Hi Pauldy,

Congratulations, that's terrific progress, finding the 2 erroneous position/limit switch actualizations !

Please review the document below for additional information regarding S84/13s1/13s2 switch group.



During opening the roof operation, until stage 2 completes, 13s1 should be switched to ground and 13s2 should not switched to ground.

The 13s1 and 13s2 actualization's you currently have would be appropriate during the initial stages when closing the roof.

The S84/13's internal micro-switches are triggered by the roof's frame as it opens or closes. In the picture, you'll see the catch lugs that enable the required toggling (i.e. seesaw) changeover action.

I would visually inspect the switch at this time. However, because It's somewhat buried, you might need to remove the interior left side rear paneling.


Since your S84/16 rear bow lock switch is currently not correctly actuated, please review the document below for additional information.


If you can't determine the reason why, re-watch the Top Hydraulics video I referred you to in post #2, which enabled you to free up your S84/16 locked latch mechanism.

Klaus also pointed out an issue, where the latch's micro-switch may fail to trigger when the cylinder rod extends and explained how to fix via an adjustment.

Let me know if you have any questions.

 

Hi Serndipity,
Yet again, thank you!I have had the opportunity to look again at the Fabric bow switch S84/16. As you suggested I reviewed again Klaus’s video with particular attention to the sticking latch mechanism. I could not really see a problem so I have attached a video clip, to see if you can see the issue.

I did a video in slow motion, but even after compression it is too large to send.

Starting with the piston in the fully retracted or locked position I used the hex key to move the piston to the fully extended or unlock position. Once I get to unlock the mechanism jams and I have to manipulate the mechanism each time before the bow can lock. The latch’s micro switch looks OK to me.



With regard to S84/13s1/13s2 switch I did not get very far because it began to rain and I have not yet been able to remove the left hand internal rear panel. However, I could not see an obvious switch, so I have attached a photo of what I could see. Could the rectangular structure marked be the switch? There is not one of those on the right hand side of the car.
Cheers
Pauldy

 

Deleted photos uploaded instead of video

 

Uploaded video clip
Trying to upload but video doesn’t appear in the post. Using .mov file

 

Hi Pauldy,

The documents in my last post (#26) only pertained to the theory of operation of the 2 position/limit switches that you found with incorrect actualization's., so I apologize for not providing some additional troubleshooting and repair suggestions sooner.

BTW, in post 25, when you checked the the A25 switch group you only determined that the compartment cover was locked with correct A25/s2 actualization, but did not verify that the cover is also closed, which in later model CLKs is done by checking the actualization of S84/5, as was explained in post #22, which is copied and pasted next.

"Should you have a later production MY 2001, the matrix showing the stages of roof operation vs.the position/limit switch actualization's is still valid (e.g. just ignore checking the A25s1 cover closed switch which is no longer exist). However, the controller still needs to know if the compartment cover is open or closed, which is now done via only the S84/5 switch. Strangely, this switch is not listed in the matrix, as it's installed in both early and later production CLKs. To check whether the controller sees the compartment cover as being opened or closed, measure the ohms resistance between pin 50 and vehicle ground. The cover is completely open at 0-5 ohms and closed at greater than 20K ohms.".

Specially, during the early stages (e.g. 1 through 4), when opening the roof, the controller should see that the compartment cover is closed as well.

However, at this time, let's assume the controller sees your cover is locked and closed and instead further explore both of the position/limit switches that you found being incorrectly actualized.

Beneath the document explaining the S84/13 switch group theory of operation, I included some additional comments, which is one word, says it currently actuated a$$-backwards (e.g. as required by the controller for when closing the roof, not for when opening the roof).

In other words, the controller thinks the soft top has already opened, so doesn't need to proceed with stage 3 and succeeding operations to open the roof.

Therefore, this would be a matter that you should further explore at this time.

Since the S84/13 switch group actualization's are triggered by the roof's framework (e.g. as it opens and closes), my first suggestion would be to check if the switch is correctly normalized (e.g. are positioned correctly, enabling them to actuate into the state that the controller needs them to be). This is done by fully opening and closing the roof manually.

Then retest the S84/13 actualization's again with your multi-meter. If after the normalization, the the actualization's are still incorrect, you're going to have to access and inspect the switch for potential issues.

Fortunately there is an existing thread, from Mercnotworking, that you need to read (e.g. post #'s 1 through 6), as it pertains to a faulty S84/13 and how he successfully diagnosed and repaired the problem.

Note that in post #6, the first TopHydraulics video, at the 2:49 time mark, shows how to remove the interior left side rear panel, the second video, at the 7:56 time mark show hot to find S84/13 to inspect it for potential damage (the first picture is just a screen shot from the second video). The last two pictures (which I believe are screen shots from eBay listings) seem to indicate that there may be 2 different part numbers for the switch, which is not true (e.g. both pictures have the same exact MB and the manufacturer's part numbers and the upper picture shows a complete and good part, while the picture beneath it, shows a broken part with the lever and its catch lugs missing from the rocker wheel.

Your S84/16 rear bow lock switch is currently not correctly actuated (e.g. it should be switched to ground, not an open circuit), so the controller is seeing the lock as already open, where it should be closed through stage 3.

You mentioned that your latch mechanism has again jammed. You'll be able to free up the jam again, using the technique explained above, in post #'s 2 - 4. Perhaps the jam reoccurred because when your manually closed the roof, the striking flange on the rear bow may not be pushing down far enough onto the latch mechanism.

However, in the video, where Klaus explains how the controller, cylinder's rod and micro-switch work, is 100% correct, but for only for the R129 class models (e.g.produced between 1990 and 2002).

The rear bow latch in our CLK A208 class (P/N 1247700426), is from the W124 E class model (e.g. produced between 1993 and 1995).

