Large Scale Central

White LED

I am in unchartered territory, for me, here.

I would like to put a white LED in a radio conrolled engine that is powered by 14.4 volts run through an RCS system. Wondering if I can get that at my local Radio Shack.

I’m sure I can bug Dave or Tony about this, but was wondering if someone else has that info.

Ric,
E-mail me your address and I’ll send you the LED and resistor.
Ralph

Cool, thanks Ralph. Email on the way.

You could have picked it up at RS, but would probably paid 4 times what it’s worth and no one there would have been able to help you pick it out or answer your questions. White LED’s have really gotten very cheap as of late. I don’t bother with the math, just use a 220 ohm resistor in one leg which usually works for me.

A 220 ohm resistor would cause 57ma to flow through the LED when runing with a fully charged 14.4 volt battery. The LED’s are rated at 20ma max.

Thanks Bill -

Didn’t say it was right, only that it works for me most of the time. Difference being I have none wired directly to a 14.4 battery. On track power locos, which is where all of my white LEDs are, my power supply can only get a little over 13 volts and I never run full throttle.

That is probably OK for your track power.

Yep, but if he was running SG, and 20v on the rails, it would have been real bright for a real short time. 20 / 220 = 90 milliamps, usually enough to destroy any LED.

The calculation is simple take voltage, divide by .020 (20 thousandths of an amp, or 20 milliamps) and you get your resistor value…

There’s also lots of online calculators for LEDs that take more things into consideration, but the above formula is conservative and works fine.

Regards, Greg

Greg Elmassian said:
Yep, but if he was running SG, and 20v on the rails, it would have been real bright for a real short time. 20 / 220 = 90 milliamps, usually enough to destroy any LED.

The calculation is simple take voltage, divide by .020 (20 thousandths of an amp, or 20 milliamps) and you get your resistor value…

There’s also lots of online calculators for LEDs that take more things into consideration, but the above formula is conservative and works fine.

Regards, Greg


I do not use fixed resistor values anymore on variable voltages or DCC. Way to risky of blowing your LED.

Now I use a current limiting circuit, only 2 parts needed and works up to 36 volts.

Good idea. I’ve used a circuit with a transistor and resistor to make a current source. Can’t remember it now and wikki isn’t being much help.

Tom Ruby said:
Good idea. I've used a circuit with a transistor and resistor to make a current source. Can't remember it now and wikki isn't being much help.
Here is what I used. The LM-317 also comes in a TO-92 package rated at 100 ma. Just the right size for powering a LED or series string around 20 milliamps

Use a 68 ohm resistor, 1/4 watt is fine.

http://users.telenet.be/davshomepage/current-source.htm

Dennis, what do you do to drop the input voltage of the regulator down so you do not overdrive it?

Regards, Greg

Greg Elmassian said:
Dennis, what do you do to drop the input voltage of the regulator down so you do not overdrive it?

Regards, Greg


The input to the regulator will take up to 36 volts. Add a rectifier to the input of the regulator and this will allow the light to come on only in one direction.

The current regulator circuit does not need a voltage drop circuit. It limits the current to the LED, and you will see the actual forward voltage of the LED ie 3.5 volts. The remaining voltage is not used. The regulator will not get hot like a resistor might get. That’s because the regulator voltage drop is about 1.2 volts and you are only allowing 20 ma of current so the actual wattage is only .024 watts.

The current regulator circuit output if accidentally is shorted across will not burn out the resistor, regulator or LED’s and will not get hot either. This is why I like this circuit over using a resistor only or a voltage regulator circuit, both will get hot and may do damage.

I have used this circuit on DC & DCC lighted USA passenger cars running 12-14 hours a day without any problems. The circuit has a couple more components to make sure the LED’s do not flicker if you have dirty track or wheels.

If you are interested you can take the USA passenger car lamps out and replace them with LED’s using this same circuit, the current goes from 3/4 Amp per car to only 1/10 amp per car. Take a consist of 6 cars (6 X .75A = 4.5 Amps) just to light the interiors of the cars to (6 * .1A = .6 Amps) That is a big savings in power, over 7 to 1 savings.

Dennis, something is wrong here… say you have a led and the current regulator connected across 20 volts DC… say the led is got 20 ma, and the forward voltage drop is 3.5 volts… then the regulator circuit MUST be dropping the rest of the voltage… MUST…

Even the link provided says you may have to limit the input voltage to avoid overheating the regulator…

so 16.5 volts is dropped across the regulator, and at 20ma, that is 1/3 of a watt, so you will probably have to begin to think about heat sinking.

If you use a high current led, like 40 ma, you HAVE to heat sink the regulator.

I know how much less the current draw is, I tripped a breaker running my USAT passenger train with 3 E8’s.

The point here is that the regulator WILL drop the REST of the input voltage, and it’s easy to get close to the power rating of the unit and require heat sinking.

Regards, Greg

Just a comment … My G-Scale Graphics Enhanced Critter Controls and Enhanced RailBoss controls have “LED drivers” built in. i.e. No resistor required. It provides a constant current source of 11ma, bright enough for most LEDs. If not, you can still have directional control using the battery voltage and a resistor of your choice. I also have warm white LEDs available.

