Andy's Freshwater ATS - part 2

[Andy_cy]

Hi guys, just to recap on my setup, I'm running an approx. 95sqinch screen (8.26" width) ATS in my freshwater tank of ~200L with 1 arowana and 4 datnoides all juveniles. Had some problem with nitrates, which is why I've turned to ATS.

ATS was growing well with green hair before, but nitrates weren't going down. It was rising but at a slower rate, until my weekly water change comes and it will drop again.

This was how the screen looked like with 20w led floodlight + small diffuser to prevent bald spot. Lotsa green hair algae, supposedly the best type to grow on an ATS.
http://farm8.staticflickr.com/7317/8855103893_4ac83499db.jpg

However, things changed when I was experimenting with different photo cycles, and also removing the small diffuser covering the 20w led chip. And it turned out like this:
http://farm4.staticflickr.com/3747/9083726731_3607a27229.jpg

It was green in the middle and turning brown/black at the periphery. Not sure if it was algae dying off, or another type of algae taking over.

So I've decided to use CFLs instead. Made a new fixture with 2 x 23w cfl on each side of the screen, 92w in total. Photo cycle was 20-4. These are the results:

Day 1
http://farm6.staticflickr.com/5467/9148304436_76494cffa0_n.jpghttp://farm8.staticflickr.com/7315/9148325544_f546ccccb1_n.jpg

Day 3 - skipped day 2 photos
http://farm4.staticflickr.com/3678/9146104045_ec8c6fb7ea_n.jpghttp://farm4.staticflickr.com/3830/9146114343_a0e605a3d0_n.jpg

Day 4
http://farm6.staticflickr.com/5524/9146148603_7e545bfec1_n.jpghttp://farm3.staticflickr.com/2826/9148380280_f1551a8da9_n.jpg

Day 5
http://farm3.staticflickr.com/2882/9148387886_b27221240f_n.jpghttp://farm6.staticflickr.com/5451/9146167291_54d16de6a9_n.jpg

Day 6 - colour is a little off here, camera problem.
http://farm3.staticflickr.com/2881/9148400906_7496a2a841_n.jpghttp://farm4.staticflickr.com/3774/9148401798_1283ce37a1_n.jpg

Day 7 - cleaned the screen on day 8, did not photograph.
http://farm6.staticflickr.com/5519/9146176721_60cd67abfd_n.jpghttp://farm8.staticflickr.com/7327/9146177719_66f464609f_n.jpg

At day 3, I already knew these lights were not gonna work out. Because I was once told, 2 days is too short to see any difference, I decided to just let it go on for a week and see the end result. The reason why the lights failed is not because it's too weak. My glass tank design has a glass barrier between the lights and the screen, which separates them by 4", and so it causes the CFL's brightness on the screen to be significantly less. It would be good if it's just 2" away from the screen, perhaps then more greens would grow. The CFLs also produce a ton of heat and the glass barrier was very very hot, it burns to touch it.

At the end of 8 days, I got a thin mat of very dark green, almost black algae. A little of it was brownish too. Obviously a different type of algae. My nitrates tests also disapprove this kind of algae. Right before my weekly water change, my tank had 80-100mg/L of nitrates, still not getting any better.

I have switched back to the old 20w led floodlights with diffuser on, to even out the light on the screen. And will update the progress of my screen and the nitrate levels too.

I believe the LEDs will work better than the CFLs in my setup. I'll use them until I build my own led panel with 3w reds. I feel that these lights I'm testing out with just aren't efficient enough to increase nutrient uptake by the algae.

Please guide me along on the things I'm missing out on. Thanks for all the help so far, greatly appreciate :)

[SantaMonica]

Yes just go back to what gave a full screen of green; that's going to be as good as it gets. If you need more filtering than that, you'll need additional screens.

You might try letting the algae pool at the bottom.

[Floyd R Turbo]

Well, at least you know that CFLs didn't do the trick. You would not have known unless you tried.

Ill be interested to see if going back to the original setup returns the screen to the original growth.

I do want to note (mainly for anyone else reading) that it looks like you are still getting green growth but with brown slime growth on top, not the black oil growth.

This is the black oily growth

http://i611.photobucket.com/albums/tt191/FloydRTurbo/00%20Algae%20Scrubbers/Algae%20Scrubber%20Client%20Pics/Kusyairi/26thMarch_zps5786a0a0.jpg

And this is what is looks like after cleaning

http://i611.photobucket.com/albums/tt191/FloydRTurbo/00%20Algae%20Scrubbers/Algae%20Scrubber%20Client%20Pics/Kusyairi/526thMarchafterclean_zpsc6d928ed.jpg

[Andy_cy]

@SantaMonica

I went back to my old setup. Things aren't looking quite the same. The same kind of dirty green algae growth is seen. There are small spots of bright green growth, but they exist only in little spots. I feel that the dark green growth is sort of inhibiting the growth of green hair. I might do a thorough screen cleaning tomorrow.

