My new tank!

[der_wille_zur_macht]

Making a sump though does give one some practice before getting to the expensive stuff.

Yep, and that's exactly why I did it. And, I got a sump in the exact dimensions I wanted. I'm going a little "unconventional" as far as filtration/nutrient export so I wanted these exact dimensions, rather than waiting around for just the right used tank to show up.

I've made some more progress this week but it's hard to photograph - I ran three new 15A circuits in the basement to the spot under where the tank will go. One of them will be shared with some general-purpose utility outlets in the basement, the other two will be dedicated to the fish tank. Now, once I build the stand, I'll just have to pull wire up through the floor. Next up is configuring the RO/DI and water change station - then the house will be "ready," so I can just build the stand and slap the tank in.

I'm probably going to get the tank, sump, and return pump in, then aquascape and get water in, and let the whole thing run like that to "break in" for a few months before I bring the lighting and other major systems online. During this period I'm hoping to get my ATS nicely populated with algae so the tank will be ready to stock when everything else is up and running.

 

[Conesus_Kid]

Is this going to be a skimmerless tank, or is the ATS a supplement?

 

[der_wille_zur_macht]

We'll see. :lol: Don't want to give away future punchlines for this thread.

 

[der_wille_zur_macht]

For those who don't venture into the DIY forum, I have the prototype for the "final" LED driver/controller working:

 

[5380]

What kind of capabilities are you going to have with this controller? Could you provide a link to your DIY article?

 

[solRNY]

that is wasy cool, your lighting system gonna rock! are you an EE?

theres no heatsink, or copper pours, do the ic get hot?
how many leds can it drive?

is that a pic micro?
are they programmable dimmable?

thats so awesome!

 

[der_wille_zur_macht]

I shoulda spoken more clearly - the final prototype of the LED driver is done; the controller is at very early prototype stages.

Controller:

http://www.reefcentral.com/forums/showthread.php?t=1783536

LED Driver:

http://www.reefcentral.com/forums/showthread.php?t=1783536

The controller is an Arduino-compatible board that will have two 328p AVRs, a pH amplifier, a port expander for a 20x4 LCD, a buzzer (for alarms), an IR remote sensor (so you can use a generic TV remote to control the tank's functions), connections for temp sensors, float switches, etc. and an Ethernet interface onboard. There will be a separate "module" with relays to control devices on the tank. Right now the "prototypes" of the controller are just breadboarded circuits based on an RBBB Arduino clone, but the functionality is all pretty much verified.

In terms of LED-specific functionality, the controller in the photo above has software that just randomly fades the LEDs up and down so I could test it. The final product and other prototypes I've rigged do more useful things, like fade the LEDs on in the morning and off at night, plus things like random "storm" simulation, i.e. dimming the lights to simulate cloud cover, etc.

The DIY driver is in a final state. It's three CAT4101 ICs on a PCB, so you get three channels capable of 6 or 7 LEDs each, running on a 24v power supply and PWM dimmable. It's basically like three buckpucks stuck onto a single PCB. I'm going to have 9-10 of them on the final tank to drive the LEDs, and the PWM pins will all be tied back to the Arduino controller.

In terms of cost, the DIY driver plus a cheap power supply will cost around $1.00 - $1.25 per driven LED, vs. a buckpuck and power supply or Meanwell driver, which cost around $2.50 - $3.50 per driven LED. For a small project it might not be significant but since I'm going to have ~200 LEDs, it'll save me a fair amount. That was a pretty important factor for me.

In terms of the controller, it's about saving money, but also flexibility. The entire controller will cost maybe $150 but it'll pretty much be capable of anything I want it to do, since it'll have plenty of I/O and be networkable.

 

[der_wille_zur_macht]

are you an EE?

That's the best compliment I've had all week. :lol: I'm just a self-taught hobbyist.

theres no heatsink, or copper pours, do the ic get hot?
how many leds can it drive?

Answered some of this above. There's a heatsink under the lampshade that the LEDs themselves are attached to - that part is critical. The driver PCB you see in the photo doesn't get warm enough to need one. In production, each of the ICs will be dropping a volt or so at 700mA, so less than a watt. They're able to handle 3 - 4w so it's no big deal.

is that a pic micro?
are they programmable dimmable?

It's an Arduino clone called a Really Bare Bones Board (RBBB) so it uses an Atmel ATMega 328p. I use these for development because they're cheap. It's programmable through the Ardunio environment (which is loosely C based). The whole thing is open source, on the hardware and software side, which is nice because it lets me use their technology in my own PCB (see thread linked above).

