My Theory on Toyota's Unintended Acceleration Problem

I do not think you will find a single cause.

1. The Lexus with the CHP driving had inccorect floor mats which were not secured.

2. Sikes certainly appears to be malicously and deliberately fakeing a hardware problem.

3. The Dr with UA in reverse is a classic pressing the wrong pedal scenario. (since the ICE does not even power the wheels in reverse, sticking throttle plates can be ruled out)

4. Some accident somewhere may perhaps be kinda due to a system fault, maybe.

Toyota is in a no win scenario, with no cases that are system failures to examine, no fix can be offered. (Optional snotty remark: Yet internet 'experts' can start social networking memes based on wild speculation)

 

"Closing post for the weekend"?? What, do you only theorize
during business hours? Well, here's a little overtime for you:

The pedal signals are not tiny amplitude. They use a 0 - 5 volt
range, and stay within defined limits over that range where actually
hitting 0 or 5 volts triggers an error condition [showing a short].
In some other cars the tolerances may be a bit more sloppy, but
I and my scope can assure you that the level tolerances in a Prius
are quite intact.

UA reports have been rolling in for gen II and gen III priuses, I
think.

Stepping on the brake does not "override the tiny amplitude signal"
from the pedal. It changes a *different* signal from which the ECUs
derive the braking desire of the driver. As I pointed out in another
thread, the braking demand needs to become substantial, i.e. a level
that would actually slow the car down, before the "override" condition
is determined, but it does get determined and works. That's done
in software.

It didn't happen in Sikes' vehicle because he didn't press the brake
pedal hard enough, but more likely rode it lightly for miles.

Given the prevalence of RF-emitting devices that consumers routinely
bring into their cars, Priuses and not, your theory would have orders
of magnitude more cars running away instead of a few isolated reports.
Especially people who [successfully, I'll point out] have installed
various types of higher-power transmitters and had no issues except
perhaps with their audio gear [where millivolts *do* matter].

What exactly is a "Radio Shack Ground Isolator" and what do you think
it actually does? Can you give us a circuit diagram and discussion?

The NHTSA is rapidly becoming irrelevant during this. Take the case
of Rhonda Smith's lexus. Someone *else* owned that car for the whole
intervening time after her incidnet. Where's *their* report on its
performance/safety? Now the NHTSA has it [at taxpayer expense, I
might add] and we haven't heard squat about what they're actually
going to do with it. And don't even get me started on Gilbert;
my take on his pretense of "work" is already in other threads.

I'm not even sure why I'm sticking my oar into this, but w_t_w, you're
missing a lot of the big picture. In other words you're trying to
narrow the field *too* far, and fixating on one thing.

_H*

 

I've done a lot of work in the field of "by wire" control, be it industrial control bus and especially military and some civvie fly-by-wire aviation. I will state one truth that cannot be disputed:

There is NO SUCH THING as a Perfectly Reliable system

Deal with it

Depending on how much time and money one has to throw at the engineering, one can expect three 9's, even five 9's, reliability. To go from three 9's to five 9's invokes an exponential order of magnitude increase in cost

I don't expect everybody to understand how the Agilent 7000 series Mixed Signal Oscilloscope works, nor do I reasonably expect everybody to understand how major CRC errors can cause glitches. There are other sources of glitches. I went back to this MSO screenshot and highlighted what I thought was the important data



which if one "drills down" by following the soft buttons on the side of the scope, discovers the major - indeed fatal - error was introduced by floating the reference.

