Chad December 26, 2011 Author December 26, 2011 Bob, I wish it was mine!! No, I enlisted the equipment and expertise of the guys at Avast to do the cutting and turn my drawings into a finished product.
Chad January 3, 2012 Author January 3, 2012 A bit of an update and some pictures. Making good progress! I picked up the bare fixture on Monday. It looks great. I need to do some finish work to polish up some edges, but it will look fantastic. This is the bare fixture without any other parts attached. Credit for the cutting and assembly goes to Dan at Avast, not only does our local company do rocking reef peripherals, they do custom CNC work. A slightly different view of the bare fixture without the top: Here is my modest setup for doing the soldering of Rob's DIY constant current driver boards And some in-process work. Soldering is ~95% setup/cleaning and <5% actually applying heat. I almost always use masking tape to hold parts in place while I solder. It makes for a very clean final product. And a picture of the final driver boards. After soldering for a bit, I switched to some more mundane assembly work. Here is a big pile of hardware and assembly parts: Doing some minor assembly work (fans, large power supplies, switches, etc.). And some assembly done with heat sinks sitting in place (it looks so clean right now without wires!! hopefully I can keep it that way!).
Chad January 4, 2012 Author January 4, 2012 Thanks Rob. The estimates I did when I was designing the lift points put it around 30#. I think the final assembly will come it a bit lighter than that, since my estimates were largely based on solid renderings of parts (which doesn't account for open space within the volume... this makes things like the power supplies and electronics that are metal, but mostly hollow seem much heavier), but TBD at the moment.
roni January 4, 2012 January 4, 2012 wow, this is looking great. Can't wait to see the final product!
BowieReefer84 January 4, 2012 January 4, 2012 Nice project. Following along. Is this why you now hold the par meter... haha
Chad January 4, 2012 Author January 4, 2012 Thanks for the comments, all! I didn't get much done on the fixture last night, got home from work late. It will probably be Friday before I am able to get more time do much more than some minor polishing work (it makes a mess and, as the pictures show, I commandeer our dining room table for my work... something that is only tolerated by my wife if I clean up the mess before she gets home ). I did get this thread up to date, though what is an estimated cost for the fixture? Chris, my cost breakdown is below, however, this doesn't count the literally hundreds of hours I have doing the R&D work, benchtop experimenting, random needed tools or pieces of equipment for testing, etc. I am almost a year into the project at this point. LEDs, optics, heat sinks, and heat transferring stickers: ~$400 Microcontroller and peripheral parts (RTC, protoshield, GLCD, buttons, wire, etc): ~$150 Power supplies, driver parts, fans, plugs, switches, etc: ~$150 Misc. hardware (wire, solder, desoldering braid, fasteners, eye bolts, trim, grommets, etc.): ~$100 Fixture material and CNC time: ~$350 Nice project. Following along. Is this why you now hold the par meter... haha Shhhhhh!!!
BowieReefer84 January 4, 2012 January 4, 2012 Cutter has them listed http://www.cutter.com.au/proddetail.php?prod=cut680
Chad January 4, 2012 Author January 4, 2012 There is one unit, and yes, 3' wide. Sweet! I actually looked at that yesterday, but I didn't look close enough since the picture doesn't immediately look like what I would expect it to (the picture is of the lens only). Thanks for that!
Integral9 January 4, 2012 January 4, 2012 That looks pretty good, although taller than I originally imagined. I don't know why, I guess I skipped over the dimensions. But whatever. Nice choice on the fans, btw. Should be very quiet. Get some decent dust filters if you haven't already. Can you post a circuit diagram? What is the net difference in light output with your daylight simulation? as opposed to just having a "normal" fade in / out of blue to white and back to blue again? Is that going to result in less PAR/hr for your tank? Also, I'd be wary of any parts purchased at Radio Shack. They often sell spec reject parts. Test them if you can to be sure they are delivering what you expect. But honestly, I'd ditch them and buy quality parts from a reputable source. I realize it's too late to change your design, but I think that with the fixture 5.5" tall I think it might be better to have the display on the front of the unit rather than the top. I'm almost 6' tall and I can barley see the top of my low-profile t5 fixture. Also, the Patent Office is now "First to Apply", not "First to Invent". So get your application(s) in now.
Chad January 4, 2012 Author January 4, 2012 This is probably a good time to point out that many, many, many design decisions have been made at this point. If any of y'all have questions on the "why's" of anything, please ask, because I probably wouldn't even think to put most of it in this thread. If I have thought about it, I will share my reasoning, if I haven't thought about it, I want to. It is never too late to change course for a great reason. Great insight! I agree with you on RS parts... the only things I have left are resistors, which are easy to verify and not in a value-critical roll (pull up or down resistors) so I am ok with them. I ordered replacements from future electronics or sparkfun. I have some dust covers for the fans. I also agreed on the 5.5 inches and changed it to 4.5 inches, I am not sure when I made that change (I guess it was after September although I don't remember doing it), but it was a good one. I wanted to go much narrower than that originally, but it would mean mounting the power supplies externally, which I wasn't willing to do. As it is, I had to route a pocket in both the top and bottom parts of the fixture in order to get it to fit (you can see the cutout in the last two pictures above). I think with the 2" stainless trim and smaller 1/4 square trim, that 4.5" is the right number for a visually appealing exterior. The effect of my moving Guassian gives a parabolic daily intensity distribution similar to what you would find if you put a PAR meter outside for the duration of a sunny day: A fade into the morning, a high intensity period that peaks around mid-day, then a fade into the eveing. This was verified by summing the calculated current on each channel with the peack on each of the 33 channels and then plotting that. I considered putting the buttons and display on one of the sides or the back, but decided against it for a couple of reasons. First, I wanted to maintain the smooth, clean, modern look around the exterior of the fixture (in my mind these are the most frequently seen portions, where the top/bottom are secondary) and second, the top of my current fixture (that this will replace) is about my chin level, so access won't be bad. On the circuit diagram, do you mean the full fixture diagram or individual components? I have a block diagram for the fixture at home and ckt diagrams from some of the blocks, but don't have a composite. I will post the block diagram when I get a chance to dig it up.
