Hi Art,

Correct me where I'm wrong here.

Note in your schematic the 3rd brake light inductive coil on pin S doesn't have an opposing coil on the reed switch. If the brake lights aren't being activated by the brake switch (pin 54) then there's no current in that coil, the reed switch isn't tripped and the warning light is out, like it's supposed to be. Now if the 3rd brake bulb is burned out there's also no current through that coil, so what in your schematic now activates the warning light? I've always assumed there was a separate reed switch added for the 3rd brake light with an artifical current in an opposing coil that had a resistor on it rather than a bulb. My recent inspection suggested that's not the case, leastwise in the one I just checked which may be a slightly revised design from earlier red ones.

In the one I recently examined (I'm too lazy to go remove and crack it open for a pic), there was only one reed switch for the combined brake lights. Surrounding that reed switch is one full coil for one brake bulb and there's two partial opposing coils, a small one (for the 3rd brake light) inside a larger partial one (for the other brake light), both with fewer windings than the larger coil. You have to look closely to see that extra little inductance coil. When current is running through both of the two partial opposing coils it then matches the magnetic inductance of the larger coil to keep the reed switch from tripping and no warning light. If any of the three coils carries less current than normal (normal being all bulbs on or all bulbs off) then the reed switch trips and you get the warning light.

The LEDs used in 12V-14V automotive applications generally need a resistor, either an internal one or an external one added to the circuit. Many have some kind of internal voltage limiter to handle voltage spikes. CANBUS bulbs have added circuitry with internal resistors and more stable voltage/current limiters to work with bulb sensors, electronic controllers and dimmers. The Auxito 1157/1156 LED taillighting I'm using does not need an external resistor and was not listed as CANBUS compatible. The Auxito 9004 LED headlights I'm using also don't need an external resistor, but are listed as CANBUS compatible and having internal thermal sensing circuitry so as not to fry the LEDs and transistors with both current limiters and an internal variable speed fan for cooling. These 9004s are still advertised as possibly needing a CANBUS isolating adapter with non-CANBUS cars in case of bulbs blinking or bulb sensor problems.

I usually noticed these intermittent bulb sensor problems for a few minutes during warmup after starting up driving or during cool down soon after coming off freeways after long drives. Sunny warm days are worse than cloudy cool days. Hot summer days parked out in the sun were especially bad. Keeping the contacts, connections and grounds all spotless and making sure the solder connections in the sensor are all in top shape has helped keep these intermittent problems at bay (until one 5 minute episode yesterday, I've had no bulb sensor problems for many months).

I'm now working under the assumption that what's been intermittently upsetting my bulb sensors (two 940s) is subtle variations in the currents through these LEDs as the internal circuitry unevenly warms up and cools down. While compact fluorescents get brighter as they warm up, LEDs get dimmer as they heat up and try to limit the internal voltage and temperature. Proper LED headlight specs should give separate startup and running lumens, but most mfrs opt to advertise only the startup lumens. If the bulbs thermal response is the same left/right then everything is okay, but if it varies left/right such as during warm up or cool down then that difference may be enough to momentarily upset the bulb sensor. Different bulb sensor designs may be more sensitive than others. Using CANBUS isolators would likely eliminate my intermittent problems.

So, I ask, why is our bulb sensor design seemingly more sensitive with lower current LEDs? Why is it seemingly always the taillights, not the much higher current LED headlights? Is it unbalanced left/right current that's temperature related? Do reed switches trip based on an absolute difference in current (eg. 1 Amp) or do they trip based on a proportional difference (like 10% of the Amps). If it's proportional then that might explain why they become more sensitive as the current gets lower. Eg., a .9 amp difference in a 2 amp circuit would be much more significant and trigger a warning compared to a .9 amp difference in a 10 amp circuit. I'd have to really bolt down my thinking cap to start wrapping my brain around thoughts of the strength of opposing magnetic fields inside an inductance coil under different conditions as well as the magnetic response of these little reed switches under different conditions.
--
Dave -still with 940's, prev 740/240/140/120 You'd think I'd have learned by now