Volkswagen Eos Forum banner

1 - 20 of 20 Posts

·
Registered
Joined
·
5 Posts
Discussion Starter #1
Hi everyone,
I have a roof with an error “02805” I have VCDS and that’s how I was able to check it. But the rrror code doesn’t guide me in any directions.
I tried checking the binary and it is out of sync but can’t exactly put my finger on what it’s trying to tell me. Any help would be greatly appreciated as I do t want to be bring it to VW. Here’s a video of the roof opening binary if anyone could help me. Thank you all 🙏🏼
 

·
Registered
Joined
·
5 Posts
Discussion Starter #3
yes, I know about that. But as I said I’m having trouble putting my finger on the issue and am asking for help. 🙄 🤦‍♂️
Any help would greatly appreciated
 

·
Registered
Joined
·
200 Posts
  • Like
Reactions: Jamiemag

·
Registered
Joined
·
967 Posts
The first part (which you've probably figured out) is that the senders for "rear shelf locked" are reporting that the locks for the rear shelf have moved then moved back. This isn't something I've seen before and I don't think anyone else has reported anything like it either.

This could be a mechanical issue (misalignment) or a hydraulic issue (leak or pump fault). It seems unlikely that it would be an electrical issue since the senders appear not to be malfunctioning and you can see them sending both a 0 and a 1.

If you really want to avoid the dealer, you should:
  • read everything on this forum that has anything to do with the roof - voxmagna wrote a detailed guide on how to troubleshoot the roof electrics
  • read the section about the roof from the VW Self Study Programme
  • get yourself a workshop manual (Bentley Publishers, Elsawin, or Erwin)
  • get yourself a diagnostic tool like a VAS5054 that works with ODIS, the software that dealers use that has their guided fault finding (read the link in my signature)
 
  • Like
Reactions: Jamiemag

·
Registered
Joined
·
5,693 Posts
If the locks are really closing after being opened as the sensors say, they can't do this on their own because quite large hydraulic forces are required. The controller, pump or solenoid valves must be incorrectly sending hydraulic fluid in the opposite direction to open the lock? I'm not sure if there's overload protection in the roof operation (pump?) that could stop an operation mid way if a part was jammed and try to force the roof op. into reverse (aka windows pinch protection)?

Diagnostics on its own won't solve this strange problem which needs understanding and eyes on looking at the parts to see what's moving and why the sequence isn't being followed. Further investigation could lead to trunk lockout syndrome if care isn't taken: If the shelf is unlocked the trunk locks should be on! Plenty of information here which can be searched, but it's no substitute for detail in the V.W EOS workshop manual and self study guides.
 

·
Registered
Joined
·
967 Posts
Now I think about it, these are the claws that I've observed can slowly close if you leave the roof part open with the rear windscreen on top of the roof, and they need a lot of force to open again. I don't know whether hydraulic fluid would need to be sent in the opposite direction or whether this could be caused by a leak or a sticking valve.

Having taken another look at the SSP it seems like one of the shuttle valves may be stuck - after the first step of opening the roof, which releases these claws, the roof pump then reverses direction. If the hydraulic system remained in the same configuration as it started in due to a valve not moving the way it should, the claws that just opened would close again. I don't know if it's plausible for a shuttle valve to stick in that way though.

As far as I know there's no logic to automatically reverse a roof operation if something goes wrong - it'll just stop, and it may let you reverse it if it's in a state where that's permissible.
 
  • Like
Reactions: Jamiemag

·
Registered
Joined
·
5 Posts
Discussion Starter #8
Hi!
Thanks for the reply,
yes, I read that but sadly that's just a generic code that might suggest the roof is out of sync.
I reset the roof and no joy. Nope the engine wasnt running.

Thanks mate,
 

·
Registered
Joined
·
5 Posts
Discussion Starter #9
If the locks are really closing after being opened as the sensors say, they can't do this on their own because quite large hydraulic forces are required. The controller, pump or solenoid valves must be incorrectly sending hydraulic fluid in the opposite direction to open the lock? I'm not sure if there's overload protection in the roof operation (pump?) that could stop an operation mid way if a part was jammed and try to force the roof op. into reverse (aka windows pinch protection)?

Diagnostics on its own won't solve this strange problem which needs understanding and eyes on looking at the parts to see what's moving and why the sequence isn't being followed. Further investigation could lead to trunk lockout syndrome if care isn't taken: If the shelf is unlocked the trunk locks should be on! Plenty of information here which can be searched, but it's no substitute for detail in the V.W EOS workshop manual and self study guides.
Hi mate, strange thing is, is that in basic settings mode the roof will open it if I program it to do so by entering the binarys manually.
So I'm not sure if they actually are locking back because, wouldn't the roof not open at all then?

