I currently have a BC29 that has been running for several years. It was initially running on 72w of PC lighting, but i wanted more light so I 70w MH retrofit into the hood to change things up. I was scared the MH bulb would melt/discolor the splash guard on the stock hood, so I cut out the area under the bulb and replaced it with glass. After moving PC's around and a long time in use, the glass had cracked (but still held together) and some of the rest of cover had yellowed. Anyway, i read a few builds/guides on here and decided to change things up once again.

I am by no means an expert at any of this, so i intend for this thread to be a learning experience for me and hopefully it will help some other people too. I have limited knowledge of LEDs and circuitry. I will be posting lots of pics, asking questions, and reporting problems i run into along the way.

Starting point was buying the supplies I know i will need for the build. Got them from another member who bought them during the group buy i think. Still in the mail.

Heatsink (7.2 X 14 in), 24 3w LEDs (12rb 12cw), 2 meanwell eln-60-48p power supplies --$280 shipped



First problem I ran into after reading up a bit: I should have gotten the D type meanwell drivers instead of the P type. I only want simple on/off operation along with dimming through a dial. If someone would like to trade, I am still looking for the D type drivers.

The next thing I wanted to get to work was this, the dimming circuit for the meanwell driver to get a 0 to 10vDC source. I think I read somewhere this could be built without the capacitors.
http://www.nano-reef.com/forums/index.php?...st&id=67222

So i went to radio shack and picked up some parts. Total for this stuff was about 22 bucks with tax.
Circuit board
2X LM317T adjustable voltage regulators
2X Linear taper 10k Ohm potentiometers
1 pack each of 1k and 470 ohm resistors
Random knurled knobs
5X2.5X2 inch project box




I couldn't find any AC adapters to get a 12vdc source so I turned to ebay for those and bought 2 for $12 shipped total. They were over $20 each at radio shack.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...e=STRK:MEWNX:IT

Also needed a pair of plugs for the AC side of the meanwell. Walmart for 2.88 each


This is all I have for now. I need to source a new splash guard for the biocube in the mean time, since my current one is trashed and missing screw holes. I'll start working on the dimming circuit before the rest of everything gets here so I can be ready when the parts come.

If anyone sees something that i've bought wrong or has some advice, any help is always appreciated.

Thanks for looking.

Here is what i've built so far with 22 guage wire. I know its sloppy, I haven't built a circuit in a long time. The red lead is the input from the AC-DC adapter, and the green lead is the output side going to the meanwell.

I tried a 9v 500mA AC-DC adapter to test the circuit, i know i should be using a 12v. When i test the voltage of the power supply it shows 14.3 volts with no load on it. When i hook the power supply up to the rest of the circuit, I am able to adjust the voltage with the potentiometer between about 2 volts and 12 volts on the output side

Does this mean the circuit is working? lol or does this mean i've built a distaster. I'm right about where i should be, except 2 volts above on each side.

Sounds close. Measure the input voltage with a load on it to see where the voltage sits. The input voltage must be 2v above the output voltage. It could explain why the output goes up to 12v. When the voltage difference drops below 2v, you can get some odd results. The 2v bottom end is about right though.

the same power supply that was putting out 14.3v stays at about 11.8 volts when hooked up to a ECU fan. Still doesn't really make sense to me because it says 9v on it. Will i have to worry about the same problem when i get my 12v power supplies too?

Cheap wall warts like those tend to have very poor valtage regulation. The 12v power supplies will be less of an issue. The 11.8v output under load could have been what caused the 12v top end measurement. I know that sounds odd, but like I mentioned before, once the input and output voltage difference drops below 2v, odd things happen.

LEDs came today, no labeling. I'm guessing the only way to test them is to turn them on? would 2 double A batteries in series work?

IS there another way to tell them apart that i don't know about?

Yellow dies are white, silver dies are blue. The D315 label is another giveaway. That's the bin label for the royals.

I thought the P meant dimmable?

"Z"

"P" means pwm dimmable. "D" is analog dimmable. No suffix means that it's not dimmable.

Sorry but PWM stands for?

Thanks

pulse width modulation

http://en.wikipedia.org/wiki/Pulse-width_modulation

Thanks and nice avatar!

hey evil, sorry for asking all these stupid questions, but I have one more for now.

I have a 12V AC-DC wall outlet ( same one from before)
With no load its putting out over 14VDC
When hooked up to a load its around 11.5VDC
When hooked up to the LM317 circuit with no load its putting out a range of 1.9VDC-12.70VDC
When hooked up to the LM317 circuit wired to a load (spare CPU fan from a computer) its putting out 1.3VDC-10.30VDC

Does this mean I'm good to go? Should the lower end of the range be zero, and is the upper voltage of 10.3 ok? Thanks

You're good. The lower limit will always be around 1.2-1.3 volts, as it's the dropout voltage of the regulator and can never go below that.

So the circuit you built is PWM and will work with P type drivers? I'm on the same boat as you and have been reading a lot but still feel lost. The circuit you're doing seems much easier then the others.

This circuit was for the "D" type driver (0-10v). The pwm circuit is a lot different and is explained in the Meanwell thread.

