Picking Resistors for LEDs

Hot Topics:

By John Baichtal
Posted 2012/08/30 @ 7:30 am
Category Electronics

Lenore Edman from Evil Mad Scientist Laboratories has a thorough post on picking which resistors to use with LEDs.

So… you just want to light up an LED. What resistor should you use?

Maybe you know the answer, or maybe everyone already assumes that you should know how to get to the answer. And in any case, it’s a question that tends to generate more questions before you actually can get an answer: What kind of LED are you using? What power supply? Battery? Plug-in? Part of a larger circuit? Series? Parallel?

It’s a great piece that helps makers learn the why rather than just leading them to (for instance) an online resistor calculator.

In the Maker Shed

Gift Certificate

Makey Makey Standard Kit

BrushBots

Getting Started with Arduino Kit v3.0

13 Responses to Picking Resistors for LEDs

Charles Gallo on August 30th, 2012 at 4:10 pm said:

Optimally, you DON’T – LEDs are NOT ‘voltage’ devices, but CURRENT devices, and typically have a junction drop of about .7 volts, like any diode, you use OHMs law, but you are MUCH MUCH better off using a constant CURRENT driver, which you can easily make out of a LM117, or if you want to get fancy, use a switching LED driver. The fun is, once you setup a CC driver, you can string a BUNCH of LEDs in series, and they will all get the current

Reply ↓
- fonz on August 30th, 2012 at 4:45 pm said:
  
  the junction drop of an LED is not about .7 volts, it is several volts depending on what color it is. If your supply voltage varies a lot or for big LEDs a constant current driver makes sense, for a simple 20mA LED it is wild overkill
  
  a series resistor is a kind of constant current driver, just not a very good one, unless the voltage is high and the resistor big, look up “thevenin”
  
  Reply ↓
Kent Durvin on August 30th, 2012 at 5:31 pm said:

If the voltage is relatively constant and known, a resistor is a perfect current limiter. You must know the voltage drop of the LED, but the voltage is usually about the same for any given color. Red is about 1.6V. Green is about 2.2V. Blue, UV, or white are about 3.7. Subtract that from the power voltage. That gives us how much we need to drop. For example, 9V-1.6V=7.4V. At whatever current we decide for the LED, we need to drop 7.4V. 10mA is good. So 7.4/.01=740ohms. 680 is a standard 5% value, and it would give just a little more current.

Reply ↓
sparky3489 on August 30th, 2012 at 7:31 pm said:

http://led.linear1.org/led.wiz

Done.

Reply ↓
- Randy Pyter on September 3rd, 2012 at 6:53 pm said:
  
  Red, orange, yellow, and yellow-green (570 nM) LEDs usually have a Vf rating of 1.8-2.2 volts True green (525 nM), blue, white, & IR LEDs usually have a Vf rating of 3-4 volts. Read the data sheets. The brightness ratings of the 3mm, 5mm, and small smt LEDs are usually, but not always, rated @ 20mA. The current limiting resistor, in addition to protecting the LED, can be used to achieve a different light output in different color LEDs to match the perceived brightnesses. With most low power LEDs if the current is cut by 25% the light ouput is also cut by 25%. Using a lower current will extend the life of the LED while exceeding the rating rating will shorten the life, potentially to 0. Please note that the green LEDs sold on Ebay are always true green with the higher Vf rating. Please also note that the white, blue, IR, and true green parts are also much more suceptible to damage from static electricity. Always read the data sheets or if not availaable use the numbers listed above.
  
  Reply ↓
  
  sparky3489 on September 3rd, 2012 at 7:13 pm said:
  
  You forgot to mention UV LEDs (360-365 nM, true UV) usually have a Vf rating of about 3.7 to 4 volts depending. But then all you really need to do, as you say, is read the data sheets.
  
  Reply ↓
jammit on August 31st, 2012 at 3:37 pm said:

My rule of thumb is 50 ohms per volt for a bright LED, 100 ohm per volt for indicator LED. For example, a 12V power with a 510 ohm (closest to 600 in standard sizes) resistor gives a current of approximately 23ma for bright, and 12V with 1.2K ohm resistor gives a current of approximately 10ma, which is good enough for an indicator. This only works for voltages 3V and up for all LED’s that aren’t silicon carbide (blue or white).

Reply ↓
Randy Pyter on September 4th, 2012 at 8:25 pm said:

I typed IR and meant uv, duh.Senior moment. I created a simple Excel spread sheet that does the resistor value and wattage calculations and as a reference lists the standard eia values for 1% & 5% resistors and the standard smt & through hole sizes and wattages. It also has a cell that asks for the number of LEDs in series as many hobby materials put three LEDs in series and use a 12 volt supply. Please also note that identical LEDs are not consistent in color (nM), Vf, and light output (mcd). In the industry LEDs are often bin sorted for one of these parameters to achieve more consistency. For the hobbist, light the LEDs up side by side and set aside any that look different.

Reply ↓
- sparky3489 on September 4th, 2012 at 8:34 pm said:
  
  I kind of did the same thing a while back in a program I wrote in a very old language of BASIC. I could actually write it in VB with animations and stuff….Hmmmm.
  
  Perhaps it can be mentioned that a LM317 can be used as a cheap LED constant current driver with only one additional resistor. I just happened to be investigating that very thing for work. I’m trying to find out why some LEDs are shorting out. Any ideas?
  
  Reply ↓
  
  Randy Pyter on September 5th, 2012 at 7:04 pm said:
  
  I am an ME and it is quite easy to get beyond my rudimentary knowledge of electronics. You need to insure that the reverse voltage rating is not exceeded or you will end up with dead LEDs. I have seen seasoned EE’s wire red-green LEDs up anode to cathode, swap the highs and ,lows, exceed this parameter, and wonder why LEDs quit working. The far east LEDs being sold on Ebay can be quite inconsistent and I have had several doa. Sorry I can’t be of more help.
  
  Reply ↓
  
  sparky3489 on September 5th, 2012 at 7:40 pm said:
  
  Not sure what an ME is, but I’m a level III Electronics Technician and I can see why you’ve seen seasoned EE’s wire LEDs wrong. They don’t study at the component level.
  
  The reverse voltage is ONLY important if the voltage being fed has the possibility to reverse polarity.
  
  Reply ↓
  
  Randy Pyter on September 6th, 2012 at 6:26 pm said:
  
  ME = mechanical engineer – The Maytag corporation had an appliance contoller that somehow “flopped” (ME term) the voltage and fried many LEDs due to exceeding the reverse voltage spec. Even as a mechanical guy I was amazed and somehow the blame was put on my company who never even saw their controller but didn’t warn them. Most techs that I have worked with had better hands on experience than most engineers, especially in the analog days Good luck with your op amp problem.
  
  sparky3489 on September 6th, 2012 at 6:47 pm said:
  
  Gotcha!
  Although I’m a bit confused on that last statement. I don’t have an op-amp problem.
  
  This is some of what I do – http://sparky3489.webs.com/pcinterfaceproject.htm
  
  Reply ↓