On Mar 14, 1:17 am, email@example.com wrote:
quoted text -
More BS. You stated that most new cars had one computer for the
engine, one for the body and some cars had only one.
I'll go with the definition used by the website for engineers who
design embedded computer systems, because that's what these computers
And they say any modern car has dozens of computers. You on the other
hand told us that you can use a microprocessor in a way that it is not
in fact a computer, which is ridiculous, unless you're using it for a
BS. There are lots of microprocessors in cars doing more than being
a smart switch. I could implement a smart switch with a simpler
design using a sequential state machine.
Hmmm, I remember you telling us that the airbag function, the climate
control, and the radio were all in one computer. Now you're backing
away from that. But what we have above is still ridiculous.
Explain how you have "smarts" in sensors or anything else without
there being a computer? It may be a $1 microcontroller, but it is a
computer. Perhaps you've missed the news. Want me to go find you
more stories that say your home is full of computers today besides the
one on your desk? Like the typical DVD player, microwave, fancy
washing machine, dishwasher, telephone, cell phone, etc. They are
all implemented using microprocessors or microcontrollers.
No but anything containing a microprocessor or microcontroller is.
Hence the stories that say cars today are chocked full of them.
Where is your story that says cars only have one or two? And what
is the big driving force for this computer centralization? It would
be like going back to the days of the mainframe, instead of a
distributed computing architecture.
Yes, enough said. The people who made that supposedly outrageous
claim are just engineers who design embedded computer systems.
Ohh. Stop the presses. In the other thread you made the claim that
it existed today and you knew it for sure. Now it's a coming
Obviously it's you who is the bean counter. Because if you were
involved in actual engineering design, you'd know that the cost of a
cheap microcontroller pales in comparison to the engineering and
development effort, system simplification via logical segmentation,
Then surely you are an idiot, because anyone that has knows that when
you partition a system in some logical fashion development is faster,
easier, less costly, easier to maintain, etc. Yet, your here arguing
that there is one computer controlling the airbags, the climate
control, and the radio. To what purpose? Take the radio. It
surely already has an electronics board containing the tuner, human
interface, etc. Why the hell would you not put a $1 microcontroller
in there to run it and instead rely on a computer somewhere else?
Putting it in the radio, you now have a fully functioning subsystem
that you can work with. Do it the way you claim, and you have to rely
on some computer being developed God knows where.
I see you have no answer to that one
Sure, just like were going to go back to the days of the mainframe.
You really are lost here in the wilderness. How the hell does one
powerfull CPU implement redundancy? In fact, it's exactly the
opposite because you have NO redundancy. How does it make fault
detection easier? I don't know about you, but I don't want the
computer that is controlling the airbag detonation to be co-mingled
with a bunch of other crap like the radio.
Please provide a single reference to back this up. How is the team
going to design the radio in Korea while the CPU for the thing is
being designed in Detroit? Yeah, it could be done, but anyone with
a lick of experience knows that it isn't done that way. You put a
cheap microprocessor or microcontroller on the board already in the
radio and then you have a self-contained functioning subsystem. Your
way, you save what? Nothing. You've made the design more complex.
You've made the development FAR more complex. And all you've
accomplished is to replace a $1 microcontroller with part of a
computer somewhere else that has to be MORE powerfull and hence costs
more. You haven't saved a thing. Which is why the trends are
exactly the opposite of what you claim, ie there are more computers
going into cars each year, not less.
BS. You put that radio code into the same computer that controls the
airbag and I don't care how you try to segment it, one thing can
effect the other. Hence, now to make any change to one function, you
have to revalidate the entire thing, especially when the computer can
detonate the airbags. PCs use all kinds of advanced modular
techniques too. Would you rely on yours to control the airbags while
also running God knows what?
Client/server is a classic example of distributed computing. Geez,
what percentage of the computing power today is mainframe versus 30
years ago? I've provided you plenty of links that say the number of
computers in cars today is dozens and increasing. Just provide us
with a link that says there are only one or two and that the trend is
to centralize the computing into one CPU. Link please.
On Sun, 14 Mar 2010 07:31:09 -0700 (PDT), firstname.lastname@example.org wrote:
Didn't read far enough before you started yakking back - the
It is already happening in a big way as I just explained.
I didn't say ONE computer was redundant.. I said easier to implement
redundancy - because only 2 pcocessors are required instead of 2 of
How does one company make peripherals for another companie's game
machine or whatever? They have the code, running on a computer, and
they make their part work with that code. it's done all the time in
the computer industry.
One company makes the hardware and another company writes the drivers
and the application programs to make use of the hardware. They even
"port" the applications to different operating systems and different
Yeah, it could be done, but anyone with
Except that's not how Delco, and numerous other companies are doing it
today. Many GM cars today will not function with the radio removed.
All kinds of things don't work because the radio is tied into the BCM.
And the radio can NOT be run without being connected to the BCM. That
IS how it IS being done TODAY.
