LED Array for a Pico

[Llamaguy]

has anyone ever looked into a controller for an array? so that instead of just turning some off or on early, the rows light up indivually for a cascade effect and the same for dusk to mimic the sun actually moving across the tank. I don't know enough about programming to do it though

[sammydee]

I'm looking into doing one of these LED arrays for my 20L tank. I was thinking maybe 400 whites and 100 blues. I have access to a laser cutter, so cutting the acrylic is no issue. 400 leds at 20ma each makes about 90 watts of lighting, plus about 20 watts of blue 'actinic'. Would that be the equivalent of a 250 watt MH bulb?

 

Sam

[sammydee]

Neanderthalman - you run your LED's at 4 volts right? I don't want to run them that high, but equally 3 volts is way too low. Is there any other option from a 12 volt powersupply to get about 3.5 volts without a resisitor?

 

Sam

[zachtos]

Neanderthalman - you run your LED's at 4 volts right? I don't want to run them that high, but equally 3 volts is way too low. Is there any other option from a 12 volt powersupply to get about 3.5 volts without a resisitor?

 

Sam

 

if your'e going to spend all this time and money... dont create an array w/o current limitting resistors... you can get 100 resistors for $3 at allelectronics... CHEAP INSURANCE on something you put SO much effort into.

[sammydee]

Actually I was thinking of using a voltge regulator to create a high frequency switching circuit, and adjust the voltage this way.

 

The problem with resistors is that they just dump the excess energy into the tank as heat - neanderthalman has already demonstrated that resistors are NOT needed for a stable LED array.

 

Try entering four 3 volt LEDs pulling 20ma into here: http://led.linear1.org/led.wiz - it says to use a 1 ohm resistor. Is that tiny amount of resistance going to make all that difference in current?

 

Sam

[dga]

A Note on Wall Transformers:

 

Most AC and DC Transformers are not regulated unless specified. They provide a specified voltage at a specified load. If the load is lower than specified, the voltage will be higher then specified. With a higher load, the voltage will drop below that specified. An example would be: A 12 volt 500ma transformer may read 20 volts with no load or 9 volts with a 600 ma load.

 

this is important info on why many builders arrays, burn out there LED's all transformers must be metered WITH a load b4 hooking up to any given amount of LED in series.

[sammydee]

I do not think that is the case with computer power supplies. These supply a very accurate voltage because they are designed for high quality electrical equipment.

 

Neanderthalman - are these: http://cgi.ebay.co.uk/ws/eBayISAPI.dll?Vie...AMEWN%3AIT&rd=1 the white leds you used?

 

Also - you know about electronics right? Well I want to regulate the voltage of my entire array using a voltage regulator. Am I right in thinking that they work by switching on and off at very high frequency? In that case would I be able to hook up one inline with the power supply to my array and adjust the frequency with a sort of dial/knob thing? Or do they come premade with a certain output voltage?

 

Otherwise I will try and use a PIC chip and see if I can find a 1khz+ 12 volt transistor that can take 10 amps. Reckon they even exist?

 

sam

[neanderthalman]

Computer power supplies are regulated, which is why I recommend them. Zachtos does bring up an excellent point, resistors are cheap, and it doesn't hurt to use them as insurance. Some wallwarts are electronic, and do not have a large iron-core transformer. Such wallwarts are typically regulated to some degree, though probably not as well as a computer power supply.

 

Sammydee, those are the exact LEDs I used, and I'm already working on a PIC controller for my array. I'm not sure how voltage regulators work. As for the power FET that can handle 10A, yeah, they make them. I ordered some a few months ago that have a logic-level gate voltage and can handle 92A. 92 A. Not joking here. Apparently, the case limits them to 80A, however. Personally, I wouldn't feel comfortable pumping 80A through one of these little buggers, but they'll certainly handle 10A. The max Drain-Source voltage is 30V, so it's fine there too. The total cycle time On-Off is about 530ns, so you should be able to get it up to almost 2 MHz switching frequency. Datasheet

 

If you want 3.5V from a 12V power supply, you're going to have to use resistors or do some sort of PWM to lower the RMS voltage. The part number is FDP8896. I got them at Fairchild Semiconductor's website, www.fairchildsemi.com. You can even get them to send you free samples.

 

 

 

 

Now, to change the tone of my post entirely. I'm going to recommend that you not spend any money on LED lighting at the moment. This was a bit of an experiment to see if corals could be kept under LEDs, and I'm questioning whether the experiment was successful.

