Who wants a cheap, simple, Arduino-based LED controller?

[der_wille_zur_macht]

If you don't have a batter the RTC is probably not functioning correctly. It won't operate without SOMETHING there. If you want to run it without a battery, put a bit of bare wire or something else conductive in the battery holder to short out the leads. Most of the time they'll run like this (though of course you'll loose memory when it loses power).

pandimus, that's the way the battery clip is designed to operate. I used that clip as a holdover from when I designed the Hydra. On that design there were other components/pads UNDER the battery so we needed it raised up. On this design that isn't the case. In version 2 of the Typhon I'm using a different clip, fwiw.

Speaking of version two, what do people think about the FTDI chip being onboard? It would mean you would not need a separate FTDI breakout to program the device, but it would bump the price up by ~$5. Right now, I'm thinking I may put the parts in the design, along with a header for the serial port (as there is now) which would let people choose. If you wanted it onboard, buy the parts and solder them in. If you already have a cable or BUB or other breakout, leave the parts off and just plug your breakout into the header. What do people think of this?

[pandimus]

That would be a good idea, to have the option. I obviosly have a bub, but for those who are only doing this one option, it would be preferable (and cheaper), to just put the usb onboard.

Thanks on the battery tip. After i started putting components together i just left those dowel pins on, makes it easier to insert the battery.. I just haven't bought a battery yet. But now that you mention it does make sense because the clock isn't functioning at all. well off to get a battery.

[der_wille_zur_macht]

Check ebay if you don't mind the shipping delay. You can probably get 10 or 20 batteries for what a single will cost at a local drugstore.

[pandimus]

too late, found one at my local mom and pop hardware store. 3 bucks.. works like a champ now.. thank you.

to hook it to my 3 driver 1401 board, How do i attach the leads? obviously positive lead goes to the individual pwm circuit, what about the - lead? I have the older board with the mix match etching.

[Megablue]

Originally Posted by der_wille_zur_macht

Speaking of version two, what do people think about the FTDI chip being onboard? It would mean you would not need a separate FTDI breakout to program the device, but it would bump the price up by ~$5.

What about a chip used for the pinguino boards (arduino compatible), for a board with the PIC32MX220F032D you have the usb support, or for 5$ more an integrated rtc (PIC32MX440F256H)? You can easily add micro sd flash reader or have enough memory for a web server, but that's not a typhon any more
I'm planning to use an arm STM32F4 for my next controller.

[der_wille_zur_macht]

Yeah, there are definitely many other platforms we could move to that would get us all sorts of great features but you're absolutely right, then it wouldn't be a Typhon any more.

I have some cool thoughts for the UI but I want to work them out before I talk about them in public.

[SchnitzelReef]

Well after spending about 2 hours reading thru this thread,(which is awesome!) I never found an exact answer to....WILL THE TYPHON DIMM THE ELN 60-48D?? A few people have said yes, some have said maybe....i don't understand. The typhon is supposed to be PWM, and the eln D is 0-10v analog...so technically it shouldnt...right? And if it does work will you explain how to hook it up. Also if it does DIMM, will it DIMM gradually from 5%-100%? Or will it only DIMM in steps...(I originally posted this in a diff thread by accident). Thanks for any help


Sent from my iPhone using Tapatalk

[der_wille_zur_macht]

There have been many anecdotal reports that it works just fine. It's not the CORRECT way to dim the analog version of the ELN, but apparently it works without any issues for many people. I've never heard someone report that it did not work.

You would wire it the same way as the digital version.

Dimming performance would be the same as the digital version - i.e. you'd get a response as indicated in the datasheet, albeit chopped into 100 discrete steps.

[TheFishMan65]

I don't think that is quite true. I think you get 255 step, but not linear like the P version. Some one posted a chart and 10% was not 10% of the PWM signal more like 50% (IIRC). So you may have to play with steps to get the correct gradual brightening and dimming. I have not tried this only read about it maybe in Katchupoy's thread (Another Arduino build or something close).

[der_wille_zur_macht]

To be clear, that's why I worded things like this:

Originally Posted by der_wille_zur_macht

you'd get a response as indicated in the datasheet,

For reference, the datasheet shows now (non?) linear the response is:

http://www.meanwell.com/search/eln-60/ELN-60-spec.pdf

[Patzig]

So for those us with a mix of 48D's and 48P's it might be a little hard to match up their current outputs on sunrise/sunset.

[der_wille_zur_macht]

Yeah.

