John's DIY LED/T5 Hybrid Build

[jbdreefs]

I ordered most of my parts a week ago, so I figured that it was time to start writing and describing my build.

GENERAL OVERVIEW:

My tank: 60 gallon (48 x 18 x 16 h).

It seems as though the "full spectrum" LED fixtures are the new craze. However, I'm not going this route. Why? The recent success I saw using T5HO bulbs was enough to convince me that my future lighting fixture had to include T5's. I would go 100% T5's if it weren't for the higher energy costs and the fact that corals just look soooo much better under LED's. The obvious solution: I'm incorporating both.

A couple years ago I made my own DIY led fixture with the white and blue 3w LEDs: http://www.atlantareefclub.org/forums/showthread.php?t=64188">See it here</a>. It actually worked pretty well. My 2nd attempt is going to take things up a notch in form and fashion.

Generally speaking, this fixture is going to be three [IMG]http://www.ebay.com/itm/30W-Cree-XLamp-XP-E-XPE-White-Blue-LED-Light-30v-36V-1000mA-For-Coral-Aquarium-/221314831548?pt=LH_DefaultDomain_0&var=&hash=item33876454bc">30w Cree Multichips (blue & white combination)</a>, combined with 10 3w generic royal blue LEDs and two 54w T5 bulbs of the 10000k variety.

The LEDs will be controlled via PWM coming from an Arduino board. This will be my first attempt at working with Arduino. If you are not familiar with Arduino, [IMG]http://www.arduino.cc/">click here.</a>

Honestly, the build should be fairly simple. I'll do my best to explain the methodology and take pictures along the way.

More details to come...

 

[labernathy]

Tagging along

 

[jbdreefs]

POWER TRAIN #1:

How to drive LEDs seems to be the most intimidating aspect of the DIY LED fixtures to those unfamiliar with the process. I'll try to briefly explain:

How many LEDs and I attempting to power and what is the voltage requirement per LED? For example sake, let's assume that we are trying to power 10 "3W LEDs" and the voltage requirement (usually denoted as VF or 'Forward Voltage') is 3.5 volts DC. When assessing compatible drivers, this example would require approximately 35 volts (3.5 x 10). The math is simple for voltage. We simply add the voltage requirement for each diode.

Current (or Amps) requirements can be more complicated. However, if we are talking about one string of LEDs wired in series (or one after another, connected + - + - + - and so on) then the current requirement is rather easy. We simply look at what the LED's are rated to handle. Most 3w LED's fall somewhere around 750ma (or 0.7 Amps). In this example we would need a driver that outputs a "constant current" of 750ma which is the rating for one of the LEDs. It is important to note that reliable LED circuits utilize a constant current DC source. You will find all sorts of drivers including constant voltage drivers. What you would be looking for is the "constant current" versions.

Still using this hypothetical example, we would be looking for a constant current driver that is capable of outputting approximately 35 volts DC at 750ma. Easy right?

Now, I'm going to throw a curve ball. The DC power source I have selected is a http://www.ebay.com/itm/New-AC-to-DC-24V-10A-240W-Regulated-Switching-Power-Converter-Supply-Silver-/170855720740?pt=LH_DefaultDomain_0&hash=item27c7cb6724">24V constant voltage switching power supply</a>.

BUT WAIT JOHN! YOU TOLD US YOUR LED CHIP REQUIRED 30-36V!!!!! [I]You are so clever!</em>

Technology has provided all sorts of ways to manipulate electricity. One such manipulation comes from the use of a [IMG]http://www.icstation.com/product_info.php?products_id=1626#.U1XuYqwU81s">Boost Converter</a>. A boost converter essentially does what the name implies; it boosts the power. I've ordered the boost converter found by following the hyperlink, and it is going to allow me to take my 24V source and boost it up to 30-36V as deemed necessary.

Some of you may ask, why didn't you just get a 36V power supply. The answer really boils down to the readily available and affordable relays that I am planning to use. These relays have a 30V DC maximum rating.

