Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

resisters for leds

3538 views
22 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    December 2004
  • From: Gainesville area
  • 1,396 posts
Posted by scubaterry on Monday, September 10, 2007 3:09 PM
David and Jeff....I have some more questions concerning my 3.3v PP but I don't want to hijack this thread so I am going to start a new thread called "3.3v computer pwr supply for LEDs". If you get a chance check it out. Sounds like you guys know your trons.
Terry in FLorida
Terry Eatin FH&R in Sunny Florida
  • Member since
    July 2006
  • From: Vail, AZ
  • 1,943 posts
Posted by Vail and Southwestern RR on Monday, September 10, 2007 12:34 PM
 dstarr wrote:
 scubaterry wrote:

If I could just borrow this thread for a moment.  On a similar
LED question.  I have a 3.3 v computer PP that I am using to power
my structure lights.  I have one complex that so far  has 12
white LED's.  Combination of 3mm and 5mm.  I currently have
them wired in parallel directily off the power supply.  Someone
said that I should have a limiting resistor for current
protection?  Is that a valid statement.  If so how large of a
resistor should I use.  These LEDs have about 10 hours on them as
is to date with no problems. Thanks

Terry from FLorida 

 Hmm, interesting hookup.  3.3 volts is the newer logic level, replacing the traditional 5 volts.  The forward voltage drop for most LED's is not far from 3.3 volts.  So there isn't much extra voltage to drop across the current limiting resistor.  It's not a circuit that I would use, but from what you say, the white LED's are lighting and not going poof.  You could put a 27 ohm resistor in series with each LED, but if they are lighting and surviving, adding the resistors won't change things much.  

   Was I doing is from scratch, I'd want to use a higher voltage supply, so I could drop more volts across the resistors.  What's happening here is the LED forward drop seems to be just enough to limit the current to a safe (non LED smoking) value.  Forward voltage drop on LED's can vary quite a bit from batch to batch.  It's considered better design to depend upon the resistor which is good to 5 % and the supply voltage (usually good to 1 %) to determine circuit current, rather than LED forward voltage drop which can vary by 30 % or more.  On the other hand, if it's working for you, there is no reason not to leave well enough alone.  Long as the LED's don't smoke and the brightness is reasonable, it's working.   LED's blow out in milliseconds, faster than the eye can see.  If yours have been lighting for some time, they will continue to light for a long long time.  LED's don't wear out like incadescent lamps, they will glow happily forever. .

 

I think I agree with dstarr.  I would be tempted to add 27 Ohms in series with each LED, since I think that at 3.3V you are probably at the high end of current through the LEDs, and probably reducing their life.  It's better to add a resistor for each LED, rather then for a group, since the resistor size would change depending on how many LEDs are in the group.

Jeff But it's a dry heat!

  • Member since
    August 2006
  • From: Franconia, NH
  • 3,130 posts
Posted by dstarr on Monday, September 10, 2007 11:58 AM
 scubaterry wrote:

If I could just borrow this thread for a moment.  On a similar
LED question.  I have a 3.3 v computer PP that I am using to power
my structure lights.  I have one complex that so far  has 12
white LED's.  Combination of 3mm and 5mm.  I currently have
them wired in parallel directily off the power supply.  Someone
said that I should have a limiting resistor for current
protection?  Is that a valid statement.  If so how large of a
resistor should I use.  These LEDs have about 10 hours on them as
is to date with no problems. Thanks

Terry from FLorida 

 Hmm, interesting hookup.  3.3 volts is the newer logic level, replacing the traditional 5 volts.  The forward voltage drop for most LED's is not far from 3.3 volts.  So there isn't much extra voltage to drop across the current limiting resistor.  It's not a circuit that I would use, but from what you say, the white LED's are lighting and not going poof.  You could put a 27 ohm resistor in series with each LED, but if they are lighting and surviving, adding the resistors won't change things much.  

   Was I doing is from scratch, I'd want to use a higher voltage supply, so I could drop more volts across the resistors.  What's happening here is the LED forward drop seems to be just enough to limit the current to a safe (non LED smoking) value.  Forward voltage drop on LED's can vary quite a bit from batch to batch.  It's considered better design to depend upon the resistor which is good to 5 % and the supply voltage (usually good to 1 %) to determine circuit current, rather than LED forward voltage drop which can vary by 30 % or more.  On the other hand, if it's working for you, there is no reason not to leave well enough alone.  Long as the LED's don't smoke and the brightness is reasonable, it's working.   LED's blow out in milliseconds, faster than the eye can see.  If yours have been lighting for some time, they will continue to light for a long long time.  LED's don't wear out like incadescent lamps, they will glow happily forever. .