Though, we have to remember that all of this is highly subjective. Does it really matter if the dimming is totally linear? Our eyes aren't going to perceive a linear response as linear, anyways. And, the intensity of light on a natural reef does not change in a linear fashion. Even if you want totally linear dimming, it doesn't really matter if the driver's response curve is not linear, since we have total control over the signal being provided to the driver. Editing one or two lines of code in the core firmware could make the response shape look however you wanted it.

The one thing we really can't "fix" in software is the fact that both styles of ELN "cut off" low in the response curve.

[TheFishMan65]

DWZM,

Dimming performance would be the same as the digital version - i.e. you'd get a response as indicated in the datasheet, albeit chopped into 100 discrete steps.

The dimming (as I recall) is not the same. There is no chart in the data sheet (did I miss is it) that states how a D will respond to a PWM signal. Where do you get 100 steps?

I think perhaps we are communicating poorly

[der_wille_zur_macht]

Originally Posted by TheFishMan65

I think perhaps we are communicating poorly

+1.

What I meant when I said "the same as the digital version" was basically that it'll dim according to it's curve in the datasheet.

100 steps because the Typhon works on whole-number percentage increments, even though the AVR's PWM pin has higher resolution than that.

So, if you use the analog driver with a Typhon, it'll march up and down it's dimming response curve in 100 steps - just like the PWM version will march up and down it's own curve in 100 steps if used with a Typhon.

[TheFishMan65]

Not enough knowledge of the Typhoon to remember it only has 100 steps - I need to get mine running

Yes they will both march up and down their curve, but IIRC the curves are very different.

[CaliO]

Originally Posted by der_wille_zur_macht

See the Hydra project I linked above - that's more or less our goal there. Build a "core" hardware and software library that let people construct a full-scale reef controller. Though it also has Ethernet hardware and a pH amp onboard, since it seems people who want a full-blown reef controller also want that stuff most or all of the time.

To reiterate, this project is *NOT* intended to be a full blown reef controller. It IS intended to control a few channels of LED drivers, and maybe some other stuff if you get creative (hence why I brought out an I2C header.) This definitely won't be all things to all people, and it probably won't be a very satisfying project for people who are well versed in this sort of thing and can do it themselves, because it'll seem overly simplistic. It's mainly intended for the people who have some basic tinkering/soldering experience but don't have a clue how to design and build a microcontroller - i.e. it's intended to make the entry point into the "microcontrollers for reef use" hobby a little more attainable.

Someone above picked up that I hinted at a $30 price point - to be clear, that's a very rough estimate based on parts cost. It'll be nice if it ends up being a cheap thing for people to build, but as someone else pointed out above, this isn't a cheap hobby, and price isn't the main driver here.

The design is made to plug in to the back of this display:

http://www.moderndevice.com/products/16x2-blue-lcd

There are tons of other 16x2 displays that are pin compatible, but I'm linking this one to help people visualize. The LCD will plug in to a pin header on the board, and the board has holes for standoffs that line up with the LCD's holes, so they can be screwed together. Basically, the bottom edge of the controller PCB will stick out below the LCD, and will have a row of buttons on it. The power jack and headers for programming (FTDI), I2C, and LED driver(s) will stick out the "sides" underneath the LCD.

I thought about modeling it in one of those 3d circuit rendering tools (i.e. fritzing) but I might as well just wait until the prototypes get here and I can build it for real.

So sorry but the thread is so long and it's late. But I see your still here so I was wondering if you ever built this ^ ^ ^ and if so can you PM details...


If cost is lower then -> http://www.clay-boa.com/dim4-4-po...ntroller-v1-0/ IDK if that would work? But seems like it would.. Using [2]Meanwell LPF-60D-48 drivers




Or my last option is Reef Angel


Thanks

[bigguyuk]

Originally Posted by der_wille_zur_macht

See the Hydra project I linked above - that's more or less our goal there. Build a "core" hardware and software library that let people construct a full-scale reef controller. Though it also has Ethernet hardware and a pH amp onboard, since it seems people who want a full-blown reef controller also want that stuff most or all of the time.

To reiterate, this project is *NOT* intended to be a full blown reef controller. It IS intended to control a few channels of LED drivers, and maybe some other stuff if you get creative (hence why I brought out an I2C header.) This definitely won't be all things to all people, and it probably won't be a very satisfying project for people who are well versed in this sort of thing and can do it themselves, because it'll seem overly simplistic. It's mainly intended for the people who have some basic tinkering/soldering experience but don't have a clue how to design and build a microcontroller - i.e. it's intended to make the entry point into the "microcontrollers for reef use" hobby a little more attainable.

Someone above picked up that I hinted at a $30 price point - to be clear, that's a very rough estimate based on parts cost. It'll be nice if it ends up being a cheap thing for people to build, but as someone else pointed out above, this isn't a cheap hobby, and price isn't the main driver here.

