Automatic transmission quandary.

Question is this:

When an automatic car is in gear but is stopped (like at a light), what happens in the automatic transmission to keep the engine from stalling?

I assume the auto clutch disengages and allows the engine to operate without putting load on the transmission, right?

But I also assume it doesn’t disengage entirely because shifting the car into N results in the engine operating more freely (which you can tell by the change in sound and sometimes slightly higher rpm).

Here is why I ask -

My wife’s car is a MB 2005 C230K Coupe. Very low mileage - only 18,000 or so. The car is otherwise pristine, except for the irritating subtle knock that seems to come from under the car towards the back, as if from the differential. It happens when the car is still cold and when you pull up and stop at a light. The knock is slight but we both can hear it and feel it. Shifting into N makes the noise vanish completely. Once the car is up to operating temp, the noise seems to vanish.

Since we can’t feel it while driving, I’m thinking NOT: engine, trans or diff mounts. Since we can feel it while stopped in D, I’m thinking NOT: drive shaft flex joints.

What’s left to consider? Transmission oil pump when cold? Low trans fluid? Clutch not disengaging right when still cold?

What else?

Dealership service team can’t identify the issue but I’ve not pressed past the initial consultation and inspection.

It used to have a simple answer but “automatic” transmission tech has gotten quite complex.

The basic replacement for a manual clutch is a torque converter: it has a rotating vane attached to the engine and a rotating vane attached to the input of the transmission, separated by a fraction of an inch and filled with heavy fluid. The engine spins the input vane and the force of the fluid transfers that to the output vane. When the engine is turning slowly (at idle, or below about 1500 RPM) the force isn’t enough to overcome inertia and light brakes will hold the car in place. If you clamp the brakes hard, the torque converter will work in reverse, limiting how fast the engine can turn against the static output vane. (That’s called the stall speed; it’s typically around 1500 RPM in street cars and can go up to 4000 RPM for drag racing, so that the engine can reach maximum power at launch.) These got quite sophisticated and “tight” in their coupling before newer tech came along to help.

Mercedes originated the next innovation (AFAIK), which was a torque converter with an internal clutch, that would “lock up” when slippage wasn’t desirable. During shifts and at a stop, the clutch would unlock like a manual one, allowing some slip; while driving it locked up the shafts, keeping slippage and eliminating power and fuel economy losses. I think most newer vehicles have some form of locking torque converter for those reasons.

Very new and high performance cars have something a great deal more like an automated clutch - the fluid slip coupling, if any, is a minimal part of having electric or hydraulic actuators work a hard clutch under computer control. And the highest-performances stuff of all (Lamborghini, Ferrari, high-end Porsche) use steering-wheel-mounted switches to allow the driver to manually shift using an automated clutch. Actually, the driver flips one of these paddles to “request” an up- or down-shift, and the car’s control system does the optimal set of moves to make it happen. It’s like having Michael Schumacher along just to do your shifting for you.

To answer your question, there are a dozen things that can cause a thump or knock when sitting at idle. Bad mounts, bad u-joints, bad half-shafts. At 18k, I’d bet on a broken mount rather than anything worn out.

I’m with you about the dozen possible things… but, if it’s a mount (engine, trans, diff), why would the knocking vanish when the car comes to operating temp?

The engine is coupled to an automatic transmission through a torque converter. When you shift from Park or Neutral into Drive, the transmission engages and disengages particular clutches to route power through specific planetary gearsets to form a path for mechanical power to get from the torque converter to the transmission’s output shaft. In Drive, at a dead stop, the engine is exerting a small amount of torque on the torque converter due to the speed mismatch (engine/impeller at ~600 RPM, turbine at 0 RPM); this is the torque that makes your car creep forward when you release the brake pedal. Step on the gas, engine RPM rises, and the torque communicated to the turbine increases quite a bit.

When you shift to Neutral, the transmission disengages the planetary gearsets from the torque converter. Now the turbine in the converter is free to spin up to some RPM very close to engine RPM, which means the small torque that the engine used to deliver is no longer being delivered; the engine RPM speeds up a little bit (or not much at all, if the ECU is smart enough to realize you’re in neutral and backs off of the electronically-controlled throttle accordingly).

You may want to inquire on a Mercedes-specific discussion board; there’s a fair chance someone else has seen your exact issue and knows the cause/solution.

Without any further info, I might guess it’s related to the exhaust system. Due to the torque that’s passing through the transmission when in Drive, the entire engine/transmission assembly rotates slightly in its elastic mounts (relative to where it would be in neutral). That, coupled with the slight change in engine RPM, may produce different resonant behavior in the long, floppy exhaust pipe. When you’re driving down the road, engine RPMs are different, so the condition disappears. When the driveline is warmed up, elasticities and dimensions change, so again, the condition disappears.

It may be difficult to safely diagnose, since it will require someone getting under the car to look around while someone else has it in Drive with their foot on the brake. The wheels would need to be very securely chocked to make this safe. I expect the dealer could do this, but it will take more time/$$ to do it.

ninja’d by Amateur Barbarian…

Obviously you need a Dvorak keyboard. :smiley:

My experience with automatic cars is rather limited, but it was my understanding that the lock-up clutches only activated once you hit top gear. Is that still the case?

Interesting… Thank you all.

The exhaust issue had occurred to me as well but it’s such a tough one to diagnose and I can understand why a tech would not spend hours investigating just to hand me a bill for $400 and tell me it’s a heat shield.

