Toyota Accelerator Sensor?

[MikeMl]

I own a 2006 Corolla, and have been listening to the coverage of Toyota's accelerator recall. I haven't received a recall notice yet, but I expect to. Something about the mechanical fix doesn't seem right to me, and there are folks who have been saying that it could be an Engine Control Module software problem, so I got to thinking about it. The Throttle Position Sensor is a potentiometer built by **broken link removed**

I pulled out the Corolla's wiring diagram to see how the Throttle Position Sensor is wired to the ECM. I was questioning what would happen if one of the wires between the ECM and the Throttle Position Sensor became disconnected? Could that cause the ECM to think that the pedal was all the way to the metal, and cause a "runaway"?

Here is a fragment of a wiring diagram that spans 13 pages. The Throttle Position Sensor is T1, the light blue rectangle in the lower right. The ECM is the big blue mass at the top. What follows is my speculation as to how it functions: If there are any trained Toyota mechanics that can correct my speculation, please chime in...

T1 pin 3 is the wiper of the Throttle Position Sensor pot and it goes to ECM pin VTA. T1 pin 2 is the low side of the pot and goes to ECM pin E2 via the junction block and brown wire. T1 pin 1 is the high side of the pot and goes to ECM pin VC via the junction block and yellow wire.

Interestingly, the pot high side is driven by a pin on the ECM rather than being tied to a source of 12V. Similarly, the pot low side is driven from a pin on the ECM rather than being tied to a ground lug. I'm speculating that this is so the ECM can test for an open Pot or bad connection.

During normal operation, presumably the ECM drives VC high, and drives E2 to ground. That allows the ECM to A/D what comes in on VTA to determine the pedal position. Presumably, during a Power Up test, the ECM can remove power from VC (or switch it to ground) which would cause VTA to go to ground so that the ECM can check half the the connections. By driving Vc high, and turning off E2, that would cause VTA to rise to the max voltage, so the ECM can check the other connections and the pot integrity.

I know that this type of scan is done when the key is first turned on, but is it done continuously during driving?
If the self-test is periodic while driving, say once a second, then if the connection between T1 pin 2 and ECM pin E2 became disconnected, the ECM would know that the High on VTA is caused by a broken wire, not because the pedal was floored. If the self-test only happens once during key-on, then the ECM could think that a broken wire means go fast???

So, does anybody know how often the self-test is done????

 

[kchriste]

Just theorizing here. I would suspect that they don't use the entire range of the POT either. That way, an open leg would put the voltage above the safety trigger point and the ECU would go into fault mode and close the throttle.

 

[Diver300]

Are you sure that the throttle sensor is a potentiometer? Lots of the communication in cars is now serial. My 2001 car has several serial busses as well as the CAN bus and I was always under the impression that the accelerator position was serial.

The VC wire is most likely 5 V regulated rather than 12V. If it was tied to 12 V they would have to allow for fluctuations in the 12 V as well, and it's only a few mA at 5 V.

It would be difficult to test the connections because the VC for the vapour pressure sensor and the throttle sensor are in parallel, so the current would have to be tested quite accurately to detect one of them being disconnected.

However, I guess that all of this is irrelevant to the problem. Just about all stuck throttle problems are mechanical. Just be ready to stamp on the brake, and turn off the ignition if the car takes off. Also, if it is a petrol car, find out what it is like to stop without the power brakes, as the power assistance will fail after a few presses if the throttle is wide open.

 

[BrownOut]

I recently saw Apple Computer co-founder Steve Wozniak talking about his problems with sudden acceleration on his Prius. He said he could stop the acceleration and restore his car to a sane state by tapping on the breaks. I don't know if that helps answer your question or not, but I thought it was interesting...

 

[Jony130]

Did you see this document
https://www.electro-tech-online.com/custompdfs/2010/02/h33.pdf

 

[Mickster]

Are you sure that the throttle sensor is a potentiometer? Lots of the communication in cars is now serial. My 2001 car has several serial busses as well as the CAN bus and I was always under the impression that the accelerator position was serial.

