E89 roof issue checks question on microswitches

E

Ecc33

Member
 Scotland, UK
So I have had some roof issues.

It was intermittant but now pretty much in fail mode now.

The roof works every thing , bar :

1. when fully open, the car does not beleive it is done, and I get red LED on the roof open button, although mechanically everything is in position.
2. when it was working, and I could close the roof, it also did not beleive it was fully done, but it came and went, the troubleome aspect being I couldn't operate the side windows.

However,

now i have the main issue which is I can't close the roof, it does everything until the double stack roof+window has to move the window back to the rear.
1. boot opens
2.double stack moves to roof position
3. boot closes
4. whirring, but no movement of window back down.


I have renewed the wires visible on the main hinge both sides. Checkking the replaced bit of wire (x6) showed the removed bits all seemed to have continuity.

I check the connections for the hall sensor in the boot ram, and the 2x microswitches.

I read these should be 2.7ohm when pressed, about 500 ohm when open. I'm getting 2.7 regardless, on both sides.

The internet seems to think this is very rare and unlikely.

It's £145 quid x2 from bmw for the switches -- if they dont read 500 then both a knacked, worth getting the replacements ?
 
The switch readings are almost certainly a measurement problem, not two dead switches. These are resistance-coded switches — the module distinguishes closed (~low Ω), open (~defined resistance), and broken wire (infinite) precisely so it can detect harness faults. Two switches on opposite sides failing in the identical mode simultaneously is vanishingly improbable. Far more likely: he’s measuring in-circuit, so the meter is seeing a parallel path back through the harness and module. The switches need to be unplugged and measured in isolation, with the plunger actuated by hand, before anyone spends £290.





The better diagnostic is live data, not an ohmmeter. ISTA (or Protool against the convertible top module) will display every microswitch and Hall sensor state in real time. Cycle the roof, watch which position confirmation fails to arrive at the halt point, and you’ve localised the fault in minutes. The stored fault codes will usually name the specific sensor too. This should be step one, not the multimeter.





The failure signature itself points two ways. The history — mechanically complete but “not believed” open or closed, windows inhibited — is classic position-feedback loss (sensor or wiring), and the hinge-loom repair was the right instinct; the boot-lid loom is the other known break point. But “whirring with no movement” at the rear-window stage is different: a motor running against a decoupled drive suggests a stripped gear or drive cable, which no switch replacement will fix. Worth also checking battery voltage on charge during cycling — roof logic throws phantom position faults on marginal supply.





So the answer to his actual question: no, don’t buy switches on that evidence. Isolate and re-measure, then read live states, then decide.
 
B21 said:
The switch readings are almost certainly a measurement problem, not two dead switches. These are resistance-coded switches — the module distinguishes closed (~low Ω), open (~defined resistance), and broken wire (infinite) precisely so it can detect harness faults. Two switches on opposite sides failing in the identical mode simultaneously is vanishingly improbable. Far more likely: he’s measuring in-circuit, so the meter is seeing a parallel path back through the harness and module. The switches need to be unplugged and measured in isolation, with the plunger actuated by hand, before anyone spends £290.





The better diagnostic is live data, not an ohmmeter. ISTA (or Protool against the convertible top module) will display every microswitch and Hall sensor state in real time. Cycle the roof, watch which position confirmation fails to arrive at the halt point, and you’ve localised the fault in minutes. The stored fault codes will usually name the specific sensor too. This should be step one, not the multimeter.





The failure signature itself points two ways. The history — mechanically complete but “not believed” open or closed, windows inhibited — is classic position-feedback loss (sensor or wiring), and the hinge-loom repair was the right instinct; the boot-lid loom is the other known break point. But “whirring with no movement” at the rear-window stage is different: a motor running against a decoupled drive suggests a stripped gear or drive cable, which no switch replacement will fix. Worth also checking battery voltage on charge during cycling — roof logic throws phantom position faults on marginal supply.





So the answer to his actual question: no, don’t buy switches on that evidence. Isolate and re-measure, then read live states, then decide.
Click to expand...
thanks for the message.
i will respond to each point:
> Far more likely: he’s measuring in-circuit, so the meter is seeing a parallel path back through the harness and module.

I unplug the inline plug (it's white on the right, and black on the left side of the car)
put a meter on the pins of the plug on the switch
read 2.7K
activate the switch with a small pick, i can clearly hear the switch clicking on and off
reading remains a 2.7K regardless

this is wrong method/reading/technique ?


to get live data what do i need to buy?. i dont have any reader right now.

> the boot-lid loom is
I will check this, where is best to access this ?

> But “whirring with no movement” at the rear-window stage
This suggested the hydraulic pump is pumping but no pressure due to some sensor not being in the correct state. My guess was the boot not thinking it's closed so it wont lower the glass.

I thought the boot state sensors were these 2x microswitches
 
Most BMW specfic app like Protool / ISTA and even Carly will do..plus a few dedicated units like Foxwell

Without those devices you're just flying blind..
 
AS B21 said, I think Bimmercode is for coding and BimmerLink is for Diagnostics. Personally, BimmerGeeks ProTool is better. It's tied to your google account. You can open and use it on all your Android phones and tablets. Bimmercode/BimmerLink is tied to your current device, if you get a new phone you have to purchase it again.
 
Last edited:
  • Like
Reactions: B21
B21 said:
The switch readings are almost certainly a measurement problem, not two dead switches.
Click to expand...
In this special case, why not?

B21 said:
So the answer to his actual question: no, don’t buy switches on that evidence.
Click to expand...
My recommendation is a clear YES: buy 2 new microswitches (54377228585/54377228586) plus 2 new salmon relais (12631742690 ) for the hydraulic pump, if not yet renewed.
 
Last edited:
Yes, indeed, in this special case I wouldn't rule that out.
That's the reason for my clear recommendation to get it replaced each decade and before it fails.

It will be interesting to see, however, whether it will still be available in the next decade—pure gold for those who still own a new pair then.
 
  • Like
Reactions: B21
RobbiZ4 said:
Correct is (2.200 + 560=2.760) ohms and 560 ohms, both only if unplugged.
Click to expand...
Ok. So 2.7 is correct as they are unplugged, regardless of open/closed ?


I will get bimmer link i see bimmer code does not do what i need. Looks like bimmer code can transfer to new phones just not between android / ios. i dont have android so the pro geek tool isnt so useful.


I saw that link ⌛ E89: Collection of the typical age-related defects in the roof environment and was working through the items. Have done most of the points. Should have obd by the weekend. Just not sure whether to get the 585/586 switches now or wait til after i;ve read the codes. Your post suggest defo get the switches now. so maybe I'll phone bmw back.
I just ordered them . £140.30 each from bmw

salmon relays are on order. I also swapped the ones I have, operations were the same i.e no change when relayed were position switched.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom