> Mortimer
privacy.net> wrote:>> If you remove the electronic inputs to a modern TC automatic
>> transmission -
>> those from the ECU and those from the throttle-position and engine- and
>> road-speed sensors - how will the transmission respond if you apply
>> (mechanical) power to it.
>
> You're ignoring the fact that the designers ahve chosen to cripple the
> autoboxes in this way.
"cripple"? Some would say "enhance".
What's the downside? How are they crippled? That purely ECU failure could
result in an un-operational vehicle? What's the guestimate as to whether
modern ECU controlled gearboxes have a better / worse failure rate than
their previous, purely mechanical ancestors? There's a lot more simple,
mechnical ways in which autos could be crippled, without anything untoward
happening to it's ECU.
And I don't get the ignored thing. What's happened with conventional autos
is evolution - like many mechanically complex aspects of automotive
engineering has fallen under electronic / computer control. Where these
transmissions are concerned, that's now an all or nothing thing. To be
honest, I'm not aware of any wholesale failures of gearbox ECUs that
suggests that such an evolution is a backward step.
> It's not required for operation, it's a design
> decision.
That sentence almost reads like an oxymoron. It's certainly a requirement of
how they are /currently/ designed - so as they are /currently/ designed, it
/is/ required for operation. Going back to purely hydraulic, would take
quite a bit of /re-design/ work - certainly not the same 'box. You almost
make it sound like it's optional - for current autos, it's not - it would
take wholesale redesign of them to remove such a dependency. I'm speculating
that so much so, and you wouldn't consider it the same gearbox, or certainly
the same design.
They are not the same beasts as they used to be, even if they largely appear
the same in terms of how they operate. Just because the /type/ of gearbox,
in years gone by, has been designed to be autonomous in terms of purely
hydraulic operation, hardly means that modern equivalents can easily and
simply be reworked. At the time, there was likely no other tenable means by
which it could be achieved.
In much the same way as it would be to make a DSG purely manually /
physically controlled - so much so with the DSG, you'd wonder whether it
would be truly tenable without ECU control, and electro-hydraulic actuation.
> The semi-autoboxes are incapable of being designed to operate without an
> ECU.
Did you really mean "incapable"? Or do you mean in context of them operating
automatically / autonomously?
If you conside the DSG, without an ECU, it would be quite a complex bit of
kit to design manual controls for - and no doubt, quite complex to actually
operate.
That's all smoke, though. Both modern autos /including/ the DSG, are
/designed/ to be operated by ECU. Look at it's upshift speeds - faster than
any consumer (perhaps even considering specialist, or motorsport) box that
preceded it, by quite a margin.
Are you suggesting that it would be /impossible/ to design the DSG to
operate purely mechanically, by means of hydraulics? If so, I wouldn't
necessarily argue - but it would be a bold statement.
The bottom line is that the DSG is as automatic as a modern conventional
auto - it only differs on cog layout and interaction, and using clutches as
opposed to a torque convertor. Hardly such a distinction as you'd question
it's degree of automation. Neither current conventional auto, nor DSG, are
capable of automation, nor autonomy, without the external control from their
ECU.
