I'm not asking for advice or suggestions, but merely making an observation that might affect everyone that drives a car.

I replaced my dome light a few days ago, from a halogen bulb to an LED based
light into the same receptacle. It works nice and has nice bright white light and runs even cooler than the original light!

Mose dome light have three positions. Off | Door Activated | On

I noticed that while driving around at night that when the switch position is in the "Door activated" position that there is a fairly faint glow of the LED's. If I switch to off, the glow goes out. IF I switch back, the glow returns. This just makes me wonder how many other lights (Light above the mirror, glove compartment, trunk...etc. could be draining the same small current. It's got to be small as the LED's are very dim, but not sure what sort of current is required to get a 20mA - 25mA SMT LED to begin to glow.....3mA? 5mA, or more.

 

I would guess at 0.1 mA or less per LED chain. (I imagine that there will be one or more chains of maybe 3 or 4 LEDs in series)

Even 1mA would be obviously on. Not bright, but way above "very dim"

 

If the LEDs are of good quality the light output may be approximately proportional to drive current down to much lower revels than the normal rating. That, coupled with the wide dynamic range of human vision means that not much current may be needed to see a glimmer of light.

It could however be a good idea to check up on what the level of leakage current actually is, and where it is coming from. One possibility that occurs to me is dampness getting into the switch or wiring, which if left untreated could lead to corrosion.

 

I will resist the temptation to make any parallel comments on how the glimmer might be suppressed (other than eliminating the leakage altogether), lest this be taken to be a suggestion for an automotive modification.

 

My guess is that it is designed to be like that. Incandescent bulbs that are turned on and off frequently last longer if they are kept warm when they should be off. The console lamps on DEC PDP-15 and DG Nova mini-computers were designed like that.

 

Fine if so, but a waterlogged door switch is also a possibility, in my opinion. I have had such problems myself, although leading to intermittent operation rather than leakage. But then again, I live in a damp climate. Does it rain much in Vermont?

 

Some vehicles run current through the bulbs not only as a "keep alive" voltage (to keep them warm) but as a diagnostic; if a bulb is open, no current will flow through it, and the vehicles' computer will report an error in some fashion.

We used to have a Jaguar XJ-S that had light bulb failure idiot lights on the dash; the system constantly supplied current to all safety lighting via small modules that contained a heater and a bi-metal switch; if there was not sufficient current through the heater, the bi-metal switch would stay closed and the idiot light on the dash would stay on. These modules were a PITA to replace, because you had to remove the center console to get at them.

 

Does it rain much in Vermont?

I'm guessing, but I'd say second only to the state of WA. Yes it does indeed rain here and has been quit a bit recently.

 

I will measure the current that is flowing as an informative en-devour, but will end this conversation after I report back so as to stay shy of crossing over into "No-No Land!"

 

Dome light problems I don’t know about using LEDs but I had an interesting dome light problem back in the 70’s. I got into my Charger on night and shut the door the dome light went to ½ brightness. Not much later I was pulled over for driving without tail lights. The next day I replaced a bad fuse for the tail lights and the dome light worked. It was repeatable when I pulled the fuse. Never did understand the connection. Bottom line to this day I cannot stand car and truck wiring

 

Once upon a time, stepping on the brake pedal turned on the dome light. The wires were rubbed bare inside the frame of the car. (1965 mustang)

 

Some more observations:

Voltages:
Dome light on: 12.38V (Battery is not in good shape)
Set to Door: 6.09V (With Load)
Set to door LED removed: 7.74V (Open circuit)

Tried to measure voltage across a 2.5 ohm resistor in series with LED in the light switch set in the "Door" position. I could not read anything even on the mV scale.

 

auto electrical is complicated by the chasis ground. makes it hard to troubleshoot. a short somewhere can cause totally bizzare and unexpected things to happen. My dad had a 60s mustang that had all sorts of wierd stuff going on; the fuel pump would turn on and off intermittently (like a blinker) with the key off and the left turn signal on, but the blinker itself didn't blink. turn the key and the fuel pump stops pulsating and the blinker comes on. try to drive it then the right blinker doesnt work and neither does the horn, but the both work with the key off. try to run the wipers and nothing works. and all this is intermittent and not repeatable. turns out the problem was a few of the wires in the harness that runs behind the bumper got pinched and shorted out all together to eachother and to the frame by probably the lightest fender bender imaginable.

 

Maybe I should change the title from:

Light where there should be none to
no current where there should be some!

Why doesn't my meter pick this up?
I can take the 4 LED array out and test in a bench top setup with 6V - 7V and a variable resistor till I get the aprox. brightness. At that point I can calculate the aprox. current consumption.

 

I messed up an order once and got one of these
http://www.superbrightleds.com/cgi-...e&Page2Disp=/specs/festoon_3910-x6.htm#photos

Problem was it was in ultraviolet. With the truck door open at night the only thing you could really see was anything white colored, totally useless.

 

I think the current may simply be too small for you to get enough voltage across 2.5Ω, even on a millivolt scale. I have just tried a few medium-sized white LEDs, normal rating 70 to 100mA. They all gave a small amount of light at less than 100 micro-amps.
Your dome light probably has a higher normal current rating, but even if it is 1A, on this basis you might see a glimmer by 1mA. That would only raise 2.5mV across your 2.5Ω series resistor.

Smaller LEDs can glow with even less current, especially if they are good quality. I keep a little light on my bedside table so that I can get at my glass of water etc. without turning on a bright light. It uses a high-efficiency green LED with a pot to vary the brightness: at a current of about 3μA, the LED is still just visibly glowing in a darkened room.

 

It could be displaying HF noise spikes on the 12v supply (not DC current at all). That can be common with LEDs in vehicles.

An easy fix is to put a 0.1uF greencap and 100uF electro across the LEDs (a basic decoupling capacitor). If the dim glow stops you will know it was HF noise and not DC leakage current. That will also really increase LED life.

 

That's a good point: if there is HF noise about the LEDs could be seeing nasty short-term voltages, possibly including reverse voltage which is really bad for them.