DIY LEDs - The write-up

[stugray]

tahiriqbal,

I am still confused.

In the picture you posted above, you are saying that the LED is NOT a cree that you are applying the power supply to?

I am not talking about the star board.
I am talking about the DIE.

So it is NOT one of these:



I find it hard to believe that there is another vendor out there that makes their DIE exactly like the CREE.

And the star to the right looks surprisingly like a luxeon:



Just an amazing coincidence if you ask me.....

Stu

[Soundwave]

+1 on that.

However the 100 watt led intruiges me. I wonder how efficient they are. I wouldn't think that two of them would put out more light than my entire fixture.

[Soundwave]

5000 lumens each at max seems pretty bright though.

[Dante_JoseCuerv]

Woah, I just found this thread and I wanna say, Soundwave - great design. This is definitely a great starting-point for everyone wanting to use LEDs over their tanks.

[stugray]

One thing I see that bothers me is that the power supply he is giving the specs of:

"Output Voltage (VDC) 12VDC 24VDC
Output Current (constant) 8.3 A 4.2 A
Output Power (W) 100W"

I assume the 12/24VDC is an option?

This means that to drive "banks" of LEDs, you will need several series ( 3 LEDs in series @9Watts) strings placed in parallel to get a full 100 Watts out ( 10 strings of 3 gets you ~100W ).

The problem with a Const. Current power supply & using parallel strings is:

If JUST ONE LED goes out ( they fail open ), then that series string goes away and all the CC power goes into the other strings.
This overloads the LEDs, now you just blew out ALL of your LEDs.

By driving single strings using the CC PS, when just one LED blows out, you save all the remaining LEDs from damage.

Stu

[james3370]

ok stu, that raises a question to me on my build.....

i am using (5) of the 1watt blue for moonlights & then (5) 3watt white for t5 "shimmer suplementation"

there will be a 350mA buckpuck for the 1watt & a 700mA one for the 3watt. power for each string will be a 20v/1a wall wart.

they will be wired in series. so if 1 goes out, will they all go out? if 1 goes out, will it overload & then fry the others??

i like to think i'm pretty sharp about figuring out technical stuff, so maybe i'm just having a brain fart right now LOL

[tahiriqbal]

Let me clarify few issues and questions you have raised regarding the LED diodes. They do look very similar to Cree as there are thousands of manufacturers out there who are manufacturing LED diodes very similar to Cree and Luxeon diodes but trust me they are not by Cree or Luxeon.
In order to satisfy and to answer few of your concern I will take some close up pictures of my LEDs and post out as soon as possible. Alternatively I could post out one single LED to one of you so it becomes clear that these LEDs are by our own manufacturing facility. Now you need to decide amongst yourself who should I send this LED to, so he/she can make all of you aware if I am telling the truth or not. Is this good enough to prove my point?

Regarding placing LEDs in parallel has its own problems and I would definitely agree with you to point that out. The constant power supply has its own overload protection functionality and would shut itself if any temperature or power surge arises. I need to find out what would be best way to place these LEDs, either in series or parallel in order to avoid any failure.
I am well aware of this issue therefore I have decided to divide my whole lighting fixture into four sections and if one section fails then other three would work. Each section is removable so if single LED fails which means whole section will not work then I could simply take out effected section and replace single or number of LEDs and make it work again.

Regarding 100watts LEDs they are available from 5k to 25k and prices are anything between $110.00 to $148.00 plus postage to the USA.

Thanks for the concern and positive feedback. This project will be finished by the end of April 2009 and I will post out all relevant information as said before. This lighting system design will be patented and may also be available for sale by the end of 2009.

[stugray]

tahiriqbal,

"They do look very similar to Cree as there are thousands of manufacturers out there who are manufacturing LED diodes very similar to Cree and Luxeon diodes"

NO there are 1 of EACH manufacturers that make 3W LEDs that look like the CREE or Luxeon, not "thousands".


For someone who has followed the manufacturers information enough to make $$ off of CREE stock, I think I would know if there were "thousands" of competitors.


SO, for now, I will give you the benefit of the doubt & bow down now.

Carry on then.

Stu

[stugray]

Oh and the,

"This lighting system design will be patented and may also be available for sale by the end of 2009."