Thanks again!
 

·
Registered
Joined
·
5,693 Posts
As aku suggested, this isn't the usual fault, although every EOS roof fault seems to have something different which is why there are very few simple 'Do this to fix it' solutions.

You have to tie up what the computer and roof controller tells you with what you can see actually happening physically? When you use diagnostics to manually control the roof, the sensors are either bypassed or the codes they send come from diagnostics and not from the sensors? That's how the service shop is able to move or access roof parts when a sensor or its wiring is faulty. Numbers on a computer screen don't always tell the truth or relate to the real events.

If the roof can complete a full open and close using guided fault finding, that suggests there is no issue with the pump, valves or controls. If there was an intermittent sensor wiring problem, that might explain your confusing results. What you have to do is get the locks into both states checked by looking at them either locked or unlocked, then use diagnostics to read back the sensor states, which should correlate with the physical position. It's unusual to have two sensors giving the wrong answer, but V.W do have some nasty wiring which splits up sensor ground return commons on different splices.

Before you head off into the unknown, search my posts for 'Gremlins' and have a good look at your roof controller plugs and wiring.

Fixing an EOS roof with a workshop manual, diagnostics and training should be no harder than repairing an automatic washing machine. :)
 

·
Registered
Joined
·
5 Posts
Discussion Starter #11
As aku suggested, this isn't the usual fault, although every EOS roof fault seems to have something different which is why there are very few simple 'Do this to fix it' solutions.

You have to tie up what the computer and roof controller tells you with what you can see actually happening physically? When you use diagnostics to manually control the roof, the sensors are either bypassed or the codes they send come from diagnostics and not from the sensors? That's how the service shop is able to move or access roof parts when a sensor or its wiring is faulty. Numbers on a computer screen don't always tell the truth or relate to the real events.

If the roof can complete a full open and close using guided fault finding, that suggests there is no issue with the pump, valves or controls. If there was an intermittent sensor wiring problem, that might explain your confusing results. What you have to do is get the locks into both states checked by looking at them either locked or unlocked, then use diagnostics to read back the sensor statse which should correlate with the physical position. It's unusual to have two sensors giving the wrong answer, but V.W do have some nasty wiring which splits up sensor ground return commons on different splices.

Before you head off into the unknown, search my posts for 'Gremlins' and have a good look at your roof controller plugs and wiring.

Fixing an EOS roof with a workshop manual, diagnostics and training should be no harder than repairing an automatic washing machine. :)
Thank you! I do appreciate the help. It’s honestly very much appreciated as it’s giving me heartache
 

·
Registered
Joined
·
967 Posts
That’s brave, using the basic settings to move the roof. Well done for doing so without breaking anything ;) and when you’re done fixing your problems, it would be great if you could write up how you did it to help others.

If you can tell the car through basic settings to raise the rear windscreen and it does so successfully, that suggests a very odd fault - either the sensors or claws are both going wrong in more or less the same way at the same time, or the roof controller really is telling them to move back.

Take a look at some of the other measuring blocks while you operate the roof with the roof switch - you may be able to see the state of the valves, or even what the roof controller thinks the switch is doing. I have no idea what extra data you will be able to see (it’s a bit late and rainy for me to pop out and test!) but any other data you can get could be more clues to what’s going on - or at least be able to rule some things out.

And welcome, by the way! I hope you’ll stick around after your baptism of fire with this issue ;)
 

·
Registered
Joined
·
5,693 Posts
I think his diagnostics sends the codes and these are received ok by the roof controller, but when the rear window frame is supposed to be locked or unlocked, both sensors aren't telling the truth? If a roof op. can be completed ok including the trunk lid swing out, the locks which are all linked together must be moving on and off which he can check by looking through the ski hole with a flashlight.

I can't imagine a mechanical disconnect fault which closes locks for the trunk lid swing out but keeps the rear window frame locks open, but hell, strange faults can arise and eyeballs are the best tool when the logic doesn't make sense. When the two rear hard flaps open and the rear window rises, he should be able to see the locking claw open then start to roll over and lock as the flaps close. From an earlier thread I replied to, those rear flaps are attached by pull wires to the cam connected to the locking claws. When the flaps are fully open the locking claws are right back and as the flaps close, the claws roll over to the locked position. As the flap is starting to close it should be possible to peek under and see the locking claw move. I paint the tips of all my locking claws with white correction fluid to see what they are doing. If he can't see under the flap, put some paint on the underside center of the open claws when the flaps are open, operate the roof then go back to the stage using the controller not diagnostics and when the flaps are open, see if paint has been transferred to the roller keeps. If 'yes' the locks are rolling over and doing what they should.

