What's in the workshop this week?

[MikkiJayne]

Originally Posted by HPsauce

I'd be looking at the other set of tracks for the TCU gear-specific signals then, based on my understanding of how the F125 works. I can't see them in your photo, or all the contacts.
(I think I worked out the likely logic and posted it not long ago) Edit, here: http://forum.a8parts.co.uk/showpost....98&postcount=8

Interesting. I will have a look at the other side in a bit

[HPsauce]

I suspect that if the TCU gets an invalid set of signals from the other side of the F125 then it doesn't know what to display so you get nothing on the dash display.

The reverse light circuit contact damage was probably a red herring, just burnt from the high current but still functional.

[HPsauce]

Originally Posted by MikkiJayne

Interesting. I will have a look at the other side in a bit

By the way, have you worked out yet how that side is driven? If the shifter can only go as far as D or S does the gearbox move the F125 drive shaft when changing gears automatically and/or when you use manual override (lever on earlier version without S and/or tiptronic)?
Or is it just that on non-S models that's an input (gears 3 and 2) to the TCU that doesn't exist on S models?
Given there's no "1" position I'd guess that the TCU is informed by other means what the gearbox is up to and the F125 switch is purely an input mechanism, but why in that case is it mounted on the gearbox? It must be both an electronic input to the TCU and linked to a mechanical one to the gearbox (PND?), or is it just that directly mounting removes the inaccuracies of the cable linkage?

[MikkiJayne]

The gearbox mechanically only sees PRND if memory serves (possibly one position after D - I'll check later). [edit - this bit is incorrect but left here for context later] The cable stops moving after that, and the other selections are done in the shifter mechanism which is connected directly to the TCU, so S, 432, and tip are all just software in the TCU. [/edit] The TCU definitely uses the F125 to know what mechanical mode the gearbox is in. There's no other sensing circuitry. Presumably ZF put it there as part of the gearbox design, and then left the selector mechanism and the actual software up to the OEMs.

This is why you can get S mode in an FL1 by changing over the TCU, although changing the shifter top is also important as if you put an S ecu in to 3 or 2 it gets quite cross! The only issue I've found so far with that upgrade is that the cluster gets a tad confused and tends not to display the gear selection at all. I think the later TCU sends a different set of messages for PRNDS compared to the early one for PRNDS432. I need to swap in an FL2 cluster which I'm sure will solve it.

[HPsauce]

Thanks MJ, having never looked at these in detail I'm just intrigued as to how it all hangs together.
I can understand that there has to be input from the shifter to the TCU - Tip+/- for a start, and the kickdown switch and probably many other sensors/controllers.
Presumably there's a link from the TCU to the gearbox (valve body?) to issue commands to change gears etc. does that have a feedback/control element too?
When you say "Mechanical Mode" do you mean just PRND (and maybe S/4) or also 3 and 2, which the F125 is theoretically capable of signalling?
Just wondering how/if the TCU knows when the gearbox is in First gear...…

As for the cluster, presumably FL2 has a different bitmap or coding for the gearbox display vs FL1 as it has more options to display, so it's hardly surprising it gets confused and gives up.
I don't imagine the developers ever considered providing cross-compatibility either way between FL1 and FL2 TCUs and FL2 and FL1 clusters!

[MikkiJayne]

Its quite an interesting mode of operating actually. Left to its own devices, the gearbox will be a 1-speed hydraulic automatic, with P, R, N and D (3rd / 4th depending on model). The TCU actually overrides this behaviour by controlling the solenoids in the valve body, and monitors the results via the input and output speed sensors.

"Just wondering how/if the TCU knows when the gearbox is in First gear"

The valve body does only have PRND, so the TCU sees 'D' on the F125, and based on its numerous inputs it will select 1st using the solenoids. It doesn't know a gear has been successfully engaged until it sees some output revolutions, and it knows the gear ratios so it can calculate what the output speed should be based on the input speed.

When the TCU detects a fault (such as input rpms happen but output rpms don't) it basically shuts down and leaves the gearbox to its own devices and it reverts to the 1-speed auto, which is how limp mode works.

"I don't imagine the developers ever considered providing cross-compatibility either way between FL1 and FL2 TCUs and FL2 and FL1 clusters! "

The FL2 cluster is backwards compatible to PRND432 as not all FL2 had S mode, but its fair enough the FL1 cluster isn't forwards compatible

[ainarssems]

I am not sure if I understand this right but if you look at the testing for F125 from Elsa it has different outputs between different pins for D, 4, 3 and 2. As it's being turned by gearbox there must be something mechanically moving in the box when moving lever between D, 4,3,2.

[HPsauce]

Ainars that was my thinking, but equally, just because the capability is there it doesn't means it's being used.

If, as MJ says, the gearbox is measuring input vs output speeds (and it definitely does because we've seen many references to the errors that occur when there's a mismatch) those circuits could well be unused.

On the other hand, let's have a look at my old F125 switch for evidence of movement in that area. There are photos somewhere.....
Yes indeed, definite evidence of movement through the full arc, and my car this was removed from is a (very) late FL2 with S mode....
The mystery deepens, my shifter is incapable of doing that!

[MikkiJayne]

Ainars, yes, you're correct. I was wrong earlier about the shifter cable not moving for 432. I suspect I have confused the shifter cable for the shifter lever inside the valve body, or possibly the shifter cable not moving in tiptronic. It was a couple of years ago, and I've slept since then

The selector lever on the gearbox definitely has positions for PRND432, even in the S version, however the diagrams for the valve body only show PRND and I'm sure when I had the valve body out of one that behaviour was visible too. So, the TCU is seeing PRND432 or PRNDS from the F125 (where 4 and S are the same) depending on the year, but the actual valve body doesn't care about anything after D as by then its all software control by the TCU.

Looking at the shifter, PRND432 and PRNDS are all definitely on the F125 with only the tiptronic controls on the shifter body itself (as far as gear selection). The shifter cable does not move in tip, so that operation is purely electronic. I think this is what I was thinking of in my previous post.