NO it has already been patented in the US.
That is why Solaris had to close it's doors.

The only way you will be able to market this in the US is by supplying it as a kit with no reference to "Aquarium Lighting".

If that is the case, then Soundwave has already beat you to it.

New LEDs will come along always & we will incorporate them into our DIY designs, but the underlying physics doesn't change.

All you have done is claim to have a higher Kelvin rating than is currently possible, and used COTS power supplies to drive them.... hardly patentable.

Unless of course you explain how your LEDs get 25,000 kelvin rating in a single package.

Stu

[lynxvs]

Originally posted by james3370
ok stu, that raises a question to me on my build.....

i am using (5) of the 1watt blue for moonlights & then (5) 3watt white for t5 "shimmer suplementation"

there will be a 350mA buckpuck for the 1watt & a 700mA one for the 3watt. power for each string will be a 20v/1a wall wart.

they will be wired in series. so if 1 goes out, will they all go out? if 1 goes out, will it overload & then fry the others??

i like to think i'm pretty sharp about figuring out technical stuff, so maybe i'm just having a brain fart right now LOL

If you have two strings in series with 5 LEDs you should be fine. I one goes out the whole string of 5 will go out but it will not effect the other string. However most failures I have seen the LED actually shorts out so the other LEDs are not effected but if you lose one LED with a short you decrease the voltage for the string, say the string requires 21 Volts the forward voltage drop of each LED is 3.5 V so if it shorts it's 21 - 3.5 = 17.5V this might sound better but it puts a strain on the constant current source in the form of power dissipation. The higher the voltage difference between the power supply and the load the more power is dissipated thus more heat.

Like I said before if the LEDS are wired in parallel the current is split between the two paths. If you supply 2 A each path carries 1 A so if one path opens the current has to go somewhere so all the current will go down the remaining path so now you have 2 A flowing in that string of LEDs.

[james3370]

Originally posted by lynxvs
If you have two strings in series with 5 LEDs you should be fine. I one goes out the whole string of 5 will go out but it will not effect the other string. However most failures I have seen the LED actually shorts out so the other LEDs are not effected but if you lose one LED with a short you decrease the voltage for the string, say the string requires 21 Volts the forward voltage drop of each LED is 3.5 V so if it shorts it's 21 - 3.5 = 17.5V this might sound better but it puts a strain on the constant current source in the form of power dissipation. The higher the voltage difference between the power supply and the load the more power is dissipated thus more heat.

they are listed as requiring 3.6-4 v so 4v x 5 leds, i went w/ 20v wall wart. if an led shorts, it will go out, but the rest will work....just put a strain on the power supply till rectified??

Originally posted by lynxvs
Like I said before if the LEDS are wired in parallel the current is split between the two paths. If you supply 2 A each path carries 1 A so if one path opens the current has to go somewhere so all the current will go down the remaining path so now you have 2 A flowing in that string of LEDs.

would wiring them in parallel be better???

[lynxvs]

No, because of what I stated above... you don't want 2 A of current running through that string..... that woud probably smoke the LEDs

[Soundwave]

I would love to be the taker of the free LED.

[kcress]

You should only have one driver for each string. Period.

Any other scheme risks all the LEDs as soon as one fails shorted.

Two stings in parallel will toast them all. Why?

If you are running two stings in parallel and each string is, for example, 700mA, your driver would need to put out 1400mA. Now if one LED shorts the driver will continue to drive 1400mA into the two stings. But the string with the shorted LED will have a different voltage requirement than the good remaining string. This causes what is termed as "current hogging". The good sting will either go dim or OFF completely while the bad string may have 1200mA running thru it. The remaining LEDS will fail in seconds.

Once the entire string with a short in it has blown or one of the LEDs fails OPEN the driver will then focus on driving the 1400mA thru the remaining good string. Every LED in that string will also fail within seconds in a domino effect.

One driver per string!

Drivers cost money.. How do you deal with this?

Two ways: The first is to string far more LEDs in a string. Using a 36V or 48V driver or at least 24V. 12V borders on the ridiculous.

