The Next Step in DIY LED Lighting

[mabviper]

I'm on board w/ this project all the way. Since I don't have a lighting system right now, I think it would be pretty cool to program a system that other people can use and hopefully be happy with. As far as microcontroller selection, they're all pretty much the same o.o Though, I'm still biased w/ PIC though, lol. I found their forums are pretty helpful.

 

so... I'm in the programming loop so I don't get rusty w/ C and embedded software dev. ^^

[DarkDevil]

Psychographic: if you want to program the controller, you will need a computer, but just running the setup you don't need the computer.

 

ls7corvete: could you elaborate on why you would prefer ethernet setup over LCD? do you mean if you want to change any setting, you need to turn on the computer? although ethernet/wifi/rf/bluetooth capability may be offered as an expansion module in a later time.

 

 

also, remember remember, this is probably going to be offered as a DIY project, we are just going to provide a parts list, plans on how to build it and such, and you have to build it yourself. It is more then likely that this is NOT going to be offered as a complete finish product, which means, basic soldering skill and equipment is needed; basic computer skills is also needed to finish couple simple tasks.

[ls7corvete]

Psychographic: if you want to program the controller, you will need a computer, but just running the setup you don't need the computer.

 

ls7corvete: could you elaborate on why you would prefer ethernet setup over LCD? do you mean if you want to change any setting, you need to turn on the computer? although ethernet/wifi/rf/bluetooth capability may be offered as an expansion module in a later time.

 

 

also, remember remember, this is probably going to be offered as a DIY project, we are just going to provide a parts list, plans on how to build it and such, and you have to build it yourself. It is more then likely that this is NOT going to be offered as a complete finish product, which means, basic soldering skill and equipment is needed; basic computer skills is also needed to finish couple simple tasks.

Much easier to setup, imho. setting controls with a full keyboard and mouse is easier, especially when more complicated setups are programed. I would think the user interface through a LCD would be very limiting. Everyone who builds one of these is going to have access to a computer. I am also assuming that this is targeted to nanos and most people would like a smaller package.

 

With ethernet you would still need some controls on the light itself. Feeding, on, off, etc. I am guessing we can install a group of controls and use the ethernet controler to setup what these buttons/controls do. A pre-programed set of controls should be released with the unit. More knowledgable users can setup their own commands to bind to the buttons/controls.

[evilc66]

Ethernet is actually really expensive to implement. Interfaces like serial and USB will be far better alternatives.

[ls7corvete]

I assumed they would have a similar price. I trust you guys will setup a similar effect though.

 

Is there a component set you were planning on? bus type? controller type(pic?)?

[phishroom]

Interesting idea, evil. I might have something to add. I've been only an intermittent contributor on the forums here, but have been scanning some of the LED traffic.

 

There are a couple things about me that might combine to be useful to this project, and as time allows I'd be glad to help where possible:

 

(1) I'm a professional boat captain and interpretive naturalist. (Nature tours, fishing, etc.) One of my primary interests is the intersection of my work and this hobby. I'm attracted to the eco-system in a box idea, but feel that there are many, many, many misconceptions and superstitions that reef keepers perpetuate about how things are best done in a tank. Lighting is a big part of this. The variation in light intensity, photoperiods, and the balance of solar and lunar cycles is very, very, very important in nature, and something that we've only begun to understand in application for reef keeping. This recent "LED moonlight" marketing crap is an example of poor implementation.

 

(2) My "night job" is in live music and semi-professional concert lighting. I say "semi-professional" because it's not a big paying gig. However, I'm using cutting edge equipment and techniques, including LED intelligent lights and computer interface lighting control. To be clear, I'm not a computer language programmer, but I do program lights to be manipulated live via computer. I think a way I might help here is in the interface design and logic. I might be able to help in recruiting language/programming people that might have an interest, or at least point you in the right directions for some ideas and solutions.

 

I'd like to help out, so keep me in the loop!

[Vancouver Reefer]

Hi All,

 

This is one of my first posts on here and im currently in the research phase of my own DIY AIO nano. I used to be an engineer and i have been looking at all the very inspirational posts on here and im currently in the design process of a fully automated tank. This wont be getting assembled until later in 2009 as that gives me time to design, build and test everything before it all comes together and gets wet.

 

As far as ideas for a tank controller, here are a few things im looking to accomodate:

 

Real time & date stamp.

PWM output to LED drivers.

Temp sensing of heatsink/Tank hood with pwm output to cooling fans.

LCD display for easy-to-view tank parameter indications.

PH, ORP & Temp inputs with display.

Remote monitoring from PC/internet host.

Floor flood sensor to turn off all external pumps upon water detection.

 

The more modular and plug and play we can design these units the more fun i think it will be in not only designing, but building and implementing onto our tanks.

 

Reefing is one pleasure. DIY'ing cool gadgets to add to the reefing is even better.

