On Sat, 13 Mar 2010 20:32:22 -0500, firstname.lastname@example.org wrote:
transmission, but the manual valve controls the supply of fluid under
pressure to the solenoids. The only solenoids generally acting on the
shift linkage are the park/brake interlocks that prevent the
transmission shifting out of park without the brake being depressed.
Park is at the end of the range of motion of the manual linkage, so
shifting into or out of neutral is not affected.
On Mar 14, 1:22 am, email@example.com wrote:
Uh huh. And you're so knowledgable about all the various Toyota
models that you can state that for sure there isn't even a slim
possibility that one of those interlocks is couldn't malfunction under
conditions of either
A - the computer running amock
B - the car is under full throttle acceleration
You'd testify to that in court? You're the guy who says modern cars
have only two computers and some have only one. We're still waiting
for a single reference to support that nonsense. Or should we
reserve final judgement until all these cars are looked at by
engineers who understand the design and what controls what. Ever
see the NTSB rush to pass judgement on a plane crash before a full
team consisting of representatives of the manufacturer, engine maker,
avionics maker, and the NTSB itself has had a year or more to analyze
and make sure what really happened?
Maybe you're not aware of it, but some cars today can PARK THEMSELVES
at the curb. Clearly that involves shifting the transmission in and
out of drive while applying the accelerator. I suppose you're an
expert on how that system works too.
On Sun, 14 Mar 2010 06:40:24 -0700 (PDT), firstname.lastname@example.org wrote:
Since the park/brake interlock is NOT computerized, yes I would say it
is impossible for any computer issues or throttle position to
interfere with that process.
The only possible exception is with Hybrid Synergy Drive on the
Hybrids, where neutral is electrically controlled and I am not
familliar enough with the computer system there to know what could
POSSIBLY happen with MULTIPLE computer level failures. They are
designed that just about any single failure will fail "safe". Is there
a combination of possible failures that could allow the vehicle to run
but not allow BOTH throttle control and transmission disengagement? I
suppose it is POSSIBLE, but it has never to my knowlenge been
demonstrated and the odds against it happening on even ONE car would
be extremely high.
Yes I would.
Totally aware of that.
But it is not part of any of the corollas and camrys being recalled
for the sticky throttle problem.
to be at least part of the problem on the cars in question.
You WILL find that the throttle control mottor/assembly is not
adequately sealed, moisture is getting into the throttle mechaism,
causing corrosion, which is causing rough operation of the actuator
and sticking of the throttle. it is a mechanical problem - and not an
electronic problem, on the whole.
Might be a glitch in the program that is not shutting the engine down
when the throttle sticks - that has not been proven yet.
The corroded throttle actuator HAS been documented, on several
vehicles - and NOT all Toyotas. Nissan and Infinity are also affected
for sure - and quite possible other companies, including the "big
You read it first here.
On Sat, 13 Mar 2010 07:37:42 -0800 (PST), email@example.com wrote:
IF the car has a CVT, there MAY not be a mechanical linkage. I have
not been able to find a reference for sure, and have not physically
looked at one to see.
Anything that has a plantary gear multi-ratio transmission DOES have a
mechanical way of putting it in neutral that is totally separate from
the electronic controls.
I had a minor nose to tail accident because I forgot that the vehicle I was
driving had ABS.I am so used to cadence braking in urgent situations that it
is now a problem for me.
These recalls make me ponder as to whether anyone driving today knows how
their vehicle works?
On Mar 13, 3:29 pm, firstname.lastname@example.org wrote:
More speculation and over reaching. How the hell can you know what
all cars do or don't have? You're the guy that tried to tell us that
most modern cars have only two computers and some have only one. I
provided you several credible links that say that is total nonsense.
On Sat, 13 Mar 2010 13:08:04 -0800 (PST), email@example.com wrote:
Say what you want. I've been in the business for a long time. All you
can say is nobody knows. Well, thre are peoiple out there who DO know
- and I talk with a lot of them.. I don't care what you think.
On Mar 13, 6:45 pm, firstname.lastname@example.org wrote:
I never said nobody knows. I said YOU don't know how all cars that
have been built in the last decade are built, so you shouldn't go
around making blanket statements that just make you look foolish.
Certainly the engineers that designed a particular car know.
Next time you talk with them, ask them how many computers are in a
modern car. Is it one or two, like you claim, or dozens of them like
the website dedicated to embedded computer engineering that I provided
you with as a reference says? And while you're at it, ask them if
the same computer that controls the airbags controls the climate
system and the radio. Obviously you haven't been in any engineering
development or you'd know how highly improbable that is. Despite
claiming to know so much, you can't even grasp why.
Here's a clue. Who in their right mind would want a critical real-
time system that has the capability of exploding airbags in your face
co-mingled with the radio? And what exactly would be the purpose of
building it the way you say with everything in one computer? First,
it's unlikely that those components are even designed and/or built in
the same place. More likely there are completely separate engineering
teams, maybe one in Japan, another in Michigan, a third in Ohio. So
why would they choose to share a computer and complicate things?
