2008 Boxster 987 clamp shell not closing - Rear control unit coding

[rennbod]

I committed the fatal sin with my Boxster and did not clean out the drains since this time last year and was punished by the drains clogging up and flooding the interior of the car.

The result was a very wet body control unit, all the rear lights illuminated and once I got into the car it decided to start moving the roof on its own and broke the plastic ball ends off two of the roof (soft top) arms on the roof transmissions....

So, as I have a durametric, I decided to buy a new body control unit, and coded it up as best as I can and replaced both the arms with the plastic ball ends.

Everything seems to be fine and working great, except for one, minor and confusing feature.. The clam shell at the back of the roof comes down properly when you put the roof down, but does not come down completely when you put the roof up!

I have tried adjusting everything which can be adjusted, I have spent days going through the forums searching for a similar problem, and I have even taken the roof transmissions off and inspected them and figured out how they work.

The transmissions run until they hit the two microswitches inside the transmissions, which also seem to be working properly

Nothing seems bent, other than the ends of the rods going to the clamshell, which seem to be bent from the factory.

On my old boxster (986) I once had a problem getting the roof to work properly and noticed on that on that car when the roof was all the way forward, a microswitch above the door pillar would be made and then the roof seemed to run for a couple more seconds to finish closing things and put everything under tension, on this 987 the motors stop the instant they hit the microswitches in the transmissions,,, Is this right?

The other odd thing, the durametric seems to be able to code the rear control unit, and I can change coding and see the effects, but if I try to read or clear fault codes in the rear unit it just sits there forever trying to connect... Is this normal, or an indication of a seperate or connected fault? Occasionally the Gateway and other systems are reporting being unable to communicate with the rear control unit also... Which is odd.. How can I be coding it via its communication system, but not able to talk to it for fault codes?

Anyway, I would have put up two problems here seperately (Clamshell not closing and communication errors) but thought it might be best to give you guys the full picture!

... HELP!!!

[rennbod]

There are a few items on the coding page for the rear control unit which I can't figure out.. Things like "SAD" activated or not... hmmm.. Not sure what "SAD" is on the car.. But my personal coding seems to have "SAD" enabled right now! lol

[rennbod]

As an update...

A friend has a 987 and today he popped over to help me look at my Boxster and we took the plunge into swapping parts between the cars.

As mine is a 2008 it has the rear control unit ending in 08, whereas his had the control unit ending in 04.. His has the arms that go onto the transmissions with the two keyways so the arms can be put on 180 degrees out by mistake, mine has the teeth with a notch where the arms go onto the transmission, which can only be put on the right way.

With his transmissions installed, my clamshell closes all the way!!!

But here is the odd thing.. I have taken my transmission apart, checked it out and the microswitches are undamaged, the gears are not damaged, everything is fine and I can not see how it can skip a tooth or even be assembled in a different way..

His transmissions end in a part number of 01, mine end in a part number of 02 and apparently the ones you get if you order them from Porsche end in 03.

So are the parts interchangable?

More importantly, do they have different ratios...

On the one hand, I could say my transmission is faulty.. even though everything seems fine with it on inspection... So I really do not know if I should buy an 01 transmission which I know will fix the fault, but may have a different gear ratio, meaning I have to buy two of them.

Or.. I could buy an 02 transmission, only to find out it will do exactly the same thing as my current transmission does (not closing the clamshell all the way).. but it is likely I will get an 03 transmission and I am unsure if I need to buy them in pairs.

Not sure if anyone is reading this, but I will keep it updated so anyone in the future with the same problem may find the answers they need

[rennbod]

Ok following some in depth investigation, it seems I am moving into uncharted territory, hopefully this will help someone else oneday.

The two transmissions I have are ...

987.561.179.02 from my 2008 Boxster

987.561.179.00 from the 2005 boxster

The differences between the two transmissions are relatively minor.

Microswitch assembly inside the transmissions...

