24 LED's in parallel

24 LED\'s in parallel

Hello I want to hook 24 5mm LED's in parallel powered by 4 Nimh AAA cells for a tottal of 4.8volts. Will this be safe for the LED's recommended draw of 3.6V? I should lose some voltage due to the fact that there are 24 in parallel right? How long do you expect the 700mah cells to last before dumping to around 3V?

Re: 24 LED\'s in parallel

You should use current limiting resistors, preferably a separate one for each LED. If there's 4.8 volts from the battery and 3.6 volts across the LED, that's 1.2 volts across the resistor. At 30 mA, that would be 40 ohms (39 ohms is a standard resistor value and is close enough). 24 leds times 30 mA is .72 amps, so the cells should run the led's for about 1 hour.

Re: 24 LED\'s in parallel

So what if I don't use the resistors? Would it significantly reduce the life of the LED? I'm not going to be using it a whole lot and even if it reduces it to 1% life, or 1000 hours, its all good.

Re: 24 LED\'s in parallel

If you don't use resistors, there's a high chance that some will burn out before others. And it's not 1000 hours of life a little lower type of burn out, as the LEDs are being run at higher current and heats up, their resistance dropss and even more current flows through them, so, it's a thermal runaway situation. At bare minimum, have a 20 ohms resistor.

I do have a 56 LED battery drainer here, I used a 47 ohms resistor for each LED, they light up at 3 volts and can be run all the way to 7 volts. Draws around 0.7 amps with 4 x 2100mah Sanyo AA, probably around 1.4 amps with 6 of those.

Re: 24 LED\'s in parallel

LEDs, except in very special circumstances, should not be directly driven. When driven at the voltage they need they will draw as much current as the source will provide. And in the case of rechargeables this is a lot. These would likely destroy themselves in short order unless the wiring provides enough resistance to prevent it! One resistor per LED is the best idea. However if the LEDs are closely matched, by using ones from the same batch and checking the drop voltage, you could treat it as one large LED. Some practical matters. Voltage of the fully charges cells would be about 1.35V each. 1.25 is considered the mid-point when the cells retain about half of their power. And a pack of cells (battery) should not be drained below 0.9V per cell to reduce the chance of damaging the weakest cell(s). Rated capacity of cells is figured at 1/10 the capacity. Or stated another way, how much it can supply over 10 hours. To complicate this, with a simple resistor you are going to be using a constantly varying amount of power as the battery discharges. Now some math: 4*1.35V = 5.4V, 4*0.9V = 3.8V. This last is lower than some LED makers suggested 4.1V however if is higher than the real drop volts it will still work. Next lets assume the real LED drop voltage is 3.4V and max. rating is 30ma. Total max. current is then 24 * 30ma = 720ma The voltage we need to see across the resistor is the 5.4V - 3.4V = 2V. So, using Ohm's Law R = E/I, 2/0.720 = 2.78 ohms. To see which direction to go to find a real resistor we should do our nominal calculation 5V - 3.4V = 1.5V, 1.5/0.720 = 2.08 ohms. Now we need to see how much power the resistor has to handle: P = V * I, 2 * 0.720 = 1.44W. For practical matters it looks like two 4.7 ohm 2 watt resistors in parallel is the best compromise. This will be running the LEDs somewhat over their max. part of the time. Now about those 24 LEDs, unless you're getting them real cheap, wouldn't a 1W Luxeon be easier. You're going to have to deal with the heat anyway.

Re: 24 LED\'s in parallel

The LED's, the batteries, and other components all create resistance. I'm not an electrical engineer, but I would bet that you won't need any additional resistance.

I have a green LED which I scavenged from a cheap button-cell key-chain light. It runs contentedly in a 3 D cell light with no resistance. It also runs fine on a 3.7 volt Li-ion battery with no resistance. My experience contradicts the advice of the engineers.

I have Rayovac 15 minute rechargeables. They measure 1.46 volts right off the charger. Once again, quite different than the 1.2 volts that most people in these forums attribute to rechargeables.

My suggestion is to wire the leds without resistance and see what happens. Worst case you will lose one or 2 before you turn off the power. Best case, you'll get more light and better efficiency.

Most commercial lights over-drive 5mm leds by 2 to 3 times the recommended current. Lets say you want to provide 80mA per led. 24 X .08 = 1.9A. That's probably close to the limit of what 4 rechargeables can deliver, especially considering all of the resistance in the whole circuit.