 

VR

[evilc66]

Thanks for posting VR. We have sorely neglected this thread.

 

The modular approach is something that is paramount in our plans. We hadn't gotten to the point of integrating more tank functions into the plan, as we were focusing more on the light control, but it's a natural progression.

[isidro0]

What I would do, is pick a development board (http://www.futurlec.com/DevelopmentBoards.shtml) that is wired for RTC and LCD. Do a tutorial on how to install buttons, lcd and ds1307. Provide a schematic and a program for 2 timed based pwm outputs for white and blues. The program should be done on a free compiler, I perfer Bascom.

Everybody should be able to have a basic setup for $50.

Once the program and hardware is out, people should post their software upgrades and hardware mods for everyone else to pick and choose which options are worth the headache.

[Carinya]

I like the computer interface idea, but don't make it just for PC users.

[mim37204]

i was actually throwing around the idea of using one of my spare gumstix to do exactly such a thing. but if yall figure something out, perhaps i could find a better use for this little gadget.

[neanderthalman]

Ok. First up. I like this idea. I like this project. I like the open-source concept.

 

 

However...I would like to make some possibly harsh criticisms, so I apologize in advance if I come off as rude or arrogant.

 

It cannot start out as a purely collaborative exercise. Open source is great, I love it, but one person, or a very small group of people must spearhead a basic, working prototype to the community. Only then can it become open source and community driven. It's the old adage about too many chefs. Too many people are chiming in with ideas and input in the early stages to accommodate each and every request or suggestion. This needs to be simplified in a big way.

 

Linux did not start out a community, but one man. He released it and a community formed. Build a community around the project, not a project around a community.

 

 

More specific suggestions:

 

The modular design is interesting, but it's going to be more work than the project itself. I'd recommend simplifying the project, by eliminating the modular nature.....but preserve the modular effect.

 

Each and every unit should be programmed and built as a fully-functioning unit. As many processes as possible should be hardware driven...much like the suggestions for a RTC circuit. Hardware PWM, for example, is much more robust than software (though software PWM ain't too shabby, IME). If there are features people don't want to use...it is simpler to disable them via dip switches or menu settings on the LCD than it is to add them via modules. This preserves the effect of allowing users to choose what features they have and don't have, but simplifies the design immensely.

 

 

 

Further to the simplification...start small. Really small. On/Off on a timer, based on the RTC input. No PWM, no K adjustment, no ethernet or USB, etc.

 

Then and only then add on features one at a time. Instead of On/Off....dusk to dawn via PWM. This leads to K-adjustment via PWM. It's also a springboard to moonlight cycles of phase and moonrise/moonset. Add features one at a time and build it up slowly.

 

 

From a project management perspective, it would be a good idea to set release dates for the next feature(s), so that there is a goal, a purpose...otherwise projects like this will sit on a shelf for years. There needs to be a driver, and something as simple as a small group of users expecting an update is sufficient.

 

 

Lastly, features that mimic "movement" of the sun or moon across the tank are gaudy and unnecessary. It sounds neat, I even toyed with the idea a few years ago when I planned and programmed an identical project for my LEDs. I got caught up in the idea that it was "awesome" and didn't realize that it carried a large amount of "stupid" along with it. Think about it. In nature, is a substantial difference in lighting over a space three feet long going to exist from the tracking of celestial objects? Nope. Not gonna happen. Everywhere will get lit with the same intensity at the same time. The change is the direction of the light source...which you're not going to be able to mimic without physically moving and/or rotating the LEDs. The effect will simply be a moving spotlight in the tank...real natural. Don't waste your time, efforts (and excessive wiring) on this.

[evilc66]

Ok. First up. I like this idea. I like this project. I like the open-source concept.

 

 

However...I would like to make some possibly harsh criticisms, so I apologize in advance if I come off as rude or arrogant.

 

No trouble, thats why we started this thread.

 

It cannot start out as a purely collaborative exercise. Open source is great, I love it, but one person, or a very small group of people must spearhead a basic, working prototype to the community. Only then can it become open source and community driven. It's the old adage about too many chefs. Too many people are chiming in with ideas and input in the early stages to accommodate each and every request or suggestion. This needs to be simplified in a big way.

 

Linux did not start out a community, but one man. He released it and a community formed. Build a community around the project, not a project around a community.

 

Done. There are three of us working on this project. So far it hasn't gone beyond the planning stages, but it's getting there.

 

More specific suggestions:

 

The modular design is interesting, but it's going to be more work than the project itself. I'd recommend simplifying the project, by eliminating the modular nature.....but preserve the modular effect.

 

Each and every unit should be programmed and built as a fully-functioning unit. As many processes as possible should be hardware driven...much like the suggestions for a RTC circuit. Hardware PWM, for example, is much more robust than software (though software PWM ain't too shabby, IME). If there are features people don't want to use...it is simpler to disable them via dip switches or menu settings on the LCD than it is to add them via modules. This preserves the effect of allowing users to choose what features they have and don't have, but simplifies the design immensely.