In this link , in posts #'s 56 - 59, you'll find additional details regarding how the 1247700426 latch micro-switch actuates.

Even then, it's somewhat confusing, because the included annotated pictures are of the R129 latch, which sometimes contradicts the written text, which was cleared up in Post 59.

Specifically, the controller will see that the rear bow lock is locked, when the lever sufficiently pushes down on the the micro-switch, which will connect pin 29 on the controller to vehicle ground. Thereafter, when opening the roof, the the controller should actuate a valve block solenoid to hydraulically extend the cylinder rod, releasing the lock and enable the roof opening to proceed to stage 4.

Following are some suggestions to help find the cause of the open circuit (high resistance) your test disclosed.

Reduce the hydraulic pressure, open the compartment cover, disconnect the S84/16 pigtail from the harness, connect your multi-meter leads (set to read low ohms), one to each of the pins inside the S84/16 pigtail connector. Your meter display will show either 0 ohms (i.e. closed circuit) or high ohms (i.e. open circuit), depending on whether the lock is set to an open or closed position. Remember these reading. Next, using the hex wrench, turn the latch to the opposite position and repeat the test. The meter's display should now display the opposite of the 1st test, indicating that the micro-switch is actuating properly. Note: With the S84/16 assembly still on the compartment cover, there might not be enough clearance for MB toolkit wrench, so you might need to use a 5mm L allen hex wrench to open and close the latch.

However, that's provided that the hydraulic actualization's (e.g. which are described in the last document I included in post 26).

I know that you previously mentioned that your hydraulic fluid level is proper and by inspection, see no cylinder leakage from the rod or port seals. However, the cylinder also has an internal piston, whose seal if defective, which would prevent the rod from extending or retracting properly. TopHydraulics is the only re-manufacturer on the planet that has the knowledge, uses superior materials and provides excellent customer support, to rebuild MB hydraulic components, better than new.

Another reason for an open circuit has been broken wiring or crimped hydraulic lines, which is caused by the flexing of those items, at the compartment cover hinge. These have been previously found by inspection, which requires removing the protective covering on the harness.

Alternatively, there is a method where you can use your multi-meter to test both the signal wire (e.g. between the disconnected main connector at pin 29 to it's smaller connector inside the compartment cover ) as well as the corresponding ground wire, which does not return to the main harness wiring, but terminates at the W7 ground, which is inside the right wheel well (e.g. not at the disconnected main connector), but that would require really long multi-meter test leads.

 

Hi Serndipity,
Thank you so much! This is fantastic, I have learned so much from you.
Rain is still forecast for this weekend, but I’ll see what I can do between rain showers.
cheers
Pauldy

 

Hi Pauldy,

Just a quick note to let you know that I'll be updating the S84/16 troubleshooting section in post #30 in this morning.

 

Hi Serndipity,
That's great!
Yesterday I was able to check the S84/13 switches and Pin 50.
1. With the roof down Pin 50 resistance was open circuit, thus S84/5 was closed.
2. I followed your guidance for S84/13 switches and manually opened and closed the roof. The resistances with the roof closed and locked were:
Pin 33 (S84/13s2) = Open circuit =0 from the matrix
Pin 31 (S84/13s1 = 1.5 ohms = 1, Stages 0, 1of opening operation and stage 12 of closing operation.
So that looks like good news and resolves one of the issues.

Thanks again for your patience and time and I hope that you have a good weekend.
Cheers
Pauldy

 

Serndipity,

Thank you so much for this.
I have been reading it through to understand what I need to do next. However, on the S84/16 the link to the thread and posts # 56-59 in the following section " The rear bow latch in our CLK A208 class (P/N 1247700426), is from the W124 E class model (e.g. produced between 1993 and 1995).

In this thread , in posts #'s 56 - 59, you'll find additional details regarding how the 1247700426 latch micro-switch actuates.

Even then, it's somewhat confusing, because the included annotated pictures are of the R129 latch, which sometimes contradicts the written text, which was cleared up in Post 59." opens up another window showing our current thread and not to the thread relating to the W124 E class. Perhaps posting the URL might help?

Once again, your help has been incredible and I really appreciate your time and patience.

Cheers,
Pauldy

 

Hi Pauldy,

I'm bad......the hyperlink "In this thread" wasn't opening the right thread, just did an edit and fixed it.

I wrote that part in my reply (i.e.post #30), after watching the video I posted in post #2, where between time-marks 5:48 - 9:35, Klaus explained how to fix a jammed rear bow latch, as well as adjusting the lever that actuates the micro-switch, if the latch mechanism had developed to much play.

Since the R129, W124 and A208 use a very different controller architectures, programing, how their position/limit switches actuate etc., you can ignore those sentences and focus just on how to diagnose and troubleshoot the A208 roof.

You have now had your rear bow latch jam several times, which when closing the roof, would occur when your manually close the roof, the the flange on the bottom of the rear bow may not be pushing down far enough onto the latch mechanism.

Here's a screen shot, from the owners manual, on locking the rear bow.




Note: When the hydraulics close and lock the rear bow, 2 cylinders (left and right sides), at a system pressure is over 2,000 psi take care of this.

It was good to hear that you checked your S84/5 and a simple re-normalization of your S84/13 s1/s2 switch group, both now show the correct actualizations.

Given that, my suggestion would be to:

1: Open the trunk and recheck the status of S84/16 (e.g. pin 19 to vehicle ground).
2: If it's not correct, reduce the system pressure, and using the hex wrench, turn the latch to open and then back to locked.
3. If either 1 or 2, resulted in the correct S84/16 actualization, restore full system pressure, close the trunk and try hydraulically opening the roof.
4. If neither 1 or 2 resulted in the correct S84/16 actualization, proceed to the S84/16 troubleshooting suggestions in post #30.