This dirty green growth also looks slimy and tend to grow much slower with minimal 3D growth.

@Floyd

Thanks for sharing. Mine is definitely not black like that. I'm getting this very dark olive green shade with a tinge of brown. It's kinda hard to describe it, and my photos are not showing the proper colour I'm seeing.

Anyway, the algae that is growing now is much worse at exporting nitrates. Nitrates building up quicker than when I had the green hair.

I've ordered LEDs and just waiting for them to arrive. Gonna build a diy panel of 12x3W 660nm and 2x3W 455nm, for each side. Will it be good enough to encourage GHA and also make then grow more aggressively?

My guess is, using these LEDs will increase the growth rate of the algae. If this is still not fast enough to match with the rate of nitrate production, then perhaps another screen is the only answer.

Any thoughts on this? Am I going the right direction?

[Floyd R Turbo]

I hate to give a non-answer, but one of many things could have happened here. The fact that you are getting no different result could mean that something has shifted in the overall system that causes the growth to lean in a particular direction and that the change in growth had nothing to do with the change in the light source. Just another possibility.

But the end result, no matter what the cause, is that filtration is still lacking.

Post pics or a sketch of your intended LED layout before you build it (if you haven't started already)...just in case I spot something amiss. Namely, don't run the blues at full power. I don't know if you have seen my posts regarding this. I build my fixtures such that the blue LEDs are running at half-current by wiring them up in a parallel configuration within the series string. It knocks the intensity of the blues down so they don't cause photo saturation (a bald spot right in front of the blue LED)

[Andy_cy]

I have to agree with you that it does seem that the growth is leaning towards a different kind of algae, and there are too many factors to consider. Nothing in my system changed, except that I am pushing kh and gh up. Kh used to be 1dKH, now 2dKH. GH used to be 8dGH and now it is 9-10dGH. But the changes are so slight, I doubt it's due to kh/gh. My pH sits at 7.5-7.6 now, when scrubber lights come on. I don't know how much the pH falls during the night when it goes off, cos I'm asleep.

Anyway, I did a thorough screen cleaning today, brushed the screen to get rid of more basal growth, hopefully to restart and get GHA again. This is how the screen looks after 4 days of switching back to the old LED floodlights. It's a lot greener than with CFLs. And SantaMonica is right too.. It's like brown growth over a thin layer of greens.

http://farm8.staticflickr.com/7442/9180976312_14f98266d8_z.jpg
http://farm3.staticflickr.com/2837/9180967888_5bb6399a3c_z.jpg

[Andy_cy]

This is how I intend to build my led panels. I will be using 1" U-channel bars as my heatsink. It's pretty heavy, about 3mm thick aluminium. Hopefully each 9" bar can dissipate enough heat for 4x3w LEDs. Will be using 2 fans for each panel for extra cooling. I can't find the proper heatsink here so had to DIY one myself. Here is the plan:

http://farm3.staticflickr.com/2873/9181424298_bd3e3c5ae7_z.jpg

The 660nm LEDs will have 90 degree optics while 455nm will be on 120 degree ones.

Is it advisable to use blues or can I do without them? In any case if the LEDs are too strong, I can always dim them down further. Will this work? Thanks!

[Andy_cy]

Remember I said I had been dosing a bit of potassium last time, after every weekly water change? Now that I've stop dosing K, could it be that the different ratio of NPK that's favouring brown algae more than the GHA? Just a thought.. I don't have test kits for K or P, not for freshwater.

[Floyd R Turbo]

Did you start dosing K before the algae screen turned brown?

[Andy_cy]

Did you start dosing K before the algae screen turned brown?

Nope. Actually only dosed twice during the beginning. First 2 weeks, I had GHA. Then I dosed K for the following 2 weeks, same GHA. Only recently then I encountered this issue with brown algae.

Btw, is my LED setup ok? If yes then I will place the order now.

Thanks, Bud.

[Floyd R Turbo]

The 660nm LEDs will have 90 degree optics while 455nm will be on 120 degree ones.

Is it advisable to use blues or can I do without them? In any case if the LEDs are too strong, I can always dim them down further. Will this work? Thanks!

Don't get any optics with any of these, if the optics are actually attached to the LEDs. Some LEDs come manufactured with the dome shaped so that it results in a wider or narrower dispersion angle, generally I use ones with the widest angle which is usually 120 but sometimes 140.

