The LED "Arms race".....
- Thread starter Chip Green
- Start date
Enigma
Well-Known Member
Thank you!
Where can I find a calculator so I don't have to wear out my lead on estimations?
I don't think there is enough gain to under drive that far.
700 mA is over 180 lm/W, IIRC.
Chip Green
Well-Known Member
Theres not enough gain there to justify that initial cost. I get 150+ lm/W with 11x $8 EB 560mm strips at 700ma.....50,000+ lux 12" from the panels. Simple builds.
CobKits
Well-Known Member
see the canadian thread in indoor grow section, having that discussion right now
https://www.rollitup.org/t/canada-grows-to-the-4-plant-limit.948839/page-5#post-13786892
https://www.rollitup.org/t/canada-grows-to-the-4-plant-limit.948839/page-5#post-13786892
Enigma
Well-Known Member
I was there over a month ago, less COBs driven efficiently rather than multiple COBs under-driven.
I'm more interested in a calculator so I can calculate different COBs.
I have an idea, high powered COBs may not be the best for it, I may end up with as much as seven on a custom aluminum heat sink.
In a nutshell, I think my Veros are too much unless I under-drive a lot.
Last edited:
CobKits
Well-Known Member
its hard to trust datasheets/simulators unless you can account for and adjust Tc accurately
Enigma
Well-Known Member
I always look at the worst case scenario, if there isn't a Tc of 85° C listed I move on.
Actually, I don't believe I've seen Tc values other than 25° C or 85° C.
Photon Flinger
Well-Known Member
But that's what the simulator is for - to help plan this stuff out beforehand with the magic voodoo math!
CobKits
Well-Known Member
sure. read the disclaimer about warranty of data accuracy and suitability of use
you can blindly trust the model that bridgelux tells you may or may not be accurate when it differs wildly from both datasheets and any real-world verification, but you do so at your own risk
you can blindly trust the model that bridgelux tells you may or may not be accurate when it differs wildly from both datasheets and any real-world verification, but you do so at your own risk
wietefras
Well-Known Member
Yeah, you must be using voodoo math, because real math would have given you a more realistic outcome *ba dum tss*
These COBs cost $25 a piece? You need 40 which would be $1000 for the lot instead of 8 for $200. Zo, you need to save $800 on your electricity bill. In this fast moving field I personally don't want a window of more than 2 years to make that back, so you would need to save $400 per year on electricity by saving 188W (212W instead of 400W). Even if you flower continuously all year, that would mean you need a kWh price of 49 cents to break even.
And that's assuming your calculations are correct that those 40COBs at 212W produce as much light as the other 8. Which is probably also "voodoo" math and therefore wrong.
Photon Flinger
Well-Known Member
Ah believe in what you want, it might be a free country where you are at.
Everything is somewhat relative though. SK 600s going for $2300 canuck bucks which works out to about 65 cobs. Cob option is far more versatile.
Might also want to take into consideration total LES surface of all your light sources.
Everything is somewhat relative though. SK 600s going for $2300 canuck bucks which works out to about 65 cobs. Cob option is far more versatile.
Might also want to take into consideration total LES surface of all your light sources.
Enigma
Well-Known Member
I've been doing fine estimating with my pencil so far.
If the calculators aren't reliable then why do you use them?
Or are you using something you wrote in excel?
Do you test each COB at different temperatures and you have a data log?
Or are you pencil and paper like me?
I'm curious to know how other people come up with their figures.
xX_BHMC_Xx
Well-Known Member
I'd like to get something like the canopy substrate and some CV drivers and just keep adding a handful of COBs every month or two until it's just a solid block of light...
ttystikk
Well-Known Member
The expensive ultra efficient rigs of today point the way forward. As time goes on, both diode efficiency increases and cost reductions will inevitably make such designs ever more affordable and therefore accessible.
That's how the old stuff is rendered obsolete.
Meanwhile, don't wait! The value of the product we're all interested in growing on this forum makes the cost of the lights well worth it. My by now overly expensive Cree CXB3590 based lights will be working just fine for many years, continuing to amortize their cost the whole time.
That's how the old stuff is rendered obsolete.
Meanwhile, don't wait! The value of the product we're all interested in growing on this forum makes the cost of the lights well worth it. My by now overly expensive Cree CXB3590 based lights will be working just fine for many years, continuing to amortize their cost the whole time.
coreywebster
Well-Known Member
Exactly. My last harvest numbers are just about in. looking like 34-37oz from 660 cob watts in a 4x4 space. I think my set up has paid for itself several times over. (That's an all time personal best from me in that space)
wietefras
Well-Known Member
Like I said, that's the problem. You think math has something to do with believing ...
Photon Flinger
Well-Known Member
Budies 101
Well-Known Member
The light I just built is 300watts supposed to be 170lm/w, cost about 260$ to build and if I did it in bulk I could do it for about 220$. The difference between this light and a new video card is the light will grow near the same 10 years from now, that video card if not dead will be pretty far behind. The big difference is price vs market prices.
800$ video card (top top top of the line and looking for lowest cost)
260$ LED light (DIY built 170 lm/w top top of the line not lowest cost as in no bulk)
1200$ Kind/platinum/blackdog/so on LED light (already 10-20% less good at turning the watt from the wall to usable light by the plant.)
Soon LED's will be so far beyond HPS that even a simple cheap led light will be equal to a high end DE HPS light. Many companies are already moving into better chip and away from blurple.
800$ video card (top top top of the line and looking for lowest cost)
260$ LED light (DIY built 170 lm/w top top of the line not lowest cost as in no bulk)
1200$ Kind/platinum/blackdog/so on LED light (already 10-20% less good at turning the watt from the wall to usable light by the plant.)
Soon LED's will be so far beyond HPS that even a simple cheap led light will be equal to a high end DE HPS light. Many companies are already moving into better chip and away from blurple.
CobKits
Well-Known Member
they are good guides. most are reasonably accurate for scaling purposes. i provided one example where one manufacturer's simulator was broken, to satisfy your request.
i can post some observatons of temperature effect with cobs. im not interested in current time at getting exact par/lumen output from any cob as for my purposes its not justified to have a calibrated chain of light sources from nist to me. maybe down the road when i get a bigger sphere to test fixtures
right now i am only interested in relative differences between cobs and can say with fair certainty in multiple tests that veroC isnt a magic bullet, it is essentially identical to a citi 1825 gen 5 at all currents from 0.1 to 3.6A. The test data literally falls right on top of each other.
theyve long since paid for themselves. youre in the gravy zone now