 

For reasons I've never been able to narrow down, I have a colony of zoos that have been closed since shortly after I built my array. Everything else in the tank has been happy and open, so I chalked it up to either a water source issue that I had around the same time, or to just one of those weird things that zoos sometimes do.

 

Yesterday, I put my tank under my new 150W 20K MH for a color comparison. While the lighting looks nearly identical in both color and intensity, the colony of zoos is now starting to open under the MH. For the record, my other zoos have been fine under LEDs. At best, all I can say is that you might have trouble with individual corals or colonies under LED lighting. I'm going to keep my tank under the MH for the next week and see what improvement there is. This *could* be a fluke, but it doesn't look that way at the moment.

 

Give me a week to see if the colony opens up further. If it does, it is quite possible that LEDs are missing some important part of the spectrum.

[zachtos]

Even if you do find that some part of the spectrum is missing, some LED manufacturers may cover that spectrum better then others. Or if they are brighter they may hit that spectrum better... Its tough to give good results w/o spectrographs and PAR meters.

 

- but I am interested in trying this out on a small system as an experiment also... w/ possible future scaling up plans w/ a deeper tank.

[neanderthalman]

That's a good point, other LEDs may have a better spectrum than what I used. It could also be coincidence. I WAS seeing growth in my fire and ice zoo colony, up to 9 1/2 polyps from the six I got in november. I'm just getting mixed signals from the little buggers. Some of my zoos seem to like the LEDs, some zoos seem to not.

 

Because of quantum mechanics, you can't have "upscattering", where low energy photons from the LEDs trigger the release of high energy photons from the phosphor coatings. If a part of the spectrum is missing, it'll be the high energy, low wavelength blue-violet end beyond about 460nm (in my array) or so.

[sammydee]

Perhaps some UV LEDs may be a good idea then, to ensure full spectrum coverag of the high end - there is bound to be some bleed off into longer wavelengths, so the UV, blue and white leds should between them pretty much cover the lot.

 

Maybe I should throw in some 6000K LEDs? That could also be part of the spectrum which they are missing. I still aim to use this LED array, partly because it is so economical with power, partly because they have such a long lifetime, and partly because it is very easy in a powercut to run them off a twelve volt car battery backup. Maybe I could supplement it with actinic power compacts?

 

BTW Neanderthalman, what voltage are you running your LEDs at? 4 volts or 3 volts each? I was thinking, you could use the pic chip to regulate the voltage exactly to what you want, without wasting energy as heat like a resistor would.

 

Sam

[neanderthalman]

I'm running mine at 4V each, slightly beyond the max forward voltage of 3.8V. You can't use a PIC to truly regulate voltage, but you can do Pulse-Width Modulation to reduce the RMS voltage. What it does is, say, pulse the LEDs at 4V, but pulse them so that they're only on 75% of the time. This would effectively run the LEDs at 3V. These numbers might not be totally accurate, as LEDs are non-linear devices, but the idea is there.

 

If there is an issue with spectrum, then it might be a good idea to supplement them with PC lighting of some kind. Keep in mind that I haven't yet made any conclusions as to whether or not the behaviour of my corals is a coincidence.

[zachtos]

maybe the LEDs don't penetrate water very well compared to that halide. I'm unsure as to if LUX is an indication of water penetration or not... That was one thing I wanted to find out in order to scale up in the future, Is if LEDs can penetrate 18" of water well. The other was to get the LUX and PAR readings of LEDs in and out of water. My array plans to use white LEDs and some UV 410nm range LEDs to supplement the missing wavelength, since 460nm is the major part of the whites.

 

Other possible causes I would point out either water paramaters or possible chemical warfare of leathers/softies. Extra evaporation from halide increased salinity? Maybe even increased temperature from the halide induced opening of polyps? shooting in the dark w/o any info though.

[neanderthalman]

That's Exactly why I'm not jumping to conclusions here. There could be other factors at play. The temperature has remained constant at 80F, the salinity is stable at 1.024, though I'm topping up twice as often with the MH. I don't have leathers either, nothing but zoos and clove polyps.

 

One possibility I mentioned was a bad water source. I switched water sources last summer, and it was contaminated with something that really freaked out everything in the tank. My GSP whithered and died over the course of the next few months, and that one colony of zoos has also been slowly declining. Another colony that was in there during this contamination incident was also affected, but they reopened after a few weeks.

 

Whatever toxin it was, it would have been absorbed by the colony. After removing the toxin from the water via water changes, that toxin *might* gradually leach back out of the tissues into the water column. They would then be removed by my weekly water changes. The level of toxins in the tissues of the colony might be dropping to safe levels, causing them to open. The second colony, that opened right away, could simply be a little hardier and resistant to the toxin. It could simply be coincidence that the toxins are dropping to safe levels at about the same time I switched lights.

