new 1000w led strip project

Discussion in 'Grow Room Design & Setup' started by grisbi, Feb 22, 2018.

?

what's your strip preference for a bloom room between:

  1. NICHIA 757 Tunable White 2700/6500K

    0 vote(s)
    0.0%
  2. NICHIA 757 Tunable White 2000/4000K

    1 vote(s)
    6.3%
  3. Samsung Q Series Strip @3500k

    1 vote(s)
    6.3%
  4. samsung F-Series Gen3 [email protected] 3500k

    8 vote(s)
    50.0%
  5. Bridgelux EB Series Gen 2 @3500k

    6 vote(s)
    37.5%
  1.  
    grisbi

    grisbi Active Member

    yes @lukio you right, i could take juste 660nm strip, but i prefere get mixed strip with white + red directly on it, the red and white will be spread perfectly this way.
    i know its a harder way to do, i could just take white stripes and oslon ssl redstrip, but ,first its not funny :bigjoint: and second i want a nice looking and solid diy stuff, i really think led growing is going on this direction for limited height or multilevel growing space, with those new led/driver we can have the light penetration....and finaly im kind of fetishist.... i want the best :eyesmoke:
    waiting for the @sunni @Randomblame @CobKits expertise
     
    Randomblame likes this.
  2.  
    lukio

    lukio Well-Known Member

    coreywebster and grisbi like this.
  3.  
    grisbi

    grisbi Active Member

    im totally fan with the rond pcb design witch feet with 140mm heatsink, this pcb gonna be killer for flower room too with the red dude!, and amazing penetration! i think we are about finaly to do what we expected from led and my opinion is there is a convergence: mixing the best nice white mid power led , high eficiency, good thermal issue, amazing custum covering and some extra red or burple ( depend if ur on flower or veg)

    but im wanna do strip @lukio with next generation of led, like the nichia 757 not with lm561c (hope cutter will say YES)

    ps: i asked growmau when this board with red will be released, but no answer, do know something about it?
     
    lukio likes this.
  4.  
    lukio

    lukio Well-Known Member

    i think he said about a month, bro!
     
    grisbi likes this.
  5.  
    grisbi

    grisbi Active Member

    thank you dude!
     
  6.  
    grisbi

    grisbi Active Member

    Ladddy and gentleman...
    BIG NEWS :
    here picture of my strip project handle by the master from aussie! ( i love st kilda!) the one, the seriouse guyz frooom www.cutter.au :
    THHEEEE SOL-STRIP !! :D
    size: 560mm x 25 mm

    [​IMG]

    still some work to do, arangement, and few white led to add, cant wait to get thoses beaaaaasssssst on action new 2k18 led will rock on i think ;) !!!!

    arigato gosaimasu Mark and the cutter team! <3

    i'm doing my best to find waterproof aluminium case housing + tempered glass to fit perfectly with theses baby.

    lets finish that on music :mrgreen:
     
    Randomblame likes this.
  7.  
    Randomblame

    Randomblame Well-Known Member

    Ambitious project, I like it very much.
    However, the lamp will be twice as expensive as if you build them with F-Series strips. To squeeze out +200lm/w the Nichia 757 needs to run at 65mA, which means you need a hell of lot of sol-strips. I was facing the same question a quarter year ago and chose LM561c running at ~60mA (525mA per strip / 11,8w) because F-strips perform better than Q-strips when running with the same current. Nichia strips were not available to the date and I did not want to wait.

    LM561c vs. LM301b.png

    You have to think about how much the efficiency jump is really worth to you. A 10% more efficient lamp is likely to cost twice as much and it takes longer to recover the cost over your electricity bill.
    Even my lamp is actually way overpriced when one looks at it from the standpoint of dimnishing returns. Instead of 20 I could have built it with 10, would have paid half that much but still get 175lm/w.

    Your heatsinks look good, no question, but if you actually drive the sol-strips with so little current, then they are unnecessary. A simple frame would be enough to fix the stripes without any heatsink. A heatsink is only necessary if you drive the sol strips harder, but then you get no 200lm/w. And this is where the F-Strips come in again!
    What would it cost if you want to get the same price/performance as an F-strip?
    Let's take the 187lm/w at 60mA from the example above(cost are 14$/2ft.strip, 11,8w, 187lm/w).
    How low do you have to drive the Nichia 757 to be better than that and what does it cost?
    If you plan to use them for the next 5-7 years, okay, then the costs will play a subordinate role. However, if there is something better in 2 years and you already look in that direction, then you should look at it from the price/performance point of view.
    Shipping costs, VAT and customs are also things that must be considered, usually you need to add ~25% for euro buyers.
    Do not get me wrong. I would really like to see you finish your sol-strip project. Would certainly be one of the best in the DIY market. Just be aware, you will have about twice the cost for such an efficient build.

