Headlight switch led
R = V/I (where R is the resistance you need and V is the difference in voltage of 12 and LED nominal voltage and I is current that the LED uses)
Lets take a white 3mm LED from superled's.
it has a nominal voltage of 3.4V.
it has a continuous current draw of 25mA.
so, the resistor we need can be any tiny resistor, since the current wont cause much heat (1/4 watt should work fine)
Now the value it must have is defined by (12-3.4)/ .025
or 344 Ohms
It's normal to put it on the positive side,but it really doesn't matter.
EDIT: Also, give yourself room for charging voltage (driving voltage is usually around 14 or so) if you use nominal values for the LED, you should be fine. Dont use max specs when determining the resistor unless you're using max voltage (maybe 15 would be safe there). or you could damage the LED.
Last edited by safemode; Mar 9, 2011 at 01:21 PM.
As for when you need one vs. when you don't, it depends on the characteristics of the LED itself. The reason why I'm guessing you burned out the LED you tried to use is because it was rated for 3.3V, and you applied 12V - that's why you need the resistor for that one. The bulbs you ordered as replacements for your gauge cluster were already rated for 12V input voltage, meaning they were already configured to accept that without the need for additional resistors (this is typical for LEDs that are drop-in replacements for standard incandescent bulbs).
Trending Topics
The Best V8 Stories One Small Block at Time
Here are some of my pics from the bulb install.



Here you can see the grey bulb installed. It took maybe 6-7 minutes to tear it all apart, install the bulb and rebuild it. Tuns out it doesn't need to be torn apart, lol. It's a 1 minute job.... just pop the bulb out, insert the LED and DONE!
I've always wondered why the TA's had to do more work to replace theirs as well.
The problem in this case is you are working around a design that was done when incandescent lighting was the only option, hence the waveguides being the way they are.








