Interesting. So now the "Impossible" becomes "Not Likely"??

 

The Earth's sun is going to implode tomorrow.

"Not Likely" or "Impossible" are quite often the same.

 

ROTFLMAO

 

You should read the reports on the ~late 70's / early 80's issues and fixes related to AUDI's "UInintended acceleration" issues.

 

Here's the problem with this research:

It sounds like in order to perform this test, they either removed the inverter from the vehicle, or connected 12 volts directly to the inverter rather than having it fed in by normal means (because it would be much harder to test it with everything connected normally).

The reason that's an issue is because VCFRONT would generate "Power Reduced" errors if the steering motor caused a voltage sag that severe for a certain period of time, and throw errors at you about that the car might not restart.

Of course what they could mean is that those sags are extremely short, like measured in milliseconds, and that's why VCFRONT doesn't pick it up - because it's in between voltage samples. But you know what can pick that up? The human eye. If the 12V system has a millisecond long voltage sag, you're going to notice it as a screen flicker inside the cabin, or of it's night outside, in the headlights. And if this was normal behavior for the steering motor, you would notice it a lot.

So what this research has to refer to is a steering motor with a defect that's intermittent enough for VCFRONT not to see it, but often enough for the inverter to sample the baseline voltage and decide that 0 volts is full acceleration. That's rare enough that it's probably never happened outside of that one testing lab.

Also calling into question who paid for this, because the only reasonable way anyone could profit from paying out money for this kind of research is if a class action lawsuit is being planned based on unintended acceleration claims, and some law firm planning it needs some evidence they can present to a jury that there is indeed a defect (as rare as it may be) and therefore that all of the victims of unintended acceleration accused of hitting the accelerator by accident should be compensated.

In other words, it appears the old Toyota Prius class action might be revived against Tesla.

 

I’ll try to replicate. Each of us should and share our results. Maybe a poll?

Bob Wilson

 

I’ll try to replicate. Each of us should and share our results. Maybe a poll?

You won't be able to outside of a lab environment.

 

No problem. I've downloaded the technical paper and will look at it later. Then I'll see if it can be replicated with my Model 3.

Bob Wilson

 

Here's a voice with knowledge on how these systems work

 

There are two credible sources I follow: Munro Associates and; Phil Sadow, aka Ingineer. When they offer an opinion, I listen very closely.

Bob Wilson

 

Here is what seems odd to me:
"3) the
% pedal with CREEP added is then tested to determine if the rate of movement is within the
limits for a normal driver-induced signal. If this test fails, then a CAN alert is posted, and the
torque is limited to zero. "

In other words, there is already a test that should pick up the sudden apparent (but incorrect) change in pedal position and ignore it.

Perhaps it's the journalists reporting that is faulty, but the described issue isn't really a fault with the inverter design: it's a fault with the pedal positioning sensor design (which is on the same board as the inverter).

More:
"This higher inrush current can't be supported by the "12V'
battery, which can supply a maximum current of only 100A or less, "

The car's 12V lead-acid battery can only provide 100A? Really? The article quotes a GM recall to show that battery voltage can drop to 8V, which I would have thought sufficient to keep the reference voltage of 1.65V accurate. Also, the GM car doesn't have the DC-DC converter adding another possible 200A of current to keep the 12V rail high enough, unlike the Tesla.

"this "1.65V" calibration voltage
can have large negative-going voltage spikes or dips on it as a result of the electronic power
steering drawing more current during low speed turning maneuvers (which can draw over 300
amps at times). "

So it's the power steering causing the power drop? But wait, earlier, this was claimed:
"This is because the "1.65V" supply
voltage can also have very short-duration negative-going voltage spikes superimposed on it as a
result of in-rush currents from electric motors and electric solenoids turning on that are
powered by the "12V" supply voltage from which the "1.65V" calibration voltage is derived."

So it's not the power steering? Or is it? How suddenly can the power steering motor demand power? Surely drivers gradually increase torque on the steering (gradually in comparison to the millisecond times that are being discussed).

Finally, the report shows a couple of graphs of accelerator depression in unintended acceleration cases. What I see is that the accelerator takes about a second to go from zero to full. The paper attempts to explain this by suggesting that the rise times are filtered (but there is no evidence for this, but this appears to be directly contrary to the information provided earlier that implausible rise times are rejected. It cherry picks tow graphs to show similar rise times as evidence for this suggestion, but two similar graphs could simply be coincidence. One would need to see all the graphs for this.

