60 led array what i need to power?

Thread Starter

AIRMALE

Joined Sep 13, 2010
9
I have salvaged 3 60 led arrays (each array is 30 parallel sets of 2 (5mm white) leds in series. No resistors on circuit boards. They came out of a HUsky 180 led work light. How can i power these boards with 12 vdc and what resistor or driver i would need?
 

KJ6EAD

Joined Apr 30, 2011
1,581
It's a Chinese product that goes by many names: Designer's Edge Ecozone L-1300, etc. It was "designed" to run from a rechargeable battery, presumably 12V nominal but probably had a single power resistor for each board of 60 LEDs instead of separate resistors for each pair of LEDs or any active current regulation. How stable is your power supply?
 
Last edited:

Wendy

Joined Mar 24, 2008
23,415
Are you going to try to power them up in their original circuit boards, or are you going to remove them and remount them in another board? There are pro's and con's with both approaches.

How about a picture of the front and back of the boards you have?

An exacto knife will be needed to slice connections and rewire it.

While 12V would work you would be better off with a higher voltage, something in the range of 18 to 24 volts. It would reduce the number of components you will need in the long run.

I have written a tutorial for beginners with LEDs, you need to cover chapter 1 and the 1st half of chapter two...

LEDs, 555s, Flashers, and Light Chasers

We'll also need to know what the forward dropping voltage of your LEDs are. You can measure 10 of them then average the number. This is used to calculate the resistors you will need.

I'll draw up a quick schematic for doing this, it is really simple, and post it under this paragraph...



You can replace R1 with qty 2 120Ω ¼W resistors in parallel. If the voltage drops below 4.0V with no LED then replace the battery(s).
 
Last edited:

Thread Starter

AIRMALE

Joined Sep 13, 2010
9
It's a Chinese product that goes by many names: Designer's Edge Ecozone L-1300, etc. It was "designed" to run from a rechargeable battery, presumably 12V nominal but probably had a single power resistor for each board of 60 LEDs instead of separate resistors for each pair of LEDs or any active current regulation. How stable is your power supply?

The 3 boards are wired parallel to one common plug. I don,t want to cut out and add resistors on the boards.
 

Wendy

Joined Mar 24, 2008
23,415
You may have to. Chinese junk in known for major short cuts in design, and there is probably a reason this unit is dead. If you want to repair it (which wasn't what you said in your first post, why would you want to remove the LEDs otherwise) it will last only as long as the design will allow it to.

You can do this however you wish, it is your project, but I gotta tell you one of my buttons is requests that change in the middle. We are all volunteers here, and my time has some value to me.

I designed the LED tester because it is something that is needed by everyone, not just you, but I did spend more than an hour drawing it. You will still need it for something like this. Come to think of it, I will be changing that schematic as I just thought of something (basically I had a brain fart).

**************************************

Never mind, my brain fart was thinking I was having a brain fart. :rolleyes:
 

Wendy

Joined Mar 24, 2008
23,415
One last thing, the schematic I drew is a LED tester. It will help you find bad LEDs, which is likely the root cause of your problems. One of the reasons you want extra resistors is redundancy, which this product probably didn't have. So when one LED burned out it took the chain with it, just like Christmas lights.

I explained LEDs in the article I pointed you to. I can't make you read it, but if you want to learn some reading is going to be required.
 

Wendy

Joined Mar 24, 2008
23,415
Yeah, that is what I was afraid of. Paralleling LEDs like that is a major design flaw, something only a Chinese engineer (I'm being generous and assume he had some schooling) would do. This is the kind of mistake a freshman in college would be expected know not to do. By saving around 25¢ in resistors he guaranteed the board would die in a few months. This is ironic given that LEDs in general have a lifespan measured in 10s or 100s of thousands of hours.

Likely one of the LEDs is shorted, and now you have to figure out which one. Modifying board using an exacto blade is pretty easy. Build something like my LED tester and start slicing. I am about to step out, but I will show you where and what sequence when I get back, assuming someone else doesn't beat me to it.

