Algae Scrubber Basics

[Floyd R Turbo]

You might find that the smaller scrubber works for quite a while, vs a temporarily oversized scrubber which has a higher potential to have issues

As for the fixture question, in theory that should work but if you move the fixture back away then add lenses to focus the light, you end up with the same # of LEDs. Now I'll say this, that spacing guide works well when you start out at 1/2 current/power and sometimes you can leave it there, but if you want to amp up the growth you can run them stronger if you have to load to support it. So it's #1 coverage and #2 power. If you use 1/2 the LEDs and full power with lenses from further away, that's the equivalent to 50% up close. Then you're power limited on what the light output can be, but then you can add LEDs if you find that you need them - so that's another way to offset the initial cost I guess.

 

[zachtos]

So is it 2 inches on center to every LED or 1 per 12 in sq? Every 2 inches works to 1 per 6 sq inches which is double.

 

[Floyd R Turbo]

physical spacing - a grid of LEDs in a 2" x 2" pattern, it's described on that link.

Minimum coverage: One LED on each side of every 8 sq in of screen
Maximum coverage: One LED on each side of every 4 sq in of screen

It's really the maximum coverage that is 2" O.C. = each LED covers a 2" x 2" area, another way of putting it (sorry)

24 x 14 = 336

336 /8 = 42 LEDs per side
336 /4 = 84 LEDs per side

 

[karimwassef]

I prefer to use PAR as a measure. I've been successful with ~200 PAR at the screen center. 300 is better, but 200 will do.

 

[zachtos]

I prefer to use PAR as a measure. I've been successful with ~200 PAR at the screen center. 300 is better, but 200 will do.

Good to know, I have an apogee PAR meter, and have built LED fixtures twice now. 84 LED for a 12x24 area seems extreme, that's more than double I would need for a 24*24 area in an Sps tank.

 

[Fish'InMN]

I prefer to use PAR as a measure. I've been successful with ~200 PAR at the screen center. 300 is better, but 200 will do.

Everyone should take this to heart - light intensity/output measurements are way more useful than LED number and power guidelines when it comes to recommending setups. This is the first PAR number I've seen in hundreds of pages on this topic and that is an insult to everyone involved.

 

[Floyd R Turbo]

Sorry I disagree that it's an insult to suggest a coverage configuration based on experience.

I understand that PAR can be useful but the issue as it's been explained to me is that 660nm Deep Red from LEDs does not always register properly on a PAR meter because of the nature of either the source or the bandwidth. I'm no PAR expert so don't take this as slamming on PAR. This is what I've been told by people who know LEDs better than I do...but now that they are making PAR meters that are more in-tune with LEDs (apogee) this may have changed.

Worst case: PAR can allow you to evaluate the evenness of coverage, but it won't accurately tell you what your intensity really is. What I mean by that is that let's say that your PAR reading is not accurately reflective of actual intensity due to said potential issue, so it's "off" by a factor of some kind. That factor will be common across the entire illumination area, so it can be effectively "cancelled out" because the variations in readings will all be relative to each other and therefore pertinent from the perspective of relative intensity. But that does not necessarily imply that the absolute intensity is correct.

What I know, based on a lot of experience, is that you can achieve even coverage by using a DIY array of LEDs of 660nm chips that are spaced out at around 2"-3" on center place about 1.5" +/- 0.5" from the screen.

If you can either 1) dim them or 2) diffuse them, this can help to avoid initial over-lighting (photosaturation).

It has been suggested way back somewhere in this thread that there should be a "Rule of PAR" (srusso suggested this as an idea) where your PAR of your scrubber should be matched to the PAR of your tank but this was dropped as feeding was the more relevant relationship, not tank lighting...but it might come into play in certain situations, so I wouldn't say that I have ruled it out completely.