As far as the arrangement and quantity, IIRC, your screen is 8.25 x 11.4 ish, so about 90 sq in. My current recommendation for LED arrays is one 660nm running at 700mA on each side of every 8 sq in (min) or 4 sq in (max). So:

90/8 = 11.25
90/4 = 22.5

So you're at 12 which is good and the blues will help supplement so I would use them, but wire them in a parallel configuration so they run at half current. You might add 2 more blues also, so that you have 4 "X"s, with reds at the tips and blues at the center (you have 2 of those "X"s in your diagram already...)

And I would definitely run the LEDs full out at 700mA, but having the option to dim is nice. The blue may not kick on very much when you dim them though, if they are parallel wired.

This is what I mean by parallel wired within the series

http://i611.photobucket.com/albums/tt191/FloydRTurbo/00%20Algae%20Scrubbers/DIY%20LED/IMG_0295.jpg

Close-up

http://i611.photobucket.com/albums/tt191/FloydRTurbo/00%20Algae%20Scrubbers/DIY%20LED/IMG_0298.jpg

So you would come into the first column/bar of reds, then over to the parallel blue set, then into the next column of reds, etc. Your wiring would be R-R-R-B=B-R-R-R-B=B-R-R-R where "=" is the parallel connection. The pair of blues would be "seen" as a single blue by the driver (one voltage drop) so your power needs don't change, and you get better coverage without hotspots.

[Andy_cy]

The LEDs I am ordering comes with angle choice of 60/90/120 deg for the 660s and 60/120 for the 455s. I am not actually gonna buy separate optics, just intended to get the 90deg 660s and 120deg 455s. Since my light will be roughly 3-4" away from screen, I would think that 90deg for the reds will concentrate more light on the screen. With so little blues, 120deg will spread then out.

I'm really impressed by your ingenious way of connecting a parallel between a series...why didnt I think of that?! Haha. I will order 4 blues for each side now. Connecting that way will cut down my blues by 50%. Great.

Thanks Bud, great info! Just hope these LEDs will do a better job and really want to see my nitrates <5ppm someday.

[Floyd R Turbo]

That might not be a bad idea with the 90s for the reds in your situation.

On the parallel configuration, don't kill yourself, I did quite a bit of searching on Reef Central and even asked the specific question to the LED guys about if they've done it, could it be done, has it been done, etc - nada. Apparently no one even had thought of trying it before me, no idea why that was.

The only downside I can see on it is if you have a pair of LEDs that are poorly matched, electronically. Meaning that when running in parallel like this, each LED should have the same voltage drop. What happens is that if they are not identical or close, one will pull more current than the other at the "common" voltage drop, and then that one will run a little brighter. This isn't really a big issue though. It's not the same as the "parallel strings" issue, where each string is running at full current and then if one string goes down the other gets 2x current.

[Andy_cy]

Alright, I will get 90degs for the 660nm and 120degs for the 455nm. Follow your circuit, and hopefully some real stuff starts to grow. The shipping to singapore will take another few weeks..bah! Till then.

[Andy_cy]

Any possible errors before I place my order?

http://farm4.staticflickr.com/3667/9187669713_d817c12435_z.jpg

[Floyd R Turbo]

Nope, except for the 500mA driver current, any reason for that?

[Andy_cy]

The dealer I'm getting the LEDs from said that although the LEDs are rated max 700mA, it's recommended to run them at 500mA, approx 71% brightness, to lengthen lifespan. He also said that 70% is still very bright for a 3w led.

You reckon I should just blast them at full power? My heatsink isn't the best available but I have 2 80mm fans per panel.

[Floyd R Turbo]

That's true of all LEDs, they're all going to last longer if you run them at lower than max current. I would just run them at 100%, or at least at 600-650mA. The lifespan we are talking about is 50,000 hours at rated current so that's 8 years at 16 hours/day running at 100%.

Heat is the enemy so you just have to make sure you use a good thermal grease or adhesive, and make sure your heat sink dissipates the heat adequately. For reference, there are people that build LED arrays out of U channel and put cool whites/royal blues on them and run them full out (1000- 1500mA) with probably tighter spacing than you have.

A thought regarding this, you may want to add another section of 'blank' U-channel in-between each row you have shown to allow heat to dissipate in all directions. The more aluminum, the better. If you want to see what I mean, look at acorral's EC1.0 scrubber, the one where he makes the heat sink out of U-channel as well. Also your fan would be more effective if you have a closed area for it to pull from, with gaps between the rails the air flow would be lower at the extremities. Just a few thoughts...

[Andy_cy]

I get what you mean. I will post pictures of my LEDs once it's done. Just made the order, guess it'll take a few weeks to reach me. Good to know that at least these U channels will work :)

[Floyd R Turbo]

Andy, on chriswf's thread you mentioned your LEDs and being hot, so is it this DIY array that you're referring to?

Pics? You promised!