48V/2.2V = 21 LEDs
36V/2.2V = 16 LEDs
24V/2.2V = 10 LEDs

Alternatively you could run strings in parallel but you would need to put a fuse in series with each string. As soon as an LED fails shorted that string would hog current, exceeding the fuse rating, and the fuse would blow. Promptly the full current would try to run thru the adjacent parallel strings and those fuses would also promptly blow. It would take some careful fuse selection however.

[tahiriqbal]

Originally posted by Soundwave
I would love to be the taker of the free LED.

Hi Stu,

It would my pleasure to send you LEDs. I will send you two instead of one. One in 10K and one in 25k.

Please send me your address via PM.

I will even bare charges for the postage too. Oh by the way regarding patented, I am talking about for the European market not the North American market. Also PFO is trouble with Orbetec and as far as I am aware they have stopped trading for the time being. My LED lighting fixture is whole lot different than PFO as it is divided into four sections and each section has its own power supply and dimmer controller unit.
Also our LED lighting system has much higher Kelvin rating which is also unique and new to the market and no one else has managed to achieve such high results.
By the way I will be on holidays for the next two weeks with family to Sweden and if you don’t hear much from me then you would know why. I will try my best to stay in touch but there may be a slight delay in response.

Thanks for your comments.



Kind regards

[lynxvs]

Originally posted by kcress
You should only have one driver for each string. Period.

Any other scheme risks all the LEDs as soon as one fails shorted.

Two stings in parallel will toast them all. Why?

If you are running two stings in parallel and each string is, for example, 700mA, your driver would need to put out 1400mA. Now if one LED shorts the driver will continue to drive 1400mA into the two stings. But the string with the shorted LED will have a different voltage requirement than the good remaining string. This causes what is termed as "current hogging". The good sting will either go dim or OFF completely while the bad string may have 1200mA running thru it. The remaining LEDS will fail in seconds.

Once the entire string with a short in it has blown or one of the LEDs fails OPEN the driver will then focus on driving the 1400mA thru the remaining good string. Every LED in that string will also fail within seconds in a domino effect.

Let me just clarify this a little bit.... Assuming you are using a constant current source. If one LED in a string shorts the same current will flow through the string and the other LEDs will still function. Your right in the forward voltage will change but the current will remain the same. The burden is put on the constant current source to take care of the extra voltage. It takes care of this in the form of power dissipation. On the other hand if one LED opens in a string all the current flows through the good string and all the LEDs will fail.

[kcress]

I'm sorry but that didn't clarify anything for me..

Yes, if you have a single string and one shorts then functionally nothing changes, the single string continues to operate at the rated current since the driver is a constant current source. But if you have two or more strings in parallel then all LEDs will fail eventually, unless some LED fails open, protecting, (like a fuse), the rest in its string.

[lynxvs]

No need to be rude,
I was just commenting on this statement:

"But the string with the shorted LED will have a different voltage requirement than the good remaining string. This causes what is termed as "current hogging". The good sting will either go dim or OFF completely while the bad string may have 1200mA running thru it. The remaining LEDS will fail in seconds."

which is false if it's a constant current source. The LEDs will smoke if one string opens however.

[IamBIGGUN]

Will THESE Work? ????

They also have BLUE .

Don't know if these compare to the ones you guys are using or not. They don't give a wave length for the blue......I think I read that the white was 5500K.

What do you think?

Has anyone tried them?

Thanks,
Kevin

[Soundwave]

You could give it a shot, especially at that price. But, just like cheap halide bulbs on ebay, they could be a lower quality. I would hope for a higher color temp than 5500k, though, I'm not sure where you saw that. I would give it a go and see how it works out.

[Soundwave]

One more note, I don't think those are real luxeons. They look like a different die was used. They appear square instead of round like the true luxeons.

[The0wn4g3]

Hey, I haven't been keeping up with this thread, but I finally got the driver from China and made a little prototype on a PCB board. Here are the results
I followed the instructables here and used this driver.

Some photos:

8x 100ma LEDs


My first attempt at soldering on the right.... didn't go so well. =/



About 15" above my system in a pitch dark room. Not too bad if you ask me. 3 or 4 would make a wicked refugium light at about 4W each.

[stugray]

The0wn4g3,

If you dont solder these DIEs down to a heat-sink, then they wont work for long unless you are driving them at less than ~1/2 current.

