Diy led

.Perfect. I have a 300 gallon and just want to do around 12 red with no optics to give a different effect on all the red colors in my tank. Make them pop even more. You are right. It is so cool to have the ability to tweak the lights at will. I recently made a 24 light display for night viewing with 20 royal blues and 4 -18000k white using a T5 reflector
And no optics. Looks better than the other 24 royal blue display I have in the back of the tank. It adds more shimmer
 
.Perfect. I have a 300 gallon and just want to do around 12 red with no optics to give a different effect on all the red colors in my tank. Make them pop even more. You are right. It is so cool to have the ability to tweak the lights at will. I recently made a 24 light display for night viewing with 20 royal blues and 4 -18000k white using a T5 reflector
And no optics. Looks better than the other 24 royal blue display I have in the back of the tank. It adds more shimmer

Josephv....just want to be clear I was quoting operating specs for a Cree XPE Red...they run 2.3V @ 700ma. just check the specs on whatever brand you choose and then use the math to figure out what kind of driver you or shoot a note here if there's a question.
 
The reds that I have on hand are the ones with the lens that came off of, they are XRC's. When I got them a few weeks ago I was told that they are 350mA and not 700. I will look into this further, but for now I will listen his instructions. I only picked up 5 to start with.

Blackhawk...here's the spec sheet for the XRC...says forward V is 2.4@700ma for the Red...page 3...2nd block of Characteristic data, 2nd line from the bottom above junction temp. http://www.cree.com/products/pdf/XLampXR-C.pdf The XRCs are the same as XREs in color ...700ma shouldn't be a problem.

I want to add that the Reds being discussed here only run at 2.3V forward Voltage....current is important but 3.5V that we run most of the whites and RBs at will fry them. Most of the drivers auto adjust but a constant voltage driver won't (FEDY) and you need the right amount of LEDs.
 
Last edited:
Thank you. I'm not sure what they have going on where I picked them up. But since there is no indication of what exactly I have (only their word) I will start out dimmed and work my way up. Its rough when these guys confuse opinions with specs. I luckily have a test set that I can use on one of them and will run it up to 700 slowly. Better to blow one than several. Now that I think of it, I will start off with the one that is missing its lens.
 
We got ours from FEDY but they appear to be the same. Google FEDY LED and you get the page. Ask for 220lm 18000k led. They have min order quantities of 100 @ $1.27ea. They work fine if used within spec. I haven't lost any. Like all LEDs you have to keep them cool. Satis sells same if you look for 18000k or ask them but I have not tried them but would not be afraid....they're just older technology and work fine.
 
Now that the small FEDY FD-100-4.2 driver is working fine only one other person posted they got it working,
it time to work on the large FEDY FD-200-36 driver.

The label on the side of the driver, Output DC 36±0.5V 5.6A
36V ÷ 3.6= 10 led per string
5.6A ÷ 700mA = 8 parallel strings
I would like to run the leds under driven.
I ran 11 leds per string. 8 parallel strings.
My test numbers
Total out put of driver 2.39A / 2.35A after worm-up
each string 290mA to 310mA
Something is not right with the driver.
015.jpg

012-1.jpg
 
Now that the small FEDY FD-100-4.2 driver is working fine only one other person posted they got it working,
it time to work on the large FEDY FD-200-36 driver.

The label on the side of the driver, Output DC 36±0.5V 5.6A
36V ÷ 3.6= 10 led per string
5.6A ÷ 700mA = 8 parallel strings
I would like to run the leds under driven.
I ran 11 leds per string. 8 parallel strings.
My test numbers
Total out put of driver 2.39A / 2.35A after worm-up
each string 290mA to 310mA
Something is not right with the driver.

I couldn't find the driver datasheet, but it sounds like a constant voltage supply (as opposed to constant current). So, if you stack 11 LED's with a forward voltage of 3.6v, you'd have 39.6v. The supply will only put out 36v so the current will drop (as you're measuring). The problem is that the IV characteristics of the LED's (diodes) is very "sharp". That is, a very small voltage change results in a *large* current change. If you want to use this type of supply in this configuration, then you'd really be better running at least one less LED in series (the string should probably be 9 versus 11), but have a resistor in series. The resistor must be chosen to handle the power it will be forced to handle (I*V) with some margin (2x?). If you're going to run these strings in parallel, then I'd recommend using some sort of "ballast" resistor anyway. It would make it easier to "trim" out any current imbalances in parallel strings.