 

[jbdreefs]

POWER TRAIN #2:

John!!! You have already contradicted your original example twice. What about using the "constant current" drivers? Everything you have ordered so far appears to be constant voltage. </em>

The basic power plan includes the use of one more http://www.ebay.com/itm/150-1500mA-Buck-Regulator-LED-Driver-for-1-50W-High-Power-LED-/161026417634?pt=LH_DefaultDomain_0&hash=item257dec23e2">component</a>. This is a DC to DC constant current driver that will take the constant voltage source and turn it into a constant current source for the LEDs. I will have three 30w multichips paired with three of these drivers wired in parallel.

[I]Why parallel? </em>

The source voltage going into the LED driver will be somewhere around 36V. As we noted above, if I wire everything in series (one after another), then we have to add the three voltages: 108V (36 x 3). I'm sure there is a power supply some where in the world capable of this DC voltage, but it's probably very expensive. I didn't even bother to look for one. Rather, I'm going to wire the three drivers in parallel which allows 36V to flow through to each driver.

[I]Why wouldn't we do this in your example? </em>

When we string diodes in parallel the current becomes divided. Even through each driver will receive 36V, each driver will be receiving less current (if we assume the current is constant). The good news is that the boost converter I have selected will automatically adjust the current such that it maintains the constant voltage that it has been set to produce.

[I]How does changing the current have anything to do with the voltage? </em>This is where I declare that I'm not an engineer and refer you [IMG]http://en.wikipedia.org/wiki/Ohm%27s_law">here</a>.

 

[jbdreefs]

Look what came in the mail today! A smashed 24 volt power supply!

alt="" />

I bent the enclosure back into shape and inspected the PCB for obvious issues. It doesn't appear that the damage was bad enough to disturb the guts. I'll plug it up tomorrow and make sure it works.

 

[mikesmith34]

That is a real pisser.

 

[jbdreefs]

Mikesmith34;951443 wrote: That is a real pisser.

Tell me about it! I hope the remaining parts arrive in better shape. When it take 14 days for shipping, a bunch of back and forth could seriously delay things...

 

[jbdreefs]

More parts arrived today!

I've received my http://www.icstation.com/product_info.php?products_id=1617#.U1yS1qwU9T0">microcontroller</a>, [IMG]http://www.icstation.com/product_info.php?products_id=1626#.U1yTD6wU9T0">boost converter (to 36v)</a>, [IMG]http://www.icstation.com/product_info.php?products_id=2135#.U1yTgKwU9T0">real time clock</a>, and some miscellaneous items like a breadboard, jumper wires, etc. The size of these things is a little mind boggling. The only item that was about the size I expected was the boost converter. Everything else was smaller. I included a few coins for size perspective.

[IMG]http://i1198.photobucket.com/albums/aa459/jbdiaddigo/Mobile%20Uploads/20140426_211141.jpg alt="" />

I'm waiting for the LEDs and drivers to ship from oversea. I placed that order a couple days after ordering the above components.

Tomorrow, I'll have to see if I can get the arduino on-board LED to blink.

For those not familiar with Arduino, you can check these channels out on youtube:

Jeremy Blum has a pretty nice tutorial series.



I also found this tutorial series very good. This guys does a great job explaining things in ways that make sense to you and I, the average person.



Hoping to start building next weekend, but we'll see.

 

[jbdreefs]

Well, those links to the videos didn't work. Try these...

Jeremy Blum:
<span style="color: #000000"><span style="color: #0000bb"></span><span style="color: #007700">[</span><span style="color: #0000bb">youtube</span><span style="color:">]</span><span style="color: #0000bb">fCxzA9_kg6s</span></span><span style="color: #007700">[/</span><span style="color: #0000bb">youtube</span><span style="color:">]</span>
<span style="color: #000000"><span style="color: #007700"></span></span>
<span style="color: #000000"><span style="color: #007700">The other:</span></span>
<span style="color: #000000"><span style="color: #007700"></span><span style="color: #007700">[</span><span style="color: #0000bb">youtube</span><span style="color:">]</span><span style="color: #0000bb">09zfRaLEasY</span></span><span style="color: #007700">[/</span><span style="color: #0000bb">youtube</span><span style="color:">]</span>

 

[jbdreefs]

Well...

