Clunk when taking up the drive and when changing gear

[striplar]

I've got a fault on my 451 which sounds like a clutch actuator issue but I can't see what the problem is. There are no codes showing on the ODB scanner.

It's most noticeable when going from a stopped engine in the stop/start situation. As soon as the engine starts and the clutch goes to take up the load there's a thump. It sounds like the clutch is being very aggressively operated initially although it then picks up the drive smoothly after that.

I've taken the actuator apart and cleaned and re-greased it with Lithium spray grease and checked the gear on the motor shaft. Everything seems fine there but the problem persists.

I've moved the actuator through its full travel on the bench and it seems ok. Is there a damper in the system or is it just a strong counterbalance spring in the unit?

The fact that it can still smoothly take up the drive once the clunk has happened makes me think that the encoder on the actuator motor is working as it should. I imagine the ECU would throw a fault code anyway if that wasn't functioning.

Has anyone had this before? I'm really stumped to think what's wrong.

Any help would be very much appreciated.
Roger

[10:45PM]

Might be the clutch release arm in the bellhousing that's binding...? or the clutch binding on the splines.?

The 451 also has a problem with reverse selection on some early production from what I remember ?... ever had that ? just won't select?

[striplar]

Thanks for that. I'm not sure but I've managed to break the motor gear in an act of supreme stupidity so I need to get another one anyway. At least I now know how it all works and I think I may have a handle on what's been wrong with it. There's a pair of rollers in the actuating arm that are forced apart by a wedge so that they lock in position against the walls. It's these that transmit the force to the clutch from the actuator. My guess is that these weren't gripping the wall of the actuator so that nothing happened until the mechanism was out of travel. That would account for the clunk as the actuator took up all of the slack before then working correctly. Unfortunately I can't prove that because it's broken... Doh!
That scenario fits the symptoms and makes me wonder about the wisdom of squirting grease into that part of the mechanism. The idea of this device is clearly to adjust for wear and it mustn't slip else it's going to do what my car does.
Anyway, I'll get another one and see if the problem is resolved.

As an aside, I'd tested the motor and the integrated encoder and that was working fine until I broke the shaft.

Testing the Clutch Actuator.

Pin	Wire colour in photo	Function
1	Black	0V
2	Green	Channel A
3	Blue	Motor 1
4	Yellow	Channel B
5	Red	+5V
6	Brown	Motor 2

Testing the motor and mechanics...

• Service the mechanics of the actuator as described elsewhere
• Make up 6 leads with small receptacles that will fit over the flat pins of the connector. Don't be tempted to try to clip wires to the pins, you can destroy the sensors if you inadvertently allow 12v to touch any of the wires
• Initially, just connect two wires to 3 and 6 so we can test the motor.
• If the plunger is retracted, momentarily touch the wire from pin 6 to the -ve of the car battery, and the wire from pin 3 to the +ve of the battery. The rod should rapidly extend.
• if the plunger is extended, momentarily touch the wire from pin 3 to the -ve of the car battery, and the wire from pin 6 to the +ve of the battery. The rod should rapidly retract. Leave it in this position because it's easier to refit to the car this way. Don't worry if it won't stay in the retracted position on its own, it's likely to bounce back to the extended position. You can prevent that happening by pushing against the plunger when it retracts. That should be enough to make it stay put.
  Note:- This procedure is identical for the 450 and 451 but I found that the 451 wouldn't return the plunger to the released position. This may be because of damage to the gear but I'm not sure.

Testing the encoder outputs...

• Remove the motor from the actuator and connect up all of the 6 wires
• Use a 5v power supply (not more) or use a 4.5v battery to create the circuit below using three resistors in series with the leads.
• The 100 ohm resistor isn't strictly necessary, I used it to protect the device in case I wired it up incorrectly. I suggest you do the same.
• The two 10,000 ohm resistors are pullups. The encoder outputs can only sink current, so left unconnected they never reach 5v
• Turn on the 5v power supply and measure the two output channels while slowly turning the motor shaft. They should both snap sharply from 0v to 5v and back again creating a square wave output. There are seven complete cycles per revolution.
• Now connect the motor to a 12 volt battery. You can run this in either direction for a minute or so to make sure it sounds ok.
• If you have an oscilloscope you will see the outputs shown in the picture. Sadly this site won't allow the pictures to be included in the text.