Stu

[james3370]

Originally posted by IamBIGGUN
Will THESE Work? ????

They also have BLUE .

Don't know if these compare to the ones you guys are using or not. They don't give a wave length for the blue......I think I read that the white was 5500K.

What do you think?

Has anyone tried them?

Thanks,
Kevin

i bought from the same seller. got (5) 1watt blue & (5) 3watt white....mine are just gonna be for moonlights & t5 "shimmer" supplementation

Originally posted by Soundwave
You could give it a shot, especially at that price. But, just like cheap halide bulbs on ebay, they could be a lower quality. I would hope for a higher color temp than 5500k, though, I'm not sure where you saw that. I would give it a go and see how it works out.

if i was doing them for other than what i posted above, i'd be concerned, but for what i paid for them, i'm not too worried about them being the best

Originally posted by Soundwave
One more note, I don't think those are real luxeons. They look like a different die was used. They appear square instead of round like the true luxeons.

yeah, i'm pretty sure mine aren't.....the board they are attached to look like the ones stu posted at the top of the page, but they are mostly silver instead of black like the ones shown & they don't say "lumileds" on them like these do



i have a ways to go on my build, but once i get them running, i'll update ya on how they work....but as i said, i'm just doing them for moonlights & to add a shimmer to the tank & not for my primary lighting source.

even though i already have the buckpucks, i'm considering ordering a 350mA dimmable w/ pot so that i can dim the moonlights. what do you think.....(5) 1 watt blue that say they are 460-470nm wavelength, mounted 5" from water surface on a 17" deep tank, spaced evenly over a 25-30 wide area & powered by a 20v/1a wall wart.....too bright at 100% & need to dim or not??

[kcress]

Stu; Those are 100mA LEDs not 1000mA. I think they are about the limit of what you can feed LEDs without a heat sink concern. I just ordered a (100) 100mA LEDs and they are in a slightly larger than normal gum-drop size. There would be no way to even heat sink them. They do have kinda large flags on the leads so that may be it for them. The0wn4g3's LEDs have 4 pins, I think, which is how they can reject some heat, though not on that circuit board.

The0wn4g3; You used the perf board upside down. You're supposed to have the copper underneath so when you solder the pins they have the copper to solder to.
That's looking pretty bright!

Originally posted by lynxvs
No need to be rude,
I was just commenting on this statement:

"But the string with the shorted LED will have a different voltage requirement than the good remaining string. This causes what is termed as "current hogging". The good sting will either go dim or OFF completely while the bad string may have 1200mA running thru it. The remaining LEDS will fail in seconds."

which is false if it's a constant current source. The LEDs will smoke if one string opens however.

Why was that being rude?? You paraphrased my original post into a less accurate statement that seemed to miss the details of mine.
Your statement is correct, if and only if, you're talking about a single string of LEDs. I wasn't. I was talking about the hazards of parallel strings. If one sting changes the other string certainly doesn't continue to have the same current flowing thru it!

Draw a circuit diagram and figure out what I stated. There is nothing incorrect in that statement whatsoever. If you think it's wrong you must not be understanding it or what current hogging is.

Don't take this wrong I am trying to help you understand this.

Draw a constant current source running two resistors. Make it a 1A source. If each resistor is 10 ohms you will have 5V provided by the constant current source and 0.5A running thru each resistor.

Now change one resistor to an 8 ohm resistor.

The constant current source will see only a 4.44 ohm load and will change its output voltage to:
1A x 4.4 ohms = 4.4V

Now you have 4.4V across a 10 ohm resistor.
4.4 / 10 = 0.44A (it now has less of the original current running thru it)

Looking at the 8 ohm resistor:
4.4 / 8 = 0.55A

The 8 ohm resistor is "hogging" more of the current.

This is VASTLY more pronounced with LEDs and results in the chain reaction failure mode I described.

This is what The0wn4g3 will see happen as soon as one of his LEDs fails since he has ten parallel LEDs all being driven by one constant current source.


The0wn4g3; I would suggest that you add one or more LEDs to the parallel ones you have so you have 11 or 12 sharing the current. Then if one fails the others still won't be over driven and you will get a chance to notice and replace a blown one before they all fry.