In my experience, the best way to introduce fatal glitches into such a complex control bus is to play havoc with grounds, either corroded, loose, or shorted elsewhere. A Martinair Boeing 767 had major glitches after a battery bus shunt was improperly serviced

http://flightsafety.org/amb/amb_jan_feb99.pdf

Now, believe it or not, there was no requirement to test for this scenario. Hence, instead of a graceful recovery, major glitches were introduced. This scenario IS now tested for, which in the application of a fly-by-wire aircraft is crucially important

The scenario of a floating reference will *usually* introduce a Fatal Error which - due to the nature of how CANBus is implemented in Critical Applications - makes every device send a Stop and go into a graceful shutdown for a brief period of time

The devices will then enter a startup again. The sequence of startup is very important, one must carefully pick the so-called Hot Restart Group to avoid a scenario where the devices restart, detect fatal errors, Stop, wait, restart, detect Fatal errors, etc, an endless Hot Restart Sequence

This happened to one of the recent Mars rovers

For non-critical applications, rather than Hot Restart Sequences I would have to guess the device will enter a Limp mode of some sort. I don't have any experience with automotive, so I can only hazard a guess

CANBus sends out tons of diagnostic data during normal communications, but its up to the end user/manufacturer to decide whether to capture and record that data. In the aviation world, many crashes and incidents would have been prevented if only more data was recorded

For example, the 737's that had weird rudder reversal problems. Two were unrecoverable and crashed, and the odds are similar to what Toyota cars are experiencing with UA

The Rudder Story

http://www.ntsb.gov/recs/letters/1999/a99_20_29.pdf

A similar design PCU is used on 747 elevator control. However, due to stricter European reporting requirements, a British Airways 747 with an elevator glitch had recorded not only basic parameters, but such things as pilot controls position, applied pressure to pilot controls, actual position of elevators, "command disagree" flagging of PCU operation counter to pilot applied force, etc

Such recording would probably be costly for the automotive industry. I really am unsure how to monitor for such a situation that doesn't dramatically increase the cost of a car

But I'm also old enough to remember when it was fairly common to have a car accelerate out of control. Usually due to the vacuum powered cruise control sticking, the carb return spring breaking, the transmission kickdown lever jamming, or the carb throttle cable jamming

 

On the contrary! Criticizing a hypothesis is the way it is tested. One may offer alternative explanations (which is what you are asking for) but it is equally valid to point out errors of fact or logic in the hypothesis.

Note, for example. Hobbit's post above. Your assertions about the internal workings of the Prius are mistaken, so your conclusions lack validity.

In order to eliminate an invalid hypothesis from consideration.

 

1:09PM EDT Sat SoFla TwiliteZoneTime:
Yesterday, I found a 3+yr old complete "Delphi SkiFi3 XM radio" setup all the car accs included. WOW, I was given this as a birthday gift & now I finally get the chance to be "commercial Free in my '07 !!!!!!!!>"
Or will I ??
This Delphi came w/ 2 ANTENNAS-both having about 20+ft. of thin wire. Yes, I know 1 is the Normal needed antenna & the other is for FM transmission over an unused FM channel--in So. Fl there R NO OPEN channels !

I am posting here 'cause I noticed the OP also has an '07 and I have great anticipation on weather is was a dealer or personal install. There are no REAL tips in the instruction book on where & HOW TO install & hide the antenna wire. So here I sit w/ baited breath & lots of coffee as I finish mowing the lawn & watching AMERICAN COLLEGE Football & hopefully someone has what i need to avoid a potential Install...Plug & Pray.

 

The unintended acceleration isn't found ANYWHERE else other than in America. So out of all the millions of other Toyotas in the world, it only affects American Toyotas.

Hmmm, why's that? Litigation hungry lawyers and some people who will blame anyone else for their own error?

Unintended acceleration is caused by people pressing the accelerator OR by getting non authorised mats stuck under it.

Pure and simple.

 

Welcome to Amereeeeca!

 

Grumpy, you had one in a parking lot. Gramps sideswiped while parking. Grandma gets out of car & stands in front of Prius to guide it into a space. Next thing you know (as per Gramps) UA takes over, Grandma is hit & Prius is stopped when it collides with another parked car.

Never seems to happen to young people.

 

Well that kind of UA happens to all sorts of cars, and you're right, the drivers usually happen to be 80+.

I have a story on this as well, involving an unnamed taxi driver in his late 70's

 

Smart pedals won’t put the brakes on driver error / UCLA Today