With microcontrollers costing as little as $1, there is no reason to
co-mingle all kinds of distinctly separate functions into one
module. You put the computer close to where it's needed, segment
the system logically, and have the computers talk to each other if
Ever do any engineering change work or code validation? Somebody
decides they need to fix a tiny bug in the radio or marketing wants to
add a new feature. The fix involves changing the program code. If
you have a $1 microcontroller functioning as the brains for the radio
and entertainment system, you can make that change, validate it with
regards to the radio and not worry that it COULD result in screwing up
the airbag system, killing people. Do it the way you claim, and as
soon as you change that program code for the radio, you have one hell
of a big system validation problem covering not just the radio, but
the critical airbags, climate control and God only knows what else
because you decided to put it all together. There is a reason for
modular design, ie breaking things up into logical units. And maybe
I'm wrong. Maybe some cars do have the same computer that runs the
airbags running the radio and climate control. But I'm betting you're
wrong. Just like the mainframe gave way to PCs everywhere, with cheap
microcontrollers and microprocessors, it makes sense in a car to do
the computing where it's needed instead of in one central place.
Just show us an example of a credible source that says the airbag,
radio and climate control are done with the same computer.
On Sat, 13 Mar 2010 17:10:25 -0800 (PST), email@example.com wrote:
In the first place - my statement about the number of "computers" in a
car was in reguards to the operation of the powertrain, primarily -
and I did also explain there are semantics involved. Depends what you
want to call a computer.
At one point, cruise control was one computer, ABS was another
computer, fuel injection was another computer, and ignition another
one.sunroof was another computer, the automatic transmission was
another,power windows was another computer, and if you had a driver
information panel or a trip computer they were both separate - as was
the compass and/or climate control. Digital dash was another computer.
If it had electrically operated seat belt retractor locks that was
another "computer" - then you added SRS (air bags) and automatic
In todays cars, the fuel injection, ignition, transmission, cruise
control, antilock brakes, traction control and even the voltage
regulator are commonly all handled by one main computer, while climate
control, driving computer, compass, doorlocks, sunroof, compass,
instrument panel and seat belts, along with wipers and numerous other
functions - in many cases even the RADIO. functions are controlled by
another main computer.
The fact that microprocessors are also used in many of the "smart
switches" that are built into things like BMW tail-light assemblies,
door and window operators etc is not under dispute, but the definition
of them as computers is playing pretty loose.
Calling the air bag computer control a computer IS accurate on some
cars - on others all the operational smarts are in the sensors, while
the "computer" is simply a monitor that makes sure there are no
defects that would prevent the air-bags from functioning in case of an
accident and turning on a light and storing a code when a malfunction
A string of switches that must all be turned to the right position to
activate a circuit is not necessarilly a computer.
And even calling each "smart switch" on a BMW a computer, 100 is a big
It WILL happen. and in the not too distant future.
The bean counters will make sure it happens unless legislators get
their panties into enough of a knot over this toyota BS that they pass
a law to stop it.
Yes, I've done some. It's a royal pain.
Ever here of virtualization? It is happening in a big way in the
computer industry - and you WILL see it in automotive computers, with
one computer running numerous "virtual machine" processes on one
processor , and each of those "virtual machines" handling different
functions, with the code for each virtual machine and each function
beng totally segmented and separately modifiable.
It all has to do with the integration and convergence, By eliminating
all the separate $1 microcontrollers and "converging" all the
different systems on one more powerfull CPU, there are MANY benefits
that can be realized. Among them is easier fault detection and
reporting, easier to implement redundancy, data verification and data
sharing, and in the long run lower costs.
You will see more centralized design come back into automotive
manufacturing as the complexity of today's cars increases. And the
programming of these processes running on these virtual machines can
still be done wherever you want it to be done - Bangalore, Karachi,
Mumbai, Detroit, Singapore, Mexico, Milan, or wherever.
Modular design in the computer world does not have to mean separate
physical computers. Opject Oriented Programming is modular program
design, and re-useable programming building blocks can be assembled to
control many different processes using the same basic code snippets,
with separate data tables to modify the logic.
And today you are seeing a large movement back to centralized
processing with client-server applications and "thin client" networks
like Citrix because it is more effective and more secure to have the
data and processes centralized.
Just like "single entry" accounting systems, where all the data gets
entered in one place, and is processed and manipulated and stored in
one place. You see more relational databases where all the information
required to run a business or produce a product is stored in a single
data file on a single system - and possibly replicated to a second
system for redundancy and capacity sharing.
This instead of a collection of numerous files, with different types
of data stored in different files all over the storage device.
It is simply easier to control and verify data that is all in the same
Many cars have the radio, climate control, and door locks, window
controls, compass, driving computer and instrument panel all on one
computer, along with part of the airbag monitoring system.
While some vehicles use a real "computer" to control the airbags,the
airbag control module in some cases is not a computer at all. Simple
boolean logic (and, or, not, nand etc gates) is all that is required.
The module gets signals from the main (Powertrain control or body
control) computer saying the conditions (speed etc) are right to
enable the supplementary restraint system, and if a front impact
(inertia) sensor gets tripped, it sets off the front air bags. If a
rollover sensor or side impact sensor gets actuated, it sets off the
side curtain airbags. No programming or computational ability is
required, so it is not a computer. It is just a couple of electronic
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.