Converting interior lights to LED

[Lou Schneider]

I think you need to check your math, John.

The LED itself drops around 0.2 volts when glowing,

Most white LEDs have a forward voltage drop of around 3 to 3.5 volts.

So you may consider the entire 13.6 to 16 volts as dropped across the resistor

Voltage dropped across a resistor is wasted power - it dissapates as heat, not light.

Using a theoretical 0.2 volt LED on a 12 volt source, 1.6% of the power will go to the LED.  The remaining 97.6% of the power will be wasted heating up the series resistor.

Using a more practical 3 volt LED, 25% of the available power goes to the LED.  75% disappears as heat.

To make a reasonably efficient light, you want as little voltage (and power) lost across the resistor as is possible.  The way to do this is to stack LEDs in series until their total voltage drop approaches the source voltage, then insert enough series resistance to create the proper amount of current through the LEDs.

The problem with this approach is since there is relatively little voltage across the series resistor, a small change in source voltage will cause a dramatic change in current through the LEDs. Which is why you need a constant current source (essentially a self-varying resistance) instead of a fixed resistor to limit the current.

 

[John From Detroit]

I agree with you Lou, however that was NOT the issue I was addressing.

The higher the resistance the less the effect of voltage variations by the way,, Which is why I went to one extreme

A constant current source is the best choice, and they are easy to build, I have made a few for battery charging.. I have one battery charger designed for AA nicads, it can do one, it can do, up to around 15 in series, it really does not care (I would have to do a lot more math to get the exact max number) it's a constant current charger and will adjust itself to the number of nicads in line

Alas, the device it was designed to charge, I don't have batteries in any more... (I got tired of replacing them every half decade or so)

 

[Gary RV_Wizard]

The person I talked to at LEDLights wouldn't indicate what would happen if the LED was exposed to 14.3-4v for any (or what) period of time.

I'm not surprised, Bernie.  They probably don't know and wouldn't accept the liability of telling if they did. LEDLights is a distributor, not a techie cobbling stuff up at his workbench.

I think Karl is on the right track (as usual). We are talking about a commercially bundled LED "bulb" here, one that is designed to function in a common 12V power environment. If that environment is a computer or electronic device, the range is probably right around 12.0 and can be fairly narrow, but if it is designed for an automotive/marine environment that range had better be quite a bit broader and be centered somewhere upwards of 12.6 vdc and tolerate at least 13.8 vdc (14.4. would be a better safe high limit).

Maybe the best way to find out is to buy one of their automotive replacement bulbs and experiment with it.

I have some G3 halogens that I would love to replace with something cooler and less power hungry, but I haven't seen any LEDs that have G3-equivalent bases. Actually I don't really care if they are LEDs - just something less hot and not too bad on the power demand.

 

[Jeff]

I have been living with a problem involving LED's for 8 months (gave up for a while waiting for a brainstorm 8) 8) ). The light switch by the door to provide entry illumination turns on a ceiling-mounted florescent tube just behind the driver/copilot seats AND a string of LEDs around the ceiling's edge.

Last spring the Zener Diode that is installed just downstream of the switch failed and melted everything around it. I was able to find a couple of the Zener diodes at at a electronics supply house in Washington but when I turn on the switch everything seems to work OK but the new diode starts getting hot as well. I do not recall seeing any resistors.

I guess I need to figure out how to disable the LEDs and use the florescent fixture.

I just HATE having something that doesn't work! ??? ???

 

[Karl]

Jeff,
Zeners typically will not handle large amounts of current and that's why a series resistor is required. Knowing the input voltage range,  output load(current), and desired output voltage, we can determine the size of the resistor required. To my knowledge, simple zener shunt voltage regulators will always have a series current limiting resistor. Maybe you just didn't locate yours yet. While 5 watt diodes are available, higher currents are usually handled by adding a transistor to bleed off the excess voltage. My recommendation would be to replace the zener with a simple voltage regulator such as the LM7812 mounted on a small heat sink. Three wires - voltage in (positive), voltage out (regulated positive), and ground (negative) which is common to both input and output. A fluorescent fixture will usually put a lot of noise back on the input power line, so you may want to add a couple of capacitors on the output to suppress it. I'd be happy to provide wiring diagram for you; it's really simple to do.

 

[Jeff]

Karl:

Thanks, that's great. We are leaving for Algodonas in the early AM so it will be a couple of days.

 

[John From Detroit]

Karl, you are right about the simple zener shunt regulators The diagram is always

Hot--Resistor--Zener--Load

Zener is hooked up between the resistor-load junction and ground

NOW: in this case it is

Hot-Resistor-Zener-Resistor-LED

With both Zener and LED going to ground

So it not only has a limited range... It is shunting power to ground

Thus the 3-terminal regulator is a much better way to go, You still loose some wattage to heat, but not nearly as much

Of course I know how to build a zener/transistor regulator that works like a 3-terminal if I want,  But that's way too much labor

(I did build one that is rated 3 amps, breakered 3 amps, and has delivered 10 amps with a slight bypass of the breaker)