Those not happy with the 12b37 update

[Kabigon]

Originally Posted by Timthom62

Part of the benefit of the dual clutch transmission is that when one shaft is engaged, the PCM and TCM have already enabled the next expected gear on the disengaged shaft. So if you're in 5th and accelerating, 6th gear is already engaged on the opposite shaft and waiting for the clutch to engage. The point being to prevent the lag between shifts that was normal with a standard automatic transmission.

The algorithm for pre-selecting the next gear is not quite so simple.

For example, if you are in 5th gear, 6th gear will not be pre-selected until you hit ~74 MPH.

Gear selection (for the next gear) often takes place at the time of the upshift.

[Kabigon]

Originally Posted by suss6052

If the car improperly learned the clutch touch points after a hard reset where the battery is removed long enough for the TCM to lose its memory that could cause the clutch pack to slip while it learns where the engagement is. This is why the drive cycle is so important if and when the TCM's keep alive memory is actually wiped rather than just updated.

As previously mentioned...

1) After a couple of days of normal driving. it doesn't matter if the drive cycle was done correctly or it was never done because...

2) The TCM is ALWAYS "learning" (for OBVIOUS reasons).

[Timthom62]

Originally Posted by Kabigon

The algorithm for pre-selecting the next gear is not quite so simple.

For example, if you are in 5th gear, 6th gear will not be pre-selected until you hit ~74 MPH.

Gear selection (for the next gear) often takes place at the time of the upshift.

You might want to let Ford know so that they can update their service manual:

Section 307-01 Automatic Transmission > Description and Operation > System Operation:

"Control Strategies

Operating principle of the transmission.

In this transmission, the use of a dry double clutch in conjunction with an electromechanical control system means that two gears (transmission ratios) are engaged at the same time.

One of the multi-plate clutches is engaged in driving mode, the other is already preselected when approaching the next gearshift with the clutch open.

Depending on the position of the accelerator pedal and the demand issued by the driver, the clutch of the previously activated gear is opened, while at the same time the clutch of the pre-selected gear is closed. As a result of this overlap in clutching, only minimal losses in tractive force are encountered during gearshifts."

[suss6052]

Originally Posted by Kabigon

2) The TCM is ALWAYS "learning" (for OBVIOUS reasons).

If the TCM is always learning why do some people seem to have almost no issues, or none at all. Where as others have a period where its better after the update then slowly devolves to being as bad as before? Would this be due to the information that was fed to the TCM, or is this just bad logic/ programming?

Just curious, since some people are specifically reporting a large improvement after an update or a removal of power, until after the learning window resets and the car implemented what it learned over the previous cycle.

[Kabigon]

Originally Posted by Timthom62

You might want to let Ford know so that they can update their service manual:


One of the multi-plate clutches is engaged in driving mode, the other is already preselected when approaching the next gearshift with the clutch open.

What I am saying is that the pre-selection for the next gear often takes place immediately before the gear change. This does not contradict the manual.

FWIW, I have logged the "internal" gear selection and I often monitor it in real time. I'll post a log in the next few days.

[Timthom62]

Originally Posted by Kabigon

What I am saying is that the pre-selection for the next gear often takes place immediately before the gear change. This does not contradict the manual.

FWIW, I have logged the "internal" gear selection and I often monitor it in real time. I'll post a log in the next few days.

I wasn't saying that it did contradict the manual, I was saying that it was more than the manual had to say on the subject and therefore the manual needed updating.

I'd be very interested to see the data, as it does not appear to be available elsewhere.

[Kabigon]

Originally Posted by suss6052

If the TCM is always learning why do some people seem to have almost no issues, or none at all. Where as others have a period where its better after the update then slowly devolves to being as bad as before? Would this be due to the information that was fed to the TCM, or is this just bad logic/ programming?

Just curious, since some people are specifically reporting a large improvement after an update or a removal of power, until after the learning window resets and the car implemented what it learned over the previous cycle.

I am referring only to the clutch enagagement.

I have blanked the adaptive memory several times (with an IDS) and whether I do the adaptive learning "by the book" or don't do it at all, the results (after a couple of days) are always the same. The clutch engagement is smooth and I have zero problems with the transmission.

The reason that I say it is always learning:

The TCM monitors input shaft speed (for both input shafts) and the output shaft speed so it "knows" exactly how much clutch slippage there is at all times. Keeping that in mind:

1) Clutches do wear over time. If the TCM wasn't "learning", it couldn't compensate for the clutch wear.

2) If you think about it, there is absolutely no reason why the TCM should ever stop "learning".

[Kabigon]

Originally Posted by Timthom62

I'd be very interested to see the data, as it does not appear to be available elsewhere.

Fair enough.

You can see the gear selector drum angles (1 and 2) for each gear in the service manual. This is what I monitor.

In manual mode, it is easy to determine if the next gear has already been pre-selected as there is a big difference in the time for an upshift/downshift.

[Strix]

Originally Posted by Timthom62

"As a result of this overlap in clutching, only minimal losses in tractive force are encountered during gearshifts."

Oh, how I wish this was the case.

I feel an obvious loss of acceleration when the transmission shifts, especially under anything more than minimal throttle. It feels like a manual transmission that is being operated by a three-toed sloth.

Not only can you feel it... you can also hear it. The ECM closes the throttle between shifts, and the engine practically shuts down for a fraction of a second while the DCT switches clutches.

"Minimal loss of tractive force" my [bootyshak!

[Timthom62]

Originally Posted by Strix

Oh, how I wish this was the case.

I feel an obvious loss of acceleration when the transmission shifts, especially under anything more than minimal throttle. It feels like a manual transmission that is being operated by a three-toed sloth.

Not only can you feel it... you can also hear it. The ECM closes the throttle between shifts, and the engine practically shuts down for a fraction of a second while the DCT switches clutches.

"Minimal loss of tractive force" my [bootyshak!

But the crux of the problem is that it is just as Ford describes for some of us, but not all.

In both of the cars we have, the shifts are smooth as silk - if it weren't for the change in sound and the tach needle, you couldn't be sure it had shifted sometimes. Doesn't matter if you romp on it, baby it, or drive it normally, it's smooth. Ford says that's normal.

Other people have jerky shifting, shuddering, loss of acceleration, etc. A good portion of them have had at least one Ford dealer tell them that this is normal.

Same transmission, same software, two different outcomes.

Well, which one is normal? Is it a hardware issue, a software issue, or different outcomes from the adaptive learning function? I don't know, but I know I'd be pretty pissed about it myself.