The VC wire is most likely 5 V regulated rather than 12V. If it was tied to 12 V they would have to allow for fluctuations in the 12 V as well, and it's only a few mA at 5 V.

It would be difficult to test the connections because the VC for the vapour pressure sensor and the throttle sensor are in parallel, so the current would have to be tested quite accurately to detect one of them being disconnected.

However, I guess that all of this is irrelevant to the problem. Just about all stuck throttle problems are mechanical. Just be ready to stamp on the brake, and turn off the ignition if the car takes off. Also, if it is a petrol car, find out what it is like to stop without the power brakes, as the power assistance will fail after a few presses if the throttle is wide open.

In answer to MikeMI, I'm a trained Tech, but not Toyota.

Unfortunately, training, along with troubleshooting procedures, do not really go into the finer details of how a management system performs fault monitoring and reporting. The monitoring and reporting is normally continuous though, resulting in MIL illumination as soon as values meet the criteria to generate a DTC (Diagnostic Trouble Code)

When a DTC is generated, which could cause powertrain damage or become a potential safety issue, normally a 'limp-home' mode is activated, where engine output is reduced.

DTC's are only a pointer toward a potential source of a fault, not a definitive answer and that is where a lot of techs (and others with code-readers) can fall down. The DTC says it's this.....let's change it..

Troubleshooting procedures, when followed to the letter, can sometimes give incorrect results. A DTC which is classed as 'Not Present', usually ends up with the respective ECU being blamed, and replaced, when in actual fact the problem is usually due to intermittent contacts or water ingress to wiring harness connectors. ('Not Present' refers to a DTC which has been generated previously, but the criteria for generating the code is no longer met....These code usually self-delete after X amount of drive cycles without any re-occurence.)


Anyway, I've just seen the BBC News and a report including a Toyota representative holding a pedal assembly, worded along the lines of "We will be introducing a precision cut metal bar...." which sorta leads to a probable mechanical problem.

See here:
BBC News - Toyota defends car recall delay


In answer to Diver300 (quote above):

AFAIK, potentiometers are still the predominant means of sensing throttle position. Where three wires and a single pot were previously used, you now normally have six wires and two pots. The pots are normally classed as 'Sensor1' and 'Sensor2', each having a separate supply and ground from the ECM, with the wiper of each also connected to the ECM. I haven't actually measured the pot resistance values, but the voltage signals returned to the ECM are 1:2 (or 2:1). Each pot is fed 5V & ground, the wipers move in tandem, the voltages returned are 0V to 5V on one sensor, 0V to 2.5V on the other.




Although the gut-reaction if the accelerator sticks is to go for the brake pedal, in this case trying to pull the pedal back up would be a solution, IF it is a mechanical defect....

Wouldn't like to be in the situation to have to find out though...



HTH.

 

[Diver300]

I stand corrected on the throttle position sensor. I know that the 1994 Fiat Cinquencento had a potentiometer throttle position sensor, but the throttle was opened with a Bowden cable.

The problem in a petrol car with a stuck throttle is that there is no inlet manifold vacuum when the throttle is wide open. It's the inlet manifold vacuum that powers the brake servo.....

The BBC program showed the recording and transcript from the USA crash of a Toyota with a stuck throttle. I don't understand why phoning 911 seemed more useful to the driver than turning off the ignition, shifting to neutral or pressing the brake pedal really hard.

 

[Mickster]

I stand corrected on the throttle position sensor. I know that the 1994 Fiat Cinquencento had a potentiometer throttle position sensor, but the throttle was opened with a Bowden cable.

That's the way it still should be....less wiring between the TPS and the ECU - less chance for water ingress or resistance from terminal contacts.

