Brake And Front End Advertise | Subscribe | Contact Us | About Us

TOYOTA Recall: No Codes=No Problem: Testimony of David Gilbert

February 24, 2010
Here is the testimony of David Gilbert a professor of Automotive Technology at Southern Illinois on the Toyota unintended acceleration recall. His basic contention is that if the ECM is not able to detect a problem with the throttle how can it enter into a limp or safe mode. This is great reading, I have underlined the interesting parts if you are in a rush.
Bookmark this website Bookmark this website
David W. Gilbert, PhD.
Professor of Automotive Technology  
Southern Illinois University Carbondale
 
Testimony for the Committee on Energy and Commerce,
 Sub-Committee on Oversight Investigations
Toyota Sudden Unintended Acceleration
February 23, 2010

 
Chairman Waxman, Sub‐Committee Chairman Stupak and the honorable members of the Committee on 
Energy and Commerce, thank you for holding this important hearing and allowing me the opportunity to 
testify before you today.  

I have been a technical educator involved with automotive diagnostics and trouble  shooting for almost 30 years.  I have been witness to many evolutionary changes over that 
time.  When I first began teaching in 1981 at Northeastern Oklahoma A&M College, 
electronic fuel‐injected vehicles were relatively new technology.  Over the years, automotive 
technology has continued to progress from fundamental mechanical systems to more 
sophisticated electrical and electronic systems.  Now, as an automotive technical educator 
at Southern Illinois University Carbondale (SIUC), I have found electrical diagnostic skills to 
be supremely important diagnosing and repairing modern vehicles.  And, I have spent many 
hours studying and analyzing new electrical circuits and components.  Based on my 
knowledge of real world failures of components, I purposely duplicate multiple types of 
electrical problems in donated vehicles for my students to study and diagnose.  This 
provides my students an opportunity to analyze wiring schematics and service information, 
and actively solve diagnostic problems.  SIUC Automotive Technology graduates have found 
employment in virtually every aspect in the automotive industry.  Students graduating from 
SIUC have the technical skills to work closely with automotive design engineers to ensure 
reliable vehicle service in real‐world situations.  I believe the exemplary student placement 
record, is a result of the academic rigor of the program and the emphasis on technical 
problem solving. 

It stands to reason, that my daily teaching responsibilities would include the 
application and understanding of electronic throttle control diagnostics.  I have the unique 
perspective in my employment, to research and study multiple vehicles and electronic 
throttle control system designs.  In this preliminary report, my initial findings question the  integrity and consistency of Toyota Electronic Control Modules to detect potential 
electronic throttle control system circuit malfunctions.  The absence of a stored diagnostic 
trouble code in the vehicle’s computer is no guarantee that a problem does not exist.  I 
instruct all my automotive students with this fundamental statement: You can have a code 
with no problem ‐ and a problem with no code.

   
My curiosity in the Toyota electronic throttle control system began simply with a 
search for the truth concerning sudden unintended acceleration.  I recently purchased a 
2010 Toyota Tundra, and with the growing attention in the media to what seemed to be 
increasing events of sudden unintended acceleration, I made the decision to investigate the 
foundation of these claims on my own.  Based on my working knowledge of electronic 
throttle controls, I did not expect to the system to be easily  fooled without detecting a 
circuit fault and setting a diagnostic trouble code.  It was late one evening when I made a 
startling discovery; electrical circuit faults could be introduced into the electronic throttle 
control system without setting a diagnostic trouble code.  This discovery opened a window 
of opportunity within the electronic throttle control system for a potential problem with no 
code. 


