Toyota fault is maybe electronic!

[Pommie]

According to this page, the problem of runaway acceleration is electronic.

I caught the tail end of this story on the radio yesterday and the above expert was interviewed. The scariest part was a story of a woman whos car accelerated to 160kph and even though she selected neutral and then reverse it still kept going. It appears that anything the computer thinks is dangerous it refuses to do. So at 100kph it will not select neutral or turn off the engine. This latter bit might be wrong as it was during this discussion that I started to pay attention.

Any thoughts?

Mike.
Edit, title should say maybe but I can't edit it.

 

[Mikebits]

Sounds more like software, or more aptly named firmware to me Sounds like a scary situation though. So breaks and everything quits?

 

[Oznog]

I saw the Dave Gilbert claim before. I couldn't find any technical description of what he claimed to be doing, so it's pretty much meaningless. I mean, I could take apart an ignition switch and demonstrate that when I short the wires together the car starts without a key. Doesn't mean the car's in danger of starting by itself without a key.

I've not heard where they've taken apart the pedal and found anything amiss, unless they're covering it up. Hypothetically speaking, condensation might cause a problem and disappear before the technical evaluation. There could be a code glitch, or one of those infamous "silicon errata" where, under some odd case, the ECU or pedal will not do what the spec says at all.

**broken link removed**
Dramatic footage of uncontrolled acceleration

 

[Hero999]

What you do in that situation?

It's like a scene from a film you're approaching a roundabout at 100mph and you can't slow down, do you try to go round it, even though you have no chance of making it or do you go off road into a field?

I'd choose the field, even though there's a good chance I could flip the car and be killed, there's less chance of me killing anyone else.

 

[transistor495]

Got to love engineers...They tend to over t-shoot issues! This is not a hardware problem. Hardware fails and continues to fail until repaired or replaced. This is a logic problem caused by trigger conditions within the vehicle. It's not a short curcuit condition either! Mr. Gilberts duplicated a phenomenon but its not the true cause. The intermittent nature of this phenomenon calculated by the amount of cars on the road vs. the phenomenon indicates a software logic condition. It's a simple software bug! The vehicle computer does not understand the operating condition of the vehicle and is taking over the vehicle. Just ask the people who experienced the issue. Who was in control of the vehicle? Them or the vehicle?

 

[Nigel Goodwin]

What you do in that situation?

It's like a scene from a film you're approaching a roundabout at 100mph and you can't slow down, do you try to go round it, even though you have no chance of making it or do you go off road into a field?

I'd choose the field, even though there's a good chance I could flip the car and be killed, there's less chance of me killing anyone else.

Well first off you shouldn't be doing 100mph - but while you're hurtling off in to the filed getting killed, I'd turn the ignition off and stop the engine, or depress the clutch and knock it out of gear (or more likely both).

 

[Hero999]

while you're hurtling off in to the filed getting killed, I'd turn the ignition off and stop the engine, or depress the clutch and knock it out of gear (or more likely both).

You obviously haven't read the first post: she tried doing all of that and it was probably an automatic so depressing the clutch is not applicable.

 

[Pommie]

The really worrying thing about this is that the car is programmed to protect itself. Try to select neutral at high throttle opening and it just refuses to do it as it will damage the engine. The ignition switch is no longer a on/off switch but a request to the computer to turn off the car which it will not do at high speed.

Another interesting thing about this scenario is that Australia have different requirements of cars. Cars that go into limp home or service mode are not allowed. If you have a problem in the outback then you really don't care if the engine survives or not, you just want it to get you out of there as the alternative is just too unacceptable. There have been no incidences of runaway cars in Australia.

The other thing that people don't appreciate is that in these conditions the brakes don't work. This is because they are servo assisted and require a vacuum. With the throttle full open there is no vacuum.

Mike.

 

[smanches]

They are using manifold vacuum for the breaks? I thought that was done away with decades ago?

 

[cobra1]

Yup they still use manifold vacuum in nearly all petrol cars, deisels dont create vacuum so they have a seperate pump.

 

[Nigel Goodwin]

Yup they still use manifold vacuum in nearly all petrol cars, deisels dont create vacuum so they have a seperate pump.

That's intersting to know.

 

[Mickster]

Having just watched the video, there are a couple of points which almost-immediately don't quite seem right to me, in this 'simulation', 'reconstruction' or 'lab test'.

At 1:32 to 1:38 we see a screenshot on a laptop, displaying what is likely to be APP (Accelerator Pedal Position) sensor 1 & 2 signals.
(Before we go any further, I'll state that I am not familiar with the actual Toyota setup, but similar setups on a variety of other manufacturer's products currently in use today)

The screenshot shows the signals suddenly snap from low throttle to WOT (Wide Open Throttle) but the ratio between 1 & 2 doesn't quite seem right, given the scale of the graph. Normally, the lower line would be 0V and the top line would be 5V. Sensor 1 normally produces voltages of exactly half that of sensor 2. So if sensor 2 shows 5V at WOT, sensor 1 should show 2.5V. Any other value (within some tolerances) should flag up a 1 & 2 correlation fault and store a DTC. (Diagnostic Trouble Code)

However, on the screen you will notice that on WOT, the upper trace is where you would expect 5V to be located, whilst the lower trace appears to be quite higher than the expected 2.5V....maybe 3.2V assuming a 5V scale. Again, a correlation issue should be detected and a DTC stored.