The design is made to plug in to the back of this display:

http://www.moderndevice.com/products/16x2-blue-lcd

There are tons of other 16x2 displays that are pin compatible, but I'm linking this one to help people visualize. The LCD will plug in to a pin header on the board, and the board has holes for standoffs that line up with the LCD's holes, so they can be screwed together. Basically, the bottom edge of the controller PCB will stick out below the LCD, and will have a row of buttons on it. The power jack and headers for programming (FTDI), I2C, and LED driver(s) will stick out the "sides" underneath the LCD.

I thought about modeling it in one of those 3d circuit rendering tools (i.e. fritzing) but I might as well just wait until the prototypes get here and I can build it for real.

This is the whole idea of the project and yes it would be nice to have more features maybe as modules that can be bolted on later on
not understanding most of this thread but do fancy having a go
if anybody in the uk whats to help sourse the parts i am willing to have a go

[der_wille_zur_macht]

Originally Posted by CaliO

[COLOR="DarkOrange"]


[COLOR="Blue"][B]So sorry but the thread is so long and it's late. But I see your still here so I was wondering if you ever built this ^ ^ ^ and if so can you PM details...

I did build it and at least 120 other people have, on my count. That's what the thread is about. Read through and ask if you have questions. The post you quoted is essentially accurate - cost depends on where you buy parts, what you already have on hand, and so on - but it will almost always be cheaper than any commercial product.

I am in progress with a redesign, but it will likely be a few months, so if you want something quick, the current version might make sense.

[pandimus]

A sub menu to test lighting percentages on the new version would be nice. i/e a menu where you could go in and adjust how bright each channel is.. step back and look at tank, that way you dont have to mess with your stock programming.

[shark boy]

If more pwm channels were added wouldn't the interface have to change to something like an I2C input for the screen?
That would effect the layers of the board or the size of the pcb, also if the screen size increased to say a 20x4 the footprint on that is a lot larger as well.
Seems like either way the new Typhon will not look like the old Typhon.
I can't wait to see the new design.

shark boy

[der_wille_zur_macht]

Originally Posted by shark boy

Seems like either way the new Typhon will not look like the old Typhon.

This is true. It won't be the same size and the interface will be improved overall. It will still fit in the 10x10 category so the cost for the board (from places like iteadstudio) will not increase. Cost overall will probably be up about $10 but I'm pretty confident it's justified.

[trodrigues]

does anyone have any extra typhoon boards, I am looking for 1 bare board

[saf1]

Question.

What exactly does this version / thread allow to be controlled? I'm a bit confused now since I think I'm following two or more different build threads

I keep thinking this one is the LED dimmer and controller (on/off), temp probe option, ph option, and network? Or...was it just to control LEDs.

The hydra is what I'm thinking but there are so many pages I'm not sure where 'go' or 'start here' is anymore. Now it looks like another build thread is separating PH? Right now our LED's come on and off for x amount of time. I think corals and the one retro kit's power supply would be happier if it was on at x time, then slowly increased, all on, then slowly decrease, until off. Temp and network would be ideal, as a option for turnning on/off stuff but I'm not one to be able to take a basic design and add.

I need to follow a recipe sadly

[Loonytic]

This is specifically for controlling/dimming LEDs. Others have incorporated Ph and temp probesfor controlling a fan, but this is just a basic, simple LED controller. Once you get the components it's pretty easy to assemble.

[byee]

GREAT job to everyone who has helped move the Typhoon forward to where it is today!

I've been following this thread as well as the Hydra for quite some time. I've device to build the Typhoon sooner than later simply because I'm getting a little tired of paying $400 bi-monthly for electricity. I've even thought of shutting down the tank to same on my monthly expenses since both my kids have lost interest in the aquarium..........a side effect of Finding Nemo!

I've downloaded the necessary files from the google code site.

Installed Eagle on my laptop.

Thank to DWZM, I'm following his Eagle instructions to get the Gerber files created for the BCB house. When I run the cam by clicking on the build film icon, I get the following which is very different from what DWZM posted in thread #264.

I've attached a screen shot of what I saw when I selected the CAM button. Is this correct? If not, what am I doing incorrectly?

I entered TYPHOON_CONTROLLER as the output file. I see 2 files created after I clicked on PROCESS JOB

1] typhoon_controller_pcb (no extension)
2] typhoon_controller_pcb.gpi

Are those the only 2 files created which needs to be sent to a PCB house like iTead Studio?

How do I verify the Gerber files are correct?

Thanks in advance to any advise.

Thank you!!