Without a diagnostic trouble code set, the vehicle computer wil l not logically enter 
into a fail‐safe mode of operation.  All vehicle manufacturers have recognized the 
importance for electronic throttle control systems to perform exactly as they intended.  
Since the vehicle computer will only react to defective sensor inputs outside of the range of 
programmed limitations if the circuit is not defective; it must be good.  Knowing that 
properly operating electronic throttle control system circuits and components are vital to 
safe vehicle operation, I proceeded to investigate the problem  with more urgency.  Because 
of its important role to accurately convey vehicle driver demands for throttle opening, 
accelerator pedal sensor voltage inputs need to be confirmable by the vehicles computer as 
absolutely correct.  A complete or partial failure of these electrical circuits, sensors, wiring, 
or actuators in combination with an absence of fail‐safe strategies could potentially result in 
a runaway engine.   

The importance of these issues raised in the electronic throttle control system fail‐
safe strategies should not be underestimated.  Sudden unintended acceleration of a vehicle 
is a very serious safety concern that should be addressed without delay.  
 
Vehicle manufacturers clearly recognized the important requirement for ETC 
systems to perform exactly as they intended.  A failure of the  electrical circuits, sensors, 
wiring, or actuators could potentially result in a runaway engine.  Electronic Throttle 
Control (ETC) systems needed the added redundancy of certain sensors and electrical 
circuits to ensure safe and reliable operation.  In addition, the ECM’s were programmed to 
detect operational abnormalities or defects in ETC components and their related electrical 
circuits.  The intent was to build an ETC system that would always fail‐safe in the event of 
potential problem.   

The purpose of my research study was to contribute to a better understanding of 
electronic throttle control system malfunctions and the fail‐safe detection capabilities of 
selected vehicles equipped with electronic throttle controls.  More specifically, this research 
examined the fail‐safe detection capabilities of electrical circuitry designed to prevent 
sudden or unintended acceleration of electronic throttle controlled vehicles manufactured 
by Toyota Motor Co.  The Accelerator Pedal Position (APP) sensor was identified in the 
review of manufacturers’ service literature as a significantly important ETC input for all 
vehicles used in the study.  Since vehicle driver demands are electrically conveyed through 
this high priority sensor, basic testing was focused on the APP sensor, voltages, and 
associated wiring circuits.  A secondary purpose was to identify areas of further research of 
ETC fail‐safe detection capabilities of Toyota Motor Co. vehicles and other vehicle brands.  
This limited analysis attempted to identify and characterize potential safety concerns of 
Toyota Motor Co. vehicles, as well as other vehicle manufacturers using electronic throttle 
control systems.   

After completing preliminary tests for Accelerator Pedal Position (APP) sensor 
signal voltages for the Toyota Electronic Throttle System I examined, it was determined that 
Electronic Control Module (ECM) malfunction detection strategies were not sufficient to 
identify all types of fundamental APP sensor and/or circuit malfunctions.  
Some types of 
Electronic Throttle Control (ECT) circuit malfunctions were detectable by the ECM, and 
some were not.  Most importantly,  the Toyota detection strategi es were unable to identify 
malfunctions of the APP sensor signal inputs to the ECM.  APP sensor signal circuits must be 
undeniably correct to electrically convey the appropriate driver commands to the ECM.   
With the two APP sensor signals shorted together through a varying range of 
resistances, all four Toyota vehicles tested thus far reacted similarly and were unable to detect the purposely induced abnormality. 
 

The types of signal  faults introduced into the APP circuit should have triggered the vehicles’ ECM to illuminate a warning lamp within 
seconds.  The ECM should have then set a Diagnostic Trouble Code (DTC), entered the 
vehicle fail‐safe mode, and reduced engine speed and/or power.  When the two APP signal 
circuits are shorted together, the redundancy of the APP circuit design is effectively nullified 
and lost.  In other words, neither of the shorted APP signal circuits can be verified by the 
ECM as either; correct or incorrect.  The condition then exists for a serious concern for 
driver safety.  In the tested Toyota ETC vehicles, incorrect or corrupted APP sensor signal 
inputs could potentially result in unwanted engine speeds.  Additional research should be 
done to determine if other vehicle manufacturers may have similar inconsistencies in ETC 
circuit fault detection.  
Using shorted APP signal circuit fault conditions purposely installed on the test 
vehicles, and with known resistance values that would not set a DTC, vehicle operational 
behaviors were also noted.  It was observed that all test vehicles could be operated without 
the ECM detecting the induced malfunction.
  Depending on the resistance value of the APP 
signal circuit fault, a vehicle may or may not experience noticeable changes in accelerator 
pedal operational behavior.  Observed accelerator pedal operational characteristics 
included: normal response, sluggish response, and travel with i nconsistent engine speeds.  