 

  • Member since
    December 2004
  • From: Gainesville area
  • 1,396 posts
Posted by scubaterry on Sunday, September 9, 2007 6:57 PM

If I could just borrow this thread for a moment.  On a similar
LED question.  I have a 3.3 v computer PP that I am using to power
my structure lights.  I have one complex that so far  has 12
white LED's.  Combination of 3mm and 5mm.  I currently have
them wired in parallel directily off the power supply.  Someone
said that I should have a limiting resistor for current
protection?  Is that a valid statement.  If so how large of a
resistor should I use.  These LEDs have about 10 hours on them as
is to date with no problems. Thanks

Terry from FLorida 

Terry Eatin FH&R in Sunny Florida
  • Member since
    June 2006
  • From: Coquitlam BC
  • 629 posts
Posted by fsm1000 on Monday, July 16, 2007 12:21 AM

Wow thanks guys. Lots of great info and links here.

Gonna be reading for a while I see :D

Thanks again :) 

My name is Stephen and I want to give back to this great hobby. So please pop over to my website and enjoy the free tutorials. If you live near me maybe we can share layouts. :) Have fun and God bless. http://fsm1000.googlepages.com
  • Member since
    January 2004
  • From: Tarpon Springs, FL
  • 331 posts
Posted by cmarchan on Saturday, June 16, 2007 8:33 AM
 dstarr wrote:

   LEDs are not lightbulbs.  A lightbulb wants a constant voltage drive, and it will limit its current all by itself.  LEDs want a constant 20 mA current  and do nothing the limit that current.  Think of an LED as a rectifier that just happens to emit light when it's conducting.  Rectifier are either OFF or ON, like a switch.  If you were to hook a switch from plus to minus on your power pack and turn it ON, the power pack will see a dead short circuit.  Current thru a short circuit goes to infinity.  In the real world something melts before you get very close to infinity.  Hooking up a LED to power without a series resistor is just like closing a switch, you get a short circuit.  LED's are fairly tender and they blow out in microseconds (Faster than the eye can see) thus protecting the power pack at the cost of their own life.  Very noble minded components,  LEDs are. 

   So, LED's always need a series resistor.  Find the needed value with Ohm's law, R = V/I. I is always 20 milliamps (0.02 Amps) and V is the driving voltage from whatever you are powering the LED with.  This calculation makes a short cut, we are assuming that the voltage drop across the LED is zero. 

   If you want to be a little more accurate, subtract the voltage drop across the LED from the driving voltage.  LED's usually drop about 2.75 volts at full brightness.   I don't pay much attention to the 'voltage rating' of individual LED's from the catalog or data sheet.  It usually a worst case number ("All my LED's will be less than 3.5 volts at max current on a cold day").  When you measure the actual voltage drop across a glowing LED, its pretty close to 2.75 volts no matter who made the LED in question.   If you are driving the LED from something like 12 or 14 volts, the LED voltage drop is not very significant.  If you are using 5 volts, then it is.

   The 20 mA current is the max current that the average LED will tolerate before going "poof". Most LEDs will glow plenty bright enough on less, say 10 mA.  So the resistor value isn't very critical, you can go 20-50% high on the resistor and the LED will glow just fine.

 

Well done! To reinforce LED characteristics, the current flow is the factor to concentrate on as the voltage drop is based on the LED type and semiconductor material it is made from. Once you reach the IR drop of a LED, diode or transistor, the current flow can vary with the voltage drop changing only slightly, if at all. As Dstarr indicated use a resistor that will give you less than maximum rated current. The difference in brightness is slight and your LED will not be subjected to spike currents that could damage it.

 

Carl in Florida - - - - - - - - - - We need an HO Amtrak SDP40F and GE U36B oh wait- We GOT THEM!

  • Member since
    August 2006
  • From: Franconia, NH
  • 3,130 posts
LED's need a constant current, not a constant voltage.
Posted by dstarr on Friday, June 15, 2007 12:35 PM

   LEDs are not lightbulbs.  A lightbulb wants a constant voltage drive, and it will limit its current all by itself.  LEDs want a constant 20 mA current  and do nothing the limit that current.  Think of an LED as a rectifier that just happens to emit light when it's conducting.  Rectifier are either OFF or ON, like a switch.  If you were to hook a switch from plus to minus on your power pack and turn it ON, the power pack will see a dead short circuit.  Current thru a short circuit goes to infinity.  In the real world something melts before you get very close to infinity.  Hooking up a LED to power without a series resistor is just like closing a switch, you get a short circuit.  LED's are fairly tender and they blow out in microseconds (Faster than the eye can see) thus protecting the power pack at the cost of their own life.  Very noble minded components,  LEDs are. 