The 1st test checks the triggering of the micro-switch and begins with "Reduce the hydraulic pressure, open the compartment cover, disconnect the S84/16 pigtail from the harness, connect your multi-meter leads (set to read low ohms), one to each of the pins inside the S84/16 pigtail connector." If your uncertain as to how to connect the multi-meter to the latches disconnected pigtail, connect a test lead to the pin corresponding to the brown wire and the other test lead to the pin corresponding to gray with a pink stripe wire. My instructions to conduct the test was for a person operation and required 2 steps. If you can recruit a helper of 5 minutes, just continue to holding the test connections onto the connector, while the helper turns the hex tool to the open/closed positions. This test will show whether or not the lever is electrically triggering the micro-switch, opening and closing.

With regard to which may be broken or shorting, here's a picture showing the wiring harness and hydraulic lines at the compartment cover hinge.



This is where you're most apt to find the wiring breakage or shorting, of the signal wire that goes back to pin 29 on the main controller connector, or the ground wire that that goes to the vehicle W7 ground, inside the right wheel well.

If the multi-meter had longer test leads, or at least one could be extended, one could perform an electrical test first, which would see whether or not the wiring actually has an open or short circuit condition, which would trump the necessity to disturb the protective cloth tape and perform a visual inspection.

You'll smile after seeing the next picture.

The picture of the S84/13 s1/s2 switch group, which shows that the malfunction, in addition to the lever breaking off the rocker wheel, a catch lug missing from the lever, could have also been due wiring.

 

Serndipity,

This is fantastic! Thank you so much!
I'll try this is in the next day or two, weather permitting, and let you know how I get on. You are clearly one of those people with an infinite amount of patience to help with people like me!
Cheers for now,
Pauldy

 

Pauldy,

Here's the electrical schematic diagram, showing the wiring between the controller and the S84/16 on the tonneou cover.



When you disconnect the x1 pigtail connector, don't pull on the wires to separate the pigtail section from the harness section, which could pull the wiring out of the connectors. Use the method to separate the connector sections, as explained by Tophydraulics .

First test that the S84/16 micro-switch is triggering correctly.

If it's not, there is an alternative method to adjust it, that does not require bending the lever.

If that's not the culprit, to avoid needlessly opening up the protective coverings, I would first perform an electrical test, which would determine if the the wiring (e.g. back to the controller's pin 29 or to the W7 ground) is broken.

If you have some small gauge wire available, you can extend a multi-meter test lead by removing 2 to 3 inches of insulation from one end and wrapping the exposed wire tightly around the end of the meters test lead. From the other end of the wire, just remove a small amount of insulation, sufficient for the exposed wire to make contact to a test point. You only add the extension to one on the meter's test leads, which keeps the meter nearby and viewable . The total length of the meter's test points, needs to be able to reach from the X1 connector to the controller's connector.

Testing the ground connection for breakage:

From the harness side of the disconnected x1 connector, connect the meter to pin 1 (which is the brown wire) and the extended lead to the main vehicle ground at the negative battery connection. If the display shows near 0 ohms, that wire is not broken.

Testing the signal wire connection for breakage:

From the harness side of the disconnected x1 connector, connect the meter to pin 2 (which is the gray wire with pink stripe) and the extended lead to pin 29 of the disconnected controller connector. If the display shows near 0 ohms, that wire is not broken.

You'll probably need a helper, because it's critical that both of the test leads are making a good electrical contact in these tests.

Because the testing of micro-switch involved opening and closing the latch, it must be in the open position before closing the rear bow manually (e.g. the cylinder's rod should be extended) and don't forget to push down on the rear bow after it's lowered, prior to locking the latch.

 

Serndipity,
Today I have followed your advice as follows:
1.Opened the trunk and rechecked the status of S84/16 , Pin 19 to vehicle ground = 0.7ohms which is 1 in stages 0,1, 2, 3 on the Opening matrix. Therefore, Correct.
2. tried hydraulic roof opening: windows opened. But with front bow unlocked and windows down I could still open the trunk from the console switch - Is that OK? I thought that once the opening sequence had started the trunk could only be opened manually with the key.
3. Tested the S84/16 pigtail:
a) with piston retracted (Locked) = 0.6ohms
b) with piston extended (Unlocked) = 1 = open circuit. Therefore the lever is electrically triggering the micro-switch, opening and closing.
4.Tested continuity of S84/16 between connector and controller:
c) Pin 1 brown to negative battery connection = 0.6 ohms
d) Double check with pin 1 brown to W7 wheel arch earth (ground) =0.6 ohms
e) Signal wire pin 2 gray/pink to pin 29 =0.6 ohms
Therefore wiring intact with no shorts.
I then completed the manual opening and closing and tried to open hydraulically but with no success. It looks as if I have tested everything and aall are good. What have I missed?
Thanks again,
Pauldy

 

1.Opened the trunk and rechecked the status of S84/16 , Pin 19Pin 29to vehicle ground = 0.7ohms which is 1 in stages 0,1, 2, 3 on the Opening matrix. Therefore, Correct.

Great...... even though it took several months of effort, to finally have all of your prerequisite stage 1 -> 3 switches set properly. where the controller should now actuate stage 3, which requires, which requires a valve block solenoid to actuate, hydraulically extending the S84/16 cylinder rod, releasing the latch's lock, enabling the roof opening operation to proceed onto stage 4.

Appears that that my suggestion of, when closing and locking the rear bow manually with the 5mm tool, to make sure that the latch was normalized correctly (e.g. in the open position where the cylinder's rod should be extended) and pushing down on the rear bow after it's lowered, did the trick.