Thanks Evil. I saw the pwm circuit you posted up. But being electrical illiterate, it's hard for me to grasp. I'm working on it though.

So I basically have no interest from anyone in buying the P drivers from me, and nanotuners will be out of stock on the D drivers for a month...so back to the drawing board. Since I already have half this circuit, I am thinking about trying to build this setup.

From the meanwell thread:


This obviously complicates things for me a little, does anyone know if i will have to programming code for the arduino if i do this? Also isn't the PWM output of the arduino 0-5v?

You will have to program it to make it work, and yes, the Arduino is 5v. That's why there is a transistor switching the 10v source to the driver. There is a pwm circuit listed in the Meanwell thread that doesn't require a controller.

so after reading up on arduino basics and writing/understand code for it, i decided that it is total overkill for my needs. Sure, it would be nice to set up a 24 hour lighting cycle using the arduino and simulate sunrise/sunset, but at my level on the learning curve its just not worth it at this point.

So i turn my attention to the PWM circuit (courtesy of evil):



My questions/concerns so far:

1. What type of diode is acceptable?
http://www.sparkfun.com/commerce/product_i...roducts_id=8589
http://www.sparkfun.com/commerce/product_i...roducts_id=8588

2. Is a 12v power supply still acceptable?

3. Could this whole circuit be done legitimately using a breadboard so long as the components are tied in very securely? I don't really feel like soldering this many components twice (although soldering is sort of fun)

4. Why do they even make PWM models on these drivers? The D models can be had for the same price and the dimming circuit for those requires about 1/3 of the parts the P model does. I realize that the PWM may have certain advantages for other applications, but who else even really buys these things besides reefers?

1. Second link

2. Yes

3. Do yourself a favor and solder it. It will be infinitely more reliable.

4. The "P"s were the only option for a while. They are also used in commercial lighting applications and can be used for large lighting displays that require digital control (signs and color shifting wall washers for example).

Art
What did you do different from the first test when the readings were 2 volts and 12 volts on the output side. I copied the same things you did and I'm getting 2 volts and 11.30volts with a cpu fan. How can I get both reading lower?
Thanks
Kenny

What are your resistor values?

Not sure???? How can I check?? If you are asking the resistor i use was 1k and 470ohm. I did measure the input voltage (17.5) and output voltage (11.30)

No, that's what I was looking for. Variance in the values of the parts could bump it high, but not up to 11.3v. You could change those two resistors to a 1.5K one to drop the voltage a little more.

Can i add 1 more 1k ????

Too much. Add another 100 ohms if you want.

The only thing that differed between the 2 times was that 2-12v time there was no fan hooked up to the circuit. With the fan i got the lower voltages

I bought a whole biocube hood from another member and gutted it. i really want to go with 2 120mm fans for cooling and want to mount them above the heatsink for cooling. The only problem is theres not enough room between the hood and the splash shields. I test fitted the whole thing with the LEDS on top of the heat sink and there is almost no space between the LEDs and splash shield, maybe only 1 or 2mm in certain spots.

Both of the other bc29 builds on the site have cut away portions of their heat sink to provide clearance for the fans, but don't have access to tools to cut away the fins to make room for fans. I honestly wouldn't mind mounting the fans on top of the hood (outside) if there was a way to cover them somewhat and not just have them look like fans bolted onto the top of the hood.

1. How hot do the tips of the LEDs get? what would be an adequate clearance between them and a plastic splash shield

2. I have absolutely no room for optics. The other 3 builds didn't use them. The bottom of the splash shield is about 1" off the water. Are optics necessary?

Optics aren't required. See if you can find a local machine shop that will cut the pockets out for you for a few bucks. Many places are looking for work, no matter how small.

I didn't think there were any machine shops in my area, but apparently there are. One told me the job is too small for them to bother with. Another shop told me $65 for two 25mm deep 120mm square slots for the fans. Seems like a ripoff but I guess it may be the going rate around here. Another place told me they charge $80/hr, i can't image this would take too long.

$65 isnt too bad, but I'm sure there are better deals. Just need to find that one little hole in the wall machine shop that will do it cheap.

Looks great so far, will be nice to see how it turns out. I feel in over my head with this project - luckily I don't have the cash budgeted to do it all at once so I will have plenty of time to read up on other builds as I go.

Did you say you were making a splash guard, do you have any details on that?

Nope not making any splash shield , just had to get a new one because my old one is in pretty bad shape. Its better to take the whole project one step at a time, especially because it will keep you from buying the wrong parts and wasting money.

Buying online will save you money even with added shipping cost in most cases.

Got the heat sink back from cutting, came out ok. I had them cut the slots more than 120mm because is was concerned that the fan might not fit and i would have to go back, but i didn't have to. Theres room for exactly 2 more fins on either side of the fan, it would have been a perfect fit if i had them cut 11 fins out instead of 13.

The next thing I feel stupid about: Getting the fins cut all the way to the base....wow, no point in doing this. It gives me the most amount of clearance between the LEDs and the splash shield, but mounting the fan flush against the heatsink stops the fan from moving any air. Should have gotten the fins cut halfway or 12mm.