> Your way, you save what? Nothing. You've made the design more
Hey, I don't know where you are coming from and what you know or think
I work with computers every day - and not just as a user. I also have
a professional knowlege of automotive mechanics and systems.
I know what embedded systems are - and how they are used.
You may call everything that has a microprocessor in it a computer. I
You may think that the future is a computer in every little component.
I happen to dissagree. The cycle that started with the large mainframe
and dumb terminals has shifted to a totally decentralized network of
connected computers, and is now shifting back to much smaller and
cheaper but infinitely more powerfull central computers with "smart"
terminals. The Client-Server model, and centralized processing is on
of the more common computer models of the day - along with the much
different and opposing "cloud computing" model.
Yes, the "smart terminals" are computers. But the whole scenario IS
shifting back to centralized processing in MANY regards.
What is driving this change? In large parts it is dollars. One
powerfull computer can now be built more economically than many
smaller computers. But that is not the whole, or even the main cause.
Security today is a very expensive part of IT spending - and it is
much easier to secure a single system than a spread out system - which
is why you will find the "cloud" to be a passing fad.
Just my opinion, but I've been around the computer world for quite a
long time (and the automotive world longer)
Will I be proven wrong? Perhaps. Would not be the first time.
But you are also wrong.
On Mar 14, 2:04 pm, email@example.com wrote:
Show us a single link that says that is true. I've asked for this
about 5 times now. I supplied you with a link to an embedded
computer design website that says exactly the opposite. Which is to
say that cars today have more processors than ever before.
So show us a reference that says any car manufacturers are
implementing redundancy. You really are going further and further
out on a limb. AFAIK, in every car made today if the engine control
computer fails, the car is kaput. Show us one where they have
redundancy and why the hell would anyone need redundancy for a radio?
In other words, you have no reference. Your just continuing to make
They make the peripheral as a logical self contained unit. Lets take
a hard disk drive. That drive has an onboard microcontroller, ie
computer, that spins up the drive, runs diagnostics on it, then waits
for commands, etc. It responds to commands from another computer.
No different than a radio in a car having it's own microcontroller
instead of relying on sharing some central CPU, as you claim. Still
waiting for a reference on that too.
The whole idea of a radio relying on a central computer is silly.
The radio obviously has buttons you push, maybe a dial to turn for the
station, volume, etc. It has a display to show the station name.
Now where does it make sense to put a microprocessor to take those
inputs, and act on them. Right there in the radio, or to send God
knows how many inputs, outputs off to some central computer?
They typically have THEIR OWN CODE, runnning on their peripheral, that
gives it onboard diagnostics and the ability to communicate with other
system components, accept commands, figure out what the command says
to do, and then do it. That typically requires a microcontroller in
the peripheral, eg the disk drive example.
No shit sherlock. But that has nothing to do with the
microcontroller and program for the peripheral itself, which is itself
an embedded computer. It's located on the peripheral, not
somewhere else. The drivers and applications are located in the main
CPU for the overall system. Got that? The Application program is
running on the Pentium. It makes a call to the disk driver which is
part of the OS. The disk driver breaks that call down into commands
to the disk drive. The microcontroller which is itself another
computer, takes that command, figures out what it tells it to do, eg
read sector 24539. Then it steps the motor 50 tracks and reads the
data. If it's an invalid command to the drive, it instead responds
with an error code. All that involves an embedded microcontroller
running a computer program on the disk drive itself. Capiche?
Kindly provide a link that says this is so, And if, as you claim,
the radio function is located in another computer for redundancy, it
seem pretty stupid that a dead radio would disable a car. If you put
the microcontroller in the radio, the car wouldn't die.
Sure, I believe that. The radio doesn't work so all kinds of things
don't work, Link please.
I have a degree in Electrical Engineering from MIT. And I spent 17
years in engineering, marketing and sales with Intel. And I say,
first, I never said everything that has a microprocessor or
microcontroller in it is a computer. I said anything that has a
microprocessor or microcontroller embedded in it, eg, a microwave
oven, a cable box, a cell phone, or many of the various modules in a
car, CONTAIN an embedded computer. That is the industry
definition. That is why the embedded computing site I provided as a
reference says there are dozens of embedded computers in a car
today. That is why there are frequent media stories of how computers
are today part of many common items and your house has dozens of
them. A computer in the cable box, furnace, microwave, dishwasher,
Obviously you are totally out of touch and discredited at this point
because that has been the continuing trend for 3 decades.
Client server is the classic example of distributed computing. You
have many clients and many servers replacing ONE CENTRALIZED CPU in a
- along with the much
Yeah, Larry Ellison, Scott McNealy and others have been promoting this
for about 15 years now. We were supposed to all be using just
intelligent terminals and downloading all apps from the network
cloud. You wouldn;t need local apps anymore. Yet, just about every
app I run today is on my PC.
BS. The whole industry has been driven in exactly the opposite
direction because you can buy a powerful PC or microcontroller so
cheaply. A whole microcontroller that could run a dishwasher or the
radio in a car can be had for $1.