While the roof can be manually be opened and closed by only one person, two would be preferred, which is why the hydraulic system uses a pair (e.g. left and right side) cylinders to open/close both the rear bow and main lift framework.

Note: This was likely the cause of your S84/13 switch group, which is actualized by frame movement, being out of kilter.

2. tried hydraulic roof opening: windows opened. But with front bow unlocked and windows down I could still open the trunk from the console switch - Is that OK? I thought that once the opening sequence had started the trunk could only be opened manually with the key.

Your referring to what I said in post #21......."By design, to prevent a collision between the trunk lid and the lid of the soft-top compartment, while the roof is being opened or closed, the key fob or center console remote trunk release switch will not allow opening the trunk, so you'll have to use the mechanical key to open the trunk.".

More specifically, during soft top operation, it's control module cyclically sends the following CANBus digital messages to the central locking module, "SOFT TOP CLOSED", "SOFT TOP UNLOCKED" or "SOFT TOP OPEN".

As these messages are received, the central locking system will prevent the soft top from opening/closing if the trunk is unlocked, as well as the trunk lid from opening, (e.g. with the key fob or center console remote trunk release switch), until the roof's opening/closing operation is completed and locked.

You can still release the trunk electrically during stage 0 -> 2, or 3 if it has not correctly actuated hydraulically to released the rear bow, because the "SOFT TOP UNLOCKED" message is not sent until stage 3 completes.

3. Tested the S84/16 pigtail:
a) with piston retracted (Locked) = 0.6ohms
b) with piston extended (Unlocked) = 1 = open circuit. Therefore the lever is electrically triggering the micro-switch, opening and closing.
4.Tested continuity of S84/16 between connector and controller:
c) Pin 1 brown to negative battery connection = 0.6 ohms
d) Double check with pin 1 brown to W7 wheel arch earth (ground) =0.6 ohms
e) Signal wire pin 2 gray/pink to pin 29 =0.6 ohms

Perfect!!!!

For some time, your S84/16 actualization has indicated you had an open circuit, during stages 0 - 3, which is incorrect. This malfunction could be hydraulic, due to the controller not signaling the valve block solenoid to actuate and release latch, faulty latch and/or cylinder, or a wiring issue.

The technique of removing the connector from the controller and checking the appropriate pins, against vehicle ground, at the stage of failure for correct actualization's as shown in the matrix, will not only test and disclose a wiring problem, but is exactly the same technique that a trained MB tech would use.

Glad to hear that you made the extra effort to lengthen a test lead, which allowed you to test the harness wiring branch, inside the compartment cover, for possible hinge area breakage.

I then completed the manual opening and closing and tried to open hydraulically but with no success. It looks as if I have tested everything and all are good. What have I missed?

So far so good......your doing great.

<><><><><><><><> See additional comments below<><><><><><><><>

After you completed the manual opening and closing of the roof normalization, all of your required stage 1 - 3 limit switches were properly actuated, but you were still unable to hydraulically open the roof.


Previously you had mentioned that your S84 main operation switch was remaining illuminated when not pressed (see the below hi-lighted explanation of possible reasons).



Included below is a functional description of the S84 switch.



S84 has 3 positions and an internal resister, that actuate as the switch is operated, which provides the necessary voltage encoded levels to the controller.

This can be verified with the ignition off, the harness connector disconnected, the multi-meter black test lead is connected to pin 52 and the red test lead to pin 55.

When S84 is at rest (not pressed), you should see an open circuit.
When S84 is pressed to open the roof, you should see approximately 200 ohms.
When S84 is pressed to close the roof, you should see approximately 1 ohm.

Note: Pin 52 (a.k.a. circuit 31) is an internal ground point in the controller's electronic circuitry, meaning that the test only checks for a possible defective switch or a problem within the controller.

To check for a S84 wiring line discontinuity, repeat the above test with a test lead connected to vehicle ground (e.g. the negative battery post) and pin 55.

Note: This will test S84 for a possible wiring harness defect, as well as the S84 W18 ground point, which is located at the left front seat crossmember.

If the above checks out OK, the power roof controller is fused in 5 places.

In fuse block F4, located inside the trunk, there are 4 circuit 30 fused circuits. Specifically, the always present circuit 30 power is fused by the 15A fuse in position 11, the pump motor is fused by the 40A fuse in position #13 and each rear window motor are fused positions 10 and 12. Additionally, in fuse block F1, located in the engine bay, circuit 15 power, which is only supplied to the controller by turning the ignition switch to position 2 -> 3, there is a 5A fuse in position 15.

When checking fuses, never just trust by just viewing the sight glass, also pull each fuse and check that the blade contacts and socket do not have any corrosion, which would provide additional resistance to passing the required power to successfully operate the control module.

I did not mention checking the fuses earlier, because your windows rolled down and your hydraulic pump operated and it was assumed that the N52 controller was 100% functioning properly, as It analyzed the positions of the limit/position switches and then correspondingly powering the solenoid valves to actuate the hydraulic cylinders.

As mentioned above (e.g. in item #1) , as well as several times prior, the controller should now actuate stage 3, which requires, which requires a valve block solenoid to actuate, hydraulically extending the S84/16 cylinder rod, releasing the latch's lock, enabling the roof opening operation to proceed onto stage 4.

There is a 126 pin socket box and cable, from the R129 era (P/Ns 129-589-00-21-00 and 104-589-00-63-00), that still corresponds to the 55 pin A208 controller, which could be inserted between the controller and it's harness. This would make it possible to further diagnose the controller, with ignition on multi-meter measurements that would test the controller's internal circuitry, for power, output signals voltages and CANBus communication.

However, it's cost is prohibitive at around $3,000.

You'll find this thread of great interest.

Let me know if you have any questions.

