Prev: Fire and Drive
Next: Algae-powered vehicles within five years
From: Vic Smith on 28 Jan 2010 22:02
On Thu, 28 Jan 2010 17:47:00 -0800 (PST), john <johngdole(a)hotmail.com>
wrote:

>I think there may be multiple causes to this sudden acceleration
>problem. I do believe electronics is a primary suspect, according to
>the Alberto lawsuit.
>
Probably. I'm a bit surprised by all this concentration on "gas
pedals." I suspect that gas pedals and linkage aren't too
mind-boggling. They've been around a long time.
I had a an '85 throttle body 2.0 Cavalier that would "unintentionally
accelerate." Wife was using it for her commute and complained about
it, but I shrugged it off until it happened to me when I was driving
it. Had to keep it braked hard at a stop, and it would accelerate to
about 40 mph when you released the brake. It came and went.
I was going to throw a new throttle body on it but the car just died
on her on the way home from work the next day.
Right in front of a garage, and a couple guys pushed her off the
street and into the garage. That's how I found my mech, just in time,
as I knew nothing about the new electronic controls then.
New ECU fixed it.
There's stuff going on with modern cars that is *really* hard to
diagnose. That's why I don't like drive-by-wire, and ignition key
lockouts or interference with a mechanical switch engine shutdown.
Fail-safe can't be controlled by microchips. They are too flaky.

--Vic

From: Vic Smith on 28 Jan 2010 22:27
On Thu, 28 Jan 2010 22:12:53 -0500, dr_jeff <utz(a)msu.edu> wrote:

>
>Tell that to the pilot of the Airbus A-320.
>
>I have to disagree that fly-by-wire is too flaky or that microchips are
>too flaky. There are certainly bugs, but those bugs will be worked out.
>Computers can react far faster than humans. Computers control a lot of
>critical function; car engines are just one of them.
>
>Jeff

I'd have to see the cost and quality of the A-320 control chips, and
their engineering and programming to buy that comparison.
For all I know the chips controlling the acceleration in drive-by-wire
cars are sourced from the same outfit making chips for Mr Coffee.

--Vic
From: Hachiroku ハチロク on 28 Jan 2010 18:20
On Thu, 28 Jan 2010 21:27:58 -0600, Vic Smith wrote:

> On Thu, 28 Jan 2010 22:12:53 -0500, dr_jeff <utz(a)msu.edu> wrote:
>
>
>>Tell that to the pilot of the Airbus A-320.
>>
>>I have to disagree that fly-by-wire is too flaky or that microchips are
>>too flaky. There are certainly bugs, but those bugs will be worked out.
>>Computers can react far faster than humans. Computers control a lot of
>>critical function; car engines are just one of them.
>>
>>Jeff
>
> I'd have to see the cost and quality of the A-320 control chips, and their
> engineering and programming to buy that comparison. For all I know the
> chips controlling the acceleration in drive-by-wire cars are sourced from
> the same outfit making chips for Mr Coffee.
>
> --Vic

Same aerospace mfgr as in the other post:

We made fuel controls for Pratt and Whitney and, where spec'd Rolls Royce.
This was a pretty amaziing piece of equipment for the early 1980's. It was
two complete engine controls married together in one big housing; Channel
A and Channel B. Exact identical units.

One day I had nothing to do, and a friend of mine gave me the Theory of
Operation manual, about as thick as a bible. Reading through it, I
discovered this:

It was a dual-redundant engine control. What this means is that there is
an A channel and a B channel. However, on a failure of a component in the
main channel (A), operation did not swicth over to the B channel. Instead,
a supervisory module detected the failed component and switched operation
of that FUNCTION to the B channel. Back then they called the system RAM
the 'scratch pad', so if a failure were detected in the scracth pad of
channel A, scratch pad operation was transferred to the scratch pad of
channel B. So you still had A channel operating primarily with one
function of channel B.

It took a massive failure of channel A, or three separate failures to
cause control to pass entirely to channel B.

In case of a massive failure of the engine control where neither channel A
nor channel B were operational, the engine was sent into "limp mode" and
fuel delivery was performed by a fail-safe valve that opened on failure of
the engine control. This was a hammer-head unit with rudimentary fuel
regulation; it basically dumped fuel into the engine in a failure mode
metering method that kept the engine running as long as there was fuel in
order to make it to the nearest landing field.

I would imagine Airbus and the Boeing models employing Fly By Wire have
backups of the backups.



From: Al Falfa on 30 Jan 2010 08:58


"dr_jeff" <utz(a)msu.edu> wrote in message
news:hrudnVnZoopfu_nWnZ2dnUVZ_uRi4p2d(a)giganews.com...
> Vic Smith wrote:
>> On Thu, 28 Jan 2010 22:12:53 -0500, dr_jeff <utz(a)msu.edu> wrote:
>>
>>> Tell that to the pilot of the Airbus A-320.
>>>
>>> I have to disagree that fly-by-wire is too flaky or that microchips are
>>> too flaky. There are certainly bugs, but those bugs will be worked out.
>>> Computers can react far faster than humans. Computers control a lot of
>>> critical function; car engines are just one of them.
>>>
>>> Jeff
>>
>> I'd have to see the cost and quality of the A-320 control chips, and
>> their engineering and programming to buy that comparison.
>> For all I know the chips controlling the acceleration in drive-by-wire
>> cars are sourced from the same outfit making chips for Mr Coffee.
>>
>> --Vic
>
> So? I haven't seen reports of runaway Mr. Coffees. I am sure the chips are
> made to high quality standards. The issue really is the programming and
> quality assurance (which is testing of how well the programs work,
> particularly in unexpected situations). I don't know what Toyota and other
> car makers do in that regard.

The advantages of the electronic throttle control are readily apparent in
the 2010 Prius. The gas pedal works differently as you switch between the
Economy, Normal and Power modes. Still, the pedal is mechanically
interfaced to the foot and from what I read the problem lies in the
mechanical interface.



From: Mike Hunter on 30 Jan 2010 13:26
If you don't know, dr_jeff, what Toyota and other car makers do in that
regard, why are you commenting on something you admit you know nothing
about, again?

"dr_jeff" <utz(a)msu.edu> wrote in message
news:hrudnVnZoopfu_nWnZ2dnUVZ_uRi4p2d(a)giganews.com...
> Vic Smith wrote:
>> On Thu, 28 Jan 2010 22:12:53 -0500, dr_jeff <utz(a)msu.edu> wrote:
>>
>>> Tell that to the pilot of the Airbus A-320.
>>>
>>> I have to disagree that fly-by-wire is too flaky or that microchips are
>>> too flaky. There are certainly bugs, but those bugs will be worked out.
>>> Computers can react far faster than humans. Computers control a lot of
>>> critical function; car engines are just one of them.
>>>
>>> Jeff
>>
>> I'd have to see the cost and quality of the A-320 control chips, and
>> their engineering and programming to buy that comparison.
>> For all I know the chips controlling the acceleration in drive-by-wire
>> cars are sourced from the same outfit making chips for Mr Coffee.
>>
>> --Vic
>
> So? I haven't seen reports of runaway Mr. Coffees. I am sure the chips are
> made to high quality standards. The issue really is the programming and
> quality assurance (which is testing of how well the programs work,
> particularly in unexpected situations). I don't know what Toyota and other
> car makers do in that regard.


 |  Next  |  Last
Pages: 1 2
Prev: Fire and Drive
Next: Algae-powered vehicles within five years