    BTW, the link to the sol-strips is not working ;-)
     
    grisbi likes this.
  8.  
    grisbi

    grisbi Active Member

    im agree with your opinion @Randomblame and you get some points:
    - the price wont never be as affordable as lm 561c s6 they are such efficient for their price , same for bridgelux, but i wanted a "diy kind of high end strip" i love the fluence and progrowtech form, but i really think we can do a little bit better with new 3030 led, or at least cheaper :)
    - to get 200lm/w need to go down so much the curent tension , will need lot of strips, but this is the only way with all led stuff.
    but @Randomblame im not only obsesse by the ratio lm/w , its also about be able to push hard the strip then you can put them higher from cannope and leave your plant for a week end for exemple. I finnaly just wanted to do strip with the newest chip, and nichia spectrum look very nice!! if i did the same experience last year i would take lm561c with no hesitation, but now, news mid power led are comming out, this is not the same game ( not really conviced buy lm 301b yet, think we have to wait some next "bin" to get the full potential of that led). nichia / luxeon are comming like the winter :fire:
    -for the heatsink part: hihihi i know its useless, its just for me, i wanna a good looking structure, no risk to put water sprinkle on it. and like i said before, i wanna keep the option "fire in the hall" on pushing hard the strip, and i think heatsink will be cool to have
    cheers.

    theire is no horticultur strip anyway, could be nice to get this option no?
    lets finish on music :D

     
    Last edited: Mar 1, 2018
    Rokus and Randomblame like this.
  9.  
    grisbi

    grisbi Active Member

    @Randomblame :
    [​IMG]
    solstrip preview, can you seed the jpeg ?
    and thank you for the good words man!
     
    coreywebster and Randomblame like this.
  10.  
    Baudelaire

    Baudelaire Well-Known Member

    Hmm. SolStrips are already available, in three CCTs: 2700K, 3500K, and 5000k, plus a deep red/blue mono augmentation strip. All interchangeable and with custom fitted and tapped heatsinks, for 30% less than QBs. Check them out here. Lots of discussion and grow shows at OverGrow. -b420
     
  11.  
    Serva

    Serva Well-Known Member

    These are different solstrips you are talking about. They are using Samsung LEDs not Nichia and white and red isn‘t combined on one strip, so the spread isn‘t as good as the customize option from cutter.
     
    coreywebster, Randomblame and grisbi like this.
  12.  
    grisbi

    grisbi Active Member

    yes finalyy! mark told me! ill order asap!

    im going to do 2 kind:
    nichia 2000k + xd16 3500k
    nichia 5000k 97cri + xqe 660

    and mix that on my big pannel <3
    edit: not really the same product dude ;)
     
    Last edited: Mar 14, 2018
    Rokus and Serva like this.
  13.  
    grisbi

    grisbi Active Member

    yes thoses solstrip project was about to reproduce what cutter did with theire new pcb but on strip!
    it was about give people something different, with brand new led model serie, they are not the cheapiest way , but still affordable ( to me)
    i still believe on mixing meduim and high cri, and want to try so bad the 5000k cri 97 on veg!
     
    Last edited: Mar 14, 2018
    Rokus likes this.
  14.  
    Rokus

    Rokus Active Member

    You know bigger heatsinks always pays off, because in the end the efficiency is directly related to heat, so in theory if you ran even bigger heatsinks you could get away with half as much diodes but still be almost as efficient!
     
  15.  
    Randomblame

    Randomblame Well-Known Member


    Half as much diodes because of a better heatsink...? Nonsense!
    5°C less Tjunction would be only 1% more light. Difference between 25 and 85°C is ~6-9% depending on diode specs!
     
    Aolelon likes this.
  16.  
    Rokus

    Rokus Active Member

    I am not saying it's more efficient, it's less efficient, but not so much people don't go that route, look at Spectrum King for example...
     
  17.  
    Randomblame

    Randomblame Well-Known Member


    The Spectrum King guys should learn how to built growlights.
    They throw a hell lot of diodes onto a small PCB's which creates only a huge hotspot in the center. To make matters worse they use an additional reflector which increase the hotspot effect further.
    I think I read that they are planning a strip- or board-design, but I could be wrong...
    If you want to measure lower Tjunction temperatures, the individual diodes need more space. A bigger heavier heatsink needs only longer to heat up, surface area is what matters. Also on their 450w lights with massive heat pipe heatsinks the diodes run on the hotside(~80°C Tj).
    If you want to use the light more efficiently, you need to better distribute it. The same goes for the heat on the other side.
    Look at how efficient LED strips like the Q-series from Samsung or the EBgen.2 from Bridgelux can dissipate the heat. At nominal power, you do not even need a heat sink. That's efficient!
     
    ANC likes this.
  18.  
    Rokus

    Rokus Active Member

    There is also good things to say about spotlight designs. If you have high ceilings it might even be the better choice for working with taller plants. Quantum boards, with lower power diodes function better if you hang them close over an even canopy.

    I found an interesting yt series about this I had not seen before
     
  19.  
    Randomblame

    Randomblame Well-Known Member


    Yeah, higher ceilings would be the only reason. But usually these do not appear in standard tents, so rather something for commercial greenhouse growers! High distance also mean more loss for that reason the 640w Nextlight(640w) beats the SpectrumKing(646w) on avarage PAR readings(less distance) and uniformity(better spread, more diffuse).
    While the Nextlight puts out an avarage PAR of 381μMol/s/m² the SpectrumKing is only 325. Uniformity is 3,1 on the Nextlight while it's only 5,7 on the SpectrumKing. That means the center hotspot is "5,7 times" higher as the lowest reading. The Nextlight unit also allows a lower distance which means the avarage PAR redings would be ~450μMol/s/m² and still having better uniformity(3,9).(see video, min. 24).
    If SpektrumKing offered only for tall greenhouses, that would not be a problem as the light is usually higher. But they recommend them for tents, although they are rather inappropriate for that. Hotspots cause problems like uneven growth or bleaching. In a 4x 4' tent the Nextlight would yield much better and it would be easier to keep the plants healthy.
    Good penetration comes from diffuse light not from strong light. The SK acts like a point source which means lots of shadows while the NL delivers much more diffuse light and almost no shadows means the light reach deep into the canopy.
     
    Rokus likes this.

Share This Page