I’ve been on multiple MB boards and am a member of one but as good as they sometimes are, the signal to noise ratio is so high that even searches don’t reveal too much information specific to this issue.

The earliest ones - such as the original Ford AOD - did. The newer ones lock any time the controller detects no need to allow slippage. I can’t tell you who uses what or when they started it, but the progression has been from full-fluid coupling (which always slips a bit, losing power and efficiency) to “locked whenever possible” couplings - which includes locking for brief periods during acceleration. The ones that use a fluid coupling only for brief periods (to smooth power transfer and shifts) or no fluid at all (computer-controlled clutch) are used from middle-luxury and sports models all the way up to the maximum supercars. (I’m pretty sure many of the top supercars don’t even have a traditional manual option any more, because power and torque exceed manual clutch capacity and computer-controlled shifting is far more efficient… and tunable from street to race.)

That’s why, as in the recent threads about auto-v-manual, an auto is an equivalent or better choice than manual unless you really like shifting. Good autos have none of the downsides of the past - weight, slop, power and fuel economy losses - and once you get to maybe the $40k class, are often better than all but the most expert manual driver.

Ii drove automatic trucks for years. They have no fluid coupling at all, but a dry plate clutch exactly the same as a manual. Thee big difference is that the clutch operation and gear changes are all controlled by computer.

It takes a little while to learn the best way to drive them, but the benefits to the owners(economy) and to the drivers (easy driving)are so great that they are now standard in almost all EU trucks.

For the OP - Would it be worth just going into an exhaust place and asking them to check it out?

Foolish as it sounds, I had not considered that. :smack:

There can be two things with that car

  1. engine mount defective… a bubble in the rubber.

  2. weak timing chain tensioner when its cold . So the timing chain rattles when its cold.

The $40k class? My current car isn’t even in the $2k class. :smiley:

I did get to test drive a 2013 Kia Optima a while ago, which was an automatic with the option to choose your own gears, but I found that kind of clunky, considering I’m used to the H-pattern of a traditional manual gearbox, so whenever I tried to go from first to second, or third to fourth, I ended up pulling the lever down (which tells it to downshift, not upshift) out of habit. :smack:

Considering the old beaters I tend to buy, the ever-important question I have to ask is this: can the newer automatic cars be push-started in the event of a starter or battery failure? That has always been one of the biggest reasons I go out of my way to buy standard transmission cars; that ability alone has saved me a lot of money in tow bills.

To the OP a Mercedes has hydraulic motor and trans mounts. With a 2005 car I would not be surprised that the fluid has leaked out and the mounts have lost some of their vibration absorption ability.

Someone please correct me if I am wrong, but there hasn’t been a (US market) automobile manufactured in the past 50 years that, if the battery was completely dead (0 volts) that you could push start, even with a manual transmission. The alternator needs some voltage (actually current) or it won’t be able to provide the electricity for the spark plugs.

Prior to 1967 or so, Chrysler transmissions were designed so they could be started by getting them up to 25 or 30 in neutral and then putting them in drive. A bit fast to be pushing by hand, but it could be done. It wasn’t an intentional part of the design, however. That is, the transmission was not designed specifically to so you could push start it, but a result of how the pump was designed.

Finally, I don’t believe modern car (say, built in this millennium) with even a halfway dead battery can be push started, anymore, even with a manual transmission. That is, if the battery doesn’t have enough juice to turn the starter, it doesn’t have enough to run the computer, so it won’t start.

On the good side, the quality of the battery, starter motor, alternator, etc…, are so much more improved over those of 40 years ago that if you maintain the vehicle, there isn’t nearly the need to either push start it or have it towed.

I would go with the exhaust rattle as well, but I think the rough idle is really the issue. I would have the dealer address both issures.

With a completely flat battery, the main issue is that the alternator needs a certain amount of current to work (in contrast to the old DC generators). It doesn’t require much, though, so on older FI’ed cars, so long as there was even the tiniest bit of charge left in the battery, once you got rolling the alternator would provide enough power to run the ignition and fuel systems.

Also completely dead battery is a fairly rare situation. There’s lots of situations where the battery is less than completely dead and still can’t turn the engine over. Or you might just need to roll start because of a starter problem. Newer manual-equipped cars seem to be less forgiving of low battery voltages, so roll starting works in fewer situations, but I’m fairly sure it’ll still work with most of them if you’ve got a full battery charge (so, if you had a bad starter or something.)

Most automatic transmissions have an internal pump that is driven by the front fluid coupling or clutch setup that the engine is always turning. This provides all the hydraulic action in the system, including those that couple the engine to the drive wheels.

Without the engine turning, no coupling can take place, so you can’t “bump start” the car. Some older transmissions (with Fords this would be the old late 1950s-mid 1960s FMX transmission), there was also a rear pump which was driven by the driveshaft, so those would probably be powered enough by the cars movement to allow the coupling in the transmission to work and spin the engine over.

Also, stated earlier, the original Ford AOD did not have a clutch type torque converter. A separate shaft was coupled directly to the outer shell of the converter and went into the transmission. The lockup that coupled the engine directly to the rear wheels was provided by hydraulic couplings inside the transmission, and worked in 3rd gear and 4th (overdrive).

As to the OP, I would certainly check for an exhaust rattle. My 1983 T-Bird does this sometimes, and it seems quite random. Some combination of ambient and operating temperature and engine smoothness makes this condition an occasional nuisance.