Reef Tank LED Lighting - High PAR - Low Cost DIY

[ryaneogilvie]

Hello all,

I'm new here but have had a 75g (+20g sump) reef tank for the past couple years, and been constantly reading posts from afar for many years.

I just built a lighting setup for it. I wrote a DIY guide (easier to put on instructables than a forum because it is a bit long). But would love to share it here and get feedback for improvement. This is my latest build with hopefully more iteration to come so would love to get feedback or thoughts. I plan to make it look a bit nicer, but this is the first working prototype of this version.

https://www.instructables.com/Reef-Tank-LED-Lighting-High-PAR-Low-Cost-DIY/

I can already tell it's way better than my last lighting solution since my anemonoe has gotten the biggest I have ever seen it even in just like 2 days of testing this new light.

PS: I've also done a lot of other DIY for my tank, like building the entire base, wiring/programming an auto top off solenoid valve connected to a float valve and RODI system, 3D printed filter sock holders, and others. I'm happy to share if there is interest in any of those.

- Ryan

 

[HappyFm]

Hello all,

I'm new here but have had a 75g (+20g sump) reef tank for the past couple years, and been constantly reading posts from afar for many years.

I just built a lighting setup for it. I wrote a DIY guide (easier to put on instructables than a forum because it is a bit long). But would love to share it here and get feedback for improvement. This is my latest build with hopefully more iteration to come so would love to get feedback or thoughts. I plan to make it look a bit nicer, but this is the first working prototype of this version.

https://www.instructables.com/Reef-Tank-LED-Lighting-High-PAR-Low-Cost-DIY/

I can already tell it's way better than my last lighting solution since my anemonoe has gotten the biggest I have ever seen it even in just like 2 days of testing this new light.

PS: I've also done a lot of other DIY for my tank, like building the entire base, wiring/programming an auto top off solenoid valve connected to a float valve and RODI system, 3D printed filter sock holders, and others. I'm happy to share if there is interest in any of those.

- Ryan

I also enjoy DIY. Completing my aquarium by hand and watching it grow based on those devices is really fascinating.

I have a bit of feedback about the light design, mainly concerning the spectrum.
You are only using blue and white lights, right?
While the PAR may be quite high, the focus on wavelengths suitable for corals could be reduced.
The lack of proper wavelengths (even though white lights can provide a full spectrum, they are not concentrated at specific wavelengths) could lead to a long-term reduction in coral pigmentation.
This also decreases the energy absorption of corals since chlorophyll ABC does not only use blue light for photosynthesis.
Overall, this could increase operational costs because the energy generated may be wasted.
Which LED brand are you using? Do you have specific specifications?

 

[HappyFm]

That's impressive! The Cree V3 LEDs are indeed known for their high efficiency. The ability to achieve 50-70 PAR with just 8 LEDs at 50W power consumption is a great result, especially without using reflectors. It shows the potential of well-chosen LEDs for efficient light output and low energy usage.

Your estimation with the mobile app is a good way to compare, even if there might be some margin for error. It’s also interesting that you've compared it to the M029 + K7V3 light, which may offer useful context for understanding the light intensity at different setups. Cree V3 LEDs, especially in the blue spectrum, are well-suited for aquariums as they offer strong light output while being energy-efficient.

 

[VintageReefer]

I would add in those UV clusters

 

[ryaneogilvie]

I also enjoy DIY. Completing my aquarium by hand and watching it grow based on those devices is really fascinating.

I have a bit of feedback about the light design, mainly concerning the spectrum.
You are only using blue and white lights, right?
While the PAR may be quite high, the focus on wavelengths suitable for corals could be reduced.
The lack of proper wavelengths (even though white lights can provide a full spectrum, they are not concentrated at specific wavelengths) could lead to a long-term reduction in coral pigmentation.
This also decreases the energy absorption of corals since chlorophyll ABC does not only use blue light for photosynthesis.
Overall, this could increase operational costs because the energy generated may be wasted.
Which LED brand are you using? Do you have specific specifications?

1. 100W Blue (440-450nm) LED Array: https://a.co/d/4gEHhWZ
2. 100W Cool White (10,000-15,000K) LED Array: https://a.co/d/4R5cDx6

So the LEDs I have running right now are 4 Cool white (10,000-15,000K) (2 - 50Wversions , 2 - 100W versions) and 2 Royal Blue (440-450nm) 100Win between. So 3/5 White (10,000-15,000K), 2/5 Blue (445nm). The company didn't publish spectrums, but looking for similiarly marketed device. I expect something like these:

white light spectrum: https://www.reef2reef.com/threads/20000k-and-30000k-led-spectrum.725445/post-7543880

Blue 450nm LED: https://www.moon-leds.com/product-royal-blue-450nm-3535-smd-led.html

It appears they may be a little lower than I'd like in the 460-500nm range, favoring the ~440nm range but I was considering adding some blue 460-470nm LED and/or replacing the 440nm with those to improve that range. I think the rest of the spectrum is covered well based on the chlorophyll absorbtion charts I have seen, but not completely sure.

 

[jsmkmavity]

These look like they would compare to something like kessil style lights in size and output, is that correct? I have considered doing a diy light for some time, but wat thinking mor of the multi small led array spread out to 12"-16".

 

[Gumbies R Us]

Hello all,

I'm new here but have had a 75g (+20g sump) reef tank for the past couple years, and been constantly reading posts from afar for many years.

I just built a lighting setup for it. I wrote a DIY guide (easier to put on instructables than a forum because it is a bit long). But would love to share it here and get feedback for improvement. This is my latest build with hopefully more iteration to come so would love to get feedback or thoughts. I plan to make it look a bit nicer, but this is the first working prototype of this version.

https://www.instructables.com/Reef-Tank-LED-Lighting-High-PAR-Low-Cost-DIY/

I can already tell it's way better than my last lighting solution since my anemonoe has gotten the biggest I have ever seen it even in just like 2 days of testing this new light.

PS: I've also done a lot of other DIY for my tank, like building the entire base, wiring/programming an auto top off solenoid valve connected to a float valve and RODI system, 3D printed filter sock holders, and others. I'm happy to share if there is interest in any of those.

- Ryan

This is really neat! Congrats on being able to fully make a DIY light like this

 

[HappyFm]

1. 100W Blue (440-450nm) LED Array: https://a.co/d/4gEHhWZ
2. 100W Cool White (10,000-15,000K) LED Array: https://a.co/d/4R5cDx6

So the LEDs I have running right now are 4 Cool white (10,000-15,000K) (2 - 50Wversions , 2 - 100W versions) and 2 Royal Blue (440-450nm) 100Win between. So 3/5 White (10,000-15,000K), 2/5 Blue (445nm). The company didn't publish spectrums, but looking for similiarly marketed device. I expect something like these:

white light spectrum: https://www.reef2reef.com/threads/20000k-and-30000k-led-spectrum.725445/post-7543880

Blue 450nm LED: https://www.moon-leds.com/product-royal-blue-450nm-3535-smd-led.html

It appears they may be a little lower than I'd like in the 460-500nm range, favoring the ~440nm range but I was considering adding some blue 460-470nm LED and/or replacing the 440nm with those to improve that range. I think the rest of the spectrum is covered well based on the chlorophyll absorbtion charts I have seen, but not completely sure.

LED PAR Test • Orphek Reef Aquarium LED Lighting

LED PAR test results of all Orphek LED Products including chromaticity charts can be found here.

orphek.com

I referred to the information published by Orphek, which states that a 10k light only produces half the energy compared to blue-only bulbs.
White light also lacks UV spectrum.
Would the excess energy in green, yellow, and red cause an algae outbreak?

Theoretically, we should reduce white light and replace it with bulbs of specific wavelengths like UV, violet, cyan, etc.
Sorry about these things, but I am really concerned about the cost effectiveness / energy efficiency / spectrum.

 

[ryaneogilvie]

These look like they would compare to something like kessil style lights in size and output, is that correct? I have considered doing a diy light for some time, but wat thinking mor of the multi small led array spread out to 12"-16".

They appear close, of course a little less control on individual light appearance/spectrum, but the 100W one I have outputs a little more than the A360X Kessil (90W), and I have an equivalent of ~6 of those on my 75g tank now. I'm running it currently at about 1/2 power in my tank as to not burn my soft/LPS corals till I can move them around an assess the option of adding SPS safely now.