 

[Floyd R Turbo]

For instance here is an intensity reading that someone sent me, this is an array of DRs at 2" on center with HVs in the center of each "square" of reds, with the HV at 50% relative to the DRs and the entire array dimmed to "50%" and then at full intensity

I'm not sure how this translates to PAR but what he told me was that this was measured at dead center in front of the fixture, and there was little to no difference at any other point in front of the fixture or at the end even. So I know that the coverage technique works - because hotspots cause major issues, and it doesn't take much to create a hotspot.

I found out early on with LEDs that this is the case. Take an array for a 4x6 screen that is 6 deep reds in a 2x4 rectangle pattern and then put a royal blue right in the middle of that rectangle, and put one such array on each side of the screen. The result is you will see a bare spot dead center on the screen about 1.5" in diameter that won't fill in at all for months and inhibits a greater area to a lesser extent. The same happens with multichips.

Also another reason I focus on the config/coverage aspect is because one could take the PAR guideline and say "I can get better PAR with a WW CW array and use less LEDs" etc etc and you would be wrong, because there is a difference in bandwidth.

My recommendation are based on not just PAR / even coverage (however you look at it). There is also a "punch" factor. When it comes a thick mat of algae, what matters as you progress further through the growth cycle each week is how much light is penetrating deep into the algae mat and keeping the base of growth alive.

Now consider that when you compare 2 different scenarios which produce the same measured intensity on the growth substrate:

1) an array of higher intensity and lensed/focused LEDs that is something like 10" away from the screen, and

2) a dense array of unlensed LEDs that is more like 1" to 1.5" away

I would argue that the latter will penetrate deeper into the algae mat and this does, IMO, make a huge difference.

Punch without burning is a balance.

 

[zachtos]

For instance here is an intensity reading that someone sent me, this is an array of DRs at 2" on center with HVs in the center of each "square" of reds, with the HV at 50% relative to the DRs and the entire array dimmed to "50%" and then at full intensity

I'm not sure how this translates to PAR but what he told me was that this was measured at dead center in front of the fixture, and there was little to no difference at any other point in front of the fixture or at the end even. So I know that the coverage technique works - because hotspots cause major issues, and it doesn't take much to create a hotspot.

I found out early on with LEDs that this is the case. Take an array for a 4x6 screen that is 6 deep reds in a 2x4 rectangle pattern and then put a royal blue right in the middle of that rectangle, and put one such array on each side of the screen. The result is you will see a bare spot dead center on the screen about 1.5" in diameter that won't fill in at all for months and inhibits a greater area to a lesser extent. The same happens with multichips.

Also another reason I focus on the config/coverage aspect is because one could take the PAR guideline and say "I can get better PAR with a WW CW array and use less LEDs" etc etc and you would be wrong, because there is a difference in bandwidth.

My recommendation are based on not just PAR / even coverage (however you look at it). There is also a "punch" factor. When it comes a thick mat of algae, what matters as you progress further through the growth cycle each week is how much light is penetrating deep into the algae mat and keeping the base of growth alive.

Now consider that when you compare 2 different scenarios which produce the same measured intensity on the growth substrate:

1) an array of higher intensity and lensed/focused LEDs that is something like 10" away from the screen, and

2) a dense array of unlensed LEDs that is more like 1" to 1.5" away

I would argue that the latter will penetrate deeper into the algae mat and this does, IMO, make a huge difference.

Punch without burning is a balance.

Well, from my last LED build (3-4 years back), I used 36 LED on a 6x14" fixture with 60 degree optics and got great coverage on 12"x24", so I find it hard to believe I need double that for a turf scrubber.

The LED turf scrubber I built was about 6x6" heat sink with 9 red/blue LED (wrong spectrum, lighter red and higher blue, but it worked for chaeto, still in use). That gives 900+ PAR when only a few are on optics, it made dead spots on my test turf scrubber. If mounted farther away, it works great with lest waste light.



Not sure What I will do now since 84 LED/side is so expensive, seems odd that it comes out to +100% more than estimated. I think I will try to make LED fixture 10" from screen with wide optics to minimize light loss and just use correct spectrum and measure with my PAR meter. Can always buy more LED I guess if need. I can maybe make multi channel so I can only turn on lights on the right/left/middle so I can add more screens as my bioload grows. I have a 30" wide sump so I got the room.