The exception is if the locks are only part closing on their tips due to insufficient hydraulic pressure or stiffness then springing back open? That would produce the change in sensor state shown by diagnostics, but I'd expect the flaps to behave oddly? I've moved those lock claws manually without hydraulic assistance and a surprising amount of force is needed to rotate them open or closed past their tips even moving over rollers. There is a critical point in the arc near the tip needing most force. There are no instructions to lubricate EOS roof hinges, but in view of my finding, I now put lithium grease on the inside of all my lock claws and their roller keeps.

I'm still puzzled by the 02085 general error because vcds which I'm more familiar with would tell you if there was a sensor malfunction and which sensor it was. I hope the original complaint that rear window frame locks or sensors were the problem is still correct?
 

·
Registered
Joined
·
967 Posts
02805 is the VW fault code that gets shown in VCDS, which just tells you "convertible top - malfunction". You would need a freeze frame of the measuring blocks to know what happened - the fault codes aren't sufficiently fine-grained to identify which sensor it is.
 

·
Registered
Joined
·
5,693 Posts
I took this example from a scan posted on the Ross-Tech Forum which is what I'm used to seeing after vcds has done an Auto Scan with roof sensor errors. Do your alternative diagnostics tools do the same?

Address 26: Auto Roof Labels: 1Q0-959-255.clb
Part No SW: 1Q0 959 255 B HW: 1Q0 959 255 B
Component: Verdecksteuergeraet 0622
Revision: 0032000K Serial number: 062820597
Coding: 0655362
Shop #: WSC 00020 000 00000
VCID: 3728236C23462D9F99F-8062

1 Fault Found:
03054 - Right Roof Pillar Lock Sensor (G559)
008 - Implausible Signal
Freeze Frame:
Fault Status: 01101000
Fault Priority: 5
Fault Frequency: 1
Reset counter: 210
Mileage: 91870 km
Time Indication: 0
Date: 2000.00.00
Time: 20:22:13

Freeze Frame:
Voltage: 14.10 V
Voltage: 14.20 V
Voltage: 4.90 V
Error Code: 0


That would send me straight to the sensor wired to the roof controller connector, test it there and look for a faulty sensor (unlikely) or most likely a connector or bad wiring splice in the loom. If I got the general 02805 error with no sensor faults found, I might be looking for more physical problems that could cause the roof to abort its operation. At this level diagnostics only sees 02805 as an 'unexpected' failed roof op. but doesn't say it could be due to parts being out of sync. Another example is when sequence timing doesn't fit the program profile - a roof part gets jammed and doesn't complete in the time window. The sensors are all good & telling the truth, but the roof part didn't get to the next position and trip sensors inside the time window - output the general 02805 fault code.

The O.P can easily test this with his 'alternative' diagnostics. When the large front flap is fully open, reach down inside and disconnect the flap sensor. Clear the codes, ignition off and on then rescan. The flap sensor fault should come up straight away. Remember to reconnect the plug!

If sensors are out of sequence e.g a lock has opened unexpectedly after closing, I wouldn't expect a sensor fault, it's telling the truth but because the roof op. has a sequence error at that point, it will abort.
 

·
Registered
Joined
·
967 Posts
Ah, good point - "implausible signal" (when it's outside the range for a 0 or a 1) from a sensor does get a different fault code. But a sensor returning a 1 when it should be a 0, or not changing state when it should, will just return 02805.

OP has VCDS - and my OBDeleven would report the same codes as VCDS. I think your suggestion of peering through the ski hatch with a flashlight to see what the latches are doing is the next step though.
 

·
Registered
Joined
·
5,693 Posts
I don't know how a Hall sensor can return a 1 instead of a 0 or vice versa other than being close to the magnet or steel and correctly saying it has moved? The system at key on looks for a logic level in a valid range. This is how diagnostics puts up a sensor fault when sensor wiring is faulty or the sensor itself is shorted to ground, + 5 Volt or open circuit. The 'key on proving' expects a sensor voltage to be about 0.8V or 4.3 volts for logic low and high. The controller checks all roof sensor output states when the roof is open or closed. An immediate dash warning is given if they are wrong.

The OPs roof starts to open so we can assume the sensors all proved o.k at the beginning. The Hall sensor output is changing between logic levels and he has 2 doing the same thing so I think the sensors are telling the truth. There is no electronic explanation for 2 sensors giving inverted logic? I accept a faulty sensor could 'stick' at logic high or low, but these both change state? There is another possibility. The sensor loom wiring is bad news and if water has got on to the cloth loom it might bridge the splice taps to an adjacent voltage source?