 

Serndipity
I’ve checked my multimeter and touched the two leads together and the readings changed from 1 to scrolling through various random readings from 1 to 64 ohms and at times showing -negative numbers! I replaced the battery with a new one (the old one was reading 8.87v) and repeated the test but still got the random numbers. Then replaced the leads with those from the other multimeter and some others which are new. Looks as if the multimeter is kaput!

 

Just a quick note to let you you know that I completed post #39, with additional comments.

Also I'm not convinced that your multi-meter is kaput.

I looked at the owners manual for your meter.

From the specifications regarding the measurement of resistance, as shown on page 18 below.

When measuring the limit switch actualization's, the correct range would be 200 ohms, where the resolution is 0.1 ohms and the accuracy is 0.8%.

When using a higher range than necessary, both the resolution and accuracy could easily provide erroneous results and make no sense.

The 8.87 battery voltage appears to be OK, but you likely measured the unloaded voltage (e.g. when removed from the meter) vs. while it was loaded by the meter's circuitry, where the voltage would be lower. However, your meter will display a battery symbol in the display window, when the battery needs to be replaced.

Also noticed that the meter has 2 fuses (F1 for measuring ohms, voltage and low current and F2 for measuring high currents). I would check F1 to make sure it's making a good/clean electrical contact in its holder.

 

Serndipity,
this is great and I’m working through the tests for S84.
Just a quick question to your Note: This will test S84 for a possible wiring harness defect, as well as the S84 W18 ground point, which is located at the left front seat crossmember.
Is S84 W18 under the carpet where the 3 wiring looms for the seat emerge? It is a crossmember to which the front seat bolts attach to the floor with star bolts

So far the ohmmeter tests check out as desired:
1. S84 at rest = open circuit
2. S84 to open = 196.4 ohms
3. S84 to close roof = 1.0 ohms
4. S84 Pin 55 v battery ground = open circuit

Cheers
Pauldy

 

Your S84 results posted in lines 1 -> 3 are perfect, however I'm confused by the 4th line (e.g. makes no sense).

In the first test, where the meter is connected to pins 52 and 55 on the disconnected harness, there should only be 3 test results, as shown below.

When S84 is at rest (not pressed), you should see an open circuit.
When S84 is pressed to open the roof, you should see approximately 200 ohms.
When S84 is pressed to close the roof, you should see approximately 1 ohm.

Note: Pin 52 (a.k.a. circuit 31) is an internal ground point, inside the controller's electronic circuitry, meaning that the test only checks for a possible defective switch or a problem with the internal ground on controller's circuit board.

The reason for the second test, where you connect a test lead to pin 55 and the other test lead to a known good vehicle ground (e.g. the negative battery post), is to verify that S84 is also being connected to it's W18 vehicle ground point as well.

Note: The roof system uses 3 separate vehicle ground points (e.g. W6, W7 and W18), which then consolidate/merge into a complete unified electrical ground, as required by the controller.

When you perform the second test, if all is good, there should only be 3 results, which should be approximately the same as you previously measured in the first test.

The W18 ground is most likely beneath the floor carpeting but I can't be more specific to your questions regarding exact location.

The best I could offer is a link to the EPC for the CLK 208.465 cabriolet, which contains numerous diagrams of all the electrical wiring harnesses, but doesn't provide specific location details.

However, if both S84 test out good, it should not matter.

 

Sorry, I’m a dumb student! I totally lost the plot with the S84 test! So I repeated it with the pin 55 and battery ground and the results in #1,2,3 in post 43 were almost the same as the 1st test: S84 at rest = open circuit, S84 open= 196.7 ohms, S84 closed = 1.2 ohm. So, all good!
checked fuses #11, 13 and 15 in the trunk and engine compartment and all good with no corrosion and tight fitting.

I now need to review and understand the link that you provided in post 39.

I replaced the fuse F2 10A/250V in my multimeter, which had blown. F1 2A/250V was Ok, but when I checked across the terminals with the fuse out I got open circuit using the other multimeter. so looks as if the circuit board has shorted out.

cheers and as ever, thank you for your help.

Pauldy

 

Hi Serndipity,

I bet you thought that you had heard the last of me? No such luck, but I thank you for your patience. For a number of reasons I have not been able to follow up on your last guidance until today.

I read with interest the link to your posts with CapitalB666 back in 2019 and I replicated a number of the tests that he completed.

After all these tests I must admit I am still at a loss as to the cause of the problem.

Today I repeated the following ohm-meter test starting with 1. Roof fully closed, then 2, then 3, then 4. finally closed: (See attached Table in pdf)

When the roof is fully opened manually, using the hex key, the rear bow lock is not locked, ie the ohmmeter reads infinity or >20 kohms. Should it be locked ?

On the basis that when I close the roof manually, S84/16 always sticks when using the hex key. To lock the bow I have to manually manipulate the locking pawls / mechanism to release the cylinder piston and allow the hex key to operate.

From the information in the link to CapitalB666, should I be thinking that there is a fault in the N52 controller and that should be replaced?

The other point to add, when I try to open hydraulically from SW84, when the manual lock in the front bow is opened the windows retract (lower) but when I’ve closed the roof manually and locked the front bow the windows do not go up on pressing S84 but do when the four individual window switches S21 and N72 are operated.

Having got this far is there anything else I can try?

Thank you so much for your help and extraordinary patience.
Cheers
pauldy

 

Hi Serndipity,
I’m not sure if you have seen my last post #45, but I’ve got to the point where I do not know what else to try to fix this problem! All the tests seem to be positive and there seems to be no obvious diagnosis.
Unless there’s something I’ve missed is the likely cause to be the controller that needs to be replaced ?