It is conceivable that a driver of an ETC vehicle may not notice that an APP sensor and/or 
circuit malfunction currently exists.  Without the aid of an illuminated MIL, a driver could be 
unaware of electrical problems within the ETC system.  In addition, the shorted APP signal 
circuits were connected momentarily to the sensor’s five‐volt supply circuit with the vehicle 
in drive.
  In all test vehicles, the ECM did not set a DTC and  the engine speed increased 
rapidly to full throttle. 
 This result shows that unusual or sudden unintended acceleration of 
the vehicle was possible in the ETC test vehicles.  It should be noted that in all test vehicle 
cases, the electronic throttle valve instantaneously moved to wide‐open position when the 
fault was introduced.  More research should be done to determine the extent of Toyota ETC 
vehicles that could be affected by this condition.   

In review of the Toyota service information, collected vehicle  data, and performance 
observations; some general assumptions can be drawn from the research completed to date.  
The inability of the Toyota ECM to detect certain types of short circuit malfunctions could 
fall back to the basic design of the normal APP signal voltage limitations.  The parameters for APP signal short circuit fault detection are apparently too lenient.  In the Toyota ETC 
system, the APP sensor signal voltages rise simultaneously in direct response to accelerator 
pedal depression.  With this design, interconnected signal circuits could be more difficult to 
identify with a circuit fault detection strategy that uses only threshold voltage limitations.   
In this preliminary report, the initial findings question the integrity and consistency 
of Toyota ECMs to detect potential ETC system circuit malfuncti ons.  The importance of 
these issues raised in the ETC system fail‐safe strategies should not be underestimated.  

While the small sample of Toyota vehicles cannot be representative of all, these primary 
findings most certainly warrant further investigation and study.  Additional Toyota vehicles 
of different build years and models should be evaluated for their capabilities of ETC system 
circuit malfunction detection. 

A second recommendation should be a thorough technical investigation and 
evaluation of ETC fail‐safe strategies of Toyota, and possibly other vehicle manufacturers, 
that experience sudden unintended acceleration that do not appear to be caused by floor‐
mats or sticking pedals.  Priority would be studies of identified vehicles with a high 
incidence of ETC system related incidences, concerns, or failures involving sudden 
unintended acceleration. 

Submit a Comment   Comments (16)
Comment by:
carl blackman
3/26/2010
7:23 PM
I have a 2006 Toyota Innova [ built late 2005 ] left front wheel [ 32,000 kl on wheel ]

the wheel failed, split almost exactly in half. by looking at it can see air bubbles

in the broken area. a recall for innova's in Vietnam is in effect for wheels made by

TMV, not sure who built these wheels as only marking is TMR?? need some feed back

to see if more innova's or other Toyota's have this problem.

thank you

carl f blackman

an American living in the Philippines

063 088 309 0036 some times depending on country calling from

you have to drop the [ 0 ] on 063 and 088



carl@filipinavisa.com photos at www.filipinavisa.com

for info on Vietnam recall go to www.lookatvietnam.com/2008/toyota-vietnam-recalls-innova

have heard from insurance agent that another Innova owned by a famous basketball

player just this week had the same problem. Philippine gov. lab now has my 5 wheels

for testing, keep your eyes and ears open for more of this, will advise of test results, others

may be at risk?
Comment by:
Larry Carley
3/9/2010
4:05 PM
For Hugh: There have been numerous instances of Toyota cars with electronic throttle controls experiencing unintended acceleration, even after recalled vehicles have been "fixed."



