Chris Fowler's Blog

[+dapettit]

What size are those containers?

[chrisfowler99]

50 gallon.

I need some new pictures. It's finished up and working, finally.

[chrisfowler99]

Build shelving.

This thing is probably over-engineered, but if that top tank is full it's 350ish lbs. Better over-engineered than under-engineered.

My time this weekend was somewhat limited due to watching my daughter while the wife was at work...and watching football, but I did get the shelving built.

It's slightly oversized length-wise so I can get the bottom tank in and out. I wanted the width as narrow as possible.

Next step...plumbing. Hopefully I can get through that in the next week.

2/18/10 update:

Finished!

It was actually (mostly) complete a few weeks ago, but I finally got the ATO setup working the other day.

A few pictures and a description of what's going on.

The tanks are 50 gallons each.

The top tank is RO/DI and will fill to about 40 gallons with a float switch to cut it off.

The bottom tank is for mixing. I mix up about 30 gallons at a time. My water changes are 10-15 gallons, but having spare salt water makes me feel better.

Using a Mag 9.5 pump for mixing and sending saltwater to the tank for water changes.

The "gift wrap" container is an idea stolen from Melev's Reef. It holds about 8-10 gallons of water for ATO.

Each function requires only one ball-valve to be opened.

The ball-valves, left to right, top to bottom perform the following functions:

Top row:

1) Fill ATO

2) Fill mixing tank

3) Fill "container" - allows me to get RO/DI for "other" uses...like thawing food

Bottom row:

4) Fill tank

5) Mix saltwater

This is actually two lines. One PVC for water changes and a 1/4" line running to an Aqualifter pump for ATO.

My tank is in a room on the other side of that wall.

[DaJMasta]

Very nice! An elegant solution to an otherwise complicated problem. Now I want to see an automated water changer come out of this

[chrisfowler99]

At this point I have no solution to automate the draining for water changes.

I need a larger sump with a 15 gallon return area so I can pump out 10-15 gallons and then fill it right back up.

But this has greatly simplified the issue of running a hose out the garage, through the front door and to the tank or ATO container...especially the ATO container...

[mFrame]

Where did you get the containers and how much did they cost?

[chrisfowler99]

http://plastic-mart.com/class.php?item=912

More expensive than the barrels...but translucent and only 19" wide, which means I could keep them closer to the wall.

(and they look cool)

[+mcallahan]

makes me want to move to a house where the tank backs up to the garage

[chrisfowler99]

Does simplify things a bit.

Otherwise you're looking at having to plumb through the attic.

I'm just glad my wife didn't scream when I cut a hole in the wall.

[chrisfowler99]

So, I'm building a DIY LED lighting setup for the 265g tank that I picked up from Chad and Belinda.

There are several reasons that I'm doing this, but it mostly comes down to: "Because I want to."

When I say DIY, I'm talking about:

- Building an Arduino controller

- Building LCD drivers

- Drilling heat sinks

- Mounting and wiring LEDs to the heat sinks and controllers.

- Mounting optics to the LEDs to penetrate a tank with a 31" depth.

- Building a frame to mount the system

- Wiring and mounting the whole setup

The initital setup will be 144 total LEDs, but I'm designing with the idea of a possible expansion to 216 if I'm not satisfied with the color balance of the setup when complete.

The initial distribution of LEDs will be:

72 Royal Blue (Cree XP-E)

60 Cool White (Cree XP-G)

12 Neutral White (Cree XP-G)

It will be built as 24 strings of 6 LEDs.

12 heat sinks with 12 LEDs per heat sink. Each heat sink will have a string of 6 Royal Blue LEDs and a string of 5 Cool White and 1 Neutral White LED.

If I expand to 216 LEDs it will be another string of 6 blue LEDs on each heat sink.

I may end up, at a later date, replacing one Royal Blue LED per string with a Cyan, but I'll see how it comes out this way first.

I've been working the past few days on building the Arduino controller and as of Saturday (11/20) evening I was able to boot and install a controller program.

Without the LCD and a few components:

With the LCD, programmed and booted:

What you're seeing in that image is a clock and the configuration of the four PWM (Pulse Width Modulation) outputs on the controller.

The PWM outputs are adjustable from 0-255, with is the number of pulses per second. 0 is off, 255 is full on, and it can be adjusted anywhere in between to affect the "brightness" of the LEDs.

The controller, as currently programmed, allows for setting each PWM output to control when a group of LEDs come on, initial brightness, maximum brightness, how long to ramp up from initial to full brightness and then the same controll for the way down.

Initially I will only be using two of the PWM outputs. One to control all of the blue LEDs and one to control all of the white LEDs. If I add another group of LEDs I will use a third PWM output.

[DaJMasta]

Looks like it will be a nice setup. I've been wanting to go with LEDs for a while.... and the automated controller would help a lot (a little basic programming and you could easily have your lights fade in and out when turned on and off, and near complete control over the color spectrum).... but it's just so much of an initial investment. Since I already have decent lighting, I can't justify it simply as an upgrade

As for the light colors, I hear that people almost always like more blue than white, so I would expect the blues to be at a higher level. People also complain about the relatively tight spectrum of LEDs, so I'd consider adding some more variety. Maybe drop 12 of the cool whites for 4 red and 8 purple? Not that the current setup is an issue, just with so many LEDs the potential for a lot of variety is there. I think 144 LEDs will be enough provided they're driven at over 700mA.... I could be wrong but that's a hell of a lot of lighting power

[chrisfowler99]

Right now my blue to white ratio is 1:1, which is the reason for the additional capacity. The 72 whites should be enough for growth, but the 72 blues might not be enough for the color balance that I want. So I have the option of adding another 36-72 blues.

As far as color goes, from what I'm reading, red is overpowering, and you get what you need out of that spectrum with the natural white. And violet does add a lot of pop, but much of the violet LEDs color is out of the visible spectrum. And most of the visible spectrum is present in the royal blues.

The color that I may consider adding if I end up adding another string of blues is cyan.

Though it isn't outside the realm of possibility that, if I add another string of blues to each heat sink, that I will include a red, violet and/or cyan.

I'll probably pick up a couple of each if/when I do expand to test.

[DaJMasta]

Yeah, I don't see them as much in DIYs, but some of the prebuilt few hundred watt hoods often include a few reds. The cyan is a blue that is a higher wavelength than the royals?

Either way, you'll be able to see if it's not enough blues with the dimming controls, then it's just a matter if you want to run the whites at full blast then will there be enough blues.