 

As for the penetration, light is light, it doesn't matter what the source of the light was. A 460nm photon from a fluorescent tube will have the same penetration as a 460nm photon from a MH tube. So long as the intensity of each wavelength is the same (number of photons), then there should be no way to differentiate between the two light sources. If you can get LEDs to output the same spectrum and intensity of MH, it will penetrate to the same depth.

[zachtos]

try a skimmer or running activated carbon to rule out water quality if you haven't already.

[neanderthalman]

I would, but because of space limitations (and $$), I can't run a skimmer, nor do I have anywhere to put carbon, besides loose in the tank. I've got about a 6" square footprint to work with for the tank size. That there is one of the big reasons I'm getting a bigger tank. I'm considering plumbing this one into the same sump.........

[Llamaguy]

too bad this is "nano" reef, otherwise I would say to use a 4ft tank with an MH on on side and an LED array for the other to completely rule out water.

 

Hopefully you guys figure this all out, color combinations and everything, so that by the time I'm ready for a 75 or 90 I can use them

[sammydee]

Spectrum and intensity are the deciding factors, whatever the light source. Most likely the LEDs didn't give out enough UV light. A real reef in a tropical zone gets blasted with huge quantities of very strong UV light - this probably gets replicated far better by a MH bulb than LEDs which produce a very clean visible spectral curve.

 

Perhaps swopping back the LEDs for a bit and seeing what happens will help us find the root of this problem.

 

On another point, I ordered some of those power MOSFETs you suggested Neanderthalman. My dad can get hold of £1 picaxe chips from his school that I should be able to use.

 

I was thinking of using the chip to pulse the LEDs on and off very quickly (say about 2khz). I could change the length of time on/off using a variable resistor to an input pin on the chip (or I could have the chip automatically time night/day cycles. You could even use a LDR to automatically adjust the brightness of the LEDs depending on the ambient light.) so I could manually change the brightness of the LEDs, setting it to a nominal average 3.5 volts or whatever. Looking at the specs for those FETs, it's difficult to tell exactly what voltage is need to turn them on or off. At one point it says 4.5 volts, at one point it says 10, and another it says min and max are 1.2 and 2.5 respectively. What is the minimum and maximum switching voltage for these FETs?

 

I've just realised, you've got a 150 watt MH over a 1.6 gallon tank. That is insane lighting.

 

Sam

[neanderthalman]

Sounds similar to what I'm working on. I decided to have the intensity change automatically, with a sunset and sunrise.

 

There's no need to pulse at 2kHz either, the human eye has a threshold around 12-16Hz. You could do 100Hz and you'd never see it flashing.

 

It will also help with propagation delays screwing with the timing - I've anticipated problems with trying to make my array flash too quickly, after a month or a year, the time of sunrise or sunset would gradually shift, and be noticable. An extra 1/100th second propagation delay becomes an error of three days by the end of the year if not taken into account. Just a thought.

 

Hey Llamaguy, I'm building a 10g with a sump. Once it's set up and running, I'm going to be tearing down my 1.6 and rebuilding it. I've made some mistakes as it was my first tank, for starters, I used Florida Crushed Coral....blah!. When I do, I'm going to drill it for a couple of 1/2" bulkheads. One will drain to my sump, the other will be a return, using the same return pump as my 10g. Eventually I'll have corals under MH and corals under LED, using the same water, at the same depth. That should rule out water quality, but this is a project fairly far in the future.

[sammydee]

Apparently inverts have very good eyesight so a 100hz flicker would be visible to them - just something I read somewhere.

 

Sam

[Gomer]

Yesterday I got my 300 white LEDs and 50 blues. I got a piece of scrap acrylic and drilled out a bazillion holes. Already have 1/2 of the soldered up. I hope to get the other 1/2 today as well as some photos.

[neanderthalman]

Can't wait to see those pics Gomer

[Gomer]

Wiring would be much faster if I decided to have the left and right side be independent as opposed to alternating rows. I did rows so that I can have dusk/dawn effects with two timers.

[Gomer]

I wired it up in the lab, and I sure have a bright spot light LOL.

 

I'll bring it home along with a 2x13AH supply delux as a comparison when I shoot some photos.

 

EDIT:

I happen to have the 2x13 in the lab also (I have a tank in the office). The LED setup visually appears a fair bit brighter. This is part of the beauty of focused lighting. at about 15ft away, the spot light is about 3ft diameter (guessing these numbers).

[neanderthalman]

Awesome! How does the color look?