When ECU's started to appear on vehicles, the TPS was indeed operated by a bowden cable. The TPS was mounted on the throttle body and was in an ideal position for a tech to rev the engine from the engine bay. These days, with 'drive-by-wire', it takes two people to do some troubleshooting steps......and they call it progress.

The problem in a petrol car with a stuck throttle is that there is no inlet manifold vacuum when the throttle is wide open. It's the inlet manifold vacuum that powers the brake servo.....

That's also correct, but those lucky folks with a diesel engine still get braking assistance, since the vacuum is produced by a pump driven by the alternator, camshaft, or other means.

The BBC program showed the recording and transcript from the USA crash of a Toyota with a stuck throttle. I don't understand why phoning 911 seemed more useful to the driver than turning off the ignition, shifting to neutral or pressing the brake pedal really hard.

Panic situation I suppose...although talking on the phone could likely be an additional distraction.



Dip the clutch (or shift to N), coast over toward the hard-shoulder & then switch off the ignition....

 

[DirtyLude]

The TPS cannot actually cause a runaway anything, only the butterfly valve being physically stuck would do that. The TPS isn't actually used for any full time calculations in the ECM, it's used for short term fuel/timing corrections when you modulate the peddle. The MAP/MAS/MAF sensor, whatever it's using, isn't fast enough to detect if you've just floored the gas pedal and the engine can lean out for a split second while that sensor catches up, so the TPS is used as a quick fuel burst to compensate for that. It's also used for idle detection.

All the important sensor use the ECM's ground and regulated 5V sensor power rather than just any 12v from the regular power harness. You can't get a proper calibrated reading if you don't have a hard reference ground and power.

 

[Mickster]

The TPS cannot actually cause a runaway anything, only the butterfly valve being physically stuck would do that.

Agreed. However, if there were to be a hypothetical situation where the pot wiper were able to move, indicating a higher throttle demand than actual, on a modern system the throttle-positioner could be driven to reflect the false demand. Older systems, where both the TPS and bowden cable were directly connected to the butterfly, couldn't suffer from this example.

The TPS isn't actually used for any full time calculations in the ECM, it's used for short term fuel/timing corrections when you modulate the peddle. The MAP/MAS/MAF sensor, whatever it's using, isn't fast enough to detect if you've just floored the gas pedal and the engine can lean out for a split second while that sensor catches up, so the TPS is used as a quick fuel burst to compensate for that. It's also used for idle detection.

The TPS is used primarily to indicate demand, in conjunction with MAF and engine temp. Fuelling corrections are adjusted by feedback from upstream O2 sensors, along with engine temp and short-term trim is adjusted accordingly, via fuel injector pulse-width. Ignition timing considerations are monitored and corrected jointly by the Crank Angle Sensor (which also serves as the main engine speed sensor), and the Knock Sensor. TPS & MAF almost mirror each other when the pedal is stamped upon creating WOT. There is a slight dip in MAF when the butterfly snaps open.

Incidentally, the Crank Angle Sensor is also used for misfire-monitoring. Complete combustion results in an acceptable acceleration rate during the power stroke. Assuming no other electrical concerns which would produce a different specific DTC, repeated cycles producing a lack of crankshaft acceleration for a particular cylinder could be indicative of other problems, such as low compression, sticking valves etc. and the generic P0300, P0301, P0302....... codes can be generated.

All the important sensor use the ECM's ground and regulated 5V sensor power rather than just any 12v from the regular power harness. You can't get a proper calibrated reading if you don't have a hard reference ground and power.

Agreed. The important sensors use 5V and actuators can use either 5V or B+. (Or even higher voltages when Peak & Hold Common-rail diesel injectors are concerned)

 

[killivolt]

Dip the clutch (or shift to N), coast over toward the hard-shoulder & then switch off the ignition..../QUOTE]

I think you can just push up because it won't go into Reverse or Park. I've accidentally done it on my Camery. My wife will ultimately park her purse to close to the shift lever or something and it pops out of drive into Neutral of course it's not on the Column 94 XLT.