The overall theory appears to be one that the author has been promoting for over a decade now:

Dr. Ronald A. Belt’s Sudden Acceleration Papers

%

www.autosafety.org

 

This is excellent news. If there's a problem that's been discovered, it can now be fixed.

 

Thank you for "Dr. Ronald A. Belt’s Sudden Acceleration Papers".

I have an intermittent message about 'Unable to shift because accelerator is pressed.' A couple of quick flaps clears the condition and I can then shift and drive off. I typically see this about once per week and figured eventually to swap out the accelerator to clear the problem.

I was driving a first generation Prius when the 'sudden unplanned accelerator' brew ha ha erupted. The first Prius version used a dual potentiometer sensor that could easily wear to a point car would go into 'safe home' mode. The next generation used dual Hall effect sensors that were not subject to this wear problem. But limping home in 'safe hold' was only dangerous from other traffic.

Once people learned the brake controller recorded the last "n" seconds, the reports of unintended acceleration rapidly disappeared. Accident investigators have access to these readers.

There was a latent defect traced to unsecured floor carpets and mats from the death of the Saylor family in a Lexus. My second Prius was delivered without floor clips to hold the driver side, floor mat down which I corrected by driving back to the dealer two days after taking delivery.

There was a 'tricky' way using a decade resistance box to show a potential problem to non-technical reporters. This identified a problem of dual sensors using both ascending or descending voltages from the dual sensors could command acceleration. A better approach in German cars was one sensor ascending and the other descending.

Uncommanded acceleration can occur:

• 99.98% - stomping on the accelerator instead of the adjacent brake pedal
• 0.01% - unsecured floor mats or loose objects (aka. shoes) in the driver footwell
• 0.005% - fraud trying to blame the car for the driver's screw up
• 0.004% - previously undiagnosed, latent defects, usually takes more than one OR cosmic ray
• 0.001% - act of God or otherwise unreproducible and undiagnosed problem(s)

WHAT TO DO IF IT HAPPENS TO YOU

​

Shift to "N" by raising the shifter lever for a couple of seconds.​

​

Regardless of what the car is doing, you can always shift into "N" and the car will no longer accelerate. To be safe, you MUST practice this at least once! It also works if you have a driver problem who needs a 'time out.'​

The criticism of the Belt paper sound reasonable to me. Especially about the "100 amp" overload of the 12 V lead-acid battery.

Perhaps an end-of-life lead acid battery or the newer LiON battery might lead to the reported voltage sag. I haven't read the paper and other than going to the parking lot and rapidly swinging the steering wheel while monitoring the 12 V battery buss, I'm not willing to make a call. My 12 V battery was replaced less than a year ago so it is still fresh.

In my work and home history, I have seen multiple cases where marginal power supplies have made digital electronics behave badly. Whether the "Belt" hypothesis can be replicated determines my course of action.

Bob Wilson

 

WHAT TO DO IF IT HAPPENS TO YOU

Shift to "N" by raising the shifter lever for a couple of seconds.

Regardless of what the car is doing, you can always shift into "N" and the car will no longer accelerate. To be safe, you MUST practice this at least once! It also works if you have a driver problem who needs a 'time out.'

I would sugggest that in a Prius, those frighteningly quick little buggers, that this may be the aopprorpiate action. However in a MUCH MORE POWERFUL vehicle such as a Tesla Mdl 3 (AWD), that is NOT a good action as "a couple of seconds" is beyond deadly at full power.

I would suggest to push the Park button as that will disengage drive faster AND (after some undefined time but much less than "a couple of seconds") apply the rear brake calipers,

As far as the motive for the author, who knows? But I would like to mention Ralph Nader of past automotive fame. He is/was a total freak and a very weird person whle the Corvair was indeed a total death trap.

 

 

Still sitting here ROTFLMAO

 

@garsh That twitter response illustrates why I think the entire "study" was done just to file a class action - because that kind of technical confusion is going to impress a jury in a courtroom. No one is going to spend that kind of time and money because they "found a flaw and they want to have a clear conscience". It was done for the same reason why it was to the Prius - because it's new electronic technology that people are scared of, and it's so easy to turn up the volume on that fear and make some money off of it.

I suppose it could also be big oil or big auto money behind it to prove how much more dangerous EV's are, but that's more of a long shot.

 

Or a California billionaire who hates Tesla.

 

I like PARK too but always felt bad about the gear lock that PARK engages ... at the appropriate time. I had not tried that in my Model 3 so I may do the 'forbidden experiment' to find out.