 

I agree it will simplify things a great deal, but if we have to include a lot of features right off the bat, it's going to raise costs. We have set ourselves a practical limit for the base hardware that will be usable by 90% of the people that want to use this. Adding on additional modules will be more of a challenge in the software end of things than the hardware. Most of the addition can be taken care of by simply addressing modules on a simple buss like I2C or SPI. Doesn't need to be high speed or high bandwidth.

 

 

Further to the simplification...start small. Really small. On/Off on a timer, based on the RTC input. No PWM, no K adjustment, no ethernet or USB, etc.

 

Then and only then add on features one at a time. Instead of On/Off....dusk to dawn via PWM. This leads to K-adjustment via PWM. It's also a springboard to moonlight cycles of phase and moonrise/moonset. Add features one at a time and build it up slowly.

 

Agreed. The features that we have agreed on for the base hardware will lend itself to a great starting point.

 

From a project management perspective, it would be a good idea to set release dates for the next feature(s), so that there is a goal, a purpose...otherwise projects like this will sit on a shelf for years. There needs to be a driver, and something as simple as a small group of users expecting an update is sufficient.

 

This is one thing that will prove to be a little difficult. While we have dreams of getting this off the ground, i think it's going to be a slow process just because of the little amount of time that any of us has to dedicate towards it. I would like to see it happen though.

 

Lastly, features that mimic "movement" of the sun or moon across the tank are gaudy and unnecessary. It sounds neat, I even toyed with the idea a few years ago when I planned and programmed an identical project for my LEDs. I got caught up in the idea that it was "awesome" and didn't realize that it carried a large amount of "stupid" along with it. Think about it. In nature, is a substantial difference in lighting over a space three feet long going to exist from the tracking of celestial objects? Nope. Not gonna happen. Everywhere will get lit with the same intensity at the same time. The change is the direction of the light source...which you're not going to be able to mimic without physically moving and/or rotating the LEDs. The effect will simply be a moving spotlight in the tank...real natural. Don't waste your time, efforts (and excessive wiring) on this.

 

Features like that are something that I have always thought were frivilous. No real point to them. We are going to stick to the real core features that will be useful. Truely useful. Features like that can be added by the community, if thats what they really want, and when we get to that point.

 

The point of this thread was to get ideas from the people that might end up using this, but at the same time, reaffirming our concepts that we had come up with. Some features make sense, some don't.

[Marteen]

I would really like sunup sundown features in a controller. I think that would be my priority in a project like this. As hobbyists it's pretty easy and cheap to get something to turn on and off when we want it but to be able to dim it on and off is a whole other league. Just my $.02.

[not_sponsored]

lets bump this back up

[Urchinhead]

My $0.02... Since this is an open source project perhaps this would be an appropriate platform:

 

http://www.makershed.com/ProductDetails.as...oductCode=MKSP4

[evilc66]

I am already experimenting with the predecesor of that.

[KDris]

Just popped in to ask how everyone's projects are going. What features have you included in your arduino controller Evil? I too am working on a light/reef controller for my 5.5G with LEDs using an arduino. My controller probably won't fit everyone's needs or be very intuitive for people who cannot code C, but it hopefully will include the following.

 

-LED controlled by pots with sunrise/sunset

-Temperature monitoring

-Temperature control (turn on fans/heater)

-pH monitor

-Feeding timer

-LCD Readout

-Dosing pump control

 

For me keypad control isn't really necessary as the values shouldn't change too much and I figure I can fit LED brightness %'s, Temp, pH on a 20x4 LCD

 

Optional:

-Real Date/Time

-Thunderstorm (possibly with speakers)

-Keypad control

-Lunar Cycles

[BLoCkCliMbeR]

i dunno how much help this would be but check this guys thread out

 

hes got this thing programed to control his tank......hes a programer and he could probably be big help if you can get in contact w/ him

[evilc66]

I haven't been able to put any effort into it over the last few months. Been very busy with lots of other things, including leaking roofs.

 

KDris, you are going to need both a display and a keypad to adjust things aren't you? Can't adjust much if you don't have buttons.

 

BLoCkCliMbeR, he hasn't been online in over a year. Nice project though.

[KDris]

KDris, you are going to need both a display and a keypad to adjust things aren't you? Can't adjust much if you don't have buttons.

 

I'm going to have a display but no buttons to start, just potentiometers to control the brightness of the LEDs. If I need to adjust any of the other values I'll change the code. The LCD I plan on buying does have a keypad matrix built into it but for me its a non essential luxury.

[evilc66]

If brightness is all you are going to adjust, that will be fine. For the most part, these controllers will have timers that will have to be adjusted. Granted, it could be hard coded, but this will make it easier to adjust on the fly. Six of one, half dozen of the other