
I don't believe there is a reliable correlation between lumens and PAR. Plants don't see lumens.
That's like saying infrared is useless because it gets zero lumens per watt or ultraviolet is useless because it gets zero lumens per watt. The human visible light spectrum and the plants photosynthetic spectrum are two entirely different things.
 
fanleaf said:
Lets not forget Jimmy that I'm the one who found and sent you a link to the 1750k vero. After that we went over it together and found it interesting.
The whole thing is photons guys. The purpose of most cob builds is to save power to end up with the same result. If you think you can do that by using chips that provide less photons per watt than hps you have some learning to do.

Will the 1750k chip grow plants? Hell yes it will. Will it do it efficiently? Hell no it wont. A plant uses photons to grow and when a light source puts out half the photons at the same power as another source that makes it inefficient.
Cannabis is proven to adapt to whatever spectrum is provided to it to grow within reason. Thats why some growers go start to finish under a 6500k and some with great results.
It always goes back to this.....we spend a bunch more money going cob for a reason. If not for efficiency then why not just go hps or blurple LED's?
A chip producing 180+ lumens per watt inside the par spectrum with a decent spectrum will outgrow a chip that produces 90 lumens per watt every time.
Click to expand...

Also I'm not just using the 1750k. That would be like using a mars. Lol. I got 200w of 1750 over 6 V18's and 200w of 4000k on the 29's.
The 1750 is for sunrise sunset simulation and to give my 29's a boost in the reds for flower.
 
JimmyJames905 said:
I got 1200 micromoles at 8" on a Vero 18 (half the power of the 29). So the photons are coming from somewhere. Let's not forget that lumens are for humans and red is lowest energy photon in the bunch. But that being said 83 lumens per watt is a disgusting number. Lol
So how does a chip that only has 83 Lm/watt get over 1200 umols? Especially a tiny Vero 18?
I don't believe there is a reliable correlation between lumens and PAR. Plants don't see lumens.
That's like saying infrared is useless because it gets zero lumens per watt or ultraviolet is useless because it gets zero lumens per watt. The human visible light spectrum and the plants photosynthetic spectrum are two entirely different things.
Click to expand...
Jimmy, if lumens provided by a light source are within the par spectrum then there is a very direct correlation. Lumens that are made from a 630nm source will provide par right? There is the correlation. And for that matter if you seen no correlation between lumens and photons or par then why does your website advertise almost nothing but lumens?
 
JimmyJames905 said:
I got 1200 micromoles at 8" on a Vero 18 (half the power of the 29). So the photons are coming from somewhere.
So how does a chip that only has 83 Lm/watt get over 1200 umols? Especially a tiny Vero 18?.
Click to expand...
Well, a cheap meter is how. When testing a tiny cob where all the light comes from 1 tiny source and only at 8" is exactly how. A bigger chip sprays more light from a wider area and at wider angles. Hell, I can hit rediculous umol numbers with a 3w led flashlight if I put it right to the sensor on my Apogee.
 
Fanleaf's Huge 42 Cree CXB3590 COB Array Build Plus Other Builds

You misunderstood the point. How does a 83 lm/w chip get 1200? You made a direct connection between luminescence and PAR so how is it possible I get such a high reading from such a poorly Luminous chip. And if you watch the video I get 500μMols everywhere else, not just under the cob.
And do we have to resort to slandering each other's equipment? Like really Fanleaf?
 
Photons photons photons. The more photons you can produce evenly over an area with the least power from the wall is what its all about. Spectrum is secondary to that and does play a role but "a role".
Why build a light that uses as much power as hps and costs 3 times more?
Nobody seems to ever answer that.
 
fanleaf said:
Photons photons photons. The more photons you can produce evenly over an area with the least power from the wall is what its all about. Spectrum is secondary to that and does play a role but "a role".
Why build a light that uses as much power as hps and costs 3 times more?
Nobody seems to ever answer that.
Click to expand...

So all green photons will grow a plant right through flower correct? As you say spectrum is secondary.
 
JimmyJames905 said:
You misunderstood the point. How does a 83 lm/w chip get 1200? You made a direct connection between luminescence and PAR so how is it possible I get such a high reading from such a poorly Luminous chip. And if you watch the video I get 500μMols everywhere else, not just under the cob.
And do we have to resort to slandering each other's equipment? Like really Fanleaf?
Click to expand...
Im not slandering anything man. Stating nothing but the truth. Nothing more, nothing less. If anyone can prove what I said in the last few posts inaccurate let me know and I will retract it.
Like I said man, I can get crazy readings by beaming a flashlight into the sensor. Does that mean I could grow really well with a flashlight? Nope. Simply because theres a much bigger part of the picture to look at.
 
JimmyJames905 said:
So all green photons will grow a plant right through flower correct? As you say spectrum is secondary.
Click to expand...
I said it plays a role but is secondary Jimmy. Green photons will infact grow a plant. There is 1nm that wont.
Contrary to popular belief plants do use green photons. They wont be triggered to flower under them though.
 
