Reef Angel Forum

Being a forum member for about a year, I'm surprised of the lack of discussion about the Reef angel "core" software and hardware here in the forum...

The Reef Angel is an open source / hardware project with many users, but it looks like it does only have one or two persons being involved in the development? All users buy the "plug and play" reef angel and no one builds (or improves) the hardware themselves...

I would think the whole Reef Angel community would benefit from a more active community of DIY:ers... What do you think?

I'd love to see others contributing with the libraries improvement and even hardware

One thing to do is to improve the documentation of the software and the readability of the source code, to make it less difficult to make changes to the software.

Another thing to do, is to make it easier to use the software on a 3:rd party Arduino mega board (as main unit). Today lots of skilled DIY:ers start from scratch with an Arduino board instead of the Reef Angel, since they dont want to buy an of-the-self unit (as the Reef Angel) ... Having some of them (us) involved in this community by making the Reef Angel software a bit more generic, could increase the community development work and broaden the development of the Reef Angel.

Pny wrote:One thing to do is to improve the documentation of the software and the readability of the source code, to make it less difficult to make changes to the software.

Another thing to do, is to make it easier to use the software on a 3:rd party Arduino mega board (as main unit). Today lots of skilled DIY:ers start from scratch with an Arduino board instead of the Reef Angel, since they dont want to buy an of-the-self unit (as the Reef Angel) ... Having some of them (us) involved in this community by making the Reef Angel software a bit more generic, could increase the community development work and broaden the development of the Reef Angel.

If your coding skills are good enough that you could contribute it's easy enough to look through the RA libraries and get them working on standard arduino hardware. The methods and classes are easy enough to understand. Everything is setup in globals.h and then all the objects are instantiated through ReefAngel.h and ReefAngel.c. You also need to understand the way the ReefAngelFeatures.h turns the different code blocks on and off. Once you've got that figured out it's pretty easy to add things in, take things out, and make the modifications necessary to run on standard arduino hardware.

Truth be told though, once you get all the hardware required to accomplish everything the RA has you'll end up paying just as much if not more than the RA costs. You can get a dev platform going with just a mega 2560, RTC module, and color LCD shield. No good for actual tank control though because you're missing the 2 most important things: PH and relay box. Plus at best you'll end up with a huge ugly box next to your tank at worst a big jumble of wires and boards. RA is a much nicer package for a very good cost once you start looking at the individual component costs.

KRavEN wrote:If your coding skills are good enough that you could contribute it's easy enough to look through the RA libraries and get them working on standard arduino hardware. The methods and classes are easy enough to understand. Everything is setup in globals.h and then all the objects are instantiated through ReefAngel.h and ReefAngel.c. You also need to understand the way the ReefAngelFeatures.h turns the different code blocks on and off. Once you've got that figured out it's pretty easy to add things in, take things out, and make the modifications necessary to run on standard arduino hardware.

Truth be told though, once you get all the hardware required to accomplish everything the RA has you'll end up paying just as much if not more than the RA costs. You can get a dev platform going with just a mega 2560, RTC module, and color LCD shield. No good for actual tank control though because you're missing the 2 most important things: PH and relay box. Plus at best you'll end up with a huge ugly box next to your tank at worst a big jumble of wires and boards. RA is a much nicer package for a very good cost once you start looking at the individual component costs.

I have good enough coding skills to do it, but if I have to do lots and lots of changes to the Reef Angel code base, there is not much of use to it compared to put something together myself, since it will not follow the reef angel code for future releases...

I wish you where right on the cost issue, but after adding costs for shipment to Europe, customs fees and 25% tax on top of that, the Reef Angel costs more than you can imagine... Still, it will need adapters to be compatible with our electrical plugs, and I don't want that, so I will have to make my own power bar anyway...

I hadn't considered the cost for non-US people. Sorry to hear it's so prohibitively expensive for you.

Dev setup with 2560 board, RTC board, color lcd shield, and joystick board only needed a few changes in globals.h to move the pins around.

To build your own relay box should be fairly staightforward. Look at the eagle schematic. There are I2C pin expander boards for arduino that use the same chip as the relay box and then you hook that to one of the 8 port relay boards they have all over on ebay. I know this is possible because I'm using a similar setup to control a 4 port relay board that runs 3 dosing pumps and my ATO pump. I made it look like a relay expansion to RA.

The making of a relay box is a non-issue... I've going to send the board below to a manufacturer... It's a tiny board with a Arduino Pro mini ($3 microcontroller), space for a xbee adapter board (on the backside) and two USB connectors. making it look like an RA expansion with the standard i2c-interface (instead of the bulky HDDB15 cable), but also with the option to interface it wireless using xbee protocol... It is powered by a very small AC/DC board and controls carefully chosen relays (with low input current requirements). All boards are so tiny they will fit the standard power bars I'm using.

Don't forget the snubber diode
I don't see one in there.
When turning relays, you want to have a diode to consume the feedback voltage of the coils when they open.

rimai wrote:Don't forget the snubber diode
I don't see one in there.
When turning relays, you want to have a diode to consume the feedback voltage of the coils when they open.

Thanks, but the relays is complete boards, with the extra components for the relay, including optocouplers... Each input only requires 5 mA trigger current, so I guess I can just connect the Arduino outputs directly to the relay boards.

Ahh.. you are all good then

These are pretty small too but no xbee.

Yes, thats a neat and simple solution... I actually considered that first, and it still is my backup plan, but my solution has a number of advantages (at the cost of just a little more complexity), (if I manage to implement them!):

1, Can be extended with xbee wireless control
2, Can be extended with another extension that plugs in to this unit... Very neat if combined with xbee. (The unit has to USB plugs: 1, In from main unit, 2, Out to extension)
3, Can be made to operate in stand alone mode with configurable default states on the plugs (even after power off/on cycle). This means the unit does not turn off all outputs if I need to unplug the main unit for some reason.

If you look at the libraries closely, you will find that I have already implemented this in the libraries, so you can take advantage of that too. I called it RANet
https://github.com/reefangel/Libraries/ ... l.cpp#L459
The receiver is pretty much what you have, except it is in a different form factor, so it is a little bigger, and it's got a few more components for status indication and voltage translation.
I just sent the latest revision of the receiver to the board house a few days ago.
The RANet payload will have status of all relay boxes and their respective fallback state, as well as status of the dimming module.
Eventually, I plan on appending the payload with all the parameters too and possibly have a wireless remote screen, if I ever find a decent screen to use. Who knows

rimai wrote:If you look at the libraries closely, you will find that I have already implemented this in the libraries, so you can take advantage of that too. I called it RANet
https://github.com/reefangel/Libraries/ ... l.cpp#L459
The receiver is pretty much what you have, except it is in a different form factor, so it is a little bigger, and it's got a few more components for status indication and voltage translation.
I just sent the latest revision of the receiver to the board house a few days ago.
The RANet payload will have status of all relay boxes and their respective fallback state, as well as status of the dimming module.
Eventually, I plan on appending the payload with all the parameters too and possibly have a wireless remote screen, if I ever find a decent screen to use. Who knows

Sounds super cool Robert. I'll have to take a look.

As far as a screen goes, I think a simple small status LCD is really all that is needed as long as there are good browser and phone apps. Like you've already found out, dedicated touch screen is costly and almost everyone already has an android or iphone.

I do think a detached stand alone status screen like the apex display module would be a welcome addition. Maybe use one of those cool 128x128 OLED panels. Then you can stash the main RA unit wherever you want and only have a small 4 wire cable running out to the status screen that you can stick wherever is most accessible.