 

[solRNY]

I'm just a self-taught hobbyist.

degrees are way overrated, your a EE to me , very impressive!

being PWM dimable, tieing the input high to get 100% duty, the leds are at max current, so i would only need ur driverbrd? do you have any spares setup i can buy and try out on my 30gal?

In production

you mention "production" are you going into business?

 

[der_wille_zur_macht]

being PWM dimable, tieing the input high to get 100% duty, the leds are at max current, so i would only need ur driverbrd?

Yeah, if you tie the PWM pin high it just goes to 100%. The CAT4101 IC needs a separate supply voltage, it doesn't leech power from the LED line so you need to provide +5v to the PCB in addition to GND and the LED supply (24v). But 5v is easy to get - a wall wart or simple regulator circuit. The Arduino runs on 5v so I'm all set. But it would be easy to set up to run without an Arduino. Or even with a simple 555 circuit to act as a PWM generator based on a potentiometer, which would give you a knob to turn to adjust brightness if you didn't have a microcontroller to do it automatically.

But that said, a simple Arduino implementation can be trivial. The RBBB I use to develop is only like $13. Add an external clock, project box, etc and you're at maybe $20 - $30 for an automated LED controller that's expandable if you want to do other things down the road.

You can control what 100% means in terms of current during assembly of the board, since the IC uses an external current sense resistor. I have resistors on this test board to run at 700mA since that's a common current for HP LEDs. I've tested it on Philips Rebels and Cree XR-E and XP-E.

do you have any spares setup i can buy and try out on my 30gal?

you mention "production" are you going into business?

"Production" in the sense of making a bunch of them for the big tank. I have no aspirations of doing this as a commercial gig. Though I am definitely interested in helping other people get set up with LED stuff or lending tools/help to build LED rigs so if you're interested we should chat some time.

FWIW I released the PCB for the LED driver under an open source license (not as if it's anything brilliant or original) and will do the same for the Arduino-based controller once it's done (we're calling the controller the "Hydra" project.) I set up Google Code project sites to store the material for the projects so if anyone's interested in doing this stuff on their own they can grab the files from those sites, get the boards made, etc. Or just drop me a line and I can help get you started.

The focus of this thread to date has been the plywood tank but it's practically a means to an end for me. The part I'm the MOST excited about is the lighting and control system.

 

[der_wille_zur_macht]

Glass is going in:

2 out of 3 done. Just the big one left, then I'll let it cure for a week, move it here, and water test.

Also, some progress on the DIY controller:

 

[solRNY]

wow, your tank is looking pretty awesome! cant wait to see the water test
just curious, your front viewing pane being so large, how thick of a piece did you end up with?

nice looking diy controller, you have lots of great things going on in this build, am enjoying watching its progress

 

[der_wille_zur_macht]

1/2" glass. Depending on what numbers you plug in, that gives a safety factor around 3, which is what most commercial tanks end up at. I used half inch on the other two panels too, though it probably wasn't required. It's only a few bucks more anyways.

The controller is a prototype still - when finished it'll be more presentable. (In a case!) But it's exciting to be testing actual functionality.

 

[superkat]

DIY controller???? OK....details????

 

[der_wille_zur_macht]

http://www.reefcentral.com/forums/showthread.php?t=1783536

It's based on the Arduino platform: http://www.arduino.cc/

It's intended to be a platform for hobbyists who want to DIY a reef controller, so the functionality is somewhat open ended. Rather than configure a fixed set of functions, we tried to design it so it would meet the needs of the widest audience. By default it'll have the following hardware:

-LCD display as in picture above
-One high accuracy pH monitor (could be converted to ORP)
-An Ethernet interface
-As many temperature probes as you care to buy
-A remote relay board, similar to the switched power strips most off the shelf controllers have, for turning regular A/C outlets on and off
-A crapton of user-configurable input/output (roughly 20 free digital channels and 7 free analog channels). Float switches, relays, buttons, microswitches, inputs from other controllers, etc. could all be used on these channels.

Again, it's really intended to be a platform, rather than a closed, finished product. But there's a lot of functionality possible that wouldn't require much effort on the user's part. For instance, sensing when the temperature of your lights goes above a certain threshold, or if your return pump shuts off, or when water spills, etc. and sounding an alarm. Or even posting to your twitter account or sending you a text when something bad happens.

Plus, things like control over dimmable LEDs or T5s, control of Tunze pumps or wavemakers, moonlight simulation, simulating variable sunrise/sunset times throughout the year, monitoring for power outages or equipment failure, etc.

 

[superkat]

thank you for the link! totally doable! interested in cost comparison...by the time all is said and done, is doing this yourself cost effective?