For the case above, at 9 LED's, you would get 9*3.6v = 32.4v. So, you'd have about 3.6v left over. So, if you want to run 0.7A, you'd need R = (3.6v/0.7A) = 5.14 ohms. The resistor is pulling about the same power as one LED, so you're losing about 10% power efficiency in this scenario, for the added benefit of better bias control since you have a constant voltage supply.

The more voltage you drop across the resistor, the more "stable" the overall system will be, at least up to some point where it will provide diminishing returns. The down side is that you're burning power in the resistor that isn't making any light output. It's less efficient.
 
Steven,

Let me start off by saying, I have not built a DIY LED fixture. I am in research mode. My day job is an electrical engineer. So, my answer (and any other related to this) are based on that background and not on a DIY build. Good to know what your information source is...

I found the Meanwell datasheet. This one is a constant current supply, so it is better suited to what you're doing. LED's (diodes) will be easier to drive with constant current because their IV characteristic is so sharp. Basically, this supply is saying it will provide you with 187W of power if the load voltage is within 18 - 36V. Your original equation will work better with this driver.

5.6A / 0.7 = 8 parallel strings
36v / 3.6v = 10 LED's.

I would back off this by 1 LED in each string and add the ballast resistor & fuse. I probably wouldn't run them in parallel, but most people are. If you do, use the fuse and I would suggest adding the power resistor to allow you to 'balance' the strings. If the forward voltage of the LED's is different between parallel branches, you can fix this by changing any given branches series resistor. I haven't looked at what resistors are available to do this...

Having said all that, I think if you have a fixture setup and the original supply, you could make it work. May not be ideal, but it should function. Just drop down the number of LED's in each string to 9 and add a series resistor (5W probably) and you should be able to get it running. The fuse is always a good idea with parallel strings with either supply.

Good luck.
 
Some pics of my fixture, keep in mind the 88 white leds are running at 320mA and the 36 blue are at 690mA. The blue leds have two diodes and the whites one. FEDY does not tell witch one is Epistar/Bridge lux Chips. The small driver is a constant current driver and works fine for 30 or 36 leds.
Blue only
005.jpg

006.jpg

White only
007-1.jpg

008.jpg

All led
009-1.jpg

010.jpg

If anyone wonts to stop over and see the fixture send me a pm. I would like to get the par reading of the setup as it is running now.
 
Steven,

Let me start off by saying, I have not built a DIY LED fixture. I am in research mode. My day job is an electrical engineer. So, my answer (and any other related to this) are based on that background and not on a DIY build. Good to know what your information source is...

I found the Meanwell datasheet. This one is a constant current supply, so it is better suited to what you're doing. LED's (diodes) will be easier to drive with constant current because their IV characteristic is so sharp. Basically, this supply is saying it will provide you with 187W of power if the load voltage is within 18 - 36V. Your original equation will work better with this driver.

5.6A / 0.7 = 8 parallel strings
36v / 3.6v = 10 LED's.

I would back off this by 1 LED in each string and add the ballast resistor & fuse. I probably wouldn't run them in parallel, but most people are. If you do, use the fuse and I would suggest adding the power resistor to allow you to 'balance' the strings. If the forward voltage of the LED's is different between parallel branches, you can fix this by changing any given branches series resistor. I haven't looked at what resistors are available to do this...

Having said all that, I think if you have a fixture setup and the original supply, you could make it work. May not be ideal, but it should function. Just drop down the number of LED's in each string to 9 and add a series resistor (5W probably) and you should be able to get it running. The fuse is always a good idea with parallel strings with either supply.

Good luck.

Steven I agree with Michael in backing off by one. You want to leave a little headroom for your driver. Using the HLG 36v will drive 10 just fine though because I notice most of my FEDYs run right around 3.45V. Definitely use the fuses as suggested.

The dicey thing about the constant voltage drivers is that as heat builds it will try to maintain the voltage by raising the current and its a viscous cycle potentially. I wish they would properly label that driver.

