Volkswagen Eos Forum banner

1 - 20 of 30 Posts

·
Registered
Joined
·
22 Posts
Discussion Starter #1
Hi
Last year at the end of summer we had a problem with the roof of our EOS, the roof initially was very slow to open/fold away then it stopped completely, I did a bit of diagnostics but without any access to VSDC I didn't get very far, (I did start to put together a board with LED's to hook into the roof control module to read the I/O but that did get finished) I found that sunroof would open and the hyd pump would start, nothing would move then the pump would shut down. I checked the voltage for all three solenoids, which reading through the diagnostic paperwork I have, they should all fire together when the roof opens. Well they all had voltage on them when the roof tries to open. That was about as far as I got back then.. Oh, I found a replacement pump on Ebay in the winter with a view to changing it in the summer the following year.

Well its now the following year, and the problem still remains, I swapped out the pump but still the same fault, I've now got myself a VCDS (horray!) And looking at the measuring blocks I have all the right 'ones and noughts' in all the right places with the roof in the fully up state, (I have been watching the excellent video on YouTube by RossTechVCDS) but again nothing changes when the roof is operated. I think first I should see bit 3 and 4 change state at the beginning when the rear screen swings up.

Any ideas as to what I should look at next..
 

·
Registered
Joined
·
5,627 Posts
I think you said that both windows are down and the sunroof is right back for the start of roof op.? The next stage is hydraulic although I think the trunk lid is opened then - trunk lid lock and pull up/down are part of a roof op.? The two silver rams either side of the hinge should have moved first to unlock the roof, allowing hydraulics to move the parts and raise up the rear window. Of all the hydraulics, the rear window requires the least effort. You haven't said anything about the trunk lid locking you out? If the roof unlock was working at the first stage and the roof op. aborted, that's the state I would expect it to be in - roof unlocked and the trunk lid corner 'C' claw locks holding it shut.

If the roof 'C' locks are closed, no roof parts will move. vcds only shows the state of roof sensors when the parts positions have changed and if nothing has moved hydraulically, then it shows them all o.k for a closed roof and ready? You said roof action was slow before, I assume your battery is good, but it's best to always operate the roof with the engine running anyway. If you changed out the roof pump that still leaves the roof controller. This uses a (30A?) mosfet bridge to drive the pump forwards and in reverse. If one mosfet in the bridge has failed, your pump may not be getting the full current. You can easily check that by hanging a voltmeter on the pump and checking its voltage when it runs even briefly during a roof op. which should be around 12V. If it's much less than that, then have a good look first at the wiring and high high current pump connectors (in fact all of them!) on the roof controller. If water has got in them that will be part of your problem. Also pull each 50 Amp fuse in the engine bay fusebox and check their contacts are clean. I think you can work out where I'm going - the low level control may be working, but there's not enough current getting to the pump and that won't be measured by vcds The roof controller does have a built in watchdog timer and if the first hydraulic stage doesn't happen (sensors change state) , it times out.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #3
Thank for that Voxmagna, gives me something to look at. Yes all the windows drop down and the sunroof goes fully back, you can hear the pump start near the end of the sunroof travel, it sounds a bit strained, I'm thinking the pressure reducing valve is reliving due to nothing moving, and then shuts down or times out because nothing has moved. I can check the pump supply and also the solenoid valves again whilst I'm there. The boot lid opens and closes ok in normal operation (access to the boot) and you can hear it lock again after its shut.
 

·
Registered
Joined
·
5,627 Posts
It would be bad luck to have both pumps with the same problem? The trunk lock and lifter has a second control coming from the roof controller during a roof op. Try opening the pump manual relief valve 2 turns and repeat the roof op. The roof won't do anything after sunroof open just as now, but at least you should hear the pump motor running normally without a load as the fluid is bypassed.

Be careful because if you do get a some operation from the hydraulics then abort, you could be left with the trunk lid locks on and it's a crazy problem to get out of jail.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #5
Tried current and voltage reading today of the pump motor. I'm not sure what the current draw should be but it doesn't seem that high for the cable size that feeds it, I've made a short video clip that's on youtube.

There's also one from September last year when I started investigations, this one is the voltage readings of the solenoid coils. This one has the original roof motor.

I wasn't brave enough to open the manual relief value :)
 

·
Registered
Joined
·
5,627 Posts
I wasn't brave enough to open the manual relief value
Why not? If you are going to get any further and try moving roof parts to see if they are jammed, the relief valve must be open. It's there for that reason. You aren't going to get fluid squirting all over the place, it should allow fluid to bypass the rams with the pump effectively run off load. No roof parts should be moving.

