Boston area with DIS? Problem with clutch bleed adaptation on SMG

[eacmen]

That's not great news considering I used their seals.

I'm about ready to get a used full pump unit off eBay TBH.


Sent from my iPhone using Tapatalk

My car is a track car so I always have a spare full pump unit with me. The pump is relocated so it's easy to swap. Right now I have an OE used unit in there and my spare unit is a rebuilt unit from MLR.

I'm honestly tempted next time to just bite the bullet and buy a new OEM unit from FCP with the warranty.

[trigtm]

The unit I had an issue with was a rebuilt unit with MLR seals FWIW

That's not great news considering I used their seals.

I'm about ready to get a used full pump unit off eBay TBH.


Sent from my iPhone using Tapatalk

[eacmen]

Interesting. I've replaced all the seals in the pump unit itself and torqued everything to spec. Hopefully a seal isn't bad or installed incorrectly? I went through it several times to make sure it was right.

The unit I had an issue with was a rebuilt unit with MLR seals FWIW

[trigtm]

High pitched squeal or squeak while bleeding was an early warning sign that the internal seals in my pump were going bad.

Try elevating the front of the car while bleeding, there might be a trapped air pocket.

Interesting. I've replaced all the seals in the pump unit itself and torqued everything to spec. Hopefully a seal isn't bad or installed incorrectly? I went through it several times to make sure it was right.

[eacmen]

High pitched squeal or squeak while bleeding was an early warning sign that the internal seals in my pump were going bad.

Try elevating the front of the car while bleeding, there might be a trapped air pocket.

[trigtm]

You are in unchartered territory, but at least SMG ECUs are cheap. It certainly isn't a common failure. Have you tested the accumulator? It can be tested with DIS. I do not think this is the issue but one more thing to cross off the list since you've haven't found a resolution yet.

I was just about to post about the accumulator actually since I'm running out of things to test TBH. I've done a bunch more bleed procedures on the actuator and clutch slave with 0 changes. I was able to check the codes using the Schwaben BMW tool and pulled SMG codes 55 and 58. Not much detail on code 58 on forums but 55 seems to be pretty general with a fringe case of the accumulator. I then tested the accum and pulled a measured pressure of 27.58 bar when the target range is 29-41 bar. Also when bleeding, I geta high pitched squeal or squeak (hard to describe but I would equate it to air leak maybe). I think my next course is to replace the accumulator and the pressure switch. Another week down...

[oceansize]

So short of reverting the tune, I've now run three different software solutions and have the same results in all (INPA, ISTA and the Foxwell BMW tool). I'm leaning towards a wiring fault now so I'll need to trace the slave and the wiring on the pump (pump comes back out).

If all that traces out OK, I guess it's the SMG computer after that?

You are in unchartered territory, but at least SMG ECUs are cheap. It certainly isn't a common failure. Have you tested the accumulator? It can be tested with DIS. I do not think this is the issue but one more thing to cross off the list since you've haven't found a resolution yet.

[trigtm]

So short of reverting the tune, I've now run three different software solutions and have the same results in all (INPA, ISTA and the Foxwell BMW tool). I'm leaning towards a wiring fault now so I'll need to trace the slave and the wiring on the pump (pump comes back out).

If all that traces out OK, I guess it's the SMG computer after that?


Sent from my iPhone using Tapatalk

[oceansize]

Something I just thought of, I have the CSL SMG tune from HTE. I haven't pinged him about if this is a known issue or not but I think it's worth a try.


Sent from my iPhone using Tapatalk

Almost all of us are running the SMG with CSL software, however, it might be worth returning to stock and giving adaptations a shot.

[trigtm]

Something I just thought of, I have the CSL SMG tune from HTE. I haven't pinged him about if this is a known issue or not but I think it's worth a try.


Sent from my iPhone using Tapatalk

[trigtm]

Yeah I think it will drive with failed adaptations but I'll have lights on the dash due to code 85 that won't clear (well, clears then comes back).


I'm gonna try getting everything hot and then run it all again.


Sent from my iPhone using Tapatalk

[eacmen]

I installed the pump, made sure to confirm the twist lock connector was connected correctly (did this a few times until I felt a positive click at the bottom) and then filled the reservoir (remote res since I have a Karb airbox) and the started the first bleed (actuator) and the the second (clutch slave). I ran them both probably 3-4 times to make sure I had all the air out.