 

[ryaneogilvie]

LED PAR Test • Orphek Reef Aquarium LED Lighting

LED PAR test results of all Orphek LED Products including chromaticity charts can be found here.

orphek.com

I referred to the information published by Orphek, which states that a 10k light only produces half the energy compared to blue-only bulbs.
White light also lacks UV spectrum.
Would the excess energy in green, yellow, and red cause an algae outbreak?

Theoretically, we should reduce white light and replace it with bulbs of specific wavelengths like UV, violet, cyan, etc.
Sorry about these things, but I am really concerned about the cost effectiveness / energy efficiency / spectrum.

I am unsure if it will cause an algae outbreak, I plan to monitor it for a couple weeks and see if anything happens (I'll post back here with results as they come), I did note some growth in my return siphon fixture earlier today causing the water level to be a bit higher than normal, but no unusual growth in the display tank other than my anemonoe opening/getting bigger, but I guess time will tell.

I would raise this topic here that suggests blue white LEDs are better than metal hallide and metal hallide has been in use for years with success: https://www.reef2reef.com/threads/metal-halide-vs-led-rates-of-photosynthesis.893227/. It's not clear on the exact spectrum with Blue/White LEDs, but does bring up an interesting point.

I've struggled a bit to determine what point of the spectrum are the most important, I see a lot of different approaches, but not as many studies, so I am still unsure on which wavelengths are the most important.

 

[HappyFm]

I am unsure if it will cause an algae outbreak, I plan to monitor it for a couple weeks and see if anything happens (I'll post back here with results as they come), I did note some growth in my return siphon fixture earlier today causing the water level to be a bit higher than normal, but no unusual growth in the display tank other than my anemonoe opening/getting bigger, but I guess time will tell.

I would raise this topic here that suggests blue white LEDs are better than metal hallide and metal hallide has been in use for years with success: https://www.reef2reef.com/threads/metal-halide-vs-led-rates-of-photosynthesis.893227/. It's not clear on the exact spectrum with Blue/White LEDs, but does bring up an interesting point.

I've struggled a bit to determine what point of the spectrum are the most important, I see a lot of different approaches, but not as many studies, so I am still unsure on which wavelengths are the most important.

The principle remains that we don't reinvent the wheel. We only remake it because we can't buy it. =)) I really love the way Orphek shares the information they have. You can learn more from the link.

https://vi.orphek.com/correct-corals-spectral-needs/

How light affects coral coloration

Coral Color Management｜BlueHarbor

This is a coral color management theory created by scientifically understanding the color of coral from an optical perspective.

www.blueharbor.co.jp

Keeping a coral reef tank alive is actually quite simple, but I want it to be stable, thriving, and growing – that truly is a monumental task.

 

[oreo54]

Hello all,

I'm new here but have had a 75g (+20g sump) reef tank for the past couple years, and been constantly reading posts from afar for many years.

I just built a lighting setup for it. I wrote a DIY guide (easier to put on instructables than a forum because it is a bit long). But would love to share it here and get feedback for improvement. This is my latest build with hopefully more iteration to come so would love to get feedback or thoughts. I plan to make it look a bit nicer, but this is the first working prototype of this version.

https://www.instructables.com/Reef-Tank-LED-Lighting-High-PAR-Low-Cost-DIY/

I can already tell it's way better than my last lighting solution since my anemonoe has gotten the biggest I have ever seen it even in just like 2 days of testing this new light.


- Ryan

The method you use to drive the led cobs is considered poor practice.
It may work ( for some for awhile only) but there is no control of current to the leds.
First sind " trivia"
The cobs are 10 x 10 arrays and the LEDs can each run at 3 0- 3.4v
Considering the series/ parallel design thats 10 x 3.4 for each series string and 3000mA / 10 so 300mA for each parallel string.
So each ind led ( you have a hundred in each cob) runs at 1w assuming the 3.4v figure and assuming they draw 300mA at said 3.4v.

The higher the voltage the more current the led will draw.
With the next exception followiing

Now the problem is if the LEDs have poor thermal management the chips will start to increase their amp draw at the same voltage
This can set up a thermal runaway where the LEDs draw so much current as to burn out.

Thus the need for current regulation by using a constant current " proper" driver.

Relying on a normal switching power supply to contain the current is problematic as well

Unfortunately the " fix" is either a ac/dc constant current driver which has its own issue IF you hook too many cobs in parallel and one starts to lose them ( fail open) thus sending more current to the remaining cobs.
Or a DC step down driver which aren't made in a 3 amp version afaict.

