MKZ Transmission

[Extreme4x4]

The 5R55 has absolutally nothing in common with the Aisin trans, or the new 6F transmission.

 

As a matter of fact, the 5R55 is just a very updated (extra gear and fully electronic valve body) and strenthened A4LD from the late 80's/early 90's.

 

Most new transmissions, have adaptive learning programs in their CPU's. They are fully integrated into the engine and vehicle CPU's, so they can easily adapt to different peoples driving patterns. It will usually take a week or so, for the transmission to fully "learn" your driving style.

[waymondospiff]

I'm more concerned about the lack of midrange punch in the 3.5. I drove a Camry SE V6 the other day, and the Toyota 3.5 feels far stronger in the 2000-3500 rpm range. Part throttle gives you a solid push in the lower back. Open the Ford 3.5 up and it performs well over 4,000 rpm, which is most obvious on the highway, but I'd like to see more punch lower down.

I think the Aisin transmissions connect to Ford's PCM, and Ford's PCM features a transmission controller that does indeed learn over time--and become eerily predictable, as long as your driving is similarly predictable.

The 6 speed Aisin in our Montego is one nice piece. With Fords electronic controls it almost reads my mind....for the way I drive the car. Another feature I like is it's serviceable at regular intervals according to owners manual...and not sealed for life. I like being able to change trans fluids.

 

I consider the 3.0 in our Montego adequate for way we use this car...especially considering fuel milage. However, for those who live in mountains or hilly area, the 3.5 probably is for you. Now some of you think 3.5 is not enough.

I just felt the need to add two things:

1. Since the 6F was "intended" for use, but was changed to the Aisin, is it possible that the 3.5L is torque limited at lower speeds to prevent transmission damage? I know other vehicles (notably the Bugatti Veyron, not that it's comparable to a MKZ...) will electronically limited torque to prevent overstressing of the transmission's components. Just an idea I had.

2. I had a rental Jaguar X-Type awhile back. Obviously it had been driven by ninnies, because on the drive from the airport it felt that the dumbest transmission, hanging onto revs, shifting up then rapidly downshifting under constant pressure. It was a mess. It took about 30 minutes on a curvy backroad to learn my driving style and from then on it was a willing partner. I was amazed at the difference and how quickly it learned. Night and day, I tell you, night and day.

 

And Ralph, some people won't be happy until every car is a 6.0L 48V DOHC Supercharged V12...for the base engine. I really have a hard time believing any serious 500/Montego shopper is going to not choose a 3.5L model because the engine isn't strong enough. And truth be told, the current 3.0L is "good enough" for 90% of the drivers out there, but it's low end weakness & higher NVH give the impression of an underpowered car. Not so, but combine that impression with reading a review online about the 500/Montego where it is GUARANTEED to be called "underpowered," and people will believe the 500/Montego are slow.

 

Scott

[akirby]

The 5R55 has absolutally nothing in common with the Aisin trans, or the new 6F transmission.

 

I didn't say it did. But the shift control is usually in the PCM software - so it's logical to think that over vehicles would use the same strategy. I know the controls for the 6F are in the software - that's one difference between the Ford and GM version of that tranny.

 

 

Most new transmissions, have adaptive learning programs in their CPU's. They are fully integrated into the engine and vehicle CPU's, so they can easily adapt to different peoples driving patterns. It will usually take a week or so, for the transmission to fully "learn" your driving style.

 

I think transmissions that actually adapt their shifting to the driver are rare. I think you may be overestimating the meaning of the word Adaptive in advertising. Do you have any real examples that clearly say the tranny adapts to the driver?

[RichardJensen]

I think transmissions that actually adapt their shifting to the driver are rare. I think you may be overestimating the meaning of the word Adaptive in advertising. Do you have any real examples that clearly say the tranny adapts to the driver?

http://media.ford.com/article_display.cfm?article_id=2574

 

Doesn't say what vehicles have this software installed.

[mkaresh]

Whenever a transmission behaves poorly during one of my test drives, the salesman usually blames the adaptive programming. This does tend to be with high-end brands, though, especially Mercedes. So I don't know how many cars actually have it.

[akirby]

http://media.ford.com/article_display.cfm?article_id=2574

 

Doesn't say what vehicles have this software installed.

 

Adaptive shift-quality control is what I was referring to - it's common. I'd like to see examples where the Adaptive shift scheduling is actually used. I bet it's Land Rover, Jag and/or Volvo. I doubt it's in any F/L/M vehicles yet, although it would be nice.

[old_fairmont_wagon]

Sadly, given the state of the electronice that are installed in these vehicles, using adaptive shift scheduling in any of these vehicles is as easy as a new software burn to the PCM. There may be an extra sensor or two involved, but, really, it doesn't have to be that expensive.

[akirby]

Sadly, given the state of the electronice that are installed in these vehicles, using adaptive shift scheduling in any of these vehicles is as easy as a new software burn to the PCM. There may be an extra sensor or two involved, but, really, it doesn't have to be that expensive.

 

True - but actually developing the software that uses the sensor inputs to modify the shifting is incredibly complicated. Shift quality is relatively easy - there is a table of shifts by rpm with the current pressure and expected shift time. Each shift is compared to the expected shift time and the pressure adjusted if necessary. But trying to figure out when to shift or not or how hard or soft based on numerous inputs either requires software that's incredibly complicated or it requires an algorithm that's so simple that it's not very useful.

 

Some simple things like not upshifting if you're negotiating a curve can be done but it's going to take awhile for the software to catch up with the hardware capabilities.

[old_fairmont_wagon]

Well, coming from a computer engineering background, I can imagine that, once the software is perfected once, then repeating it for another application can either be almost as hard as the first application, or vvery simple, depending on how the first package was approached. I can imagine that if I was in charge of control system software development at a major automobile manufacturer, I'd want my software to be as modular as possible. In saying that, I mean that the basic framework of the software can be used in any vehicle that is currently manufactured at present, or is planned for the next several model cycles. The differences for each vehicle would be the data sets used.

 

The true cost would be developing a dataset for every engine and tranny combination that's available, and then a dataset for each platform configuration that's being used (vehicle weight and aerodynamic drag will have an effect on the ideal shift points). But, those are all just differences from an "ideal vehicle", a virtual test car that the software is developed from. So, in the end, no big deal. The true obstacle that needed to be overcome is the A/I for determining what the driver is apt to need based on the set of inputs received in the last 5 seconds, 30 seconds, 5 minutes, hour, week, since built. The second obstacle is to determine how that behavior needs to change for each platform/powertrain combination that's available. Once that's completed, then its all gravy.

 

I know I'm over-simplifying some things, but, modularizing the software makes it easier to move from vehicle to vehicle. You just develope the dataset that represents each vehicle and feed it into the program. I did that for analog control systems for years as an intern, while not exactly fun work, its very straightforward.

[akirby]

I think it would be easier to provide a user selectable setting such as normal, sport and track that would dictate shift speed, engine braking and upshift points rather than trying to "guess" what the user wants based on throttle/braking and/or steering inputs.