Bottom line is I see a guy who didn't know nothing about nothing a few months ago but now owns a lighting company. Is that sladerous? No, its curious because the same guy hasnt had time to run even a few cycles under its own lights but now knows whats up.
Just start a thread man.. Offer advice about all you want but when in my thread I will call out faults I see.
90% of people here will go a cheap route and run less cobs at higher power so your thread would be busy! But those 10% who want to smoke a double ended hps by building a light to do so will stay here. We can smoke hps with Cree,Bridgelux or Citi led's but you will not do it with 1 chip per every 3 or 4 square feet.
 
Lok man. People can build amazing lights with vero and citi so thats not my bitch. My bitch is when people spend $800 on a 600w cob array thinking it will out perform and put less heat in their room than a 600w hps and most of the time they will get dissapointed. Thats where my irritation comes from.
 
I really didnt mean to snap and Im sorry for that, it just ended up that way for me. I just get upset when I see so many people thinking they will build a light to save power and heat and then they go and buy either a premade light or build it in a way that they will save nothing, spend more upfront and create the same heat.
 
Dude, there are great Bridgelux and Citi chips out there and awesome arrays can be made with them too. Hell, I use some! But people will not build the same caliber light with 1212's and such.
They will also not build a light of the same caliber by running 1 chip per every 2-3 feet. People can go cheaper than Cree with Veros and build the same array as my Cree and get amazing results but they wont hang if going against a 1 chip per every square foot array with 1 chip every 3 foot. There is way too much exaggeration that goes on even in the cob light area. To get the most out of it you have to put more into it up front and pack the chips in there. Thats where people can save on power and have amazingly even canopy coverage.

I dont blame you for jumping into it. I just wish everyone would be honest that in most of the cases they will use the same power or more as hps and create the same heat or more unless you spend upfront and pack chips in there.

In a 4x4 tent a 6 cob setup will need to burn hard to create the same par as hps and while you burn em hard the efficiency goes way down and the heat goes way up.
While running 16 chips in the same room lets you run much less total power to get the same par across the entire canopy and because you're using less power you're creating less Heat.
Thats all. Well I'm out for now. Just lost power in town here.
 
fanleaf said:
I really didnt mean to snap and Im sorry for that, it just ended up that way for me. I just get upset when I see so many people thinking they will build a light to save power and heat and then they go and buy either a premade light or build it in a way that they will save nothing, spend more upfront and create the same heat.
Click to expand...

I understand and I do appreciate and accept your apology. I know you're not doing so well over there. I'm truly sorry for your loss.
I promise not to scam anybody. Please understand the market in Canada is huge right now. It's filling up with people selling Epistar chips in Chinese boxes claiming 1000w on four COBs. It makes my job extremely difficult.
I'm trying my best to offer an honest product at 200w in a world where everyone is being told they need 1000.
 
fanleaf said:
Dude, there are great Bridgelux and Citi chips out there and awesome arrays can be made with them too. Hell, I use some! But people will not build the same caliber light with 1212's and such.
They will also not build a light of the same caliber by running 1 chip per every 2-3 feet. People can go cheaper than Cree with Veros and build the same array as my Cree and get amazing results but they wont hang if going against a 1 chip per every square foot array with 1 chip every 3 foot. There is way too much exaggeration that goes on even in the cob light area. To get the most out of it you have to put more into it up front and pack the chips in there. Thats where people can save on power and have amazingly even canopy coverage.

I dont blame you for jumping into it. I just wish everyone would be honest that in most of the cases they will use the same power or more as hps and create the same heat or more unless you spend upfront and pack chips in there.

In a 4x4 tent a 6 cob setup will need to burn hard to create the same par as hps and while you burn em hard the efficiency goes way down and the heat goes way up.
While running 16 chips in the same room lets you run much less total power to get the same par across the entire canopy and because you're using less power you're creating less Heat.
Thats all. Well I'm out for now. Just lost power in town here.
Click to expand...
What about the 1825s that are efficient at 100 watts.

Have you ever run them in your calculator?

I'm just curious how mine compare at 2100ma

I run 4 in a 3x3 and it seems perfect.

I don't think a HOS would compete at 400 watts but I could be totally wrong.
 
newgrowr said:
Jimmy let me get a supplies list as well been reading you're stuff!!! Righteous man thanks for all the info I've been trying to catch up on. U the man
Click to expand...

If in the US or Canada try Future Electronics for your non Cree chips. If you want Cree use Jamie at Cobkits. His prices are the best for small builds.
For your heatsinks you can use Cobkits, Northern Grow lights or cutter electronics in Australia. Price around. Heatsinks are important.
For your driver's I use Mouser or future.
Unfortunately there's very few places where you can buy everything all in one shot for a reasonable price. Cutter electronics is probably the most reasonable for a all in one solution especially if you're looking at the grove mouse kits which are incredibly easy to assemble and install yourself.
Your selection of chips and heat sinks and power supplies will depend greatly on your geographic location as well as your budget and grow area size.
If you have any questions or concerns please feel free to contact me directly through a private message and I will try and help you select the best system for your budget.
Please understand that although LED is far superior to traditional lighting it is considerably more expensive to do it properly. The savings, ultimately in the long run are extremely beneficial, especially if you're going to be a lifelong hobby grower. You may pay more for lights on the start up cost but the dollars to watts savings overtime combined with all the free weed you're going to grow will more than pay for itself.
 
Back
Top Bottom