 

[Mickster]

Dip the clutch (or shift to N), coast over toward the hard-shoulder & then switch off the ignition..../QUOTE]

I think you can just push up because it won't go into Reverse or Park. I've accidentally done it on my Camery. My wife will ultimately park her purse to close to the shift lever or something and it pops out of drive into Neutral of course it's not on the Column 94 XLT.

You can't even do that on some of the newer vehicles.

EDIT: ^^^ Sorry, misread your post...

You should be able to move from D to N with an auto, regardless of ignition position. But if the vehicle has Electronic Power Steering, switching the engine/ignition off should be avoided until the vehicle is stopped.

 

[DirtyLude]

Agreed. However, if there were to be a hypothetical situation where the pot wiper were able to move, indicating a higher throttle demand than actual, on a modern system the throttle-positioner could be driven to reflect the false demand. Older systems, where both the TPS and bowden cable were directly connected to the butterfly, couldn't suffer from this example.

So, all of these are drive-by-wire? My expertise tends to be pre-2000's and my ECU experience is mostly speed density third party amateur race ECU's and modified/hacked/reprogrammed OEM's.

The TPS is used primarily to indicate demand, in conjunction with MAF and engine temp. Fuelling corrections are adjusted by feedback from upstream O2 sensors, along with engine temp and short-term trim is adjusted accordingly, via fuel injector pulse-width. Ignition timing considerations are monitored and corrected jointly by the Crank Angle Sensor (which also serves as the main engine speed sensor), and the Knock Sensor. TPS & MAF almost mirror each other when the pedal is stamped upon creating WOT. There is a slight dip in MAF when the butterfly snaps open.

Incidentally, the Crank Angle Sensor is also used for misfire-monitoring. Complete combustion results in an acceptable acceleration rate during the power stroke. Assuming no other electrical concerns which would produce a different specific DTC, repeated cycles producing a lack of crankshaft acceleration for a particular cylinder could be indicative of other problems, such as low compression, sticking valves etc. and the generic P0300, P0301, P0302....... codes can be generated.

Did you cut and paste this out of a book? Please, try to stay on topic. Yes, I know about detecting deceleration events using the crank sensor, yes I know O2 sensor reading during closed loop will generate short term trim and long term trim values, what does any of this have to do with anything here? None of this, except your specific reference to the relationship between MAF and TPS readings has any relationship to what we are talking about here.

Every third party ECU I've used and in OEM's I've seen reverse engineered, uses a correction for quickly modulating the throttle based on the TPS response. Usually called throttle pump or transient fuel correction. My actual tuning experience is mainly with MAP systems, though since MAF systems have essential disadvantages on race cars. MAF systems have quicker response, but it would still surprise my if transient fuel correction wasn't implemented.

Generally to be more specific the TPS is usually used for transient fuel correction, idle/WOT detection, and steady state driving detection for closed loop.

Since I really have no experience with the drive-by-wire throttle mechanical system I really can't add anything to whether this all relates. All I know is that Toyota says it's not the problem, not that that means much.

 

[Mickster]

So, all of these are drive-by-wire? My expertise tends to be pre-2000's and my ECU experience is mostly speed density third party amateur race ECU's and modified/hacked/reprogrammed OEM's.

Suppose it comes down to terminology when 'drive-by-wire' is used. A cable, to some, is a wire.....and vice-versa, depending upon where you come from.

Perhaps not the ideal words used to accurately describe the technology...


Should it be termed an "Anthropologically-manipulated closed-loop-servo system"?

Did you cut and paste this out of a book?

Nope, I wrote every word from memory, so if you see the exact wording anywhere else, please direct me to it so I can sue for plagiarism.

Please, try to stay on topic.

You know threads will always drift away from the original topic, some more than others. At least, this content has some relevance to the topic in hand.