The 2005 item has a rocker assembly which opperates two microswitches which opperates on the rotary cam inside the transmission

The 2008 has has just two microswitches, mounted closer to the cam, without the rocker part, which are directly opperated by the rotary cam

Hub which the shallow V shaped arms bolt to with the large bolt

The 2005 item has two slots with the V arm having two teeth which engage together to fix the position.. this V arm can be mounted in two different positions, 180 degrees apart.

The 2008 item has a multi toothed face with a single keyway tooth, with the shallow V arm having these multi teeth and a slot for the keyway tooth.

Metal frame

Although physically similar, there are slight differences in the metal frame plate which forms the main chassis of the transmissions, although they look similar, the main difference is the position of holes which locate the microswitch assembly. Otherwise they could be interchanged.

The gears inside the transmissions, seem to be identical, as is the main rotary cam which opperates the microswitches.

However, due to the differences in the microswitch assemblies, the later transmission will stop earlier in its cycle than the earlier transmission...

Now it gets interesting.. It would seem, that as the earlier transmission has a rocker assembly (2005), the microswitch being activated for either mode, open, or close, is the opposite microswitch to the one in the later transmission (2008) without a rocker assembly.. It seems the designer got around this by the multi connector being swapped around so the notch in the connector is the other way round.. So it would seem that pinouts would be the same.

Pinouts seem to be, the central pin is the common and connected to both microswitches, and the pin furthest from the notch in the connector is for the roof up position, the pin furthest from it is the roof down position.

Both microswitches are wired so you have continuity when the microswitches are at rest and go open circuit when contacted.. so as they would say in electrical circles, these are "normally closed" polarity switches. although to be precise, the switches are both 3 pin, and could be wired in another application as "normally open"

As a side note for anyone in the future with a faulty microswitch, the microswitches are individually identical and carry the part number 1055/58 (unsure about the 58, need to check with a microscope as the print is very small) and are made by a company called Marquardt which I found by doing some more digging.

Anyway, not much closer to finding out the source of my problems, as I still do not know what might be the fault with my car and why with a 2008 transmission my car does not completely close the clamshell with the roof up, but if I use the 2005 transmission the clam shell closes properly.

Which makes me wonder if on the PIWIS there is an option to select which transmissions you have when you code up the rear body control module, and therefore this module runs the motors for a short while after the microswitch has been contacted if coded correctly, but the durametric maybe does not have this option.

Or... The durametric does have the option, but it is one of the numerous options which can be selected in the coding screen for this module which do not seem to make sense, such as the "SAD Enabled" or one of the others... Really odd..

I will keep digging, and for the sake of some other crazy fool who might be in my position in the future, I will try to keep this updated.

[RFM]

As far as i know a 987 does not use the rear body control module, their roof system is electro-mechanically using two relais, one up and one down. Anyway, on MY2005 had the factory the rear body control module incorrectly programmed as convertible (997) where several error codes appeared as "communication error" since there is no connection between the module and roof system as opposed to a 997 which is hydraulic-mechanically. You need to change the indication "convertible present" in the module to "convertible not present" to start with. Only if the module is correctly programmed and there are no longer error codes present, can search and measuring work begin.

[rennbod]

Well this is the interesting thing, I was worried about this as I had read it elsewhere about coding a 987 as a "non convertable"... However if you code the rear control unit as a convertable the roof works, as soon as you code it as "non convertable" it stops responding to input the to roof switches.

On checking my friends 2005 car, I have found that its rear module is also coded as "convertable present"

I did however find that in the gateway that it is coded as non convertable (it may have been in the instrument cluster, but I am sure it was the gateway). and that was coded in both cars as "convertible not active", which would make sense as it seems the gateway would be the one to decide if the car has an extra module such as the 997 one... However, the roof switches (both the cabin and transmission/roof ones) in the 987 are wired into the rear control unit, and then power up relays to activate the motors.. But the same control unit is in the 997, where the rear control unit would have to send those signals to the roof control module by canbus, so you would think that the rear control unit would need to know if that control unit is there or not... However, once again, there is a coding entry on the durametric for "997 or 987" so maybe that is how it is coded there.. I have that set to 987 in mine and as my friends 2005 987 was also coded.