Above I thought you were advocating the cloud as being were we were or
No, I'm not wrong. You seem to have a hard time here seperating what
you claim to know to be true as fact from opinions. So, now it's not
that this trend exists, that this is what is happening today, it's
that it's the "future"
I'm still waiting for that reference that says cars can have the
airbag control, climate control and radio all done in one computer.
Here's a reference that talks about an airbag control module and it
sure says nothing about that functionality being in some centralized
CPU. It says the decision making and control are in the ACU itself.
Sound familiar? And while this particular reference doesn't spell
out that it contains a microprocessor or microcontroller, it's obvious
to anyone with any computer engineering background that reading the
description of what it must do, this is how it would in fact be
How airbags work
An ACU from a Geo Storm.
The design is conceptually simple; a central "Airbag control unit"
(ACU) (a specific type of ECU) monitors a number of related sensors
within the vehicle, including accelerometers, impact sensors, side
(door) pressure sensors, wheel speed sensors, gyroscopes, brake
pressure sensors, and seat occupancy sensors. When the requisite
'threshold' has been reached or exceeded, the airbag control unit will
trigger the ignition of a gas generator propellant to rapidly inflate
a nylon fabric bag. As the vehicle occupant collides with and squeezes
the bag, the gas escapes in a controlled manner through small vent
holes. The airbag's volume and the size of the vents in the bag are
tailored to each vehicle type, to spread out the deceleration of (and
thus force experienced by) the occupant over time and over the
occupant's body, compared to a seat belt alone.
The signals from the various sensors are fed into the Airbag control
unit, which determines from them the angle of impact, the severity, or
force of the crash, along with other variables. Depending on the
result of these calculations, the ACU may also deploy various
additional restraint devices, such as seat belt pre-tensioners, and/or
airbags (including frontal bags for driver and front passenger, along
with seat-mounted side bags, and "curtain" airbags which cover the
Here;s another reference, this from a company that repairs and sells
airbag control modules:
They call them computers. And they don't say anything about them
being related to the climate control or the radio. Which makes sense
for so many reasons that are obvious to anyone with any engineering
experience. By putting the accelerometers and computer in one module
you have a signal self-contained package that is seperate from
everything else and far more easily designable, debuggable,
maintainable, and most importantly reliable.
Your references that say the airbags are controlled by a central cpu
that also does the climate control and radio would be?
On Mar 14, 5:14 pm, firstname.lastname@example.org wrote:
Yeah, among the bizarre things CL claimed:
Most cars today only have two computers, some only one
The trend is toward centralizing all computer functionality in one CPU
You can use a microprocessor in a system in such a way that it is not
Some cars have the same computer that runs the airbags also running
the radio and climate control and they did that to save money instead
of putting a cheap microcontroller in the radio.
All of that is clearly wrong. I especially like the idea of co-
mingling the airbag smarts with the radio. Anyone that has done ANY
engineering, product development, etc knows how that would complicate
things enormously and greatly add to the design complexity. So we
have the guys developing the radio in Japan and they have to work with
how they are going to implement the code, share the computer
resources, with the airbag team in Detroit or wherever? Instead of
using a $1 microcontroller like you'd find in a microwave and putting
it on the electronics board that is already in the radio and right
next to the pushbuttons, tuner, display, etc. Yeah, right! And
then you have the safety implications and system validation issues of
having the same computer control the airbags that can explode in your
face also running the radio and God knows what else.
And we're still waiting for a single link that backs any of this up.
On Mar 14, 1:23 am, email@example.com wrote:
I didn't see SA claiming any such device existed. All he's been
saying is that unless a thorough analsis is done of the design of the
cars involved, no one knows what exists or doesn't exist. That's why
there needs to be a thorough investigation of all the cars involved.
It's YOU who have been making many false statements and claiming to
know with 100% certainty how these cars are designed, what interlocks
are involved on the transmission, how they work, etc. As for proof,
I'm still waiting for any references from you for a number of your
A - New cars typically have only two computers, and some have only one
B - The trend today is toward using one centralized CPU in cars to
C - The airbag, climate control, and radio are done with one computer
on at least some cars
D - You can use a microprocessor in a system and have it not be a
On a basic level an abacus and the old mercury bulb thermostat on the
wall can be called a "computer". It would really help if everyone would
agree on what constitutes a "computer". Back in 1960's when I got into
playing with computers, the school had an "analog computer" the size of
a half dozen lunch boxes of the era. In 1965-1966 I started playing with
what was the modern conception of a computer. Univac and IBM mainframes
that took up whole floors of buildings. The Univac had glass doors and
lots of blinking lights to entertain you. Ask any school kid today...
"What is a computer?" and get back to me. Perhaps a six year old can
settle the argument?
On Sun, 14 Mar 2010 00:18:36 -0500, firstname.lastname@example.org wrote:
Sorry, but you are 100% wrong. Do you understand the concept of a
deadbolt on a door preventing it from opening no matter how much you
twist and pull the mechanical door knob mechanism which is completely
separate from the deadbolt, but normally would allow the door to open?
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.