Gary’s Garage June 2024

Convertible Top Issues – Again!

Different car this time – a 2002 CLK55 cabriolet (A208). A somewhat simpler arrangement than on an R129, with only 7 hydraulic cylinders compared to the 11 or 12 on the 129.  This car also has a bypass on the hydraulic pump that you are supposed to open to manually operate the trunk. Thankfully I have not had too many issues with this top, but the latest one gave me a learning point I thought would be worth sharing.

But first, I’ll mention the first problem I had with the top.  Shortly after I bought the car, I started getting a top alarm when the top was up. This is intended to tell you that the top isn’t properly locked in place, which just might be an issue while driving..…. I was able to determine the fault only by hooking up a Mercedes specific scanner (a Foxwell NT530 with Mercedes software installed).  Since the fault only occurred when the car was moving, I had to drive the car with the scanner hooked up to see the fault.

On the 208 chassis, the OBDII connector only provides engine information.  All other information is available only on the 38 pin connector in the engine compartment. Fortunately, my NT530 setup had long enough cables so I could hook it up, shut the hood, and lead the cable through the window.  While driving, this allowed me to immediately see the intermittent fault when it showed up - the fault was with the bow lock.

A note on terminology – the rear bow is the absolute rear portion of the top.  It is lifted by hydraulic cylinders to allow the top storage compartment cover clearance to open and shut.  The rear bow has a single prong that engages a lock centerline at the back of the cover.  The lock has a hydraulic cylinder that closes a latch which goes through a hole in the prong, pulling down and securing the rear bow.

The bow lock cylinder is accessed by opening the cover and removing the upholstery pad under it.  This cylinder is then easily accessible. What I found was a little shocking – the wires to the microswitch that reported the latch position had been soldered to the microswitch due to previous damage. The soldering was poor, and the electrical contact was intermittent. This was easy to fix.

Flash forward to 2024, we were at a driver education event at Barber Motorsports Park and Cheryl was driving the car.  She came in from one session and reported that the front of the top wasn’t in its normal position.  We tried to cycle the top but it stopped mid-cycle.  This presented a problem, not only with completing the sessions on the track, but we also had to get the car home!

I attempted the procedure to raise the top manually, which I had never done on an A208.  The owner’s manual provides detailed directions, but if you’ve never done it before it’s not obvious where the components are.  I was finally able to work through the manual raising until I got to locking the rear bow.  You are supposed to be able to put an Allen wrench in the bow lock and turn it to lock the bow down – BUT IT WOULDN’T MOVE!  This meant that the car was safe to drive home since the front was latched and it wouldn’t get rain in, but the rear bow was not secure, and I couldn’t send it out on the track like that.

There was nothing I could do to operate the bow latch, even with the hydraulic lines disconnected, so I ultimately removed the latch from the car. Even then, I couldn’t lock it manually.  I studied the linkage for some time and finally found that if I slightly raised one of the pawls, I could manually latch it. I fashioned a plastic piece to hold this pawl slightly elevated and reinstalled it into the car.  I was then able to fully lock down the top and the car was able to return to the track.

I ordered a replacement hydraulic cylinder with the latch assembly from Top Hydraulics.  Before installing it, I noticed it had the same issue with manual operation and I installed it with the little plastic piece that held the pawl slightly elevated.  After I installed it in the car the top wouldn’t complete the cycle. It would raise the rear bow, stop, and provide a fault that essentially said “illogical latch position”.  I hooked up my STAR clone system and watched the live values of microswitch position during top operation.  I noticed that the rear bow latch was closing when it should have remained open (it should be open until the point in top raising where the rear bow extends, and the long prong on the back of the rear bow goes into the lock). 

I removed the rear bow lock again and did some more analysis.  I finally figured out what was going on – this rear bow lock has an open latch that holds it open until the long prong on the rear bow goes into the latch and hits the moving part of the latch.  Slight downwards motion of this part of the latch releases the open lock, and the hydraulic cylinder immediately closes the latch. Closing hydraulics are supplied to this cylinder very early in the cycle, way before the latch needs to shut.  The open lock prevents the closing hydraulics from shutting the latch, until the rear bow is extended, and its prong contacts the latch.  My little piece of plastic was releasing the open lock, causing the rear bow latch to shut without the bow being extended, resulting in an illogical latch position fault.  All I had to do was remove my little piece of plastic, and it worked.

A note of explanation for those that didn’t grow up in an environment of electrically controlled hydraulic systems, to explain more fully what I’m referring to in the previous paragraph. In these types of systems (the R129, A124 and A208 are examples), an electronic control system sends a signal to a solenoid control valve which ports pressure to the correct side of a hydraulic cylinder causing it to operate in the desired direction. Proper sequencing of the electrical signals to the solenoid control valves causes the top to operate in the desired fashion. One would expect that these signals would be sequenced in the intended order of operation of the hydraulic cylinders.  The thing that I found odd here is that the shutting hydraulics of the bow lock cylinder are pressurized in the sequence well before the cylinder needs to shut the bow latch.  The only reason that the bow latch doesn’t shut prematurely is due to this open lock, which is released when the prong of the rear bow pushes down on the latch – this unlocking allows the latch to immediately shut by hydraulic pressure. Perhaps Mercedes did this to save money by eliminating a solenoid control valve?  I don’t know, but it certainly wasn’t obvious to me that it operated like this.

I never found the smoking gun for the issue we had at the track.  I believe that the stresses of cornering flexed the car enough to allow the rear bow lock to unlock, so the rear of the top was not secured in position.  This allows the front of the top to sit in the wrong place.  I think the rear bow lock was very worn, and this contributed to the failure and perhaps my inability to manually latch it. I couldn’t even remove the hydraulic cylinder from the lock as I couldn’t get the retaining pin to come out of the cylinder – an indication of extreme wear.  No surprise I didn’t get much of my core charge back on this one!

This was a lot more work than I care to do at the track, and the repair once at home was nowhere near as straightforward as I expected. The real key was the ability to read microswitch positions in real time – comparing that to the expected sequence from the workshop manual finally allowed me to figure it out.  And of course, if I had just installed it as received, without my plastic shim, it probably would have worked correctly, though I would not have had this great (but time consuming) learning experience!

Photo Caption: Bow lock assembly, with a screwdriver lifting the open locking pawl.