 

[jbdreefs]

 

[jbdreefs]

A few 'firsts' tonight:

1) I soldered pin headers to a PCB and wired something up to the Arduino board (if you notice in the pictures, I'm actually using a clone). I was a little nervous about the soldering. However, it was a breeze. Watching youtube videos really helped.

2) I uploaded the 'Blink' code to the Arduino and it worked!

3) I downloaded a library for the DS1307 real time clock chip that I purchased. After some tinkering with the file location, the library is functional. I was then able to connect to the RTC using I2C via a sample code.

4) I then combined the blink code and the RTC code together and through in a couple 'if' statements to make the on-board LED come on at a certain time and turn off at another.

I can tell that this project is going to be VERY FUN!

RTC with headers and wired to the microcontroller:

alt="" />

This is the computer reading data from the Arduino which is telling me what time it is:

alt="" />

 

[jbdreefs]

I guess now we wait. I'm hopeful that my LEDs and Drivers will arrive this week.

The next steps for me are to:

1) Become more familiar with the Arduino environment and writing code. I'll probably pick up a few goodies tomorrow. I'm thinking that I should get a couple potentiometers and some little LEDs to play with. While the ultimate goal is to have an automated/programmable lighting period, I think it will be a good idea to start with writing a code for manually dimming the LEDs. These components will let me practice the concepts.

2) When all parts arrive, I'll hook it all together and pray that it works (it should but shipping can be rough sometimes).

3) Integrate the Arduino into the LED circuitry.

 

[northgahillbilly]

Tagging along, Ive had a MH LED fixture in mind for a while and the Arduino got me thinking about all the possibilities out there.

 

[jbdreefs]

NorthGaHillbilly;952905 wrote: Tagging along, Ive had a MH LED fixture in mind for a while and the Arduino got me thinking about all the possibilities out there.

The options are only limited by your willingness to learn and your imagination.

More parts arrived! Just waiting on my driver boards.

alt="" />

I also picked up a 'dishwasher' plug from Fry's today and tested my 25v power supply. All okay so far..

alt="" />

I am sure that my wife will insist on me staying awake to monitor the storm, so I may get around to playing with arduino a little more. Planning to write a code that will fade a couple LEDs based on readings from potentiometers.

 

[jbdreefs]

I found a few useful calculators today:

LED resistor size: http://led.linear1.org/1led.wiz">http://led.linear1.org/1led.wiz</a>

Wire gauge calculator: [IMG]http://www.solar-wind.co.uk/cable-sizing-DC-cables.htmls">http://www.solar-wind.co.uk/cable-sizing-DC-cables.htmls</a>

 

[jbdreefs]

Drivers arrived...

alt="" />

 

[jbdreefs]

Well, the boost converter isn't boosting. It's just passing the 24v source through to the LED driver. I did a little searching, and found that sometimes a faulty boost converter can be caused by a faulty potentiometer. I read about testing a pot. I'm supposed to be able to test it with a multimeter set to read ohms. No matter which three pins I touched, the reading was zero. I don't know that this is even possible because I could not find any information about what this means. My gut tells me that the pot was not soldered correctly and the pins are effectively soldered together which is therefore bypassing the pot completely.

Tomorrow, I plan to drop by Fry's and buy a replacement pot along with a few things to make removing the old pot easier. If replacing the pot doesn't fix the boost converter then I will probably older another from a different seller (and different model) located in the US so that I receive it faster.

Wish me luck. If you have experience with the pot issue, I'd love to hear about it.

 

[jbdreefs]

FWIW, the 24v fired up the LED chips, but they were very dim which is to be expected. I didn't run the power to the LED chip very long for fear of screwing something up. If I'm not mistaking, I believe low voltage causes problems just like when any other parameter is out of whack.

 

[jbdreefs]

I had to travel to Cornelia last night, so I didn't have a chance to swap out the potentiometer. I did it this evening and I am happy to report that the boost converter is now working. I'm about to fire up the driver and LED. I'm hoping for copious amounts of blinding light.