   So, LED's always need a series resistor.  Find the needed value with Ohm's law, R = V/I. I is always 20 milliamps (0.02 Amps) and V is the driving voltage from whatever you are powering the LED with.  This calculation makes a short cut, we are assuming that the voltage drop across the LED is zero. 

   If you want to be a little more accurate, subtract the voltage drop across the LED from the driving voltage.  LED's usually drop about 2.75 volts at full brightness.   I don't pay much attention to the 'voltage rating' of individual LED's from the catalog or data sheet.  It usually a worst case number ("All my LED's will be less than 3.5 volts at max current on a cold day").  When you measure the actual voltage drop across a glowing LED, its pretty close to 2.75 volts no matter who made the LED in question.   If you are driving the LED from something like 12 or 14 volts, the LED voltage drop is not very significant.  If you are using 5 volts, then it is.

   The 20 mA current is the max current that the average LED will tolerate before going "poof". Most LEDs will glow plenty bright enough on less, say 10 mA.  So the resistor value isn't very critical, you can go 20-50% high on the resistor and the LED will glow just fine.

 

  • Member since
    October 2006
  • From: Western, MA
  • 8,571 posts
Posted by richg1998 on Friday, June 15, 2007 11:36 AM

 reklein wrote:
I just happened to read this AM  In the sondtraxx manual,that certain soundtraxx decoders prefer incandescant bulbs in the 12- 16V range.

Here is info concerning this issue. I use this fix with LEDs in LC conversions for my HO steam engines.

http://www.tonystrains.com/technews/soundtraxx-lcleds.htm 

http://www.members.optusnet.com.au/nswmn1/LEDs_DSDs.htm

http://www.members.optusnet.com.au/nswmn1/Lights_in_DCC.htm

rich 

If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.

  • Member since
    July 2004
  • From: Lewiston ID
  • 1,710 posts
Posted by reklein on Friday, June 15, 2007 11:03 AM
I just happened to read this AM  In the sondtraxx manual,that certain soundtraxx decoders prefer incandescant bulbs in the 12- 16V range.
In Lewiston Idaho,where they filmed Breakheart pass.
  • Member since
    August 2005
  • From: S.E. Adirondacks, NY
  • 3,246 posts
Posted by modelmaker51 on Friday, June 15, 2007 10:56 AM

You're  best off with 1k-1/4watt resistors. Many decoders produce a surge-current for lightbulbs when powering up, (some do have a CV setting just for LEDs). This surge-current can cause an LED to fail over time, even if you have the "correct" calculated value resistor. The 1k will protect the LED and extend it's life.

All the calculations are correct in theory, but not neccessarilly right for the real world. The 1k gives you a good fudge factor and works with any LEDs in the 2 to 4 volt range without affecting the brightness much and you don't need to keep a whole bunch of different resistors around.

Jay 

C-415 Build: https://imageshack.com/a/tShC/1 

Other builds: https://imageshack.com/my/albums 

  • Member since
    October 2006
  • From: Western, MA
  • 8,571 posts
Posted by richg1998 on Friday, June 15, 2007 10:10 AM

 mmartian22 wrote:
what you are saying a 14 vt with a 450 ohms woulld work for most apps. i was told a 12 vt with a 750 ohms would work .but wasn't sure

The values I previously mentioned were just on the work bench measurements. I would rather not run at the current limit of 20 ma. I use 750 ohms in my engines just to be cautious.

Rich 

If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.

  • Member since
    July 2004
  • From: Lewiston ID
  • 1,710 posts
Posted by reklein on Friday, June 15, 2007 9:40 AM
If you take RMC they had a great series on DCC I collected all the articles and compiled them into a little binder for reference. My point is If its possible take a lokk back through those issues. I got a set of 3mm LED's from Eberle trains ,now defunct I think that also supplied 1K resistors with em. I think you'll do alright provided you know for sure the values of your components. Guessing in electronics kinda equates with smoke and mirrors. Good Luck
In Lewiston Idaho,where they filmed Breakheart pass.
  • Member since
    January 2004
  • 166 posts
Posted by mmartian22 on Thursday, June 14, 2007 9:31 PM
also i will look at links that was sent they look qute helpfull thanks guys , when i started this project i told by several people do it  one or another way in which got i confussed about it
  • Member since
    January 2004
  • 166 posts
Posted by mmartian22 on Thursday, June 14, 2007 9:14 PM
what you are saying a 14 vt with a 450 ohms woulld work for most apps. i was told a 12 vt with a 750 ohms would work .but wasn't sure
  • Member since
    October 2006
  • From: Western, MA
  • 8,571 posts
Posted by richg1998 on Thursday, June 14, 2007 10:46 AM

I have some 1.6mm OD white LEDs that are rated for 3.2 volts and 20 ma. calculated resistor is 470 ohms.