The alternative is unsolder one LED, test it and its mate, for every chain until you find the problem. You could get lucky and find it first LED, or you could have to unsolder 30 LEDs before you find the problem.

Both ways some soldering will be required.

One last thing, I mentioned it in my article, but it is possible there was a cascade failure with this design. If a LED opened up, causing a second failure, etc, you could have over half the LEDs in that design blown. See why I am so negative for this kind of sloppy design workmanship? However, the odds are in favor of a simple short, in which case the other 59 LEDs would be OK. If you were to buy these LEDs individually it would work out between $20-40, so they are worth salvaging IMO.
 
Last edited:

Thread Starter

AIRMALE

Joined Sep 13, 2010
9
Bill, your help is greatly appreciated. These boards came out of a unit that did not work the leds are fine none burned out, tested each with resistor and 12vdc based on white led vf 3.3 and 20ma if. I just want to find out what i need in the way of a diver to light each board individually with out adding a resistor to each led. thanks again for your help.
 

KJ6EAD

Joined Apr 30, 2011
1,581
There is no magic driver that can solve the problem of LEDs in parallel without some individualized current regulation. A compromise design would be to use a voltage regulator to supply the boards and add a surface mount resistor between each pair of LEDs as shown in this edit of your photo. Use a stripe of chemical solder mask stripper to remove the solder mask followed by a Dremel with a ball head engraver to cut the trace.
 
Last edited:

Thread Starter

AIRMALE

Joined Sep 13, 2010
9
Bill and everyone else, Let me first say your help is greatly appreciated. All the leds on the boards are working. I tested each one individually and all are working. Vf for the leds tested at 3.125 average. what i would like to know is what i need in the way of a driver to light them with 12vdc. thanks again.
 

SgtWookie

Joined Jul 17, 2007
22,230
Bill and everyone else, Let me first say your help is greatly appreciated. All the leds on the boards are working. I tested each one individually and all are working. Vf for the leds tested at 3.125 average.
The real question is, what is the variance per 2-LED string when using a constant 20mA current source across 1 LED at a time?

If the LED pairs in the strings were matched up very closely for Vf, you may not have to do all of the trace cutting/adding of resistors.

what i would like to know is what i need in the way of a driver to light them with 12vdc. thanks again.
If you want to light them cheaply and inefficiently, use a 10 Ohm 20 Watt power resistor.

If you want to light them reasonably efficiently, use a 600mA to 700mA BuckPuck.
 

John P

Joined Oct 14, 2008
2,025
Generally I would call driving LEDs in parallel a bad idea, but if they're an identical type it might work well enough to be usable. They won't obey the law of Herr Ohm, but they will have some current-versus-voltage characteristic that makes them share the load, more or less. I'd be interested to hear what, if anything, has failed on those boards.
 

Thread Starter

AIRMALE

Joined Sep 13, 2010
9
I took 2 leds off the one board and put them in series and solo and got these measurements

supply voltage= 9.05 vdc
resistor value measured = .670 ohms
Vf across the 2 leds in series = 5.610 vdc
------------------------------------------------
supply voltage= 9.05 vdc
resistor value measured = .670 ohms
Vf across 1 led (solo not in series) = 2.885 vdc
 

Wendy

Joined Mar 24, 2008
23,415
You won't be able to measure supply voltage. LEDs don't work that way, they use current, not voltage. This is why the resistors are needed. You need the Vf to calculate the resistance to set the proper current, once you have the correct current flowing through the LEDs the voltage is incidental and not important.

As for getting identical types, sorry, there really isn't such a thing. You can get close, but over time they will drift badly (which is also why you can get a cascade failure).

I'm still betting one of the LEDs is dropping a lot less voltage than the others. With the resistor in each chain it doesn't matter, but if you approach a volt difference the others would dramatically dim out while one would be really bright. In other words, two LEDs (in their chain) would be getting way too much current.

What little I'm hearing the battery either went south or a couple of LEDs. Take your pick.
 
Top