I can't make much sense of the left meter if that's current? It seems very low, but different rams and roof parts need more pressure (current) to move them. What's strange is the voltage drop to 9V without a large current, which makes me think there's resistance either in the wiring or the 'H' mosfet bridge driver? Another curious thing is you have the engine running (good) and I would have expected to see the voltage up to about 14v if the battery was being charged? In fact it should be pretty close to the voltage measured at the battery terminals. Having told you they use a high current H bridge driver in the controller, you should check the pump voltage if it can run in either direction. If for example the pump voltage is 12V or higher running in one direction, but only 9V running in reverse, There could be a faulty chip in one side of the H bridge.

The first roof open stage is the unlock and rear window lift which shouldn't demand much pressure/current. The trunk lid and parcel shelf swing out demands a lot more and the main hinger operation for the side bars even more. The hydraulics are moving VERY heavy parts and the initial forces at the start of the quadrant are extremely high until the rams and parts get to their mid (high) points and move down under their own weight. I modified pairs of spring compressors to manually do the job of the hydraulics without the pump and the torque needed on the 20mm screws was a lot more than I expected. So much so that I only used them to move parts about 10% to give more working space. This explains why V.W advise using 2 people to move parts!

I don't know how much deeper you want to go? If you want to limit the problem of getting the trunk lid locked you can do 2 things: 1.) Temporarily remove both chrome hoops in each corner of the lid (2 x Torx each). and 2) Drill a 4mm hole through the upper lock plastic cover so a pin can be pushed in to hold the roll over claw open. Then, provided you don't try and do a complete roof open, you can part open the boot lid and see what happens in the early stages with the plastic covers over the hinges removed. The two silver rams should move fully forwards first to unlock everything, which is when the trunk lid is normally locked down at the back and those 'C' claws should be closed. But you removed the lock hoops so that won't happen. With the boot lid part open normally only a few inches to look, I don't think there are issues watching the first stage and rams when the rear window frame comes up, but if the rest of the package starts moving, the front edge of the lid must be pulled back or it will obstruct.

I've uploaded 6 jpeg attachments for a roof open sequence which I cobbled together from different sources. Each stage file shows the pump status, its direction according to voltage, each solenoid valve whether active or not and what position the rams should be in. I would expect the pump voltage to hold close to battery voltage, whatever it was doing. Opening the 'C' claw locks for all the roof parts needs quite a lot of ram pressure. I struggled to do it manually by pushing the ram pistons forwards. Each roof part lock is a very substantial steel 'C' claw rolling over a 'keep' metal to metal and with down large force holding the parts solid against buffer stops. They are the one part of the hinges that should be greased if you can find them, but V.W say hinges don't need lube!

"There's no such thing as an EOS roof quick fix"
 

Attachments

·
Registered
Joined
·
22 Posts
Discussion Starter #7
I'll open the relief valve tomorrow and report back, Unfortunately digital meters are not the best at monitoring changing values. I do have a 10 amp analogue meter somewhere I'll dig it out. Yes, the amp readings I think are a bit low only around 1 amp then peaking at 8 for a very short period. I'm wondering if this is pointing towards a failure in the driver circuit. I could hook up an analogue meter for the amp draw and monitor the volts on the 3 solenoids, after that I'm running out of meters.

Another thought, the plug for the pump is not the original plug, it's been connected into the loom a few inches back. Not by me I'll add. I'll have that apart and check that connection, it's all taped up.

Now that I have the original pump out I'll give that a spin up in both directions, I could probably do that to the installed one with the relief screw open and just feed it with 12v in either direction.

Can the controller be taken apart ? are the motor FET's replaceable.
 

·
Registered
Joined
·
5,627 Posts
The controller board is easy to remove with care. I took mine out to give it some extra coats of acrylic lacquer. I also bought some spare mosfets just in case as it seemed to me that any future problems would not be the low level interface devices but the high current motor or solenoid drivers. But they are d2PAK if I remember and their heatsinks are soldered direct to the board during reflow soldering. Reworking any of them will be tricky needing a lot of careful desoldering heat, because there are plenty of small SMD's close by that might desolder (!) at the same time. If you are any good at hacking games consoles and working with ball grid array chips, you should have the right skills. I think the polarity was correct in your vid. but make sure the wiring back to the roof controller connector is correct, because the most obvious fault would be they got the polarity crossed over and the pump is running the wrong way when the roof cycle starts. I think even 10 Amp is a bit low. They fuse the roof electrics with a 50 Amp fuse. I've already measured just over 10 Amps peak for the sunroof motor and my meter was over range. Very few multimeters are more than 10A before blowing their internal fuse. I would expect up to 30Amp for the roof pump on full load but sorry I can't confirm. Others testing their pumps off the car on a battery use a 30A safety fuse in series and report a large crack when making the contact.