I used at 470 ohm resistor and 12 .2 volt battery and measured 18.5 ma with 3.2 volts on the LED.

I have some 5mm inverted cone white LEDs that measure the same.

I use 1/4 watt resistors. Below is a wattage calculator link.

http://www.anderson-bolds.com/calculator.htm

rich 

If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.

  • Member since
    July 2004
  • From: Lewiston ID
  • 1,710 posts
Posted by reklein on Thursday, June 14, 2007 9:11 AM
I had my multi-meter on my DH123 yesterday and it was putting out 13.68V on the lighting wires. So it sounds to me that the 14V figure is correct. To remove doubt go to the above website and do the math and save your LED's and decoders.
In Lewiston Idaho,where they filmed Breakheart pass.
  • Member since
    October 2005
  • 1,047 posts
Posted by betamax on Thursday, June 14, 2007 4:42 AM
 cacole wrote:

If you're certain your LEDs are rated for 12 volts, you don't need a resistor other than one of maybe 10 Ohms to protect the decoder against a power surge, because most decoders have an output of 12 Volts on their light function output.



That will work for about a second.

The whole point of the series resistor is to limit the current flow. Ten ohms is a waste of time, and of LEDs. LEDs are not light bulbs, and do not share many characteristics of them either.
  • Member since
    January 2004
  • From: Tarpon Springs, FL
  • 331 posts
Posted by cmarchan on Wednesday, June 13, 2007 9:14 PM
 cacole wrote:

If you're certain your LEDs are rated for 12 volts, you don't need a resistor other than one of maybe 10 Ohms to protect the decoder against a power surge, because most decoders have an output of 12 Volts on their light function output.

12 LEDs - are you referring to LEDs with resistors presoldered to them? LEDs are 2.5V to 3.5V typically; BTW you should calculate using 14 volts and not 12 volts for DCC.

In any case most LEDs for DCC locos work well with a 1000 ohm 1/4 watt resistor. Light output is not linear as with a light bulb; using a lower resistor value in most cases increases the current with not much difference in light; any spikes through the LED working at maximum current may cause it to fail. I've had to replace a few under these circumstances; in each case the installer used the recommended resistors values from Miniatronics and others.

Carl in Florida - - - - - - - - - - We need an HO Amtrak SDP40F and GE U36B oh wait- We GOT THEM!

  • Member since
    July 2003
  • From: Sierra Vista, Arizona
  • 13,757 posts
Posted by cacole on Wednesday, June 13, 2007 8:37 PM

If you're certain your LEDs are rated for 12 volts, you don't need a resistor other than one of maybe 10 Ohms to protect the decoder against a power surge, because most decoders have an output of 12 Volts on their light function output.

EDIT:  Sorry for the mistake.  I meant to type 100 Ohms, not 10.  In order to drop 2 Volts at 30mA, you would need 66 Ohms, but 100 Ohms would be better to protect the LED and decoder.

Thanks to betamax for pointing that out.

  • Member since
    September 2004
  • From: Ontario, Canada
  • 95 posts
Posted by CMLewis on Tuesday, June 12, 2007 10:47 PM

LEDs have a spec known as forward voltage;  it describes the optimum voltage across the LED and is usually paired with an optimum forward current.  This represents a point on the operating curve that is usually safe.  For white LEDs, forward voltage is in the 3.0 - 3.5 volt range, normally 3.3 volts.  Optimum forward current is typically 20mA; these ratings should be printed on the LED package.

In order to achieve this using a 12-volt power supply, you will need a 470-ohm resistor in line with the LED.  If it's too bright, put in a larger resistor.  If you put in a smaller resistor the LED will be much brighter, but won't last very long.

Hope this makes sense.

Chris

  • Member since
    October 2006
  • From: Western, MA
  • 8,571 posts
Posted by richg1998 on Tuesday, June 12, 2007 7:37 PM
 mmartian22 wrote:

Lm begining to install leds in my loco 's what i need to know is .i m using 12vt leds for rear and front lites . what i don't know is what resistor is needed for them .using digitrax decoders for them.wattages .ohm's

 Here is another link. There is also a online resistor calculator.

http://led.linear1.org/1led.wiz

I normally use 750 ohm since I run 14 volts for DCC and I like to keep the current a little under 20ma.

http://led.linear1.org/

rich 

If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.

  • Member since
    October 2005
  • 1,047 posts
Posted by betamax on Tuesday, June 12, 2007 7:22 PM
  • Member since
    January 2004
  • 166 posts
resisters for leds
Posted by mmartian22 on Tuesday, June 12, 2007 7:06 PM

Lm begining to install leds in my loco 's what i need to know is .i m using 12vt leds for rear and front lites . what i don't know is what resistor is needed for them .using digitrax decoders for them.wattages .ohm's

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!