Toyota has yet to identify the true cause of their problem, and is spending more effort trying to discredit Dr. Gilbert than to investigate the electronic glitch in their throttle control system that is causing the throttle control motor to go wide open.

Comment by:
Hugh Farber
3/9/2010
3:42 PM
I had sudden acceleration occurence in a nearly new 2003 Camry when pulling up behind a parked vehicle. I applied the brake, which caused the rear wheels to skid on the concrete,as the engine raced until I hit the vehicle ahead at which time my vehicle indeed stopped and engine rpm retuned to idle. A series of activities followed with Toyota, the local dealer, and myself. The local dealer finally replaced the master cylinder at no cost, I paid for vehicle repair-- and Toyota disclaimed any resposibility. The story is long and detailed, but I believe this was caused by a faulty brake/cruise control/accelerator system which is electronically controlled.



FYI-- pass on to Prof Gilbert//
Comment by:
cds3154
3/8/2010
8:01 PM
Where are the detailed schematics for the hookup up of all the circuits ? Can anyone show me where this is ? I have never analyzed anything as an electrical engineer without that.
Comment by:
Joe
3/8/2010
12:24 PM
A known resistance between 2 inputs is a single point failure - pinch the 2 wires together under a clamping pressure (screw/bolt) and you will break down the insulation and provide exactly what Dr. Gilbert found.



Also, a linear relationship with offset is, as Mike mentions, is strange. Most redundant sensors that are safety related use a crossed pattern - both are linear, but one goes from .5V to 4.5V, and the other from 4.5V to .5V (never to 0 or 5V). In this way, the failure above would be detected quickly - they would tend to track each other depending on the resistance.



Toyota has a BIG problem and needs to fess up, or they will see more sales erosion. Stop denying it so hard and admit there may be unknown electrical issues.



Also, get a REAL consulting firm, not a testing firm. These problems are never solved by testing - only by good analysis.
Comment by:
S.E. Olson
3/8/2010
9:11 AM
Apologists for Toyota say the Gilbert has not proven anything because he hasn't produced the proverbial smoking gun. This is true to a certain extent, because he hasn't tested one of the electronic control systems in a car that ran away and discovered the fault in that system.

That's a good point, and certainly his findings are more suggestive than conclusive. Research should continue, maybe concentrating on vehicles that had failed.

On the other hand, the investigating congressmen have asked a very telling question of Toyota, essentially asking 'you (Toyota) say that it isn't the ECM, but you don't present any records of the tests that you did to arrive at that conclusion. Why not?' In other words, brings us the test results and records. If they don't, I think that Toyota is in big trouble.

Vehicle operating conditions are very "real world." Snowy, salty shoes operate those accelerator pedals, cars splash through water, high-resistance connections vibrate a little bi
Comment by:
Jean Kennedy
3/7/2010
12:26 PM
My toyota camry has had this problem f0r many years - but not any more - last summer, I had 2 new circuit boards replaced due to a short circuit after driving through deep water - the problem is gone, the sudden acceleration must have been electronic.
Comment by:
carquestions
3/3/2010
11:28 AM
This is the Kane and Gilbert Hoax - please see the video on youtube. He set out to stop a system from recording a fault and succeeded - so what? put bubblegum over the smoke alarm sensor and see if that works - It could happen
Comment by:
Ray Rogers
3/1/2010
9:05 PM
I am a retired EE and spent about 5 years working at Chrysler; and several other years designing instruments.

I know the Engineers at Chrysler Jeep took a very professional pride in their work and the end product.

The point that should be remembered about automotive usage is that they must run on millions of units over tens of years. In that case even very remote conditions can develop. The failures can be vendor, manufacturing, or design (heaven spare us).

Normal electronics design includes single point failure analysis (all), and depending upon the severity of a failure two or three point analysis.

Failures will occur and I am sure that Dr. Gilbert "tuned" the failure; the lack of detection and saving the error code is very disturbing. Also disturbing is that earlier failures apparently didn't evoke "root cause" analysis ( a term referring to finding the real reason for a failure). Not many people can be blamed for deliberately crashing their car at 60; although I have done some really dumb things:)

I have no real stand on the issues. I don't own a Toyota and know only one person the problem happened to; although he is not a happy camper.

