LED problem - what's wrong?
 

I installed a 12V LED in the dash today to indicate whether my DRL's are on, but it burned out after about a minute. The LED had a 10k resistor soldered in series on the + lead, other than that it was just hooked up to the car's electrical system (12.4V engine off, 13.8V engine on). The resistor was to make it dimmer. After a minute of testing with the engine on it suddenly burned out. Earlier it worked just fine on the workbench with a DC supply of 12V-15V.

Any ideas as to what could have happened? Could it be fluctuations in the car's electrical system (I wasn't revving the engine)? Any suggestions as to how to wire the next LED, so it lasts a little longer? Remember: I'm no good with electronics, so be nice when explaining.

Thanks:)

LED's are cheap - try another one in the same configuration, and if it blows, you know it's a wiring problem.

Also, I'm not sure, but is the resistor supposed to be on the anode leg, or the cathode leg?

If you soldered it to the longer leg of the LED, that's Anode, and it's the input side... I was under the impression that the resistor should be on the Cathode side, but I don't know why.

EDIT: Doesn't matter what side it's on, so ignore that question. Are you certain your LED's are rated for 12V? You might try limiting the current to them even more than the 10k resistor...

Check out this led calculator: LED series parallel array wizard
You should check your led specs for voltages/current, typical voltage is 3.3V, and current is 10-20mA
Seems that 10K resistor is resistive enough - more resistance, dimmer the LED. So minimum 1.2K

Source voltage 14
diode forward voltage 3.3
diode forward current (mA) 10
number of LEDs in your array 1

Solution 0: 1 x 1 array uses 1 LEDs exactly
+----|>|---/\/\/----+ R = 1200 ohms

The wizard says: In solution 0:
each 1200 ohm resistor dissipates 120 mW
the wizard thinks ¼W resistors are fine for your application
together, all resistors dissipate 120 mW
together, the diodes dissipate 33 mW
total power dissipated by the array is 153 mW
the array draws current of 10 mA from the source.

Before you replace the diode, maybe you should drive around a bit with a volt/ohm meter attached to the leads to double check what the leads are getting.

+1 on thatguitarguy's suggestion, because just like in carpentry, in electronics it's "...measure twice, cut once..."

THREE times I cut the board -- and STILL it is too short!

If there's an inductive load, it might be causing voltage spikes. Connect a diode in inverse parallel with the LED after the resistor.

That calculator says my 12V forward voltage is suspiciously high, but I swear it was bought as a 12V, and it worked OK with 12V on the test bench.
I found Ohm's law somewhere in my highschool notes:
R=(Vsup-Vfor)/I
So if I take Vsup=15V (just to be safe), Vfor=12V and I=2mA (very liberal) then R=600. Gee, a 10k resistor seems like overkill, but I could still handle 15-20k, since I don't mind if the led is dim. I guess I could just use a 3.3V led, since they seem to be standard. It's the max amperage I'm concerned with (Mein Ammeter ist kaput), I'll have to ask in the store.

You might be right. From what I remember, the led burned out right after I pushed the button to close my grille block - the closing mechanism is a power lock motor. I wonder if that motor could effect anything else in the car?

I'll test the voltage this afternoon, maybe I'll see if the meter registers voltage spikes when fiddling with the grille block.

EDIT: Checked and don't see any surges. The reading is 13.95V with engine on, and it drops to 13.4-13.6 when the grille block motor button is pressed (<half second). Other than that - nothing. Just after starting the engine I may get 14.05, but I guess that's normal.