I am able to activate the clutch in INPA and all shift paths and select all gears there. I can also select gears without INPA plugged in using the stick. The INPA version I have is Mikes Easy Tools and an Amazon OBD to USB cable with a bridge button which is depressed.

I really feel like this is some dumb thing I am missing and not something big. I never removed the gear selection springs but not sure that's relevant here.

Thanks for everyone's help so far, good to sanity check myself on this journey. I fully plan to keep the SMG and get this working and will continue to document here.


Sent from my iPhone using Tapatalk

Will the car drive with the failed adaptations? Only thing I can think of as this point is driving it around while logging all the data points the SMG will give you.

That or try swapping with a known good SMG computer.

[trigtm]

When you re built the pump, how did you bleed it/adapt it back then? if INPA is activating the solenoids, i would assume it is communicating properly and can carry out the adaptation. If anything, google for "Mikes easy bmw tools" its a stripped down version that works well with e46 m3 SMG and can carry out all routines.

Out of curiosity, have you checked that the 20PIN twist lock connector that powers the pump is locked all the way?

If you put the key in position 2 and try to shift (N,1,2,R) does it execute the shifts? (not using INPA to do this) - If it does not, im leaning more towards the clutch solenoid being bad. If i recall correctly, by you getting to manually shift via INPA it just moves the actuator not the clutch slave. Also, in INPA somwhere, you should have the ability to engage/disengage clutch (or at least view if its engaged or disengaged, you should be able to see if its closed or open when in position 2 and trying to shift).

Best,
Irhase46m3

I installed the pump, made sure to confirm the twist lock connector was connected correctly (did this a few times until I felt a positive click at the bottom) and then filled the reservoir (remote res since I have a Karb airbox) and the started the first bleed (actuator) and the the second (clutch slave). I ran them both probably 3-4 times to make sure I had all the air out.

I am able to activate the clutch in INPA and all shift paths and select all gears there. I can also select gears without INPA plugged in using the stick. The INPA version I have is Mikes Easy Tools and an Amazon OBD to USB cable with a bridge button which is depressed.

I really feel like this is some dumb thing I am missing and not something big. I never removed the gear selection springs but not sure that's relevant here.

Thanks for everyone's help so far, good to sanity check myself on this journey. I fully plan to keep the SMG and get this working and will continue to document here.


Sent from my iPhone using Tapatalk

[Irhase46m3]

When I rebuilt the pump with all new seals and a new motor, I made sure to remove the solenoids one by one and test them using a 12v source and they tested out all the same/fine. I used a guide online to confirm they were put back in the same spot also.

The car drove fine in between now and after I rebuilt the pump. the clutch space bleeds out fine and I can hear it working correctly. All gear changes using INPA are good. Activating the solenoids in INPA are also successful which is why this is so frustrating.

I'm leaning now into thinking either my INPA installation is somehow causing the procedure to fail or there is a wiring fault somewhere that only presents itself in this way since the clutch works otherwise.


Sent from my iPhone using Tapatalk

When you re built the pump, how did you bleed it/adapt it back then? if INPA is activating the solenoids, i would assume it is communicating properly and can carry out the adaptation. If anything, google for "Mikes easy bmw tools" its a stripped down version that works well with e46 m3 SMG and can carry out all routines.

Out of curiosity, have you checked that the 20PIN twist lock connector that powers the pump is locked all the way?

If you put the key in position 2 and try to shift (N,1,2,R) does it execute the shifts? (not using INPA to do this) - If it does not, im leaning more towards the clutch solenoid being bad. If i recall correctly, by you getting to manually shift via INPA it just moves the actuator not the clutch slave. Also, in INPA somwhere, you should have the ability to engage/disengage clutch (or at least view if its engaged or disengaged, you should be able to see if its closed or open when in position 2 and trying to shift).

Best,
Irhase46m3

[trigtm]

When I rebuilt the pump with all new seals and a new motor, I made sure to remove the solenoids one by one and test them using a 12v source and they tested out all the same/fine. I used a guide online to confirm they were put back in the same spot also.

The car drove fine in between now and after I rebuilt the pump. the clutch space bleeds out fine and I can hear it working correctly. All gear changes using INPA are good. Activating the solenoids in INPA are also successful which is why this is so frustrating.

I'm leaning now into thinking either my INPA installation is somehow causing the procedure to fail or there is a wiring fault somewhere that only presents itself in this way since the clutch works otherwise.


Sent from my iPhone using Tapatalk