I went with gentle typhoon fans (sort of pricey, $40 shipped for both) but they look and sound great. Almost silent from only a foot or two away at 16dB and move a good amount of air. I've used Silenx fans in the past and have had a couple fail so I stayed away from those.

Put a 1/4" nylon spacer under the fan. You can usually find them at most hardware stores in the screw department.

How much did you end up getting the heatsink milled for?

Ended up paying the $65. I probably will use spacers like you suggested and the same fan filters used in the other bc29 builds to mount the fans into the hood with. With all 8 screws going through the hood, fans, then the heat sink, I won't have to make another bracket to hold the heat sink, it will just be held by those 8 screws.

I finished one PWM circuit based on your sketch, evil. Parts were really cheap from allelectronics.com . I got than enough of everything for only 12 bucks. Wish i hadn't gone to the shack the first time around.

Whats the best way to test if this thing is working without hooking up the LEDs?

So i hooked everything up (driver to wall and pwm circuit, with 12v supply to PWM circuit) and using a voltmeter set to 50v instead of the LED circuit and I'm getting a steady reading of 26volts. Turning the pot doesn't do anything. When uplugged the voltage goes up to around 38v and then drops to zero after a few seconds.

I'm stuck. Checked all my connections on the PWM circuit and everything looks good. Do the 1st and 2nd wires of the pot need to be connected to each other?

driver is good by the way, when the DIM + and - are hooked up to just the 12v power supply for a second i get 48v output

BTW don't know if I'm using the pot correctly. im using the top 3 leads, do i need to do anything with the other 3?

Ok, first off, you won't be able to accuraely check the output of the driver with no load. Second, measuring the voltage does you no good, as that is not what changes for dimming. It's the current that changes.

Checking the pwm circuit will also be a challenge, as it's a variable frequency output, not a variable voltage. You could try setting the meter to a frequency measuring mode and see if that works, but many meters aren't set up to measure that high of a frequency. Best thing you can do there is to make sure the output is 10v.

You will have to hook up 6 LEDs to the driver to get it to operate properly. Please don't make the same mistake that RedfishSC made hooking the LEDs up to an already powered up driver. Connect LEDs, then fire it up.

On that pot, as long as you are using the same set of contacts, you should be ok. That is a dual wiper pot that is designed to seperate two circuits.

I see, I guess i was going about checking it the wrong way.

I made a chain of 6 leds, hooked everything up, and gave it a go.

LEDs light up, but are not bright at all. You can look at them comfortably and they don't look like they could light anything up enough. I don't think they're ruined or dead because when you disconnect the AC from the driver, they light up to full power and then go out. Turning the pot does nothing.

I can take a video to show what i'm talking about, but it kind of goes along with what i've been seeing before with the voltage staying at around 26v, jumping up when power unplugged and then dropping off quickly.

When you say "turning the pot", you mean on the dimming circuit?

yea on the PWM circuit. I haven't opened up the meanwell is there something i need to do in there?

It wouldn't be a bad idea to dial SVR2 down (counter clockwise) to drop the output to slightly safer levels (~975mA)

so i got 2 D models from another member, rebuilt my simple 0-10v dimming circuits which are tested and working fine, and hooked everything up to give it a go.

The voltage adjusts between 9.8 and 1.6v on the dimming circuit. I took a vid to describe whats happening, but basically between 9.8v and about 4 volts i get linear dimming, but anything around 4 or below i get the whole array blinking. Its hard to tell dimming from the video because of the camera but it definitely dims.

Also, i tried adjusting SVR2 to check the current, and it just spins freely in circles. The lower limit when the dimmer is turned all the way down is 289 mA, but when you adjust the dimmer all the way up it goes past 2A.

Is the meanwell i bought shot?

http://img192.imageshack.us/i/cimg2634.mp4/

the problem with the other is one is the exact opposite. SVR2 seems intact on this D driver, and i can adjust it so that i get about 975mA when the voltage on the dimming circuit is at its highest (9.8v). The problem with this is one that whenever i go above 4-5v on the dimmer, it starts blinking on and off just like the first one. What am i going wrong with this one?

Edit: For this driver (working SVR2), I can get the LEDs to stay on at full brightness without going on and off by adjust SVR2, but i checked the current at this setting and its about 1.3A. Would running them like this be ok ?

Either way I don't understand why i'm getting blinking with either unit. Has anyone dealt with a similar problem?

Odd. How many LEDs are you running per driver when you saw this?

2 chains of 6 LEDs each. Im guessing this isn't normal then. How do i go about properly setting SVR2?

Try adding a 7th LED and see if it shows the same results. There seems to be some inconsistancy with these drivers recently when running close to the minimum voltage. I haven't had a chance to check it myself yet.

I wish i could but they're all soldered together and i don't have any spares

Would i be ok running them all from one meanwell driver until i get another one? The only way to keep them on steadily is to run the meanwell at full current (1.3A), putting it any lower creates a constant on/off pattern. The other driver is definitely shot, it puts out steadily more than 2A