[Andy_cy]

Andy, on chriswf's thread you mentioned your LEDs and being hot, so is it this DIY array that you're referring to?

Pics? You promised!

Hi Bud.

Been really busy lately. I wasn't referring to the new red/blue LEDs that I'm yet to build. It's the led floodlights that are heating up really badly, especially without a fan in a closed environment. But I don't give a damn about those lights anymore. My led parts have arrived!

Will post the pics soon. I'm already in the midst of building it.

One question. When I'm soldering my LEDs, specifically pre-tinning, it kinda heats them pretty much. I've tested them after when they've cooled off, and they're still working. But I wonder, will it affect it's maximum light output after being subjected to such high heat?

First time soldering :( burnt my finger...

[Andy_cy]

Sneak preview of my fine work.

http://farm8.staticflickr.com/7319/9346804321_0944a75f19_z.jpg

http://farm6.staticflickr.com/5475/9346805313_a4a34a6828_z.jpg

http://farm4.staticflickr.com/3720/9346806341_3e85bb6050_z.jpg

[Floyd R Turbo]

One question. When I'm soldering my LEDs, specifically pre-tinning, it kinda heats them pretty much. I've tested them after when they've cooled off, and they're still working. But I wonder, will it affect it's maximum light output after being subjected to such high heat?

The trick is to use a hot iron and let the solder flow onto the pad quickly. Too much heat can do damage, but generally only if you leave the iron on too long. Thus the reason you don't pre-tin LED on the heat sink, because that would not heat up the PCB quickly, you lose heat to the sink so you heat up the PCB slower and longer = bad.

This might help


http://www.youtube.com/watch?v=cpxuubdCkqQ


http://www.youtube.com/watch?v=_BqwxrfnXFc


http://www.youtube.com/watch?v=xwrEkEmWUqw

...or might have helped before you already did all this :(

[Andy_cy]

Yeah I watched this halfway through when pre-running cos I was having difficulties. This guy's soldering iron seems to be so hot, or perhaps mine is too cool? The moment his iron touches the pre-tins, it melts and the wire just fuses in an instant. For me, I have to place my iron in all sorts of position and try to get the pre-tinned blob to melt so my wire can fuse in.. And because of that I can feel the LEDs heating up pretty much. Damn... I might have to re-order all the LEDs if they don't work :( so sad....

[Andy_cy]

Is that you, Bud? Bud Carlson? Lol soldering skill is over the top

[Floyd R Turbo]

Yeah I watched this halfway through when pre-running cos I was having difficulties. This guy's soldering iron seems to be so hot, or perhaps mine is too cool? The moment his iron touches the pre-tins, it melts and the wire just fuses in an instant. For me, I have to place my iron in all sorts of position and try to get the pre-tinned blob to melt so my wire can fuse in.. And because of that I can feel the LEDs heating up pretty much. Damn... I might have to re-order all the LEDs if they don't work :( so sad....


Is that you, Bud? Bud Carlson? Lol soldering skill is over the top

LOL yup that's why I posted it

I have a variable temp iron (radio shack) that goes up to 50 watts. I crank it all the way to get it to heat up until I can tin the tip of the iron, then back it off to around 35w or so. That's where you want to be, not so hot that you get rosin spatter when you tin, but not so cool that you have to hold the iron in place for a long time.

If your LED still lights up, it's fine. You may have decreased the useful life a bit, but that's won't show until one burns out.

If you put the iron right on the blob and it won't melt, either your iron is too cold, or your tip is dirty. A dirty tip (too much solder, or a solder/rosin mixture, etc) will prevent heat transfer. you need to clean the tip often.

Also you should not put the iron on the pre-tin to melt the solder then insert the wire, nor should you melt the solder on the iron and let it flow to the pad. Use the iron to heat the wire so that the wire gets hot enough to cause the solder to melt. When pre-tinning, use the iron to heat the pad and then insert the solder wire near the pad/iron junction so that solder flows to the pad as it melts (when you first touch the iron to the pad, some of the solder on the iron should "snap" onto the pad). That's at least the ideal case, otherwise you risk ending up with a "cold joint" which can release at some point.

[Andy_cy]

If you put the iron right on the blob and it won't melt, either your iron is too cold, or your tip is dirty. A dirty tip (too much solder, or a solder/rosin mixture, etc) will prevent heat transfer. you need to clean the tip often.

Best advice. I'm using a 45W normal iron. It doesn't have fancy red/green lights nor temp controller. So I guess I'll just have to make do with what I have and try to get it over and done.

Well, my LEDs are still lighting up for now, I hope it only reduces useful life and not the maximum light output. I'll finish it up by this weekend and post more pics on it.

Thank you so much, Bud. Your help is EPIC!!