Explanation....

The ECU presumably moves the motor fully one way and waits for the pulses on the encoder channels to stop. It then knows that this is the zero position. It can then count the pulses and know the exact position the arm is in at any moment from there.
I imagine that there is a speed sensor on either side of the clutch so that the ECU can tell when the clutch is not slipping. It can use that information each time the clutch locks up to make a note of the position of the actuator.
Presumably this is constantly updated by the ECU to accommodate for wear.
This is all guesswork to be honest, but I can't see how else it can work. From that I would deduce that as long as the car is driveable, the clutch with pretty quickly sort itself out, and find the exact bite point.
The reason for two channels on the encoder, is so that the ECU can count in both directions.

[striplar]

Here's what's inside the automatic adjuster in the actuator. You can see the wedge which is inserted into the 'T' slot in the part at the top RH part of the picture.
All of this can be pulled out of the actuator, the ball is just popped into the actuator. In my opinion this mechanism should be dry.

[10:45PM]

Ooooo , can I order a canvas wrap of it ? - only $18.95

I think you'll find this 'motor' not inexpensive and your face will when you see the price.

[tolsen]

Shocking £350 - £500 for a new clutch actuator.

[striplar]

Shocking indeed,, I've just ordered one for £349.99 delivered which is the cheapest I could find. I did ring a couple of dismantlers but the only one who had one said he'd call me back with a price, a sure sign he was just going to look up the price of a new one and knock a bit off. He came back with £375 for one with less than 500 miles on it which is frankly a joke so I told him I wasn't interested. He tried to ring back shortly afterwards but I can't stand greedy people so I'd rather have a new one than deal with them.
In the big scheme of things it's not a huge deal compared to what most people spend on garage repairs. These are a piece of cake to fit and it will soon learn the right bite point without having to get Mercedes involved. Such is life, I just hope it solves the problem.

[10:45PM]

Let's hope it cures the original clunk

[tolsen]

Carparts4less is cheapest. £335.94 inclusive of postage. You can get a 7.5% discount if you are clever enough. Place in your basket, log on but go no further. They'll then send you a discount code and you can buy the actuator for £310.74.

[striplar]

Ok, this has cured the problem but it throws up the question as to what was making the 'Klunk' noise which was pretty severe. I know that the encoder was working properly because I tested it and could see the square wave outputs on the oscilloscope. The motor also ran smoothly on a 12v battery so I think I can eliminate that.
Yes, I managed to end up with the end of the motor shaft shearing off, but looking at the damage to the gear teeth at the end of the quadrant, it seems entirely possible that the teeth may already have been damaged in that area as a result of wear. This would account for why some people have found the motor gear broken off because it wouldn't survive climbing up over damaged teeth if that was allowed to progress.
So whether it was the quadrant being damaged or the automatic wear adjuster slipping is hard to tell. Other than that, it's a very simple device and it's hard to see what else could be involved.
Anyway, it's working correctly now and doing the work myself softens the impact of the expensive part.

[10:45PM]

Well done you

[Stealth451]

Ok, this has cured the problem but it throws up the question as to what was making the 'Klunk' noise which was pretty severe. I know that the encoder was working properly because I tested it and could see the square wave outputs on the oscilloscope. The motor also ran smoothly on a 12v battery so I think I can eliminate that.
Yes, I managed to end up with the end of the motor shaft shearing off, but looking at the damage to the gear teeth at the end of the quadrant, it seems entirely possible that the teeth may already have been damaged in that area as a result of wear. This would account for why some people have found the motor gear broken off because it wouldn't survive climbing up over damaged teeth if that was allowed to progress.
So whether it was the quadrant being damaged or the automatic wear adjuster slipping is hard to tell. Other than that, it's a very simple device and it's hard to see what else could be involved.
Anyway, it's working correctly now and doing the work myself softens the impact of the expensive part.

Before my actuator spiral snapped, my car was also klunking between gear changes.
After opening it up and removing the motor, I could see a couple off the teeth had been worn away, but this wasn't noticeable from main opening, but could only be seen from where the motor sits.