I think the only way to confirm the drivers is by voltage measurement at the pump connector. The controller doesn't do anything smart like PWM. It just puts full battery voltage in either polarity across the pump less some small voltage for 2 mosfet on resistance in the bridge. If you want an easier way to confirm the pump current, measure current at the big 50A fuse in the engine bay fusebox with a bigger ammeter or shunted multimeter or clamp current probe. I haven't got the fuse number handy but its the 4th and last fuse down from the rear on the left. You will get the sunroof motor draw first (7-10A) followed by mostly the pump current as the controller and solenoids are much lower current draws.

I just had a crude idea: You have 2 multimeters, set them both on 10A range and common the probes together really firm in a 30A choc terminal block with a pair of 20 amp wires coming out. You should see the same current reading on both meters and then X 2 the reading. Internal BUSS meter fuses are quite expensive to replace.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #9
Further investigations, I hooked an analogue meter to the pump with a voltmeter, the amp draw is in excess of 10 amps, a better reading than the digital meter which I think was autoranging and couldn't keep up. I also cracked open the relief valve but nothing really changed..


I had another rumage round and found an very old -60_0_+60 ammeter, no idea how accurate it is but I put it across the fuse for the motor then rested the meter on the windscreen, as you'll see it peaks at around 30 amps.

 

·
Registered
Joined
·
5,627 Posts
Current accuracy per se isn't the issue as long as current is in the right ball park and 30A after the sunroof stops moving is what I'd expect. But you haven't confirmed the voltages on the pump are always 12.5volts ish and the polarity/direction is correct? You now have diagrams for the solenoid valves to check against the first stage of roof open. You haven't explained the history of both roof pumps? Was the second pump reconditioned or a used pump without confirmation it was good?

Although you can measure voltages on the solenoid valves, it doesn't mean their coils are ok unless you check their resistances are all the same. Also, even though their resistances are o.k and they take voltage with current, it doesn't mean their plungers are operating the pump valves and diagnostics won't report that error. Seized up solenoid valves is a common failure after pumps get saturated with water and the bottom solenoid is most vulnerable, being at the lowest point.

I haven't tried this but I think if you connected 12V correct polarity to each solenoid in turn (loom plug disconnected!), you should hear the same loud 'click' for each if their plungers are working. If nothing is heard on any, I'd suspect the plunger is seized solid and not operating the internal valve?

If after these checks when the pump electrics look and feel good I would move on to possible hydro-mechanical causes and check the locking rams are actually trying to move to the unlock state. You haven't said the trunk lid has locked you out yet which makes me think the roof locks aren't coming off? If the roof locks were unlocked just after the sunroof stops retracting (pump starts), you should not be able to open the trunk lid normally. Careful, I advise you to temporarily remove the chrome hoops in each corner of the lid! I can only suspect if the pump and electrics are good, your roof locks are held seized and needing too much pump force to release them. The pump usually works hardest and makes most noise later on when the heavy roof parts are being moved. If you monitor the roof status bits with vcds and don't see the roof lock bits change, that further confirms my suspicion.

The reading delay with electronic multimeters catches many out, although some have a faster bargraph and peak value reading features. I have a battery powered quad opamp/comparator driving 4 leds with pots on the reference inputs which I can preset to any voltage above which the led lights. Much faster to see and video than a sluggish dmm. One opamp has its led tied high to turn on for a minimum voltage value - handy to determine minimum battery voltage when cranking
 

·
Registered
Joined
·
22 Posts
Discussion Starter #11
Today I stripped out the boot interior trim to get a look at whats going on, I stuck a camera inside with some light and filmed the sequence up by the boot lid whilst trying to open the roof. Now i thought that when the boot was shut the hooks at the top of the boot lid locked in place to hold the boot when it swings open but in this case they don't, should they ? I always thought that was the noise you hear when the boot lid shuts and pulls down tight, but obviously not.

Boot LH side

Boot RH side
 

·
Registered
Joined
·
5,627 Posts
I think you are getting confused about how it works. The diagrams I uploaded show there are 4 pairs of rams and you need to understand where they are physically and what each pair does. The pair you have photo'd are the trunk lid opening rams and have nothing to do with the roof locks. They don't operate until the trunk lid swing out stage, which from memory is about when the rear window frame is lifted. But at the moment I think the locks are holding it and it won't go anywhere. The initial noise you recorded ("The Whirr') is the normal trunk lid motor pull up/down, but the horrible grinding noise afterwards is the pump pushing or pulling at something it cannot move.