As ever, thanks for your patience and support.
Cheers,
Pauldy

 

Hi Pauldy,

From the beginning and throughout the course of this thread, your S84/16 rear bow latch assembly has been the center of attention.

When opening your power roof, the malfunction occurs at stage 3 and although all of the prerequisite position/limit switches appear to be correctly actualized, why the roof does is not advancing to stage 4, is unresolved.

In post #16, you considered having your MB dealership diagnose your power roof problem with their STAR Software Diagnostic Software (SDS), at a cost of $215, upon which the MB tech would operate your power roof to the point of failure and you would have received a print out of any diagnostic trouble codes (DTC) currently stored in the controller's memory.

Note: The DTCs are not permanently stored, as explained in more detail next.

At the end of each of each soft top opening or closing operation, the N52 controller will send a CANBus message to the overhead control panel and you'll hear a brief 1 second chime, which confirms that the implementation of the hydraulics and position/limit switches was correct. Additionally, during driving mode, whenever the controller sees that the roof is incorrectly closed and/or locked, the warning chime will sound longer. Further output of the warning signal can be eliminated, by turning the ignition 'ON' and 'Off' again (e.g. which will erase any stored DTCs) or alternatively, by changing the state of the improper position/limit switch (e.g.by manually normalizing the switch or after successfully repairing the fault).

As previously mentioned, each time you turn the ignition switch to position 2 or 3, the N52 controller also runs a diagnostic, where the lamp in the power top operating switch will indicate potential malfunctions.

You previously mentioned that your power soft top S84 operating switch remains illuminated when it's not being pressed, which is why I hi-lighted the possible reasons why this would occur.

Had SDS been run, here's an example of a DTC that presumably would have been on the diagnostic print out, at a stage 3 failure.



A knowledgeable/trained MB tech would then troubleshoot the S84/16 rear bow latch as follows.



Note: SDS only uses a 'watch dog timer' to record a position/limit switch DTC, which only indicates that position/limit switch is not plausible (e.g. because it did not actuate within a predetermined time-frame.

The purpose of test step 16.0 is to insure that the roof malfunction is not being caused by faulty circuitry, internal to the control module.

If the specified voltages are OK, the MB Tech would then use the same 'stages of roof operation vs position/limit switch actualization matrix document' that we've been using, to accurately/completely diagnose, troubleshoot and repair a power roof malfunction, attributed specifically to a malfunctioning position/limit switch, that the controller cyclically analyses during roof operation. This procedure also verifies that the the wiring back to the controller and all of the associated vehicle ground points.

The the primary operation that occurs at stage 3 (provided that all of the prerequisite position/limit switches are set correctly), would be the controller powering the pump's valve block 'rear bow open' solenoid, which then actuates the S84/16 cylinder (e.g. where the cylinder's rod will extend and unlatch/release the rear bow. upon which the bow should then raise a couple of inches (e.g. out of the compartment cover).

The fault would be recognized by observing if the controller is sending 11 - 14 volts, which actuate the pump's valve block solenoids. If not, possible causes would be caused by the the pump not running, malfunctions of the wiring or connectors on the pump assembly or a malfunction within the control module's internal circuitry.

Unfortunately all of the above mentioned required voltage measurements can not be done without the expensive break out box and cable.

Fortunately, when operating the power roof to the point of malfunction the 'stages of roof operation vs position/limit switch actualization matrix document', provides the required diagnostic information (e.g. isolates the area that the system is not happy with the operation of something). Thereafter, it's only a matter of inspection and/or testing to determine and repair the hydraulic, mechanical or electrical defect.

This would be a good time to address your recent questions.

1. Today I repeated the following ohm-meter test starting with 1. Roof fully closed, then 2, then 3, then 4. finally closed: (See attached Table in pdf)

When you manually operated the roof (to make the measurements shown in your document), you were manually performing the work of the controller and hydraulics.

However, in the bottom row, you checked the actualization of S84/13s2 (pin 33), which doesn't actuate until the soft has been lowered into the soft top compartment.

As shown in the document included in post #26, the purpose of the S84/13s switch group, is to tell the controller if the 'soft top is open' or if the 'soft top is closed'.

You should be checking the actualization of S84/13s1 (pin 31), which the controller analyses when opening the power roof.

Given that all of your prerequisite stage 3 position/limit switches are correctly set (for the roof to open hydraulically), this would be a good time to check if your hydraulic pump energizes at stage 3, as follows

-The vehicle must be in a quiet place with the engine off.
-Open all the windows using the buttons on the center console.
-Turn the handle to unlatch the front bow.
-Press the power top switch to open the roof.

As soon as you press the power top switch, you should clearly hear the pump turn on, the latch unlock and the rear bow should release out of the latch and rise out of the compartment cover a couple of inches.

During stage 3 (provided that the controller analyzed all of the prerequisite position/limit switches shown in the matrix have been correctly setup), the operation is hydraulic, where the controller should energize the 'rear bow lock open' solenoid and unlatch the bow (as shown and required at stage 4).

2. When the roof is fully opened manually, using the hex key, the rear bow lock is not locked, ie the ohmmeter reads infinity or >20 kohms. Should it be locked ?

That is correct (refer to the matrix actualization's that were included for when 'opening' and 'closing' the roof.

When you manually open the roof, you only use the hex wrench to unlock rear bow lock at the beginning (e.g. because you locked the bow latch at the end of closing the roof).

The latch remains unlocked because the final step is to close and lock the compartment cover (if it was done hydraulically, which is when you would hear the 1 second chime).

3. On the basis that when I close the roof manually, S84/16 always sticks when using the hex key. To lock the bow I have to manually manipulate the locking pawls / mechanism to release the cylinder piston and allow the hex key to operate.

We're back to an issue that began the thread, which we covered posts #s 1 -> 4.