Integral9 January 5, 2012 January 5, 2012 I guess it depends on your "blocks". I'd like to see how you ran your LED circuit. ie: where you put the resistors, how many LEDs in series / parallel, etc.
Chad January 6, 2012 Author January 6, 2012 I think maybe the block diagram will answer your questions then. The resistors are all in the arduino circuitry and more of the 1/4W or less category. I have a 10k pull down or pull up resistor for each of the four buttons and for each of the signal lines going to the real time clock and LCD. I chose not to place a "sensing" resistor in line with the LEDs since I would prefer to put my meter in-line and not worry about excess heat lost in an enclosed place long-term to accomplish a short-term purpose. Even if it isn't very much heat, it still is energy that doesn't need to be lost and takes away from the overall efficiency. I posted my block diagram below. Each of the eleven drivers deliver a max of a bit less than 1 A at around 21 Vdc. Each of these is a separate circuit, although eight driver circuits are mounted per board (and share some components); I have two boards for a total of 16 potential channels. There are a few extra drivers that won't be used (these are pictured above). These are the boards that I got from Rob (thefishman), maybe he can post the schematic, although I modified it somewhat such that I could individually control each driver via PWM. Each driver pushes 6 LEDs in series. No LEDs or strings are wired in parallel. The arduino provides a 5Vdc PWM signal that can vary the intensity of the channel between 0 and 100%. On the block diagram, black lines are 24 Vdc lines, light blue is 12 Vdc, green is 5Vdc PWM signals, purple is 5Vdc signal lines from the arduino.
Chad January 6, 2012 Author January 6, 2012 ^Thanks! Once it is complete and opperational you should post a link on RC. Title it something like "BEST diy led fixture ever," and watch the comments flood in lol What can I say, Mike? I got excited... http://www.reefcentral.com/forums/showthread.php?p=19710509&posted=1#post19710509
BowieReefer84 January 6, 2012 January 6, 2012 hahaha. This makes me laugh. I am sure you will have your share of naysayers, but lots of knowledge on RC as well.
Mando77 January 7, 2012 January 7, 2012 This is crazy. I hope you plan to produce or sell something like this.
Chad January 9, 2012 Author January 9, 2012 ^Thanks! Maybe, but this was a lot of work for me. I know building more of them would be less, but we shall see. I did this mostly because I wanted to increase my DIY repertoire and this project was well beyond my skill in a bunch of different areas when I started. I got a lot done this weekend, but it was mostly mundane and tedious stuff like drilling holes and laying out/mounting the LEDs. Nonetheless, here are some pictures. I hope to get to wiring and soldering the LEDs within the next couple of days. LOL, looking back through my pictures... you can tell what time of the day it was based on what drink I had! Laying out the LEDs for marking Undrilled, unmounted, but laid out and marked LEDs on the heat sinks Doing some drilling of the heat sinks to mount the LEDs, but also holes to route wire and mount the heat sinks. Shout out to Steve (forrealb50) for the use of his drill press and DaveS for the use of his garage! Here are all the optics mounted. The 6 white ones are not a different angle. At some point in the design process I changed from 5 LEDs (for a total of 55) per string to 6 (for a total of 60). I originally ordered 60 LEDs and optics thinking that it would give me extras in case I wanted another string of LEDs... Later on, I forgot how many I had ordered and had to scramble to find some more (HUGE shout out to Jim for helping me out there!!). Of note is that I am using XR-E optics on XP- LEDs... I did this mostly because I wanted wide optics and liked how these looked... We shall see how they actually work out... Working through the tedious process of screwing the LEDs down. And all mounted... I noticed AFTER I took this picture that I missed cutting notches out of one of the heat sinks! Shoot!!
wvreef January 10, 2012 January 10, 2012 Looking awsome....can not wait to see this thing fire up.... I am definatly going to do leds for my 180 and wifes 90 gallon
Chad January 10, 2012 Author January 10, 2012 Thanks, Kelly! I can't wait to fire it up either! Y'all remember this picture? I wanted to add it to this thread for completeness. This is bench testing the GLCD. Since this picture, I have modified and added some finishing flare to the images and cleaned up the wiring quite a bit (though I will be completely redoing the wiring when I mount the screen to the top. Up next is wiring the LEDs.
roni January 10, 2012 January 10, 2012 Looking good Chad! I can't wait to see this up and running. Your DIY skills are impressive!
Integral9 January 11, 2012 January 11, 2012 Nice job. That's a really nice control board too. My personal preference would have been to run the LEDs in || w/ resistors, but 6 LEDs isn't too hard to trouble shoot through should one fail and break the chain. I'm betting it will be fairly obvious as well.
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now