Now i thought that when the boot was shut the hooks at the top of the boot lid locked in place to hold the boot when it swings open but in this case they don't, should they
That's not quite true. The corner 'C' locks are only on when the main roof locks are off and vice-versa. When a roof op. is complete the last operation puts the roof locks on with a ram and unlocks the trunk lid 'C' claws. The center lock in the trunk is for normal lid locking and works seperately although it does get an unlock signal from the roof controller whilst simultaneously activating the pull up motor. Anywhere in-between roof opstages or if a roof op. stops/faults mid way and your trunk lid corner 'C' locks and may be also the centre lock will lock you out.

But I'm afraid you are looking at the wrong rams! I already said you had to remove the plastic covers over each hinge fitted in 3 parts to expose the rams which do the locking/unlocking - they also have a silver body! When those ram shafts are fully extended, the roof locks are on and the trunk lock is off because they are mechanically linked, kind of like a bistable flip-flop. If your pump is pushing fluid in the right direction (?), those rams should be pulling back (or was it forwards?) and the trunk lid claws will also roll over and lock at the back. If you get this to happen at stage 1 and the roof op. aborts on another fault or doesn't finish, you will be locked out of the trunk unless you did what I suggested.

I haven't had to do it but each ram is fed with a pair of snap in lines. You must know about those if you replaced the pump and I assume you were extremely careful to get them all back in the correct holes?? The pump body plate is marked as well as the lines. It's a long shot you do carefully which I've never had to do, but try disconnecting one line that feeds the locking rams (they are all numbered each end) and have a clean pot underneath to catch fluid. Have somebody else hold the roof op. button ready to stop as soon as the pump starts or make the pump plug easy to pull and stop it. If fluid squirts out you know its going to the right place. If not it must be going to the wrong place? Check the fluid reservoir level afterwards.

Great video but sorry, wrong view.

If you are confident the electrics and pump hydraulics are o.k with no silly plumbing errors or mistakes. I think you should read the workshop roof opening procedure and try to open it manually. I'm pretty sure when you do the first step of unlocking your roof by pushing on the 2 lock rams, you will find something is jammed and stopping them from moving. Be warned: If you start operating the roof electrically without getting parts you moved back to the same start position i.e roof fully closed and locked, you risk having the roof out of sync with where the controller expected parts to be and another problem to fix. You are now getting deep into the area of a possible trunk lid lockout unless you take precautions. We have a graveyard here of EOS owners who have been locked out of their trunks and there's no easy solution without jailbreak mods.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #13
You mean these little pistons :)
22341


Oh er, there in an difficult place, Rather than taking stuff apart that I may never get back together again i'll dig out a small camera and some mirrors :)

When I swapped over the hydraulic lines the right hand side had different numbers from the pipes that were on the replacement pump, they had just cut the lines off so it left the numbering on the old ones. all the lines were different by 2 digits, so I had numbers like 34,13,33 and the numbers on the car were 32,11,31. I Put the lines in the same place as it came out of the block. The writing on the block was only legible on the top few ports.
 

·
Registered
Joined
·
5,627 Posts
When I swapped over the hydraulic lines the right hand side had different numbers from the pipes that were on the replacement pump, they had just cut the lines off so it left the numbering on the old ones. all the lines were different by 2 digits, so I had numbers like 34,13,33 and the numbers on the car were 32,11,31. I Put the lines in the same place as it came out of the block. The writing on the block was only legible on the top few ports.
Rather than trust what might have been there when you had roof problems before, why not recheck your work, not just looking at numbers but try to identify where the pipes should be going what makes you think your second pump has the fluid ports in the block in the same physical position? Are both pumps exactly the same EOS part numbers down to the last letter?

I don't understand how you can do this work without a workshop manual with 6500 pages? YouTube videos won't help you for this level of detail. Manual Section 18.7.1 covers the hydraulic pump connection assignment and you shouldn't guess at what you think is right. V.W rarely get their data and numbers wrong! The hoses on the car should correspond to the numbers on the pump block and you could have dropped a big booboo there causing all your problems?

The workshop manual refers to a changed numbering scheme for pump connection numbering and gives crossover information for a pump supplied as a genuine replacement part, which I assume could mean pumps newer than yours? .....'This number on the car means this number on the pump block!" Never assume anything goes back in the same position. If they made the fluid block different on newer pumps you could get hoses in the wrong place.