 

[der_wille_zur_macht]

Cost comparisons depend on exact desired functionality. This controller will do some things you'd have to spend ~$500+ on a commercial controller to get, but it doesn't do other things that are more or less included with basic cheap controllers.

All said and done, you'd probably be out the door around $100 - $150 for the controller, relay unit, and other basic hardware to do the stuff I mentioned above.

Besides the potential cost savings, I'm doing this as a creative outlet. This controller will be way more customizable than anything off the shelf at any price.

 

[acdraindrps]

It's an Arduino clone called a Really Bare Bones Board (RBBB) so it uses an Atmel ATMega 328p. I use these for development because they're cheap. It's programmable through the Ardunio environment (which is loosely C based). The whole thing is open source, on the hardware and software side, which is nice because it lets me use their technology in my own PCB (see thread linked above).

We used the atmega128 last year in school. You can do a lot with them, and they are cheap too. Software is cheap for them too.

 

[der_wille_zur_macht]

Agreed. AVRs are my favorite microcontrollers.

(They're also the only micros I've ever used. :lol: )

The Arduino environment is what really did it for me. Totally open source, and you get a rock solid hardware reference design to copy. Plus enough pre-developed software that on this project we really just have to piece it all together, and do very little actual development from scratch.

 

[der_wille_zur_macht]

An update for those who miss out on the DIY forum. The tank is "home!" Last weekend we brought it to my house. It's in the garage, awaiting dry-fitting of some of the plumbing to the point that I can water test before building the stand and bringing it inside.

Also, bought a bunch of the axillary stuff recently. The most exciting of which is the parts to build a VFD-controlled pump for the closed loop. For those who haven't seen, here's a great read:

http://reefcentral.com/forums/showthread.php?t=1750674&highlight=vfd

Stu has been running his for a few months, and there are a few other posters in that thread that have been running such systems for a long time. A few people locally have asked me about this part of the project so I'll summarize here.

A VFD (variable frequency drive) allows you to control the speed of large AC motors precisely and safely. Most variable-speed water movers in the hobby are either really small AC motors, where efficiency/accuracy doesn't matter (the controllable Koralias), or they're DC motors, which are simpler to control (Tunzes) but still not readily available on a very large scale. The VFD solves this by letting you control larger AC motors.

The only "problem" is that VFDs only work with three phase motors, and none of us are likely to have three phase power outlets at home, much less a pump marketed towards this hobby with a three phase motor on it. So you have to get creative. Firstly, you need to buy a VFD that has phase conversion built in - that is, one that can take regular single phase 120v AC from any old outlet in your home, and convert it to the proper voltage in three phases. Secondly, you need to retrofit a three phase motor on your pump.

Luckily, Reeflo pumps exist. They provide really good performance, and the larger pumps (basically all except the Dart, iirc) are based on standard 56J-frame motors, which means it's super easy to get an appropriate motor from any industrial supplier.

So, what you need is:

1) A VFD that takes 120v single phase AC and produces 3 phase 240v AC
2) The "wet end" (i.e. impeller, housing, seals) from a larger Reeflo pump
3) A 3 phase, 240v motor in the appropriate speed and HP for your requirements (easiest thing to do is get a motor with the same ratings as the Reeflo pump you want to imitate).

I bought this VFD:

http://www.factorymation.com/s.nl/it.A/id.193/.f?sc=2&category=30

It cost $120. It can take a 10v analog signal, or digital inputs, to vary the speed. It can be programmed such you can trigger it to ramp between different set points by flipping it's inputs on and off with a standard reef controller or a DIY controller (as I'll be doing). While many off the shelf controllable pumps (Tunze, Vortech) allow you to switch between different programs, this approach gives you total flexibility to define your OWN programs!

I ordered the wet end of a Barracuda from Reeflo. FWIW, though they have many, many models of pump, they use the same wet ends on most of their pumps - they only have a small number of housings/impellers, they just fit them to different motors for different purposes. I bought this via emailing them. It cost ~$120.

I got a motor on eBay. It's a 3/4hp 3450rpm motor. It probably won't be able to turn that wet end at full speed, but I won't need it at full speed. For reference, if I run it around 20%, it should be about as powerful as a Barracuda at full blast. I'm sure that even at 60 - 70% I'll have more than enough flow. The motor is a new old stock Baldor, and cost $140.

So, all together, I've spent about $400, and I'll have a controllable closed loop that should push more flow than I'll ever need. I'll be sure to update the thread when it's actually together and running, but just wanted to explain it now while I had the opportunity, since I just got all the parts via UPS.