Understanding LED Drivers by LEDSupply

LED drivers have so many different types and variations that it can seem a little overwhelming at times. So lets take a look at LED drivers in depth.

www.ledsupply.com

The "correct" way to run those 100w cobs
is each having a driver..
Unfortunately it isn't free

To sum up the issue. ..

Thermal runaway happens when you drive an LED with a constant voltage source, as opposed to a constant current source. The key reason thermal runaway occurs is because an LED's forward voltage decreases as temperature rises. Or put in layman's terms, it's resistance decreases with temperature. With a constant voltage source the decreased resistance results in the current increasing. This is turn increases the amount of power going to the LED and makes it hotter, decreasing the resistance further, and increasing the current, until either the power source or the LED is destroyed. Thermal runaway is best prevented by using a constant current source. A constant current source will adjust the voltage going to the LED so the current remains constant. The LED can still overheat if the heat sink is inadequate, but you won't have a runaway increase in current and heat as you would using a constant voltage source.

One solution is to switch to 50w cobs
Their " max current" is 1500w.
Meanwell ldd's come in that range and cost around $5-$12. One per cob.
I did find 50w cobs for $5.92

50 Watt High Power Smd Led Chip 50W White Rgb Green Emitter Bulb Diode Component | eBay

50 Watt High Power Smd Led Chip 50W White Rgb Green Emitter Bulb Diode Component Rated Power: 50W Chip Size:30Mil/45Mil Lighting Colors: Warm White/Natural White/White/Cold White/Red/Blue/Green/Yellow/Royal Blue/Deep Red/Full Spectrum/IR/CYAN/RGB Sealing Material: 2*99%Golden Wires, Copper...

www.ebay.com

So each 100w is now $11.84 + say $12

LDD-H Series Mean Well Step-Down Mode CC DC-DC LED Drivers

for a Meanwell ldd hw @1500mA ( note that using their max current suggestion is also problematic regardless of chip size. Running a 100w at 3000mA or a 50 at 1500mA will generally just shorten their lifespan unless extremely good heat sinking/ cooling is used)
So you "double" the cost to say $10 to $24/100 watts .
Say 2 white 2 rb 50w cobs....total $80

One catch is that step down drivers have an overhead. You lose a few volts coming out.
A 36v ps will only output about 33v.
If you believe the 36v chip V(f) a 36 v switching power supply isn't enough voltage.
So one needs the next size up or
go a bit lower for a Meanwell power supply
Lrs350-36 can be tweaked to 39.6 volts covering your losses
Of course max wattage is only 350w approx.
$37.80

LRS Mean Well Enclosed Switching Power Supplies

Or your original ps seems fine if one can set it to 39v.
.To save a few $'s and fit my " philosophy" i' d go 1200mA ldd-hw for $9...Probably reduces your per cob output to around 40-ish watts but will run cooler and a bit more efficient ( higher par/watt)

Oh and your thermal fuse is a unique feature but if your heatsink in that area gets that hot it is likely your cob is severely damaged. Temp at the led itself will be way hotter than 77c (170F)
50 c (122F) or less may work to prevent damage
Remember you are dealing w cheap Chinese cobs.

You did ask for feedback ..

 

[ryaneogilvie]

The method you use to drive the led cobs is considered poor practice.
It may work ( for some for awhile only) but there is no control of current to the leds.
First sind " trivia"
The cobs are 10 x 10 arrays and the LEDs can each run at 3 0- 3.4v
Considering the series/ parallel design thats 10 x 3.4 for each series string and 3000mA / 10 so 300mA for each parallel string.
So each ind led ( you have a hundred in each cob) runs at 1w assuming the 3.4v figure and assuming they draw 300mA at said 3.4v.

The higher the voltage the more current the led will draw.
With the next exception followiing

Now the problem is if the LEDs have poor thermal management the chips will start to increase their amp draw at the same voltage
This can set up a thermal runaway where the LEDs draw so much current as to burn out.

Thus the need for current regulation by using a constant current " proper" driver.

Relying on a normal switching power supply to contain the current is problematic as well

Unfortunately the " fix" is either a ac/dc constant current driver which has its own issue IF you hook too many cobs in parallel and one starts to lose them ( fail open) thus sending more current to the remaining cobs.
Or a DC step down driver which aren't made in a 3 amp version afaict.

