Motors will increase current draw in weak circuits and burn out.
That's valid in some cases, but not a safe blanket statement. Field wound? Permanent magnet? Loaded? Loaded to what percent of capacity? Is the load fixed? Speed dependent? Inversely speed dependent?
Food for thought only, as it's pretty much irrelevant as far as light bulbs are concerned.
And for the record Jason... I'm picking on you ONLY because you've summed up what's been touched on several times, and short of holding a complete physics class, I'd like to expand on it just a bit.
Light bulbs will just burn dim if there is high resistance. However, lght bulbs usually fail at startup and if you have an intermittent ground you are increasing the number of "startups" and chance of failure.
This is valid, with conditions. The bulb is more likely to fail at startup but it has nothing to do with the electrical current being present or not. The current causes heat, and it is due to the sudden change in temperature of the filament that it fails most easily on startup.
In automotive applications, many conventional light bulbs today are duty cycle driven to control their intensity, and it does not affect their life span.
In a simple circuit (such as deuce and a halfs are noted for) a poor ground will result in less current flow, dividing the voltage drop, effectively reducing the voltage available for heat (light) production, and you get a dim bulb.
An intermittent ground (or intermittent B+ does the same), may or may not be at fault for bulb failures. If the failure is too rapid to see in the light from the bulb, will not affect the light's longevity, as the filament does not have a chance to cool. Essentially, if it remains hot (glowing) between "shots" of electricity, then the on/off cycle fatigue is eliminated. For an intermittent electrical connection to affect a conventional light bulb's longevity, it need not have the bulb completely off and cool, but it absolutely will be a clearly, easily visible condition, and you will see that easily in the form of a flickering light bulb.
If you drive a car with a duty cycle controlled light anywhere, you probably dont' even know it. Here's another example- A conventional household light bulb screwed into a conventional socket in your home. Those are a DC device essentially no different than an automotive bulb, except that the standardized bases are different. They only work when current is actually flowing through them. If you've got 60 cycle AC power in your home (and in the US most of us do...) then that light bulb flickers on and off one hundred and twenty times each second that it is turned on, all the while having it's polarity reversed sixty times each second that it is turned on. Does not bother it a bit, because the filament never cools.
Bottom line- If you don't see a flicker, then the ONLY electrical issue that's gonna cause early bulb failure is excessive voltage.