Greg Elmassian said:
Dennis, something is wrong here... say you have a led and the current regulator connected across 20 volts DC.... say the led is got 20 ma, and the forward voltage drop is 3.5 volts... then the regulator circuit MUST be dropping the rest of the voltage.... MUST....

No: it blocks the unused voltage. you are thinking this circuit acts like a voltage regulator, it is different.

Even the link provided says you may have to limit the input voltage to avoid overheating the regulator…

Yes if you are using this circuit on high current regulation then it becomes a factor, but for around 20 ma it is not an issue.

so 16.5 volts is dropped across the regulator, and at 20ma, that is 1/3 of a watt, so you will probably have to begin to think about heat sinking.

Not true: Your actual power dissipated is 20 ma across the regulator (1.2V @ .02 = .024 watts) the resistor is 68 Ohm and the wattage is (.02 @ 68 ohms = .027 watts) Its a different story in the Voltage Regulator design, yes I would see the remaining voltage across the Regulator and the wattage would go up and need a heatsink.

You got to try this circuit to believe how well it works.

If you use a high current led, like 40 ma, you HAVE to heat sink the regulator.

Even at 40ma a heatsink would not be needed but that’s the maximum limit for most LED’s we would use and really not needed in normal operation.

I know how much less the current draw is, I tripped a breaker running my USAT passenger train with 3 E8’s.

What made me use this circuit was a train display with a USA passenger Train with F7’s A & B units and 6 Passenger Cars, total current draw on a Bridgewerks Power supply was over 7.5 Amps, burnt up a couple of power supplies as the internal circutry of the Bridgewerks got so hot the small cooling fans in the Bridgewerks could not keep it cool.
Next year I was challenged to have 2 Passenger sets on 1 track. I had to reduce the power so this circuit did it just fine. Also converted from DC operation to DCC. Used one NCE Pro 10 to run both trains and did not have one power failure.

See for yourself a short video: http://www.vimeo.com/609545

You can also take a look at my other videos on what you can really do with LED’s.

The point here is that the regulator WILL drop the REST of the input voltage, and it’s easy to get close to the power rating of the unit and require heat sinking.

Again you are thinking this a voltage regulator circuit and your statement would be true, A voltage regulator can pull a lot of current if the output is shorted until it gets hot and destroys itself. Connect a current regulator circuit and you can keep a short across the output all day long and not harm the components. Now remember i am talking low current applications here. I have tested this circuit using a 15 amp Bridgewerks Power Supply that goes to 36 Volts. Connected the circuit and ONE LED to the out put. Left it there all day long and the LM-317 T0-92 did not get hot to the touch.

Regards, Greg

hahahahahaha

It is impossible for it to “block” voltage…

Dennis, basic electronics: if you have a closed circuit, the voltage drops of the items in the circuit MUST total the voltage drop of the power source. ABSOLUTELY.

Hook up your circuit to a constant 18 volts DC, and if the voltage drop on the LED is 3.5 volts, I will give you my house and cars and my income for the rest of my life if the voltage drop across the “regulator part”, i.e. the part that is not the LED is NOT 14.5 volts.

Sorry, you cannot change the change the laws of physics and electronics yourself.

Regards, Greg

OK, so I gather no takers…

basic laws, take the voltage drop across a device, multiply it by the DC amps, you have watts. This power has to be dissipated somewhere and unless your device glows like a light bulb, creates radio waves, magentism, or nuclear radiation, it’s all in heat.

It’s that simple, just like the sum of the voltage drops in a complete circuit must equal the voltage source.

That the law of conservation of energy. The universe and this world would fall apart without it.

So just be sure to check how much power you are dissipating in your regulators, a 1 amp regulator needs heat sinking way before it gets to 1 amp, and the spec sheet on the regulator will even tell you about how much current it will handle at what temperature.

With no heat sink, it’s surprising how little heat (current) the TO-220 regulator package can handle. Also, remember there is very little cooling air flowing inside most locos.

Regards, Greg

Dennis,
Great Christmas videos!!! I spent quite a bit of time watching them. Oh yeah, the train video was good too!!!

Wanted to change my burned out classification lights with 3v 50mA LEDs. The schematic shows they run from a 5v regulator. Since 5 resistors from RS were about the same price of one of my little AVR processors, I programmed one to give me about a 60% duty cycle. Just for grins, I made it ramp up the brightness when power is applied. Fiddled with it on my desk. Pretty cool. Just what I had in mind. Took Mr. Mallet apart. The bulbs are wired to a little PC board with red and black wires all over the place. Thinking to check the assumptions, I used my little meter to verify 5v, red is + and black is -. So I clipped out the burned out bulbs, wired in my little gizmo and leds. Turned on the lights. Hmm. No light. Check wiring. Hmm, why is it running on -1.8v. Yikes! The chip is HOT. That’s not from the smoke unit. Sure enough. THAT black wire was +5 and that red was 0. There are now resistors soldered to the board running my nice new LED classification lights.

(http://www.outsidetrains.com/mls/090823d.jpg)