I'm running a HLG 120 right now in 42V and its a really nice driver. I'm running 4 strings of 10 off it right now. I have a 185 in hand for my next build for my 120g.

As for the big concern on unbalanced strings practice has proven it not to be an issue of apocalyptic proportions. 1st you always need a method to measure the current in each string...ie resistor method or just remove the fuse and put the meter in series with each string. You want to see each string within about 5% of each other. For example low string 700ma high string no more than 735ma. if something is out of this range you then go about finding the offending LED. I usually start with the low and high string. Looking for the outlying voltage level i measure each LED in those 2 strings and actually write the value next to each LED on the heatsink. Then I swap the 2 outliers for each other...1 in each string. Then measure current again and note difference. Repeat until they are reasonably balanced. My experience is that it takes no more than 10-20 minutes to find it and rectify the situation.

Note...I didn't use to do this until michael joined this thread a couple months back and raised the concern on parallel strings and got me thinking and I went and did some research. Its not a huge concern but it should be checked and corrected as needed.
 
Is it just my eyes, or these Fedy LED are not bright enough?

I'm running 112 Fedys (1 white:1 blue ratio) with 4 Mean Well drivers on an 68" heatsink.

When I set it ~8" above the water, the color is pretty dull.. Nothing compare to MH.

I glued all 112 Fedys with 60 degree optics, is this the problem?
 
Is it just my eyes, or these Fedy LED are not bright enough?

I'm running 112 Fedys (1 white:1 blue ratio) with 4 Mean Well drivers on an 68" heatsink.

When I set it ~8" above the water, the color is pretty dull.. Nothing compare to MH.

I glued all 112 Fedys with 60 degree optics, is this the problem?

First thing, generally our eyes are not that great to judging brightness.

Next, what is your wiring configuration and meter readings for each string?

I'm using 78 FEDY LEDs on my 120 gallon (was running 2 150 watt 14000 K MH) and to me it's brighter than old setup. My LED fixture has been running for a couple weeks now and I've noticed some good coral growth already. This weekend I have a fellow reefer coming by and we'll be doing some PAR readings and I'll post the results.

I should also note that my fixture is about 16 to 18 inches above the water surface.
 
First thing, generally our eyes are not that great to judging brightness.

Next, what is your wiring configuration and meter readings for each string?

I'm using 78 FEDY LEDs on my 120 gallon (was running 2 150 watt 14000 K MH) and to me it's brighter than old setup. My LED fixture has been running for a couple weeks now and I've noticed some good coral growth already. This weekend I have a fellow reefer coming by and we'll be doing some PAR readings and I'll post the results.

I should also note that my fixture is about 16 to 18 inches above the water surface.


Each Mean Well ELN-60-48D powered 28 LEDs in 2 strings, I couldn't find my multimeter to take the reading, will do that today. Drivers are untouch, I assumed each string should be around ~650mA since Mean Well set these driver at 1300mA.


truebeliever71, what driver are you using?
 
Each Mean Well ELN-60-48D powered 28 LEDs in 2 strings, I couldn't find my multimeter to take the reading, will do that today. Drivers are untouch, I assumed each string should be around ~650mA since Mean Well set these driver at 1300mA.


truebeliever71, what driver are you using?

I am running 3 ELN-60-48P drivers, 2 for the whites and 1 for blues each driving 2 strings of 13 LEDs. The drivers are adjusted to approx. 625ma to each string.

I might have missed it but what did you replace for lights?
 
I am running 3 ELN-60-48P drivers, 2 for the whites and 1 for blues each driving 2 strings of 13 LEDs. The drivers are adjusted to approx. 625ma to each string.

I might have missed it but what did you replace for lights?


Thanks for the info on driver, I'll take the reading today to see what come out.

This is for my new tank, the old one are 250w MH.
 
Thanks for the info on driver, I'll take the reading today to see what come out.

This is for my new tank, the old one are 250w MH.

No problem. I'm pretty sure that getting the equivalent of the dual 250s since I had to dial down my controller to about 80 percent output. I can't wait to see what I'm getting for actual PAR numbers when I can measure it this weekend.
 
Back
Top