Yes, I know about detecting deceleration events using the crank sensor, yes I know O2 sensor reading during closed loop will generate short term trim and long term trim values, what does any of this have to do with anything here? None of this, except your specific reference to the relationship between MAF and TPS readings has any relationship to what we are talking about here.

You already know it, but there are probably quite a few out there who don't, but might find the information useful....

Every third party ECU I've used and in OEM's I've seen reverse engineered, uses a correction for quickly modulating the throttle based on the TPS response. Usually called throttle pump or transient fuel correction. My actual tuning experience is mainly with MAP systems, though since MAF systems have essential disadvantages on race cars. MAF systems have quicker response, but it would still surprise my if transient fuel correction wasn't implemented.

Generally to be more specific the TPS is usually used for transient fuel correction, idle/WOT detection, and steady state driving detection for closed loop.

Since I really have no experience with the drive-by-wire throttle mechanical system I really can't add anything to whether this all relates. All I know is that Toyota says it's not the problem, not that that means much.

I'm not going to try to play "Who is smarter than who" Mark, I don't have an electronics background or degree. I have posted what I understand to be true and hopefully it offers someone else an insight into something they haven't had a chance to research yet.

Regards.

 

[Diver300]

You should be able to move from D to N with an auto, regardless of ignition position. But if the vehicle has Electronic Power Steering, switching the engine/ignition off should be avoided until the vehicle is stopped.

On a car or light truck, steering to the kerb is easy with no power steering. Power steering is really not needed above about 5 mph, or when you are trying to corner hard.

I had a 1978 van, over 3 tonnes max weight, over 1 tonne on the front axle, and no power steering. OK, I couldn't turn the steering wheel when the vehicle wasn't moving, and it was hard work in town, but no problem at all on the highway. The fear of losing power steering is a lot worse than the reality.

 

[DirtyLude]

Nope, I wrote every word from memory, so if you see the exact wording anywhere else, please direct me to it so I can sue for plagiarism.

I didn't mean to be terse or accusatory, my typing reads a lot more brusque than I'm trying for.

I asked about the manual because the statements seem to be straight from a service manual. I can tell you that the relationships for fuel and timing are more complex. Spark advance has many correction factors more than just rpm (crank speed) and knock retard, especially on vehicles with variable valve timing.

 

[Mickster]

I didn't mean to be terse or accusatory, my typing reads a lot more brusque than I'm trying for.

Thanks for clearing that up.

I asked about the manual because the statements seem to be straight from a service manual. I can tell you that the relationships for fuel and timing are more complex. Spark advance has many correction factors more than just rpm (crank speed) and knock retard, especially on vehicles with variable valve timing.

You didn't ask about any manual, you made an assumption, but that's by the by.

We both know that the contributing factors toward achieving perfect combustion exceed the scope of what was required from the original post, so let's leave this peeing contest for a later date.

It's late and I already have most of tomorrow lined up for troubleshooting other people's jobs, along with whatever else comes through the door as a 'waiting' job.....


Regards.

 

[DirtyLude]

You didn't ask about any manual, you made an assumption, but that's by the by.

Did you cut and paste this out of a book?

That's a question, not an assumption.

 

[RCinFLA]

I recently saw Apple Computer co-founder Steve Wozniak talking about his problems with sudden acceleration on his Prius. He said he could stop the acceleration and restore his car to a sane state by tapping on the breaks. I don't know if that helps answer your question or not, but I thought it was interesting...

I saw this interview. He was talking about the cruise control on his Prius. I had a similiar issue on a 92 Aerostar. It acted like a feedback control problem. Most of the time the cruise control worked fine. Once in a while it would start fluxuating the accelerator. The fluxuations would build up greater over/under accelerator control over several seconds then finally put the petal to the metal. Turning off cruise control released accelerator, either by directly shutting off cruise control switch or tapping brake which disengages cruise control.

 

[Mickster]

That's a question, not an assumption.

You are correct Mark.

Perhaps my defenses are up a tad high.

My apologies.