Lighting with LEDs - some circuit design confusion

[AlanMintaka]

Hi Everyone,
I've been looking into LED lighting for my HO layout.  I make use of a lot of structures made from the old Kenner's Girder and Panel building sets; cool lighting is essential because of the plastic material used in the panels.

I'm not too knowledgeable when it comes to LED circuit design, so I decided to get some sets of LED Christmas lights and take them apart to see how they worked.  I started with sets of 20 LED lights ("C3") powered by 3 1.5V AA batteries.   The manufacturer rates the LED bulbs as 4.5V 90mW. 

The 3 AA batteries are connected in series, so the total voltage is 4.5V.  A 4.7 Ohm resistor is wired to the + terminal of the batteries (resistor color code = yellow violet gold gold).  The 20 LEDs are connected in parallel to the other side of the resistor and the - terminal of the batteries.

My confusion has to do with how I might expand this circuit to include more bulbs, and how I might replace the batteries with a DC power supply. 

I ran the numbers through some of the online LED current limiting resistor calculators, e.g. the "series parallel array wizard" at http://led.linear1.org/led.wiz. 

I started by assuming that the forward current for each LED was 20mA.  This seemed like a typical value in the various places I looked online. 

Then I experimented with different values for the LED forward voltage until the current limiting resistor for my circuit corresponded to the 4.7 Ohm resistor I found in the Christmas set.  The forward voltage required to make this happen was 4.42V.

Isn't that a little on the high side for LEDs?  The values I've found online are closer to 3.0V on average for the different colors.  I tried this with a few other calculators and kept coming up with the same value for forward voltage.

So the first question is, am I doing this the right way?  For a circuit of 20 LEDs connected in parallel to a 4.5V supply with a current limiting resistor of 4.7 Ohms and LED forward current of 20mA, does a forward voltage of 4.42V make sense?

Secondly, if I replaced the batteries with some kind of DC power supply, could I use something like one of those "wall wart" DC transformers rated for 4.5V 20mA?  Would I still need the current limiting 4.7 Ohm resistor if the transformer is rated for 20mA? 

Finally, how many LEDs could I really add in parallel to this circuit?  Using the above mentioned array calculator, I added another 20 LEDs, then 40, then 80, and the solution was always the same: 4.5V supply, 4.7 Ohm current limiting resistor, LED forward current 20mA and forward voltage 4.42V.   There must be a practical limit to the number if LEDs I could add before something has to change, but what is it?

I'm missing something here but I'm too stupid to understand what it is.  I just want to find out how to calculate the right component and power supply values if I start adding LEDs to this circuit. 

Thanks for your time reading this convoluted message,
Big Al Mintaka

[rich1998]

first of all, led's are current devices. Get a multimeter also. Big help.

i keep it simple. an average led for model railroading needs .020 amps or 20 ma max. Since I use 12 volts dc, a 1,000 ohm resistor is what I use. that allows 10 ma through the led.
some people use a 750 ohm resistor which still keeps it under 20 ma.
i would suggest going with a 12 volt dc power source and figure from there. i do not like dealing with all the math.
Also, resistors are cheap even if you buy from radio shack.
If a lot of lead's are in series, one goes bad, they all go out.
Broken or bad lead, lead, all out.

lex

[Daylight4449]

here is my advice. there are many leds with many different current capabilities. that is why i try to stay consistent. not having the electrical know how of jim banner, i usually tinker with various resistors, working down showing overkill at first to reduce the number of Smelly leds as i call them. (Man did i stink up my friends basement with 30 odd blown leds). I used them in cars. however, they are fed via track which ment i couldn't use them on a dcc setup. and their lighting would vary. I will say that my best friends were a soldering iron (Duh), a Multi- Meter  , your best friend, and some sort of device that supplies varrying amounts of power and current to your circuit, telling you at what voltage and current, Current being extremelly important, it blows.  I know not much about circuit math but can tell you that you shouold not conclude that they all can take 20 amps.

[rich1998]

if you refer to my previous post, i did not say 20 amps. i said the max current is usually 0.020 amps or 20 milliamps. most led's put out good light at 10 milliamps. i use a 12 volt dc supply with 750 ohm to 1,000 ohm resistor.
i have a 5,000 ohm variable resistor i use to experiment with led's.
a large variety of led's sold for model railroading are 20 milliamps max. no doubt that will change as the led technology evolves.
many spectrums with led's have almost 3,000 ohm resistors for the led headlight.  the current is usually around 6 milliamps. i have measured them. reason for complaints about dim lights at one time.
the only time i have smoked an led was becasue i was in a hurry.
been working with this stuff before led's ever came about.
lex

[Daylight4449]

i know... meant miliamps but was in a hurry sorry however, the principle remains the same.

[Jim Banner]

Big Al,
I was having trouble making the numbers work.  Most white LEDs are rated at about 3.2 volts, not 4.5.  But then it hit me - these are probably self limiting LEDs that have built in current regulation.  That would account for the extra volt or so.  I figure it this way:
If the set is rated at 4.5 volts, 90 mW, then the current drawn is 20 mA, just 1 mA per LED.  If you multiply that 20 mA times 4.7 ohms, the resistor is dropping .094 volts.  Subtracting that from 4.5 volts, each LED is getting about 4.4 volts, or 1.2 volts more than it requires just to light up.  That is enough extra voltage for a constant current circuit.  And that 4.7 ohm resistor?  That is probably there as a very slow blow fuse.  If an LED shorts out, the resistor will burn out in a minute or two.  Better that than have an alkaline battery explode.  (Had that happen once - it blew the centre anode post across the room and embedded it in an oak door.

Jim 

[Len]

One thing to keep in mind when using multiple LEDs in a parallel or series combination - make sure they all have the same specs! Electricity takes the path of least resistance (no pun intended), so if the specs are different you get one lit LED and a bunch of dim/dark ones.

If you want to figure it out the old fashion way, the resister formulas for a single LED, and multiple LEDs in parallel are:

LED current limiting resistor formulas:

Note: If the calculated value is not available as a standard resister, use the next

highest value.

Single LED and DC power supply:

    V t – V led
R = -----------
      I led

Where:

V t = Total Power Supply Voltage
V led = Operating Voltage of LED
I led = Operating Current of LED in amps

Example 1: 12VDC power supply and a 2-volt 20mA LED gives:

R = (12 - 2)/0.02 = 500 ohms
--------------------------------------------------
Multiple LEDs in Parallel and DC power supply:

Voltage remains constant, but the current requirement goes up by the number of LEDs

used. The LEDs must have the same specifications and the formula becomes:

    V t – V led
R = -----------
    I led x N led

Where:

V t = Total Power Supply Voltage
V led = Operating Voltage of LED
I led = Operating Current of LED in amps
N led = Number of LEDs

Example 1: 12VDC power supply and 4 2-volt 20mA LED gives:

R = (12 - 2)/(0.02 x 4) = 125 ohms
------------------------------------------

If your lazy, like me, use the resistor calculators at:

http://www.quickar.com/noqbestledcalc.htm

It even tells you what the next highest standard resister value is when the calculated answer is a 'between' value.

Len