I'll help you out with a PM and some info, but you should really think about buying a manual or getting ELSAWIN. There's some recent info here on the site. I can't do much more until you have first checked your fluid lines and their connections on the pump according to the latest information for your pump. That noise from your pump sounded pretty horrible and I hope if you do find hydraulic lines are wrongly connected, that you haven't internally damaged your pump?
 

·
Registered
Joined
·
3 Posts
I have read this whole thread and with two EOS, including one that has had a couple of roof problems I think I need to contribute at this point. The problem being experienced may be one that we had. There are four swivelling latches (claws) holding the boot lid down and two that lock the rear window down. At the point where you seem to be having a problem these latches (claws) are supposed to be turned to the open position by the hydraulic rams. Our problem was that one latch was disconnected from the ram and stayed in the locked position - despite a straining effort the roof pump could not open and flip out the boot lid. Check the photo - if the pivot bolt with the white spot of paint comes loose then the lug on the hydraulic ram arm where my finer is will disengage and the latch (claw) will never turn to the open position. I just had to re-engage the latch (claw) to the stud on the ram arm and then tighten the pivot bolt. You need to check all six latches (claws) are operational. The two for the rear windows can be found in the back of the fish tank and can be checked with the boot open normally when the trim that blocks off the fish tank is removed. The front boot lid latches (claws) are in that area as well - you should be able to get a finger on them when the boot is open and confirm they are not hanging loose. Search in the Facebook Volkswagen EOS Owners group for my post of 19th January for more details. Good Luck
Latch.jpg
 

·
Registered
Joined
·
3 Posts
This is the left rear window latch as seen from the boot sitting above the start of the fish tank. The pivot bolt also drives the bowden cable that flips open the small hatch lids above this latch.
RearWindowLatch.jpg
 

·
Registered
Joined
·
5,627 Posts
We exchanged some PMs after he got locked out of his trunk. Eventually despite my warning, the lesson was learned and he removed the 2 chrome hoops on the trunk lid so those won't be jamming. I agree that a mechanical failure with the latches further forwards could be an issue and already suggested trying a manual roof open, because if the rams releasing the latches can't be moved and tested open by moving parts manually, something is wrong with them. As you say, all the latches should work togther, but I think the OP is still looking for a problem in the pump? Although when I heard the noise the pump was making at that first unlocking stage (which is a relatively small ram), it seemed to me the pump was working hard against a jammed unlock part?

The OP seems to be in quite deep and will need time to solve his problem, so any contributions will be helpful.
Now you've been that far into the locking claws, you will probably agree with me that they should be well greased, despite no regular maintenance instruction for them? The metal to metal friction on them is strange: Even though some of them come up against rollers on a plain steel pin, the claws need a lot of effort until they get about half way around and there is considerable friction or forces acting at the wrong angles? Especially at the start, where the hinge is held down against the buffer springs holding it solid to stop it rattling over road bumps.
 

·
Registered
Joined
·
3 Posts
On the Facebook Volkswagen EOS Owners group there is a recent post by a guy who bought a car with a non operational roof. Eventually he decided to fix it. The problem was very similar to what is being discussed here and in the end he found that two of the hydraulic lines connected to the pump had been swapped. The pump was never going to open the roof if it was pushing the oil into the wrong hydraulic lines. The OP should be checking this too because I think there was a mention of the roof pump being swapped.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #19
Hi Guys, thanks for all you input, I've left the car alone for a few days as it was getting frustrating. As Voxmagna said after not listening to him about removing the chrome hoops I eventually locked myself out of the boot. I think it was because after moving the pistons I hadn't pushed them fully home. Got it open with a bit of keyhole surgery through the ski hatch.. The hoops have been taken off. The pistons in the boot opening framework both move quite freely when the pressure is released, they don't move at all when asked to when trying to open the roof with the hydraulics . Thanks for the pictures Catweazle, I think I've looked at the locks lurking up there. I'll have another look.
 

·
Registered
Joined
·
22 Posts
Discussion Starter #20
On the Facebook Volkswagen EOS Owners group there is a recent post by a guy who bought a car with a non operational roof. Eventually he decided to fix it. The problem was very similar to what is being discussed here and in the end he found that two of the hydraulic lines connected to the pump had been swapped. The pump was never going to open the roof if it was pushing the oil into the wrong hydraulic lines. The OP should be checking this too because I think there was a mention of the roof pump being swapped.
The fault started before I'd taken any pipes off, it was the original pump. The fault was exactly the same after I'd swapped the pump over to another one.
 
1 - 20 of 30 Posts
Top