Understanding LED Drivers by LEDSupply

LED drivers have so many different types and variations that it can seem a little overwhelming at times. So lets take a look at LED drivers in depth.

www.ledsupply.com

The "correct" way to run those 100w cobs
is each having a driver..
Unfortunately it isn't free

To sum up the issue. ..


One solution is to switch to 50w cobs
Their " max current" is 1500w.
Meanwell ldd's come in that range and cost around $5-$12. One per cob.
I did find 50w cobs for $5.92

50 Watt High Power Smd Led Chip 50W White Rgb Green Emitter Bulb Diode Component | eBay

50 Watt High Power Smd Led Chip 50W White Rgb Green Emitter Bulb Diode Component Rated Power: 50W Chip Size:30Mil/45Mil Lighting Colors: Warm White/Natural White/White/Cold White/Red/Blue/Green/Yellow/Royal Blue/Deep Red/Full Spectrum/IR/CYAN/RGB Sealing Material: 2*99%Golden Wires, Copper...

www.ebay.com

So each 100w is now $11.84 + say $12

LDD-H Series Mean Well Step-Down Mode CC DC-DC LED Drivers

for a Meanwell ldd hw @1500mA ( note that using their max current suggestion is also problematic regardless of chip size. Running a 100w at 3000mA or a 50 at 1500mA will generally just shorten their lifespan unless extremely good heat sinking/ cooling is used)
So you "double" the cost to say $10 to $24/100 watts .
Say 2 white 2 rb 50w cobs....total $80

One catch is that step down drivers have an overhead. You lose a few volts coming out.
A 36v ps will only output about 33v.
If you believe the 36v chip V(f) a 36 v switching power supply isn't enough voltage.
So one needs the next size up or
go a bit lower for a Meanwell power supply
Lrs350-36 can be tweaked to 39.6 volts covering your losses
Of course max wattage is only 350w approx.
$37.80

LRS Mean Well Enclosed Switching Power Supplies

Or your original ps seems fine if one can set it to 39v.
.To save a few $'s and fit my " philosophy" i' d go 1200mA ldd-hw for $9...Probably reduces your per cob output to around 40-ish watts but will run cooler and a bit more efficient ( higher par/watt)

Oh and your thermal fuse is a unique feature but if your heatsink in that area gets that hot it is likely your cob is severely damaged. Temp at the led itself will be way hotter than 77c (170F)
50 c (122F) or less may work to prevent damage
Remember you are dealing w cheap Chinese cobs.

You did ask for feedback ..

Thanks for the feedback I do appreciate the review! I think that does bring to light the concept I hadn't considered as much about the current changing as they heat up.

Maybe I can switch to this driver by the same manufacturer that has a constant current control: https://a.co/d/2RqRONd and maybe add some 3-5Amp fuses in line with each array in case of failures. The power supply operates at 48V maximum and 10 amps maximum. In a way my current PS has a safety built in right now at 10Amps max output it will allow and I run close to that at right under 8amps.

I will note some observations I made during testing:
- I just ran a test, at startup it's set at 7.5 amps, after about 20 minutes or more it reads about 7.9 amps, which I assume is near stable at that point, so not a huge difference, but it is measurable. Since I'm only running at about 1/2 power I think that difference in output would not change much.
- I did try the constant current driver you linked - the 3000mA one got very hot all over after long term use ~140-150F so I was uncomfortable using that one.
- The thermal heat sinks and fans appear to be efficient, though I am only operating the LED COBs about 1/2 power and using some 100W some 50Watt ones. the heat sinks are rated/sold for 100W - they read about 80-90F on the sink nearest to the LEDs after running for about 20 minutes.
- I tested a fan failure with the thermal fuse, I didn't time it exactly, but at 100W with a constant current LED driver at 3000mA it failed within like a couple minutes of operating without a fan. hence why I added the fuse for safety, otherwise what's to tell what could happen before the LED completely burns out. It's not really for long term stability, just there to mitigate a fan failure since the fan is integral to avoiding thermal runaway. I would note, that lower thermal fuses may get increasingly more diffictul to solder without issue. I accidently ruined one when building it I think due to not being quick enough applying heat when soldering.