Bob Wilson

 

I like PARK too but always felt bad about the gear lock that PARK engages ... at the appropriate time.

I don’t think there is any gear lock on a Tesla.

 

A far as I know the steering motor runs on the HV battery not the 12V battery so the current spikes are not on the 12V battery. Also the 12V battery cannot supply 100amps.
Once you disassemble inverter the whole system works differently for interference since all the shielding is different.
Does the peddle wiring go to the inverter, my though is it would go to the front computer then it would send commands to the inverter.
Quickly reading the report its has a lot of words saying very little before it gets to a point.
Also if this really was a problem, why is it not happening more often then very rarely?

 

The power steering does run off of 12v, and has two separate rails for redundancy. I believe it runs off of VC_Front, which also ties in the 12v battery and the DC to DC converter. The DC to DC converter does all the heavy lifting, the 12v battery is just there as a buffer and backup.

 

I've got a clamp on, amp meter (somewhere.) We could use a survey of 12 V current draw as a function of various controls including daylight running head lights.

I would suspect the 'brake booster' but won't rule out the steering motor. When my car was being wrenched onto the tow truck, I was surprised at how much strength was needed to 'adjust' the wheels.

Bob Wilson

 

Thanks! I stand corrected.

Both wheels?

The BMW I just bought and former Prius used a parking paw if I remember correctly.

Bob Wilson

 

Both wheels?

Both rear wheels.

 

A new petition has been filed with the Office of Defects Investigation (ODI) requesting a reevaluation of the denial to investigate sudden unintended acceleration (SUA) in Tesla vehicles. The petition suggests that a faulty inverter design could be causing negative spikes in the low-voltage system, which can be misinterpreted as a full acceleration command. The person behind the new petition, Ronald A. Belt from Plymouth, Minnesota, suggests this design flaw could explain the reports of Tesla vehicles accelerating without the driver touching the accelerator pedal. According to Belt intermittent high electrical current demands on the vehicles’ 12VDC systems may have caused the incidents examined in the previous investigation by the NHTSA. The inverter, which uses a voltage reference derived from the 12-volt system to calibrate the accelerator pedal position, may produce incorrect calibration voltages that lead to sudden unintended acceleration. Belt says this explains why the vehicle logs show the accelerator pedal was pressed down even when drivers claimed not to have touched it. The lengthy petition, first reported by autoevolution, identifies the cause of the negative spikes in the low-voltage system as the high current motor used by the steering assist system. When the motor requires over 100 amps to turn the wheels while the vehicle is stationary, the 12-volt system experiences a voltage drop to near zero volts. If a recalibration occurs during this interval, an incorrect calibration voltage close to zero volts can be produced, leading to a spike equivalent to pressing the accelerator pedal down. According to Belt he was able to verify his theory by intentionally inputting the wrong calibration voltage into the inverter, causing sudden unintended acceleration without the driver pressing the accelerator pedal. Two solutions are proposed to address the issue: adding a second 12-volt supply line for powering the accelerator pedal position sensors and analog-to-digital converters, or modifying the calibration routine software to test the calibration voltage before use. You can read the full petition below. breaking-nhtsa-petition-shows-tesla-s-sudden-unintended-acceleration-is-real-and-curable-217525 The post Researcher claims faulty inverter design causes sudden unintended acceleration in Tesla vehicles first appeared on Drive Tesla. The post Researcher claims faulty inverter design causes sudden unintended acceleration in Tesla vehicles appeared first on Drive Tesla. Continue reading...

I have experienced this while applying my foot on the brake pedal a few times.

 

Wow! Are you able to reliably reproduce this? Maybe a video of it?

 

A new petition has been filed with the Office of Defects Investigation (ODI) requesting a reevaluation of the denial to investigate sudden unintended acceleration (SUA) in Tesla vehicles. The petition suggests that a faulty inverter design could be causing negative spikes in the low-voltage system, which can be misinterpreted as a full acceleration command.

The person behind the new petition, Ronald A. Belt from Plymouth, Minnesota, suggests this design flaw could explain the reports of Tesla vehicles accelerating without the driver touching the accelerator pedal.

According to Belt intermittent high electrical current demands on the vehicles’ 12VDC systems may have caused the incidents examined in the previous investigation by the NHTSA. The inverter, which uses a voltage reference derived from the 12-volt system to calibrate the accelerator pedal position, may produce incorrect calibration voltages that lead to sudden unintended acceleration. Belt says this explains why the vehicle logs show the accelerator pedal was pressed down even when drivers claimed not to have touched it.

