Most small and medium sized overflow systems use gratings to keep fish and debris from clogging the plumbing. This is another perfect application for 3d printing, so I quickly drew up a first attempt in CAD and printed a test piece.
This first design restricted water flow a little too much, and raised the water level in the tanks dangerously high.
The second iteration of the overflow covers allowed more flow, but had a different problem...
The revised grating separated the falling water into narrow streams, which becomes very noisy if the water level is low.
To fix this, the final iteration of the overflow grating was designed to push the water back against the surface of the overflow so it can silently slide down the glass.
The final gratings raise the waterlevel only a few millimeters compared to an unobstructed overflow and are slightly more discrete than the original design.
Many of the printed parts I have made in the past could also have been produced through other methods; however these gratings would have been nearly impossible to design and fabricate without 3d printing due to their geometry and the number of trials that were needed.
One of the final prints for the bookshelf will be dividers to separate the equipment and display sections of the sump. As usual, the design starts in CAD.
These were fairly challenging to print due to their size, but I was able to successfully make a complete set.
When assembled they hold a square of filter foam. There is also a grated opening along the top to allow emergency flow if the foam gets blocked.
The dividers make the sump look finished and strike a good balance of style and discreteness.
One of the first decisions that must be made when setting up a planted tank is how much light to use. High light aquariums have access to some of the most interesting and colorful plants, but require more maintenance and care. Low light aquariums are better at taking care of themselves, but grow more slowly and have limited plant selection.
However for those who are indecisive, adjustable lights are the best of both worlds. The lights I purchased do not provide dimming out of the box, but their simplicity will also make them easy to modify.
I was not a fan of the plastic caps that came with the lights, so I removed them and soldered new cables directly to the PCB.
The connection point was glued in place to avoid stressing the solder joints. These lights will be connected to a programmable dimmer, but first I would like to build backlights for the tanks as well.
I experimented with several backlighting arrangements but the simplest one turned out to be the best. Directing a strip of RGB LEDs towards the wall behind the tank produces an even color across the frosted tanks.
Since I wasn't sure of how well it would work, my temporary backlights were made from foam board. Quite a lot of color shows through even with the main lights at full power.
Controlling this many LEDs has the potential to be expensive, so I searched for some time to find an appropriate device. Although it would have been fun to control every channel separately, in total there are 4 main lights and 12 RBG backlight channels. Systems capable of independently scheduling 16 channels of LED lighting are just too expensive for this project.
Instead I discovered the FLS-PP, which presents itself as a phone-controllable light-bulb, but is actually designed to power four channels of LED lights.
This device allows four channels of LED lights to be remotely controlled through a phone. More importantly, the lights can be controlled by a programmable schedule.
Since the device has a limited number of channels I will be wiring all three backlights together and they will always display the same color. This is a concession I'm willing to make since having a different color behind each tank would probably look tacky in any case.
A particularly nice feature of the FLS-PP is that it has a separate white channel control, so the same device can also control the brightness of the main lighting as well.
With the lights and electronics tested, all that remains is to build a cable to power everything.
I scavenged several sections of multiconductor cable and cut them to length, then terminated them with simple connectors. Four two pin connectors will be connected to the main lights, and three four pin connectors will power the backlights.
The main power cable uses thicker wire since it will be powering all seven lights at once.
I ran out of suitable heatshrink and settled on electrical tape for a few of the connections.
After building the cable, I laid the entire system out on the floor for testing.
When installed, the cables are easily hidden behind the bookshelf posts and the system looks very sleek.
Along with the more natural hues, the RGB backlights can produce some wild colors as well.
Since the temporary backlights were so successful I decided to build a higher quality set from sheet metal. I started by marking out the sheet metal for cutting.
Then sheared it on a 3-in-1.
The backlights will be getting painted so I gave them a quick hit with an orbital sander to clean up the surfaces.
Back to the 3-in-1 machine for bending. Sheet metal is easy and fun to work with if you have the correct tools, but incredibly difficult to manipulate if you do not.
I painted both sides at the same time without masking so it wasn't perfect. Nobody will be looking at the back-lights anyways.
The endcaps were printed and contain a small loop to stress relieve the power cable.
The new backlights span the full length of the tank and are angled slightly upwards so they should produce a better distribution of light than the original version.
I purchased waterproof LED strips for this set of backlights. These strips also have more LEDs per inch, so the backlights will be brighter than the first version.
Soldering to waterproof LED strips is a slow process because the silicone coating must be scraped off to expose the contacts.
Up to this point I've powered the lights with an adjustable benchtop power supply. This is not a practical long term solution, so a more permanent power source was necessary.
I scavenged two power bricks that were producing the correct voltages: a game-cube adapter, and a laptop power adapter, and replaced the plug to match my LED system.
Typically lighting and filtering four separate aquariums would be a wiring disaster, but by connecting the tanks together I can pull it off with only three electrical outlets.
This marks an end to the hardware aspect of this project. It's finally time to start thinking about what will go inside the aquariums.
I wasn't entirely sure how I wanted the tanks to look, so I acquired a selection of driftwood and rock for experimentation.
It is common to use glue, foam, and clear plastic barriers to build complex aquascapes. Driftwood has a tendency to float away when submerged, and you don't want an accidental bump to dislodge anything.
When aquascaping, it is important to visualize how a tank will look after plants have fully grown in. Designs that look balanced dry will often get buried and hidden by plant growth, so in general you want to err towards large and bold arrangements.
I went through several revisions of rock and driftwood before finally settling on designs I was happy with. It has been quite a journey, but the bookshelf is finally ready for plants and fish.
No comments:
Post a Comment