Wrong Parameter Reference in Selespeed ECU CFC209F

[Subocean]

It seems there is a wrong reference to the actual clutch movements within the parameter data...
I'm referring to the following parameters, taken from a Fiat Stilo Abarth diagnosis:

- Clutch Pedal Position
- Clutch Pressure Plate Reference
- Clutch engaged position (self adjusted)
- PIS closed position (self adjusted)

Apart from the last parameter, which appears to be the calculated difference between fully disengaged to engaged position, stored in memory during the last clutch self-adjustment, - all the other parameters should reflect the actual readings of the clutch sensor (K99 in case of the Stilo Abarth).

This SEEMS to be the case as one can monitor changes in values during i.e., manual gear shift op’s.

BUT:

The very same values (and changes) can be observed even if the clutch travel sensor is DISCONNECTED and completely absent!

With the sensor disconnected, values should read zero – as they do within Fiats own Examiner…

This suggests that all those parameters are based on the stored values recorded during the last self-adjustment rather than reflecting current readings.

This might also explain why Multiecuscan shows no error when the sensor is disconnected or otherwise faulty. While the gearbox ECU registers this – and consequently locks the gearbox and brings the gearbox fault light up…

Martin

[yani]

Hi,

You made me thinking... and I re-checked the selespeed 209 documentation again.
I don't see any problem with these parameters. Everything is as documented.
The fault code reporting is done by the ECU and it is not related in any way to the parameters which the program can or cannot access.
Is it possible that you have disconnected the wrong connector?

Let's see some comments from the other stilo selespeed owners.

Yani

[Subocean]

@Yani:

I can positively confirm that I'm talking about the clutch position sensor, the one that is attached right on the radial clutch cylinder within the clutch bell.
On my Stilo, this sensor is definitely broken - it's presently in front of me on my desk.

Since I downloaded your superb software only days ago (because of a then unknown gearbox fault) - I have never seen your scanner in action with a functional clutch position sensor as my sensor was already broken.

However, the readings with the broken sensor (or none at all) are as described. Clutch position is still displayed and if one gets the gearbox to change a gear and - thus opening and closing the clutch - those readings do change, indicating a different value for open and engaged clutch... But where do those values come from if the sensor isn't even connected?

A friend of mine meanwhile simulated this on a functional Alfa Sele (different ECU and position sensor). He confirmed both - your and mine statement. If the sensor is functional, clutch position readings are as they should - and identical with the same parameters in a Examiner. But he also disconnected the sensor (for simulation sake) and bingo - there are still (slightly different) values that change with clutch movements. Examiner readings are '00' in this case.

The ECU will show a fault and locks the gearbox however, your scanner doesn't show any making it practically impossibly to unlock the Selespeed... The only way around this is to introduce an artificial fault - say, stick the gear lever in one direction for more than 1 minute, which causes another fault that is identified by your scanner. Subsequentially acknowledged, both faults dissapear and the gearbox becomes unlocked again...

On order to avoid any misunderstanding, the clutch position sensor (K99) connects to the gearbox ECU plug A, pins 34, 35, 46 & 47….

Martin

[devc]

Hi,
Please try this with AlfaDiag and let us know!
Very interesting!
Dev

[Subocean]

This has been done by Belli29 - he compared Multiecuscan, Alfadiag and Examiner. Results as in my post above...

Might it be that this got something to do with the fact that Fiat uses a complete different sensor for clutch position than i.e., Alfa does - it's NOT a potentiometer, looks rather like a linear resolver, transducer or something similar. Furthermore, there is confusion within Fiat documentation about this sensor. I.e. try to find it within ePer - it doesn't exists there...

Martin

[devc]

If a LVDT (Linear Variable Differential Transformer) is used it will have 4 pins on the connector.
The Alfa Selespeed unit defintely uses a Potentiometer (3 Pin Connector)
Any random value can be possible with a LDVT Disconnected, same as open circuit mos inputs in a electronic circuit
due to noise. LVDT are however very expensive sensors.
Dev

[Subocean]

It's not a LVDT but a VAC sensor but yes, it got four pins (VAC’s also got two coils but they are quiet cheap). The values don’t change randomly either - even with a disconnected sensor there is a static reading for open and closed clutch - in my case it's something around 10.x and 19.x mm. The question remains: Where do those readings come from while there is no sensor connected?

This prevented me from identifying the problem (the said sensor) for two days. I was running in circles since clutch actuation did show changes in clutch position and no fault was identified (within multiecuscan). Only when I finally took the sensor out, which requires the gearbox to come off and before that (in order to make that happen), painful removal of the solenoid unit, pump and actuator unit. Under the cramped bonnet of a Stilo, this is no fun…

Only then I could actually check the sensor by manually actuating the radial hydr. cylinder and with that, no changes in clutch position values took place in the software (which was by then anticipated since I noticed mechanical damage once I laid eyes on the sensor for the first time. This was obviously caused by misalignment of the moving magnet and the actual sensor. It's somewhat of a bad design and can happen rather easily...)

I should find out more come Monday as I expect delivery of a new sensor. Just a shame that Fiat don’t sell them singularly. So it comes together with a new cylinder, which I don’t really need. The sensor itself is said to be made by Valeo but Valeo themselves denied this. The design gets pretty close to some Honeywell VAC’s that cost around US$12 but is not an exact match… Here we go, another 300€ for nothing much then… Bonus: I get a new clutch with the package.

Martin

[yani]

@Yani:
The ECU will show a fault and locks the gearbox however, your scanner doesn't show any making it practically impossibly to unlock the Selespeed... The only way around this is to introduce an artificial fault - say, stick the gear lever in one direction for more than 1 minute, which causes another fault that is identified by your scanner. Subsequentially acknowledged, both faults dissapear and the gearbox becomes unlocked again...

Hi,

What kind of cable are you using to connect to the gearbox? Standard VagCom, Galetto, another k-line cable, or ELM327?
Some users reported that Galleto reports wrong values for some parameters.
Also, can you please make a test with an ELM 327 cable. I want to know if the error can be read with ELM 327.

Thanks,
Yani

[Subocean]

So far I was using a standard VagCom cable but an ELM327 is on order. I expect it any day now. Will keep you updated.

Martin

[Subocean]

I meanwhile got my ELM interface. There is no difference in error codes to the K-line interface.

A missing clutch position sensor will not show up as an error message - allthough the ECU brings up the gear warning light and blocks the gearbox completely.
Since I got a new and functional sensor I also know now the only parameter that reflects that sensor: It's Clutch Pedal Position - certainly no surprise there however, all the other clutch related parameters as:

- Clutch Pressure Plate Reference
- Clutch Engaged Position
- PIS Closed Position

... are not influented by that sensor and appear indeed to be stored values from the last Clutch Self-Calibration. At any case they are static until the next Self-Calibration takes place.

Martin