The lengthy petition, first reported by autoevolution, identifies the cause of the negative spikes in the low-voltage system as the high current motor used by the steering assist system. When the motor requires over 100 amps to turn the wheels while the vehicle is stationary, the 12-volt system experiences a voltage drop to near zero volts. If a recalibration occurs during this interval, an incorrect calibration voltage close to zero volts can be produced, leading to a spike equivalent to pressing the accelerator pedal down.

According to Belt he was able to verify his theory by intentionally inputting the wrong calibration voltage into the inverter, causing sudden unintended acceleration without the driver pressing the accelerator pedal. Two solutions are proposed to address the issue: adding a second 12-volt supply line for powering the accelerator pedal position sensors and analog-to-digital converters, or modifying the calibration routine software to test the calibration voltage before use.

You can read the full petition below.

breaking-nhtsa-petition-shows-tesla-s-sudden-unintended-acceleration-is-real-and-curable-217525

The post Researcher claims faulty inverter design causes sudden unintended acceleration in Tesla vehicles first appeared on Drive Tesla.

The post Researcher claims faulty inverter design causes sudden unintended acceleration in Tesla vehicles appeared first on Drive Tesla.

Continue reading...

This has happened to me on my 2019 M3 while pressing the brake pedal a few times .

 

My apologies if I gave the impression that Uncle Ralph was a bad person, but quite the opposite, while indeed being "weird" for his time. He was obnoxious, yelled a lot and pointed out things that we Americans did not want to think about at the time.

His crusade(s) brought us seat belts, shoulder belts, airbags, padded dashes, collapsable steering columns (before airbags), assorted lemon laws (where manufacturers were actually responsible for their products!), the EPA (which eventually gave us the greatest tuning tool the auto industry has ever seen - OBD2), championed universal health care, campaign finance reform, worker rights, ... Those are just the things that I recall just now.

But he was "weird" - a loud voice that, initially, no one listened to. ALL the car companies hated him, seems Dick Nixon was his only friend. A "Thorn in the saddle" in Texan. Hated airplanes and NEVER flew on one - took taxis across country for his assorted adventures. And he hated nuclear energy.

I never met Uncle Ralph in person, but I did meet Dick Nixon (that's another story).

 

Accelerator pedals in Tesla's have TWO potentiometers in them. One is inverse voltage from the other and if both signals are NOT perfect inverses, the signal is considered invalid by the system and your car won't accelerate - it MIGHT just keep at the same speed OR it MIGHT start decelerating if the voltages don't go back to 'normal' but it won't accelerate because of the mismatch. It's kind of like how noise canceling headphones work. When active, there's a MICROPHONE that 'listens' to the outside world, it then feeds the inverse of those sounds (flips them over) to your ears which then cancel the sounds out... It's kind of funny because the 'noise' from outside is still actually coming through, but, because we humans use 'electrical' signals, the phase-inverted (by 180 degrees) cancels it out as far as what we THINK we're hearing...

Can you imagine how quite a car would be if the sound system did they same thing? It could cancel out ALL the sounds from 'outside' the vehicle... Even those pesky sirens and bullhorns yelling for you to pull over!! hehe...

 

Can you imagine how quite a car would be if the sound system did they same thing? It could cancel out ALL the sounds from 'outside' the vehicle... Even those pesky sirens and bullhorns yelling for you to pull over!! hehe...

FSD already ignores those "noises".

 

By accident, I tried the 'forbidden experiment' by pushing "PARK" instead of raising the shifter stalk to "N". There was an immediate jerk like the brakes coming on briefly and the car reverted to "D". Lifting the shift lever into "N" worked as expected.

Background, there is a route to a nature trail area on Redstone Arsenal where the dogs love to run. The last 1/4 mile (1/2 km,) starts with a sharp left turn taken at low speed. From there, it is an undulating downgrade. Shifting into "N", gravity lets the car roll and I'll briefly shift into "D" as needed to keep the speed low. A smallish, two lane road with very little traffic, I just like gravity gliding.

Bob Wilson

 

When the driver presses the brake pedal, power is cut from the motors. This is not done using software. I don't know the exact method, but it seemed pretty fail safe to me. I always call BS when a Tesla drive said they were stepping on the brake. The doesn't mean there's absolutely no way to have unintended accleration, but the brake should always stop the car.