-Definitely it had to be caused by a dirty fuse. The weird thing is that even with the oxidized fuse, the standard headlights worked. I've seen this sort of thing a couple of times on different 12V DC systems when dirty contacts allowed a high current device to work, but did not allow a lower current load to operate properly. In one case, there was a diode voltage drop in the load in combination with low wattage load, and in the other case, the load was very low wattage. In both cases the load simply wouldn't draw current despite the total voltage being high enough that it should have, and furthermore, the lower load should not have been affected nearly as much by the oxidized contacts in the supply! This is still a mystery to me, but my working theory is that sometimes corroded contacts behave like a leaky metal oxide diode rather than simply a resistance. If the inadvertent 'diode' tends to block current, a low wattage load in combination with a diode having sufficient forward voltage drop (such as LED lights), or a just very low wattage load, the 'diode' can block current. In this working theory, when faced with a large enough load (such as the higher wattage original headlights) the oxide diode avalanches (breaks down completely and acts solely as a resistance). This explains why the original headlamps worked. At least that's my theory! A resistor in parallel with the LEDs, acting as a parasitic load, might make the circuit less prone to this phenomenon. (200 ohms maybe?)
-The last time this weird thing happened was on my Morgan. It had a cheap set of ignition points and the contact between the points and the wire was through a blued section of the spring. The bluing did two things. It created resistance but also formed a metal oxide diode. This caused the CDI to hiccup when it ran fine in Kettering mode that puts 10 times the current through that part of the circuit. The Morgan is positive ground, but I believe had I converted the car to negative ground that the 'diode' would have been forward biased and everything would have worked perfectly. The fix was to file away some of the bluing to improve the contact area and add a little grease to keep corrosion from forming. Dielectric grease is the best for this, but almost any grease will do.
-This phenomenon flies in the face of ohm's law which is why I believe it only happens when dirty contacts form diode junctions that block voltage readily unless the load is large enough in combination with the applied voltage that the oxide junction breaks down. It's the only way I can explain the contradiction and it doesn't happen very often. Fred
Jim Alton wrote:Alan Pershing wrote:The bright lights are now working although the "problem/solution" is completely mystifying to me. Maybe one of you electrical guys can weigh in.
To recap: LED headlights originally worked in bright mode (brights on, blue light in tach on) right after installation.
Next night, moving the bright light stalk forward turned off the headlights (no light at all, no blue light)
Swapped LEDs for original Halogens and brights worked perfectly (bright light on, blue light in tach on)
Swapped original Halogens out for LEDs and same behavior restarted when turning on brights (headlights completely off, no blue light on tach).
Jim Franzen (who supplied the rear tail lights and has been very helpful troubleshooting the headlights) suggested I take a multimeter and make sure it detected current through fuse #12, which controls one of the bright lights and (I think) the blue light on the tach.
So, I opened the fuse box and took out #12 to make sure I had the right one; the fuse numbers are on the fuse box but hard to see unless you remove the fuse.
I verified that it was #12 and put it back in for the multimeter test. I turned on the headlights and then pushed the stalk forward.
Both bright lights came on and blue light activated on tach. File under WTF?!
I had never touched the fuse before and didn't clean the contacts when I took it out; I literally just popped it out to make sure there was a "12" on the fuse box and then put it back in. I don't know why it worked originally, why it stopped, why it was working for the Halogens but not the LEDs, don't know what magic happened when I popped it back in and don't know why all the LED headlights were going out when it wasn't working.
All I know is that they're working now and I've tested them extensively for the last two evenings and they keep working correctly.
Hypotheses are welcome....
Well, I have degrees in electrical engineering and--although I spent my career in software and [aerospace] systems engineering--I'll offer a hypothesis:
You
DID clean the fuse and contacts!
Just removing the fuse and putting it back could disturb corrosion or foreign matter. Putting it back in a different orientation (and you surely did) would also help.
It's odd that this affected both headlights--the wiring diagrams show fuse # 11 carries the current for the left high beam.