Comment by:
Ray Rogers
3/1/2010
9:04 PM
I am a retired EE and spent about 5 years working at Chrysler; and several other years designing instruments.

I know the Engineers at Chrysler Jeep took a very professional pride in their work and the end product.

The point that should be remembered about automotive usage is that they must run on millions of units over tens of years. In that case even very remote conditions can develop. The failures can be vendor, manufacturing, or design (heaven spare us).

Normal electronics design includes single point failure analysis (all), and depending upon the severity of a failure two or three point analysis.

Failures will occur and I am sure that Dr. Gilbert "tuned" the failure; the lack of detection and saving the error code is very disturbing. Also disturbing is that earlier failures apparently didn't evoke "root cause" analysis ( a term referring to finding the real reason for a failure). Not many people can be blamed for deliberately crashing their car at 60; although I have done some really dumb things:)

I have no real stand on the issues. I don't own a Toyota and know only one person the problem happened to; although he is not a happy camper.

Comment by:
HD
2/26/2010
5:08 PM
I could still be a electrical/electronic problem though.
Comment by:
HD
2/26/2010
5:06 PM
Reading the Exponent report and Dr Gilbert article it seems quite unlikely that the problem occurred in a way Dr Gilbert describes it.

He states that he doesn't get a DTC (Diagnostic trouble code) when he makes his experiment.

Well, His experiment is like disconnecting the original gaspedal and replacing it with another!

Reading the Exponent report you will find out that there are two sensors in the pedals. 0-5 V. Furthermore there is an offset of about 0,8 V between them (Example: If pedal pressed halfway one of the sensors give 2 V and the other gives 2,8V to the ECU. The failsafe algoritm I would guess quite certainly checks that the offset is 0,8 V.

This way you capture the event that both inputs are shorted to +5V (or battery). What Dr Gilbert seems to have done is to have shorted sensor A to +5V and then added a resistor between sensor A and sensor B so that the voltage on B is 4.2 V. Very unlikely double failure I would say.

I could still
Comment by:
Stephen
2/26/2010
12:27 PM
Mike, would a failure to identify a sensor fault lie with the ECU software, in sensor systems outside of the ECU computer, or both? In other words, could a failure in fault-tolerance be fixed by simply reprogramming the ECM, or is the ECM getting bad info with no chance of catching the problem, or is the truth somewhere in between?
Comment by:
Mike Mason
2/25/2010
10:49 PM
Dr Gilbert's findings are very interesting. The issue seems 2 fold with Toyota's system: First the redundant APP signal is linear, and second the ECU does not recognize a 5 volt return on both of the APP signals as out of range



I have tested the first myself, the APP signals maintain approx 0.8 volts difference in a strict linear fashion. No other vehicle I have tested uses a linear differential between the two APP sensor signals. I am concerned that the ECU did not recognize that there no longer existed this 0.8V differential when the short was introduced. I am also curious as to Mr. Gilbert's stated use of "known resistance values".
Comment by:
P. Showman
2/25/2010
3:58 PM
Dr Gilbert reported that two short circuits at the same time (i.e. a double failure, with two pairs of wires touching) could cause the problem while not triggering the normal override. I don't see how a cell phone can cause a short circuit.



But using a cell phone while driving has been shown to be dangerous for other reasons!

Comment by:
Dick Abell
2/25/2010
2:02 PM
What does Dr. Golbert think causes the fault to occur. When I go into a hospital intensive care unit ; the hospital has requested us to turn off our cell phones. Could cell phones cause a fault to occur.
Submit a Comment   Leave A Comment
Name   (Required)
Email   (Required, never displayed)
Comments

Please enter the text from the above text box

 
© 2010 Babcox Media
3550 Embassy Parkway, Akron, OH 44333
330-670-1234
(FAX) 330-670-0874