 

[Floyd R Turbo]

One thing that might help explain the difference between chaeto and scrubber for that light fixture is water - any water penetration will kill off the red component very quickly, so what might work well for chaeto would burn the scrubber, that makes sense.

Regarding the 84/side, that is for a tight array that is really close and designed for maximum intensity under optimal conditions (scrubber sized to feeding, etc etc).

What I call out for the "minimum" of half of that (42/side = 84 total) on that same screen would still probably yield good results. You could probably trim that down further, like 25-30 per side and still get good results, same size scrubber, same distance to the screen (or maybe 3" instead of 1.5-2" and up the current a bit) for quite a while and that is because you are starting with a lower bioload than the end-goal.

Now take that and cut the screen size in half: now you're down to a smaller screen, and if you go with a bit more density than the last number, say 15-18 LEDs per side, you're at a decently sized screen with room for expansion both on the number of LEDs and the overall size of the system.

One thing to keep in mind is that my guideline is just that - a guideline. Most people that build a scrubber are building something moderately sized, like in the 7x7 range or maybe a bit bigger, and for that size, the spacing guideline will get you a lot of bang in that space. Even coverage is more critical with a smaller scrubber.

I would say that with a larger scrubber, while it is ideal to hit it with the same coverage concept, I think you start to stretch the envelope and the feeding guideline might start to break down a bit. The feeding guideline seems to hold very well for tanks fed 2, 4, 6 cubes/day etc, pretty well for tanks fed 10 or 15 cubes/day but I think much past that and you start to see a wide variance so it's hard to set in stone the size of scrubber you would need.

I'm guessing (total guess) that a graph of feeding to screen size would not be a straight line but more of an arc, where the larger your feeding becomes, the ratio might start to drop a bit (screen area might need to be less that the "12 sq in/cube/day" figure once you hit a certain point) if that makes sense. Again, not a whole lot of data, and that is also compounded by the fact that extremely large tanks (300g+) are for the most part do not have sole filtration, they have multiple forms, and all of these take up some of the load.

If everything were sized relative to feeding...it would be easy

 

[jrpark22000]

For instance here is an intensity reading that someone sent me, this is an array of DRs at 2" on center with HVs in the center of each "square" of reds, with the HV at 50% relative to the DRs and the entire array dimmed to "50%" and then at full intensity

I'm not sure how this translates to PAR but what he told me was that this was measured at dead center in front of the fixture, and there was little to no difference at any other point in front of the fixture or at the end even. So I know that the coverage technique works - because hotspots cause major issues, and it doesn't take much to create a hotspot.

I'm the one who took the spectrograph for Floyd. To elaborate on why and some of his points about PAR.

I trust PAR about as much as I trust API water test kits. Both have their purpose, the biggest benefit is both are cheap. The reason I use a hach spectragraph to test water is the same reason I purchased the spectrometer, Accuracy.

The problem with PAR is two completely different spectrums can give you a similar PAR number. PAR is based on the entire photosynthesis range. As we know terrestrial plants use a different spectrum as compared to corals, look at all the red plant growers use compared to all the blue of coral growers. By taking a mathematical average of the given photosynthesis range and saying that a given PAR measurement is better suited to a given plant or coral is incorrect. Unless you're looking at the entire spectrum (without the averaging of a PAR tool,) then tuning all the peaks of light to the requirements of the given object you are growing, PAR by itself is about as accurate as an API water test kit for po4.

 

[karimwassef]

My point is that PAR is better than Watts, or even Watts/in2.

No method is perfect, but as far as making photosynthetic creatures grow in light, PAR is a decent solution.

I've been surprised at how many light sources consumer W but under-deliver on light output in the photosynthetic range.

Sure, PAR misses the UV and deep red, but Watts misses light vs. heat... that's a much coarser estimation

 

[Floyd R Turbo]

My point is that PAR is better than Watts, or even Watts/in2

I agree in general, but if you can develop a solution that yields results, and then that particular solution can be translated to watts, then it works.

