Originally Posted by bustedbenz

I'm going through the EXACT same thing on my 240D. On a cold start, the starter cranks the engine a few revolutions, and then disengages. The starter free wheels until I get off the switch, and then grinds to a halt. I let it sit a second, try again, etc.

I'm up to 5 and 6 attempts before it starts on most attempts.

My first thought was that it was, as previously stated, cold grease not allowing it to fully engage the first time. But then I got to thinking... wait a minute... it's engaging the very first time. It only kicks out after it's already engaged the first time.

I'd like you to do an experiment to see if your results match mine, if you are willing. Go out one cold morning, and turn the key to start -- WITHOUT any pre heating. No heater, no glow plugs, no nothing. Just flip the key straight to start -- and hold it there until whichever of these comes first: 1) You crank for 10 seconds and the engine doesn't even pretend to start, 2) The engine starts, or 3) The starter disengages as per your original post.

IF your car is like mine --- when you do it this way, the starter will stay engaged perfectly until the engine starts or at least tries to start.

Here's my (and my dad's) current theory. We think that as the engines get older, and the temperatures get colder, we all know it's a little harder to get them to fire. Our speculation is that the engine is "running away from the starter" -- or false starting. here's the scenario: It's a really cold day. The engine is 30 years old, and some cylinders are stronger than others. So you do a nice long pre-glow to be sure it's ready to start, and hit the starter. The starter engages, the engine starts to rotate. One or two cylinders out of the four (my case) or five (yours) -- fire the first or second time around, because they are the strongest. The engine STARTS to catch. For a very brief moment, the engine reaches the speed at which the starter "kicks out" -- and so it does kick out. BUT -- it's such a frigid day that the other 3 cylinders didn't light up, and so the engine dies instantly.

Or the short version: On a cold day, some cylinders fire instantly while attempting to crank it, and so the starter disengages or "kicks out" just like it's supposed to, but then due to the cold, the other cylinders weren't able to sustain the engine so it dies instantly while the starter is still winding down.

I'd be interested to see if you are able to get the same result --- when you start without waiting on a pre-glow, ALL the cylinders have a hard time lighting up, and so the starter goes a full 5 or 10 seconds (depending on your start ability) before disengaging -- if the engine isn't even attempting to fire, there's nothing to make the starter kick out, so it doesn't.

If you don't want to or can't afford to perform the test, then it's fine. But I'm very interested in the result if you do since I've got the same thing and I'm ignoring it for the moment. I considered the "Cold grease" theory but then it just didn't make sense why it would engage the first time and then pop back out, if the grease was preventing it from extending far enough in the first place.