I am currently considering that each coding unit is actually an individual bit, a one or a zero, which is set within a byte or group of 8 bits for example, in the control module, which will have some flash ram space for these "bits", it could be that durametric codes these "bits" in sequences of bytes, so even if you set one change in the coding screen, it actually transmits the entire sequence of bytes with the changed "bit" within the stream, and there is a good chance they have not reverse engineered all the bits, and one of them could actually be "Late transmission" or "early transmission" which the PIWIS would automatically set on the basis of the identification of the control unit being in my 2008 case ending in 08, and on the earlier cars ending in 04... So in effect when I code up my later unit, one of the bits being set by the durametric is actually what Durametric have not reverse engineered and they assume it is the same in all the control units, not knowing that the different control units pair up with different transmissions, with a different range of movement.

If this guess is correct, I believe the later rear control unit when coded properly does not use one microswitch to say "Im up", or one to say "Im down", but for the roof being "up" it waits to see the "up" microswitch, followed by the second other microswitch also being made... and from seeing them activate in the right sequence, it knows the roof is up and therefore able to inform other modules like the gateway that it is up.

I have had a chat with a local Porsche specialist who is a technical advisor to the two main UK Porsche clubs, and he was a big help in trying to figure out the above, it seems the PIWIS coding is much simpler with less options than the Durametric, and it was he who suggested that maybe not just one "bit" is changed when you change an option, but a whole sequence, that the durametric may be assuming the settings for some of them, whereas the PIWIS maps them all out to be the correct ones for the right units each time it sends them. He has emailed Durametric with details of my problem, as they might be a little more helpful to him as a pro-user with a pro kit, rather than me as a pro-user (licence) but an enthusiast.

However, what you are saying about the communication errors really is resonating with me.... Setting the coding in the rear unit on this boxster really does completely dissable the roof, probably assuming the car is a Cayman at that point, so it needs to be set to "convertible active" but perhaps there is one of the settings I do not understand, such as this "SAD" option which is confusing me, which needs to be set along with "convertible inactive" for things to work... I am not sure... More experiments I guess.

Nice to have some feedback though, I was begining to think I should just keep updating this thread for any future owners going through the same pain in the future may find it via google or a direct search here.. But if nothing else what I find may be of use to lots of people in time.. However hearing from someone now with some input, or even just to say "Thats interesting" makes it all a bit more worthwhile.

:o)

[rennbod]

Bought a new 2005 type transmission and installed it into the car, everything is fine and dandy.

Can also confirm that the 2005 and 2008 transmissions have exactly the same ratios.

Will continue investigating this though and post any results... I would like to get the right transmission back in my car, but also to de-mystify the mysteries of coding and what actually goes on at a data level... Might have to investigate in a CanBus analyzer..

If anything from this thread is useful or interesting, would love for my reputation score to go up :)

Edited June 4, 2014 by rennbod

[rennbod]

Been on the forum asking an unrelated question and thought I would update this for anyone who finds the question on Google, but can't find the answer.

 

The issue turned out to be an issue that the Durametric is missing an option to set in the rear body control unit that the PIWIS does automatically.

 

From the chassis number the PIWIS knows the year, and so is able to set several other bits in the rear body memory/coding to say which transmission it has. The later transmission has the microswitches in a different part of the sweep, so the rear body control unit needs to know which year it is so it also knows if to stop the motors when that microswitch closes, or when it closes and then keeps going until it opens again.

 

I played with every option in the Durametric to get this to work, but in the end coding the roof with a PIWIS worked right away, without having to set any options, the PIWIS knew what to do right away with the minimum of information.

 

 

[jbassman6]

I realize this post is almost 3 years old but thanks @rennbod for providing all these details as it is helping me troubleshoot my top issues and decipher the details in the 987 wiring diagram.