122Views17Replies

# Help converting a Greenlite LED light Bulb to DC power? Answered

I have a GreenLite LED bulb that I took apart and cannot for the life of me figure out how to power it via DC without blowing it up.

There are 21 LED's on an Aluminium plate, after mapping out the traces there are 3 series of 7 LED's with it running at 43.9-44.0 Volts. So I figure thats 6.2V per LED but everywhere I read White should be 3.0-3.2V

I cannot for the life of me figure out the current and voltage to provide to them without ruining them.

Here is some pictures of the whole thing with the board that converts from AC to DC.  the 1,2, and 3's are me just labeling the mapping of the series.

Tags:

## Discussions

To measure current, you connect your multimeter in SERIES with the LED board. Or you could place a 1 ohm resistor in series instead, and measure voltage drop across it with a multimeter. Make sure to use a big resistor and that you know the value of that resistor exactly. Ohm's law will tell you the current through it.

First, probe things out to see how they are connected and try to draw up a schematic for how you think the LEDs are wired. double check that your schematic is drawn right with a continuity tester/multimeter. Then "untangle" the messy schematic to get a nicer looking schematic that is electrically correct. Congrats, you reverse-engineered the LED section.

From this point, you will need to know a little electronics and understand series @ parallel circuits, and how LEDs should be powered. They might be multi die LEDs, which means each SMD part have multiple LEDs inside, hence why there are 2 LED diode drops across each one. Most likely, there are a few strings of series LEDs with a resistor in that string, and the LED strings are wired in parallel. However that is not the only way they might be wired.

You could just attach a variable voltage power supply across the board and SLOWLY raise the voltage from 0V to see if it powers on. Note what voltage & current the LEDs appear to be at nominal brightness. (That's the crude but effective method.)

First the LEDs are driven by a constant current source and applying a voltage may very well destroy at least seven... Can you measure the input DC current ?

Next, top and bottom views of the circuit as well focused as the ones that you have here would help... I need to find out if the series are balanced by a hidden resistors...

Too bad there are no small traces easy to cut and and take measurements if 14 leds keep burning or all go out...

Exampled is this group of three series LEDs which are actually a combination of multiple elements in the component also run on a constant current.... Six lamps light a large wearhouse..

Mesuring current is what i don't understand. I tried doing it with my multimeter that i just got and when I tried to measure amps I got lots of sparks!

Here are the pics as requested! Narrow end of board is where AC 120V come in.

Thank You!

Good pictures... Here is how you hook up the amp-meter...

I was hoping the circled EL Capacitor would let me see the component values... Can you read the Voltage and Capacitance of the output capacitor ?

BTW Jack's analysis of the LED composition is right on track...

Thanks for helping with the way to hook up the amp meter!! i did it and it worked great at setting 10A i get 0.16

You have two options;

1] Make or buy a PS, say about 65 VDC and put a resistor in series....

The math is quite simple

[a] We need the voltage across the Resistor = 65v - 44v = 21 volt

[b] Resistance = Voltage / current = 21v / 0.16 = 131 => use 130 ohms

c] Power = V(R) x Current = 21 X 0.16 = 3.4 w => use 5 watts

-------------------------------------------------------------------------------------------------

2] Build a simple 160 ma current regulator.

[a] The regulator will deliver 0.16a for voltage input from 48v up to 78v

[b] Click the pic to see the whole image...

[c] Resistor 7.5 ohms... Power = I x I x R = 0.16 x 0.16 x 7.5 = 0.19 w

[d] Use a 1/2 watt 7.5 ohm resistor

So the 43v 200ma DC supply and a 10 ohm series 1/2 watt resistor should be a good way to start....

Plus like Jack A Lopez suggests, you can measure the resistor voltage and your current in ma will be = volt x 100....

BTW since you marked the surface already try scratching the silkscreen to expose the copper on both sides of an led and actually measure several of the led voltages...

I am guessing each of those yellow blocks is a module containing 4 white LEDs (2 parallel strings of 2 in series) and this is why each module has a voltage of around 6.2 volts. Anyway, that's my hypothesis for why you get 6.2 V per yellow block.

Regarding the question of how to run this gizmo from DC power, I am not sure exactly what kind of DC you have in mind. If you like 12 volt DC, like from a car's battery, there exist inverters for to convert 12 VDC into something resembling mains power, which I guess is around 120 VAC where you live. (Or at least this is the voltage the lamp wants, according to the product page you linked to in your question.) In many places on Earth, mains power means 240 VAC.

Anyway, the inverter would be the easiest way to make it work.

The alternative is to build an LED driver made to match that plate of 21 (3 parallel strings of 7 in series) . I am guessing, for each of those three strings, the maximum current is about 60 mA, since 3*(60 mA) = 180 mA, and (180 mA)*(44 V) = 7.92 W, and that is approximately the power rating for the whole lamp gizmo.

It might be worthwhile to actually measure the current flowing through those red and black wires, for example by soldering a 1 ohm resistor in series with the LED array, then measuring the voltage drop across the 1 ohm resistor, while the lamp is on.

Back to the subject of building an LED driver, matched to that LED array, only one that runs from 12 VDC, or from D cell batteries, or whatever. I think that would be a lot of work. In fact it would be so much work, that it would be preferable to go shopping for a 12 VDC version of this same lamp, or power this lamp using an inverter.

I understand a bit better now! Its funny I have worked with Computers for 15 years but when it comes to this basic circuit stuff I'm CLUELESS!! Funny thing is I just found a AC/DC adaptor in my vast collection of crap down in the basement, its output is 43V at 0.2 A which i gather would power this thing with the proper resistor. Its just after the sparks I'm so nervous to test AMP's again.

I think the tricky thing about measuring current, is the necessity to break the circuit, i.e. cut a wire somewhere, for to insert the ammeter in series with the rest of the circuit.

Iceng upped a picture, showing where the ammeter should go, here:
https://cdn.instructables.com/F9H/AQ06/ILB5MN1D/F9H...

Although I am not sure if this picture, by itself, makes it clear what is going on here.

The red wire has been cut somewhere, and the ammeter has been inserted in between. Essentially the ammeter now serves as a short length of the red wire, which conducts current on the positive side of this LED array on a plate.

An ideal ammeter has zero resistance. Electrically, it looks like a short piece of wire. A real ammeter, or multimeter set up in ammeter mode, is a close approximation of this; i.e. it looks like a short piece of wire.

Thus the worst thing you can do with an ammeter, or multimeter set up in ammeter mode, is forget it is an ammeter, and then try to use it like a voltmeter, for example by connecting across a voltage source. If you do that, you're essentially connecting a short across a voltage source, which might explain the sparks.

alright I bit the bullet and just did the amp test again. so with my amp meter set to 10A i get a reading of 0.16, and with it set to 200M I get 01.6. AWESOME link for the free text book!

By the way, a lot of people do not understand circuits, so I usually try to push these people towards a textbook, any textbook, that explains this subject.

For example, here's a page from a free textbook on electronic circuits,