From memory, each Hall sensor has a fairly high open collector internal pull up resistor for logic 1 and cross contamination of the wiring joint in the loom could take it low? In my big note on testing the roof with a multimeter I included Megohm insulation tests on each pair of sensor wires whilst disconnected from the controller, to other sensor wires, battery + and ground for that reason. But this may be too much for average DIYers to understand?

My bet is the OP has focussed attention on sensor states from information given by diagnostics and jumps to conclusions about sensors when he needs to look at the mechanics. He can just about peer under the rear flaps and inside the trunk with a flashlight. He could even temporarily remove a flap cover and pull wire to get a better view on the rear window frame lock.

I'm not sure if this has been confirmed, but the rear window frame can get out of sync and ends up sitting on top of the locking claws. It sits level sort of looking 'normal' but the give away is a large gap between the bottom of the glass/frame and the silver trim. On a closed roof, if the trunk can be opened, the frame locks should be closed and if two pairs of hands can lift the rear window frame up, the frame isn't sat inside the locks!

The roof cannot recover from this on its own and the locking rams or claws have to be manually pushed back to get the frame sat inside the locks, then closed. This fault scenario concerns me when 'guided fault finding' is used and bypasses roof controller protection. I thought one option with VAS was to open and close the locks without doing anything else? If the OP can do that, he can look at what the locks actually do and check the sensor status.
 

·
Registered
Joined
·
5,693 Posts
Now I think about it, these are the claws that I've observed can slowly close if you leave the roof part open with the rear windscreen on top of the roof, and they need a lot of force to open again. I don't know whether hydraulic fluid would need to be sent in the opposite direction or whether this could be caused by a leak or a sticking valve.

Having taken another look at the SSP it seems like one of the shuttle valves may be stuck - after the first step of opening the roof, which releases these claws, the roof pump then reverses direction. If the hydraulic system remained in the same configuration as it started in due to a valve not moving the way it should, the claws that just opened would close again. I don't know if it's plausible for a shuttle valve to stick in that way though.

As far as I know there's no logic to automatically reverse a roof operation if something goes wrong - it'll just stop, and it may let you reverse it if it's in a state where that's permissible.
That's a significant comment aku. We don't know what the OP was doing with the roof before getting error 02805. Either nothing at all and a one off fault event or trying to DIY a roof service or fix something with the roof open a long time and no props? I have also noticed those claws can close when the rear window frame is sat on the roof and allowed to roll back. It seems to occur when the ignition is off and the hydraulics go into a relaxed state. Some have had a locking claw close on one side but not the other. I've had both closed with the frame sat on the top. One thing's certain, if the locks are open you can't open the trunk!
 

·
Registered
Joined
·
967 Posts
The roof cannot recover from this on its own and the locking rams or claws have to be manually pushed back to get the frame sat inside the locks, then closed. This fault scenario concerns me when 'guided fault finding' is used and bypasses roof controller protection. I thought one option with VAS was to open and close the locks without doing anything else? If the OP can do that, he can look at what the locks actually do and check the sensor status.
I don't know - but since OP has VCDS and has used the basic settings to move the bits of the roof (like how VAS does it), he will be able to open or close the locks.
 

·
Registered
Joined
·
5,693 Posts
VCDS and VAS AFIK follows each roof step strictly in sequence. You cannot just jump into a step e.g unlock/lock roof locks without completing the first steps to unlock, and re lock is always at the end when the roof op. is nearly finished. My vcds only does the first step of opening the sunroof and I've not found out yet why it goes no further through basic settings?

On the Ross-Tech site they are absolutely correct to hide this feature, publish warnings and require a code to execute each step in sequence. As I said, I can see how an out of sync roof with error code 02805 and parts sitting outside the locks could be seriously damaged clicking steps on a laptop when sensors normally providing roof protection warnings are bypassed.

I've thought for a while that we could do with discovering some useful CANbus PID codes. It's not hard to build a simple diagnostics interface and the most useful pair of codes that shouldn't cause damage would be setting the locks on an off. It could also be done by disconnecting the pump from the roof controller, Power the correct solenoid valve and externally power the pump in the correct direction. If the trunk lid is kept raised to make these temporary connections, there is no risk of it being locked out and you can confirm the roof lock status just by looking at the corner locks at the back of the trunk.

I might try this out if I have an idle moment because it would be a lot easier to do a manual roof open or close on a locked out roof with a good pump, than removing the hinge covers and pushing hard on the lock rams. For these experiments I finding it difficult getting the right sex connectors e.g for the pump and solenoid valves. It gets too messy making solder splices on the OE wiring.
 
1 - 20 of 20 Posts
Top