Que led me compro

2.25 m^2 -- if photoperiods, definitely take the extra watts.. you'll appreciate it 5-7 years from now as it can still push upper boundaries of that space. you may even need a slight dim or an extra inch or two initially, but it's the better side to err on for the long-term use. even as you reach that 90% of orginal intensity after 50,000 hours, the extra oomph will come in handy toward end.. and even push beyond, reasonably.. though the degradation in efficacy only increased in speed. if autos, you need 2/3rds of that power over 18/6 cycle... and if you run 20/4 you need 60-percent of that. it is proportional. i'm gonna focus on photos below, but could easily do the math on autoflowers and 18/6 or 20/4 using DLI chart following the same steps. photoperiods mean you will use as little as 12 hours per day... let's shoot for roughyl 900ppfd and work backward (you'll want 1100-1200 PPFD if supplementing co2 and tightly controlling resulting VPD from environment) Reference a DLI chart. 900 was chosen because it is approaching 40DLI, which is probably overshooting ambient co2 type needs, very slightly. 900 * 2.25 m^2 = 2025umol/s -- a little more than that is great too. So, i'd go with the 730W FC6500. The build quality of the FC sereies is nothing like mar's cheaper models. FC-series have nice heatsinks and robust frames. ability to put the driver outside the tent if heat is an issue, etc etc... and most importantly, they don't over-drive the diodes... looks like lumatek has improved if the "2.9 umol/J" claim is true. their old zeus model was nothing special at the time.. i remember the migro guy (despite being a mostly good source for info) gushing over it at 2.5umol/J... it wasn't even using the best diodes available at the time.. it was a 2nd-tier light from the start. ---- nitty gritty stuff and reasoning... divide watts / number of diodes.. assuming LM301b or h, it should be around 0.25watts per diode for the high-quality lights. anything higher and they are likely fibbing about the efficacy... the better bin diodes can mitigate that a little bit. reds and uv bring it down quite a bit. unless they are using the new sammy reds or the cree xpg2? those are up around 2.86umol/J while most are significantly lower... e.g. osram 660 reds are 60-70% of that, give or take. but, those are only 4-5% of total, so weighted average doesn't bring it down too much. if you see "3.10umol/J" that's total bullshit with lm301b diodes, unless it's all 5000k CCT and nothing else... and run at 0.25w per diode. that would reach 3.1umol/J (or is it 6500k? whateve the samsung.com spec sheet uses, it is a cool white) the problem with uv diodes is they'll be degradating fast after 10k hours and you'll have 40k+ to go with your white diodes on the light. so, don't let those make a decision. i'd suggest a 2nd uv-only fixture you can slip in between the bars... selective use is better than constant use.. they only help in flower as far as catalyzgin any trichome production. use it toomuch early on and the plant will build a resistance to UV just like your skin does.. and then its benefit will be reduced later. i'm only familiar with teh fc6500 650w fixture... but all the other "FCxxxx" models are of similar build quality. "FCE-xxx" is using diodes that are maybe 5% less efficient (maybe2.7umol/J?). anyway, use the little watts per diode trick for any LM301-based unit... to reach those 2.7-2.9ish umol/J ranges you need to be at or near 0.25watts per diode average or they are likely lying. if you look at teh various bins and resulting lm/w table, it is proportional and can do a fairly accurate guesstimate based on specs of the 5000k diode... 226? so if another diode in tht spec sheet table is "190lm/w" ---> 190/226 * 3.1umol/J = 2.6umol/j ... going off top of head, the 5000K lm/w diode might be 232lm/w? either case that lm/w == 3.1umol/j... simple algebra from there taking proportion of 232 and multiply by 3.1. (it may not be exact, but it'll be close enough given fact not all the diodes are the same on the lights -- you can also guesstimate which "bin" they are using in your product.) luckily, thise more expensive lights are usually accurately rated, but any discrepancy from samsung.com relative to context of that light, and you know somethign is wrong with specs... you can multiple efficacy X watts to determing umol/s if it is not listed. also, another way to double-check for simple lies if it doesn't match what is advertised.