WHAT IS YOUR EXPERIENCE WITH DIY LEDs?
- Thread starter Nonya
- Start date
- Tagged users None
Alex Cataldo
Active Member
Recently I replaced the LED chips on one of those cheap AliE Poplargrow bars, I used 450 chips rather than 470 because I really like seeing SPS glow under that spectrum. One thing I want to try upgrading is the power supplies because they only put out 20 watts of power. And make sure to use thermal paste, but so far so good!
Koh23
Valuable Member
Definitive....
Best lights i had, i did two versions, one was with 3w chips, combination of royal blues and 20000k whites, manual dimming (4 dimmers, 2 for whites, 2 for blues), i made 5 bars with leds, if i remember correctly, 450w, no control beside manual dimmer.
Second version was build with 10w chips, same principle, 20000k whites, royal blues and i add some 420nm uv....
Manual dimming, 500w, no drivers, connected to pc power suply and mehanical timer.....
If u can accept that there is no spectrum adjustment, no sunrise/sunset, no phone app, then this is still great and cheaper way to add a lot of lights to tank.... I probably run these lights at 50%, maybe less, but, large number of leds secure great coverage... Aestetic is of course poor, but....
If one want to invest more time and some money, same can be done with much more "gizmos" so you can have controller, separate channels and program sunrise and sunset, on off schedule, even build a proper and good looking case.....
So, for me, still diy led is not dead, probably is not common as before, but where is will, there is a way....
Best lights i had, i did two versions, one was with 3w chips, combination of royal blues and 20000k whites, manual dimming (4 dimmers, 2 for whites, 2 for blues), i made 5 bars with leds, if i remember correctly, 450w, no control beside manual dimmer.
Second version was build with 10w chips, same principle, 20000k whites, royal blues and i add some 420nm uv....
Manual dimming, 500w, no drivers, connected to pc power suply and mehanical timer.....
If u can accept that there is no spectrum adjustment, no sunrise/sunset, no phone app, then this is still great and cheaper way to add a lot of lights to tank.... I probably run these lights at 50%, maybe less, but, large number of leds secure great coverage... Aestetic is of course poor, but....
If one want to invest more time and some money, same can be done with much more "gizmos" so you can have controller, separate channels and program sunrise and sunset, on off schedule, even build a proper and good looking case.....
So, for me, still diy led is not dead, probably is not common as before, but where is will, there is a way....
bblumberg
Well-Known Member
I have built many DIY LED setups. I am not completely certain that I saved a lot of money on these, but I am pleased with the spectral control it gives me. Lately, I have moved to Citizen 455 nm COB LEDs as my main source of blue together with Viosys 420 nm to enhance the fluorescent pop and LumiLEDs fresh fish white COB LEDs. Much easier to wire this up than individual LEDs. Add Meanwell LDD drivers and Bluefish mini together with an efficient Meanwell power supply and you are in business.
monkeyCmonkeyDo
7500 Club Member
I don't understand how it isn't cost efficient? I plan on re building some diy led lights I have when I get or find the time but I did a diy moonlight 48" bar that's dimmable and I found it to be much more cost efficient than a all blue led bar from a manufacturer.
D
D
Screwgunner
Valuable Member
18"× 6' 1/4" aluminum plate and 120 3 watt leds blue ,royal blue.x30 each 10 uv 25 6k and 25 12k two 150 watt drivers and two potentiometers got the kit from China 300 bucks. 5 years I had to replace one driver and two leds. And it is still going strong after 12 years.
Koh23
Valuable Member
It's much much cheaper.....or at least, it can be, all depends what u want to do, and what is your vision of end product, both in terms of funcion and aestetic....
If u want to use alu bars, glue some leds, solder wires, use some 12v power supply, and call it quit, yes, its cheaper than the carton box of high end lights.
If u want to add drivers, timers, dimmers, maybe ability to program lights, separate channels by colors, wifi control, then its not as cheap as first option.
If you want to go further to the rabbit hole, and use some cnc made parts, cutted acrylic plates, to achieve that factory & expensive look, it costs even more.
But, in the end, still costs less than brand unit, and in term bang for buck, gives a lot more.....
If u want to use alu bars, glue some leds, solder wires, use some 12v power supply, and call it quit, yes, its cheaper than the carton box of high end lights.
If u want to add drivers, timers, dimmers, maybe ability to program lights, separate channels by colors, wifi control, then its not as cheap as first option.
If you want to go further to the rabbit hole, and use some cnc made parts, cutted acrylic plates, to achieve that factory & expensive look, it costs even more.