No one complained when the original solution before LEDs was based off of CFL wattage and coverage - the whole 1W per square inch thing. That as well was based off of a color temperature, use of a reflector, and distance to the screen. There were similar issue with hotspots related to the orientation of the lamp, as well as for larger scrubbers where someone wanted to use one big lamp instead of multiple smaller ones.

The difference is that the wattage is roughly half of what the CFL guideline was/is, but it is still based off of a configuration, distance, and color temp.

Until there is a drastic change in the efficiency of Deep Red LEDs, I don't see how things can vary a whole lot.

What does hose it up is when you get into an off-the-shelf fixture where they claim wattage based on the LED rating, that is, saying that a fixture with 10 3W LEDs is 30W when the actual draw is more like 15W. The latter (actual) number is what must be used.

 

[karimwassef]

That works if one LED is like another or one CFL is like another.

Realistically, I've found that assumption to be flawed in many cases. Even LEDs from the same vendor vary batch to batch. It can work, but it's inconsistent and can lead to incorrect conclusions.

It's a first rule of thumb, sure. If you don't even have enough Watts, you probably won't have enough light. But once you're over the minimum, two LEDs consuming the same W, can generate significantly different amounts of light.

PAR is a better rule of thumb...

There will be better still. It's just an evolution of improving metrics over time.

 

[Floyd R Turbo]

I'm in agreement with you there. That's the main problem with LEDs is that you can have a variance even within a bin. The cheaper the LED or more sketchy the source, the wider the variance potential. This is where you tend to get what you pay for - a more reputable supplier will have more consistent product.

That of course can compound the issue of the off-the-shelf LED fixture. One might be awesome and work perfect for one person, while another may suck royal b--ls for the next person. And, while most people are going for the best bang for their buck (going cheap), you're at the mercy of the dice roll sometimes.

 

[karimwassef]

That's where a power meter and PAR meter combined provide a clear metric for performance. I look at PAR/W and PAR/$ as useful metrics of performance and value.

Also, point PAR is a poor indicator. I prefer collecting the PAR distribution and using that to gauge true output over the desired area. With LEDs, I've gone as far as breaking out the power input into power (heat) and power (light). That's a little extreme but these characteristics really inform the true performance.

 

[jason2459]

I see no reason to believe either method is better or worse then the other. The LED's are so damn close to the screen.

 

[SantaMonica]

After looking at their video again, it looks like BRS was also testing a waterfall on the same system, which appeared to not have grown much.

 

[Floyd R Turbo]

Who is to say when that section of the video was taken, how long it had been growing for, what specific conditions there were with the tank it is on (like not much bioload, not much feeding, over-cleaning the screen which you can see that they were) when compared to the stress test tank that the HOG scrubber is on?

Also, it is not on the same system. Not even in the same room. I know because I was there last month. Guess why. Anyways the HOG shown is on their stress test tank the they use to baseline test skimmers and such.

Their office it really awesome too by the way, imagine a cube farm except everyone has a reef tank in their cube. I think it's a requirement. I mean one guy has like a 90g cube in an office cubicle that can't be much bigger than 10x10. I don't think they turn the office lights on because there is so much friggin blue glow in that office you would probably go blind if they did.

 

[bif24701]

Who is to say when that section of the video was taken, how long it had been growing for, what specific conditions there were with the tank it is on (like not much bioload, not much feeding, over-cleaning the screen which you can see that they were) when compared to the stress test tank that the HOG scrubber is on?

Also, it is not on the same system. Not even in the same room. I know because I was there last month. Guess why. Anyways the HOG shown is on their stress test tank the they use to baseline test skimmers and such.

Their office it really awesome too by the way, imagine a cube farm except everyone has a reef tank in their cube. I think it's a requirement. I mean one guy has like a 90g cube in an office cubicle that can't be much bigger than 10x10. I don't think they turn the office lights on because there is so much friggin blue glow in that office you would probably go blind if they did.

I wouldn't mind working there