Powering small leds / resistor selection.

Status
Not open for further replies.

markebenson

Senior Member
Location
fl
I am putting together a small marquee letter sign. The led letters I bought look perfect but are battery powered. I need a reliable method of constant power. I am open to any efficient and reliable solution. I was thinking about using a 12vdc supply with current limiting resistors in series with each individual led. Let me know any ideas. Thx


!ltr1.jpga1.jpg

Existing:
7 led in parallel
power source - 3.0v ( 2 AA batteries)
current draw (combined) shown
 
Based on your picture of 7 LEDs in parallel drawing 50ma, each LED is drawing 7ma. This low current draw is likely due to the LEDs being supplied with only 3v which is right at the threshold of them even lighting up. We''ll use a 3v forward voltage in our calculation to maintain similar brightness.

12-3 = 9
9/.007 = 1285.72

So a 1.2K resistor in series with each LED.

.007 A through a 1.2k resistor will dissipate 0.0588 Watts so any 1/8w or 1/4w resistor is fine.

Or if you use something like a 5v phone charger instead of a 12v power supply:

5-3 = 2
2/.007 = 285.72, so a 270 ohm
 
Or leave them in parallel and use a 9/0.05 (180 ohm) for 12V source or 2/0.05 (40 ohm). That may be easier. Current is not a problem; smaller resistors (more mA) won't hurt. The slightly unmatched voltage characteristic of the LEDs MIGHT result in differing brightness, but no different from that with a battery. From the 12V source, you do get to close to 1/2 watt dissipated in the resistor.
 
Or leave them in parallel and use a 9/0.05 (180 ohm) for 12V source or 2/0.05 (40 ohm). That may be easier. Current is not a problem; smaller resistors (more mA) won't hurt. The slightly unmatched voltage characteristic of the LEDs MIGHT result in differing brightness, but no different from that with a battery. From the 12V source, you do get to close to 1/2 watt dissipated in the resistor.

Generally it is advised against supplying many LEDs in parallel from one smaller value resistor.
 
Generally it is advised against supplying many LEDs in parallel from one smaller value resistor.
I agree, but that is effectively how they are now. No rewiring. I'd try it before going into the letter and cutting leads.
 
I agree, but that is effectively how they are now. No rewiring. I'd try it before going into the letter and cutting leads.

I am speculating here, but the difference is the LEDs are being powered directly from the batteries at 3v which is right around their forward voltage threshold hence the draw of only 7ma/ea and there is no resistor. Since there is no resistor and the whole thing is naturally self current limiting we don't run into any of the issues with paralleling multiple LEDs on one resistor.
 
If you really want efficiency, then you don't want to use a resistor to drive the LEDs, rather you want a switching power supply with constant current output. However I am not really recommending this approach, because you are operating at such low power levels (7 leds, 3V, 7mA, 150mW).

The general rule is that you subtract your LED operating voltage from your supply voltage, giving the voltage difference that the resistor has to drop. You use Ohm's law, that voltage difference, and your desired operating current to calculate the resistor value.

The smaller the LED voltage relative to the supply voltage, the more drop in the resistor and the more power wasted heating the resistor. However supply and LED voltages both vary, which means that the closer you try to match supply and LED voltage the greater the variability in LED output.

You can trade off the above effects by putting more LEDs in series and using a smaller resistor.

Consider the example: nominal 13.8V power supply and 3V LED, but the supply is actually 14.4V and the LED 2.9V. The nominal efficiency is 22%, and when the voltage changes your current ends up 6% high. Now use the same components, but put 4 LEDs in series with a smaller resistor. Your nominal efficiency is 87%, but the change in voltage causes a huge 250% increase in drive current.

There is a useful and simple trick: you can use a simple linear variable 3 terminal voltage regulator with a single resistor to create a linear current regulator. See page 10 of the datasheet:


The above component ($0.35 at digikey https://www.digikey.com/en/products/detail/texas-instruments/LM317LCLPRE3/1510140 ) and a 125 ohm resistor will provide a solid 10mA as long as the input voltage is about 4V higher than the output voltage.

What I would recommend is to use the linear regulator described above with a 30V power supply and all 7 LEDs in series. This will give you even current to all of the LEDs, you can easily change the brightness by changing 1 resistor, and efficiency will be good at about 70%.

-Jon
 
What I would recommend is to use the linear regulator described above with a 30V power supply and all 7 LEDs in series. This will give you even current to all of the LEDs, you can easily change the brightness by changing 1 resistor, and efficiency will be good at about 70%.

-Jon

OP is making a marquee sign, the letters will have varying quantities of LEDs depending on the letter. The 12v regulated power supplies in question are usually spot on at 12v, some are even adjustable within a narrow range. Resistor per LED is the right way to do this one.
 
OP is making a marquee sign, the letters will have varying quantities of LEDs depending on the letter. The 12v regulated power supplies in question are usually spot on at 12v, some are even adjustable within a narrow range. Resistor per LED is the right way to do this one.

Always a trade-off, and I am not saying that single resistor per LED is a bad way to go at these low currents.

One point, however. While the power supply is probably quite tightly regulated to 12V, the LEDs will have a range of voltages and that voltage will change with temperature. The variation in voltage dropped by the resistor was the key point for why the resistor should drop a relatively large voltage for current stability.

Thus 1 resistor per LED even with a tightly regulated 4V supply is probably a bad idea, but with a 12V supply will have good even current distribution between the LEDs. My preference at 12V would be 1 resistor per 2 LEDs in series, but that doesn't work with an odd number of LEDs.

The linear current regulator approach is insensitive to changes in LED voltage, and with a 30V supply could reasonably handle anything from 3 to 8 LEDs at 10mA, so a single drive component per letter.

-Jon
 
Status
Not open for further replies.
Top