Note: The video byte begins with the latch in it's open (unlocked) position,with the cylinder's rod will be fully extended. Then he showed that when the 5mm Allen tool was unable to turn the latch mechanism, from it's open (unlocked) position to it's closed (locked) position, before you simulate the striker (which is located on the roof's rear bow) hitting/pushing (down) on the mechanism. Once the mechanism is actuated, you'll be able turn the latch to it's closed (locked) position. If all is good, as latch moves to it's closed position, you'll hear 2 ratchet clicks (provided by a pawl and torsion spring) and the cylinder's rod will fully retracted.

What I find strange, is that when you

What I find strange is that it reoccurs when you manually closing the roof and not when the rear bow assembly's micro-switch operation was being tested, which is why in post 35, I included the screen shot from the owner's manual, that stated that the rear bow striker plate needs to be pushed down into the latch, prior to locking the latch.

I have another suggestion.

Recall that in Top Hydraulics video, where Klaus discussed the roof's tension cables, there are a pair of cables, attached to both the front and rear bows.

During power roof closure, the left and right rear bow cylinders (operating at 2175 psi), lowers the rear bow and insures that strike plate engages fully into the latch, so that when opening the roof, as the rear lock unlatches, the bow tension is quickly released and the bow and strike plate will rise out of the compartment cover.

To rule out that the culprit is not stretched or broken rear bow cables, when you manually unlock the rear bow latch, does the bow and strike plate rise out of the compartment cover?

If not, this would interfere with the latch from properly resetting.

In regard to your known broken front bow cables, the $1 fix you tried and found doesn't work well. That's because when you pull down the handle to lock the roof onto the windshield, the effort needed applies a significant amount of tension to both the front and rear bows and the cable ends at the 3rd bow cable end stop buttons, need to be pressure crimped onto the cable.

A new pair of genuine MB cables will set you back between $200 and $250.

Although it's a good idea, be careful when pushing out the 3rd bow's split plastic sleeve inserts, which is required in order for the end stop button to pass through 3rd bow's hole (e.g. replacements are available).

However, it's not a compulsory repair provided that, when you unlock the front bow, push it upwards enough to open the roof framework above the passenger windows and hold it in that position, because when the rear bow unlatches and releases the rear bow tension, there's a good amount of kickback forward, which would tend close the framework above the passenger windows. Additionally, when closing the power roof, as the main lift cylinders finish their work, sans the front tension cables, gravity will slam the bow onto the windshield, unless you catch and lower the bow (e.g. otherwise damage can and has occurred, preventing the correct actualization of the S84/11 top locked position/limit switch).

Note: The purpose of the S84/11 top locked position/limit switch is only put the controller to sleep and then wake it up, as the front bow is latch and unlatched and the controller does not use it's status otherwise.

4. The other point to add, when I try to open hydraulically from SW84, when the manual lock in the front bow is opened the windows retract (lower) but when I’ve closed the roof manually and locked the front bow the windows do not go up on pressing S84 but do when the four individual window switches S21 and N72 are operated.

To incorporate the various window features, traditional (i.e. mechanical) limit switches to operate the motors are not used.

Instead the widow motors start and/or stop function uses Hall Effect sensors, which count the revolutions of the motor shaft and are programed into the N69 door controllers and saved.

The zero count position is established when the windows are fully open and the controller will recognize a rotational shaft count of zero.

Additionally, the N69 controller's firmware program then uses those limits to provide directional control and numerous window convenience features, such as express window control, stopping window closure upon contact of an unexpected obstacle, central closing/opening with the key fob when outside the vehicle or during soft top operation.

Another example is that on the later production models, to provide superior weather/wind resistance, the windows were sealed in the roof's U shaped rubber molding channel and upon opening the door, the windows will automatically drop slightly and release from the channel.

It's possible that your window limits have lost synchronization and only require normalization, which is easily accomplished when the center console power window switch is held in the closed direction for longer than 5 seconds (even though the zero position has already been reached).

Note: On the cabriolet model, the roof must be closed when performing the normalization procedure and is done for only for the driver and passenger front windows (which utilize N69 controllers). The rear window motor operation is controlled by the N52 soft top controller, which automatically switches off the motor operation after 25 seconds.


5. I read with interest the link to your posts with CapitalB666 back in 2019 and I replicated a number of the tests that he completed.

Back in post #39, after you completed the manual opening and closing of the roof normalization, all of your required stage 1 - 3 limit switches were properly actuated, but you were still unable to hydraulically open the roof.

Because you had additionally mentioned that your S84 main operation switch remained illuminated when not pressed, I hi-lighted explanations of the possible reasons.

The noninvasive testing techniques suggested, verified that the actualization's of all of the position/limit switches, the associated vehicle grounds and wiring back to the controller were perfect, as well as the roof's S84 operating switch was not defective.

The purpose of providing that link, was that when all of the position/limit switches stage actualization's prove correct (where the roof operation fails), has sometimes been caused by circuitry within the controller being defective.

Here is another thread to read, where a rear bow lock issue turned out to be a faulty controller.

Unfortunately, that can not be further verified without measuring the controller's outputs without the ignition on and the break out box.

Note: As you read the thread included above, although the poster who joined in had highly skilled troubleshooting skills, which initially were somewhat invasive and not successful. That's because when evaluating the controller, the integrity of the system must be unbroken.

 

Just a quick note to apologize for taking so long to finish replying to your questions in post #45.

At this time, please review my comments in post #47.

As always, let me know if you have any additional questions.