Normally, the ECU ECM PCM or whatever you want to call it, supplies a Vref of 5V. This Vref is closely monitored, since it supplies critical components with a known stable source. Deviation from this source should flag up a DTC...

For OBDII, there is the requirement that emissions-related problems illuminate a MIL (Malfunction Indicator Lamp) and generate a P0XXX DTC, which can be read out by any generic scan tool.

In addition to the mandatory OBDII P0XXX DTC's, there are also what are considered 'manufacturer-specific' DTC's, designated P1XXX, P2XXX, BXXXX, CXXXX and UXXXX, which usually require proprietary equipment to read out and interpret.

With the above said, at 2:40 to 2:54 where the 'scanner' was consulted for a DTC, you will possibly note that the equipment used appears to be a generic scan tool and not authorised Toyota diagnostic equipment.

Getting back to OBDII compliance, only P0XXX codes have to be available to generic scan tools.... on the vehicles I work with, APP DTC's are designated P2XXX....

Anyway, at 2:13 to 2:16, what is he doing? He sticks a banana plug into something...is he shorting the 5Vref to battery voltage? If he is, the APP sensor 1 & 2 voltages would surely be influenced but his generic scanner would likely not see the manufacturer-specific DTC's.....

 

[Mickster]

That's intersting to know.

cobra1 is correct. AFAIK, almost all petrol-engined derivatives use manifold vacuum to exhaust the brake servo (booster) and provide assistive braking. With the lack of vacuum on a diesel engine (no active throttle butterfly to create a vacuum), an alternative method, such as a camshaft-driven, or alternator-driven pump is employed.

As an aside, if a diesel engine is producing a vacuum, either the intake system has a restriction such as a blocked air filter, or a shut-down butterfly is operating when it shouldn't.

 

[Hero999]

Yup they still use manifold vacuum in nearly all petrol cars, deisels dont create vacuum so they have a seperate pump.

I'm no mechanic but as long as the engine is physically moving there would still be a manifold vacuum so the breaks should still work if you're travelling at 100mph and you turn the engine off. Even if the breaks fail when you turn the engine off, at least you're going to stop using engine breaking, especially if you shift down through the gearbox.

 

[Mickster]

I'm no mechanic but as long as the engine is physically moving there would still be a manifold vacuum so the breaks should still work if you're travelling at 100mph and you turn the engine off. Even if the breaks fail when you turn the engine off, at least you're going to stop using engine breaking, especially if you shift down through the gearbox.

During normal driving with a petrol-engine vehicle, there is enough reserve vacuum in the servo to come to a safe stop. However, this vacuum is created by manifold depression due to a restriction caused by the throttle butterfly. In a WOT situation, the vacuum is no longer being replenished and it doesn't take very many brake pedal applications to completely exhaust this additional boost from the servo. In a WOT situation, where the driver isn't expecting it, repeated stabs on the brakes can be understood in a panic...only it's counter-productive... the assistance disappears and has no hope of being replenished.

If you have a diesel engine, with a vac pump that has no association with manifold depression, but a direct connection to engine RPM, a WOT situation is going to provide maximum braking assistance.

 

[Hero999]

What's a WOT situation?

Anyway, surely it's better than preventing the engine from being switched off and them careering towards death at 100mph?

 

[audioguru]

WOT is Wide Open Throttle. The pedal to the metal. Full power. Lots of fun or very dangerous.

My car has a governor set to the max speed rating of its tires. The first time I was passing on a two-lane highway and a big truck (lorry) was coming at me, the governor turned off the engine as a safety function. Stupid. The truck almost plastered me. The tires would have survived a few seconds at a higher speed.
The engine management computer should have asked, "You are going faster than the speed rating of the tires. Do you want the engine to be turned off now? Yes or no?".

 

[Mickster]

What's a WOT situation?

Wide Open Throttle.

The 'throttle', which is the butterfly valve in a petrol engine, creates a depression in the intake manifold, where the brake servo takes a tapping and stores vacuum via a one-way-valve. WOT basically provides no restriction and therefore no manifold depression to exhaust the servo...

I'd take my chances by dipping the clutch, pressing the brakes once and steering past anyone in the way, coming to a stop and then changing my undies.

 

[Mickster]

WOT is Wide Open Throttle. The pedal to the metal. Full power. Lots of fun or very dangerous.

My car has a governor set to the max speed rating of its tires. The first time I was passing on a two-lane highway and a big truck (lorry) was coming at me, the governor turned off the engine as a safety function. Stupid. The truck almost plastered me. The tires would have survived a few seconds at a higher speed.
The engine management computer should have asked, "You are going faster than the speed rating of the tires. Do you want the engine to be turned off now? Yes or no?".

I'm quite sure that it may have felt that way, but in actual fact the sphincter-twitching moment sped up time and Der FeckingGretTrucken was bearing down upon you, much faster than was anticipated.

Those tear-aways are always speeding around here......

Unless, that is, we actually have a Beta-Tester for MS_Auto_2010....

 

[Hero999]

I was talking about whether the manifold vacuum would still be sufficient to activate the breaks with the engine switched of but still turning at high speed due to inertia?

Surely they're no chance of the breaks failing if you turn the engine off at 100mph because the movement will keep the breaks working, so the protection is stupid?