But, in the end, still costs less than brand unit, and in term bang for buck, gives a lot more.....
vtecintegra
Valuable Member
I'm still DIY. Refuse to pay almost $1K for a light like a Radion. I'm surprised more DIY people don't use what I do, since it's as plug and play as it gets for DIY. I built two fixtures to blanket 48" x 29".
Heatsink: Maker's LED 36". LED stars screw right to its slots. Any length you want.
Drivers: Maker's 5 up board with LDD drivers.
Control: Maker's controller- 5 channels of sunrise/set, clouds, and a manual dimmer.
LEDs: Rapid LED solderless. 14 Neutral/Cool white, 28 450nm, 7 430nm, 7 395nm per heatsink.
Power: Meanwell 350w power supply.
I can indefinitely replace any piece that fails. I've got extra LEDs, drivers, and controller if I ever ned them. No phone app like the new stuff, which is a plus for me. No coral growth problems I'm ware of.
Heatsink: Maker's LED 36". LED stars screw right to its slots. Any length you want.
Drivers: Maker's 5 up board with LDD drivers.
Control: Maker's controller- 5 channels of sunrise/set, clouds, and a manual dimmer.
LEDs: Rapid LED solderless. 14 Neutral/Cool white, 28 450nm, 7 430nm, 7 395nm per heatsink.
Power: Meanwell 350w power supply.
I can indefinitely replace any piece that fails. I've got extra LEDs, drivers, and controller if I ever ned them. No phone app like the new stuff, which is a plus for me. No coral growth problems I'm ware of.
ZombieEngineer
Valuable Member
Now that there are much cheaper options for lighting, it just doesn't make sense to do anymore.
A DIY fixture comparable to a Radion XR30 is like $250 in LEDs ($3 ea for 56 3W cree stars, $8-20 ea for 8 3W UV stars, $50-200 in drivers and controller, and another $100+ for lenses and heatsink. Even more to add active cooling, splash proofing etc.
You could buy a used G4 radion for like 300-400, a used Hydra 64 for like $300 or a brand new NooPsyche for $400.
A DIY fixture comparable to a Radion XR30 is like $250 in LEDs ($3 ea for 56 3W cree stars, $8-20 ea for 8 3W UV stars, $50-200 in drivers and controller, and another $100+ for lenses and heatsink. Even more to add active cooling, splash proofing etc.
You could buy a used G4 radion for like 300-400, a used Hydra 64 for like $300 or a brand new NooPsyche for $400.
ZombieEngineer
Valuable Member
How do you get 12 cents per watt unless you are getting the absolute worst quality LEDs they make? Even a 3up cree-xt royal with elliptical lenses is $6 for 9W. Your 12c figure must be chinese garbage with terrible efficiency.
Cost per watt is a useless metric. Cost per lumen is also needed as are lifetime costs of the fixture. Lenses and dome shape matter a lot as well and that is harder to quantify.
Last edited:
For many wavelengths, your arguments have some validity. The thing is, 445 nm is pretty much the default for LEDs. Most other wavelengths use phosphors to convert from 445. Even releatively poor quality leds still are good at 445.
445 nm is also just about where you want most of your light energy. Look at the output spectrum of high end fixtures and youll see what I mean.
I agree it could definitely make sense to shell out for non blue spectra, but these will definitely be in the minority of light enery. Why pay a premium for most of your light energy when you can supplement?
Lumens are definitely the wrong metric. The same amount of light energy will have more lumens as green vs blue light for example. Lumens are a metric of perception to the human eye, not light energy. You want PAR.
Ill let you know the spectrum and par of these dozen chinese 50 watt chips when I get them.
445 nm is also just about where you want most of your light energy. Look at the output spectrum of high end fixtures and youll see what I mean.
I agree it could definitely make sense to shell out for non blue spectra, but these will definitely be in the minority of light enery. Why pay a premium for most of your light energy when you can supplement?
Lumens are definitely the wrong metric. The same amount of light energy will have more lumens as green vs blue light for example. Lumens are a metric of perception to the human eye, not light energy. You want PAR.
Ill let you know the spectrum and par of these dozen chinese 50 watt chips when I get them.
Tom Gates
New Member
I am running 4 Radions 30 blue on my 300 gallon reef tank. About a month later later they came out with the new Generation 6. They have an add on to change the generation 5 to generation 6 for 399.00 per light. Does anyone have any thoughts. I know this is off the subject. But I am very curious.
Regards,
Tom Gates
ZombieEngineer
Valuable Member
The upgrade kits are absolutely worthless. It is cheaper for you to buy completely new lights and sell your previous generation. This also has the added benefit of helping a fellow reefer on a budget get nice lights they can afford.