Re: 24 LED\'s in parallel

I put 18 UV LEDs in parallel driven by a LED Supply MicroPuck current driver. It runs off 2AA and drives 375mA, you choose how many LEDs you want to feed.

Works fine for me.

PEterW

Re: 24 LED\'s in parallel

Its best to use a resistor for each LED, otherwise - if you just use one very small value resistor and run the entire string on it and should one LED go out, that would subject the rest of the string to slightly higher current, and then the process could "migrate" to the rest and start blowing LEDs

Re: 24 LED\'s in parallel

[ QUOTE ]
FNinjaP90 said:
So what if I don't use the resistors? Would it significantly reduce the life of the LED? I'm not going to be using it a whole lot and even if it reduces it to 1% life, or 1000 hours, its all good.

[/ QUOTE ]
The internal resistance of NiMH cells is very low, so in all likelihood you will be putting about 5 volts across the LED with freshly charged cells. Forward voltage varies quite a bit even among LEDs from the same manufacturer, but in most cases white LEDs have Vf from about 3V to 3.8V. Due to the logarithmic relationship between current and voltage in a diode, every 0.1 volt increase in battery voltage will result in current being roughly doubled. This is to a first approximation-at higher currents the bulk resistance of the die helps to limit current. Putting 5V across an LED that draws 20 mA at a Vf of 3.6V will result in a huge current(perhaps 200 mA). The LED would blow out in seconds or at most minutes.

Another thing-lifetime for 5mm white LEDs at rated current is more on the order of 6000 to 10000 hours because the epoxy yellows and blackens by being exposed to short wavelength blue light. Here again, lifetime follows a somewhat exponential relationship to current, only it is an inverse exponential relationship. Increasing the current to 40 mA might result in only a 1000 hour life. The exact relationship depends up the type of diode, but the same principle applies. This is why a Luxeon III has a lifetime of 50,000 hours at 700mA but only 20,000 hours at 1A, for example.

In summary, use resistors, preferably one for each LED, in order to avoid problems. Also, be aware that even at rated current, a large array may have heat problems which give shorter lifetime than expected because the leads are the main path from which heat is removed from 5mm LEDs.

Re: 24 LED\'s in parallel

[ QUOTE ]
gwbaltzell said:
Now about those 24 LEDs, unless you're getting them real cheap, wouldn't a 1W Luxeon be easier. You're going to have to deal with the heat anyway.

[/ QUOTE ]
I don't know where the guy who started this thread gets his LEDs from, but ChiWing on EBay has 10000 mcd white LEDs for $0.25 each shipped, so using a 24 LED array instead of a 1W Luxeon can make economic sense. Since the 5mm LEDs have a narrow beam, you avoid the need to purchase optics, and also avoid the "Luxeon lottery" (ChiWing's LEDs are pretty even and consistent, with a color temp of ~6500K by my eyes). An array of 20 made a nice bicycle headlight for me.

Re: 24 LED\'s in parallel

Ok this is what I wanted to use them for. I am a student with a TI 83+ Silver Edition. It has a transparent case. I was thinking about putting a lot of LEDs' at random places inside the case. I will install wiring from the negative and positive battery terminals to power the LED's and the calculator. The batteries, will, in turn, power both the LED's and the calculator processor and screen.

The calculator runs off 4 AAA batteries, and since the LED's will make it run for a lot shorter time, I will get rechargeable nimh's. Heat isn't that much of a problem, since they won't be concentrated in an array.

I don't think I have the patience to solder 24 or so resistors together, so what kind of harm will it do if I just put one on it and treat the whole thing like a big LED?

Re: 24 LED\'s in parallel

That's a lot of LED's for this application. Some things you might think about.
1. Using a lot fewer LEDs and adding some fiber optic to scatter the light in the various places. Maybe even using those color changing LEDs.
2. Running the LEDs in series or more likely several strings in series and you will only need one resistor per string. However the voltage will have to be increased. You might ask over in Electronics & Batteries for some simple inverter designs.

Re: 24 LED\'s in parallel

CPF user Jonathan might be able to help with an inverter.
I have one of his Spider LED Bike Light modifications that uses such a circuit. It takes the 3 volts from a pair of AA cells (it can run at up to 12 volts), and jacks it up to around 25 volts to run six high-flux LEDs in series. And it's regulated, so you won't lose brightness as the batteries in your calculator become used up.
I don't know if this circuit has the {vulgar term for testicles} to run 4 or 5 strings of 5mm LEDs like you want, but it's worth a look anyway.