 

Hi Serndipity
Thank you so much! I sense that you’re getting somewhat exasperated with my failure to determine the root of my roof’s fault! You have spent considerable time and energy assisting and teaching me.
It’s now wet and windy so I now have to wait for a break in the weather before I can revert to the problem solving! It hasn’t helped that I now have a fuel leak and have to replace the fuel lines!
In the meantime I’m reviewing your post to determine my next moves in a list of sequential actions.
Just one question at this stage: once we get to the point where the expensive breakout box is the only option left to me, should I consider going back to the MB dealer for a Star SDS analysis? Is the SDS essentially equivalent to the break out box?
cheers
Pauldy

 

Hi Pauldy,

Is the SDS essentially equivalent to the break out box?

No......as shown in the picture below, the break out box is just passive hardware, consisting of a socket box with 126 banana jacks and a special test cable.


The origin of the 129 589 00 21 00 break out box goes back to the R129 SL roadster, that was produced between 1990 and 2002.

The special test cable is wired, such that the 55 pin-out assignments of the A208's N52 controller, will be identical to sockets numbers 1 -> 55 on the 126 pin socket box.

The cable's 3 connectors then permit inserting the socket box between the N52 controller and wiring harness.

Specifically, when inserted, the ignition can be switched on (e.g. to either position 2 or 3), which will fully power the controller (upon which, if all of the inputs that the controller analyses at a particular stage of roof operation are are correct), the controller's voltage outputs can measured and verified.

A secondary benefit would be, when the socket box is detached from the controller, with the box still attached to the harness, it would be much easier and faster to test the position/limit switch actualization, wiring and vehicle grounds.

As example, in post 47, I included 2 documents.

The first described, had SDS been run, showed an example of a diagnostic trouble code, that presumably would occur, when opening the roof, at a stage 3 failure.

Note that SDS only indicated that the controller S84/16 wasn't plausible, solely based on the controller's 'watch dog timer', because it did not actuate within a predetermined time-frame.

As shown the second document, a knowledgeable/trained MB tech would then troubleshoot S84/16, beginning with step 16.0, where the controller's output voltages are must be correct , which would indicate that the roof malfunction is not being caused by faulty circuitry, internal to the control module.

If the specified voltages are OK, the MB Tech would proceed to step 16.1, which is the same 'stages of roof operation vs position/limit switch actualization matrix document' that we've been using, to evaluate the S84/16, wiring , micro-switches and associated vehicle ground point, for faults.

Once we get to the point where the expensive breakout box is the only option left to me, should I consider going back to the MB dealer for a Star SDS analysis?

At this time, going back to the MB dealer is not your only option.

When MB dealerships diagnose power roof problems with their STAR Software Diagnostic Software (SDS), a MB tech would attach the system to 38 pin diagnostic port in the engine bay, set up the software and operate your power roof to the point of failure, where the controller should indicate present DTCs (which will be erased when the ignition key fob is turned off/on), upon which you should receive a printout.

But that's not where they make their money and you'll likely be asked to authorize them to troubleshoot and repair the malfunction (Read: $$$$$).

Recall in post #15 included a link, where the takeaway was that a dealership replaced a S84/16 rear bow lock and pump relay, which supposedly were broken, for a total of $1500 ($600 parts, $900 labor) and when the owner returned because his roof was not repaired, the workshop's next step was to replace the pump, because "the pump does not have the power to open the lock and open the rear bow" for $8,900 and if that didn't fix the problem, their next suggestion was to replace the controller. The poster then used the stages of roof operation and found the culprit was that his S83/5 roll bar retracted position/limit was not actuating and the power roof was finally repaired for $0, by just adjusting the micro-switch to trigger properly, when the roll bar retracts.

Besides the profit motive, given the age of the A208, the MB techs who were trained and experienced to accurately diagnose and troubleshoot these cars, are likely long gone.

At stage 3, provided that just the stage 3 position/limit actualization's shown in the matrix are correct, the controller should electrically power the 'bow lock open' valve block solenoid, to hydraulically release the latch, enabling the power roof opening to proceed to stage 4.

Before you open your wallet, the reason it took me 3 weeks to finally complete post #47, was to comprehensively read though the thread from the very beginning, mull over the peculiarities, consider what has occurred (or not) and then suggested some additional troubleshooting ideas, to try first.

Just manually opening and closing the roof, since early July, is not beneficial, because you're performing the work of the controller and hydraulics, without troubleshooting what is causing the malfunction.

Because you been reporting a persistent issue with the latch jamming, where I suggested a method to rule out that the culprit is not stretched or broken rear bow cables, preventing the bow's strike plate not to rise out of the compartment cover.

Another possibility, as discussed in post 1 -> 4, could be caused by the cylinder's internal piston seal being defective.



Hydraulic pressure is always present in the piston and rod sides of all of the system cylinders and the pressure increases when the pump operates, so when the 'bow lock open' valve block solenoid actuates, if the piston seal is leaking, the cylinder no longer has the required large active area (to overcome the rod side pressure) to extend the cylinder's rod and release the bow.

A savvy MB tech would be able to test the S84/16 for cylinder piston seal for leakage, while still on the vehicle.

Each time you turn the ignition switch to position 2 or 3, the N52 controller also runs a diagnostic, where the lamp in the power top operating switch will indicate potential malfunctions



You previously mentioned that your power soft top S84 operating switch remains illuminated, when it's not being pressed, which is why I hi-lighted the possible reasons why this would occur.

Your problem is not a wiring discontinuity to the S84 power top operating switch, because that was ruled out by verifying that it was providing the correct resistance values to the controller (which as a result should see 3 unique voltage levels as S84 is operated) and you also verified that the circuit 30 fuse, located in the truck's fuse box and supply's unswitched (always present) vehicle battery power to the controller, was good.

Since when opening the roof, is the first time the pump operates, this would be a good time to check if energizes at stage 3, using the method I detailed when I replied to your first question in post 47.

Let me know if the hydraulic pump does not run, or if you have any additional questions.