Brake Throttle Override
#41
02-27-2010, 02:01 AM
Active Enthusiast
Join Date: Nov 2008
Posts: 67
Re: Brake Throttle Override
Somewhere in another thread (I'm too lazy to find it), I did the same test and found that for the TCH the engine does drop to idle and power is cut to the car when the shift lever is moved to neutral. HOWEVER, bear in mind that for the TCH (and Prius, and possibly other models) the gear selection lever is not actually connected to the transmission. It is just a big rotary switch, and sends electrical signals to the ECU, which actually does the powertrain control. Now, if a bug in the software causes the ECU to "lock up" so that it no longer reads inputs from the accelerator, the shift lever, the power button, the ABS ECU, the MAF sensor, etc., (we've all experienced this on a PC) it is theoretically possible for the powertrain to be "stuck" in a state of delivering constant power to the wheels, which, if more than needed to just maintain speed, will cause uncontrolled acceleration. The ECU is *supposed* to have a watchdog timer and other measures to prevent this from happening, but who knows? I don't think there are any Federal laws requiring such safety measures, or are there? Anybody know?
BTW, how do I know this stuff? I work with these things for a living, except they are for large trucks instead of cars.
BTW, how do I know this stuff? I work with these things for a living, except they are for large trucks instead of cars.
#42
02-27-2010, 07:04 AM
Enthusiast
Join Date: Feb 2010
Posts: 41
Re: Brake Throttle Override
I have checked the TCH wiring diagrams, and yes indeed, the power button and the main battery pack relays (only 2 things that could cut power completly) are both connected to the Hybrid ECU and not directly to each other.
In order for a TCH to accelerate, the engine throttle must be increased at the same time voltage to MG1 and MG2 is increased. Since MG1 and MG2 are 3 phase, this involves switching power in a timed sequence. If the ECU were "locked up", I wonder if it would even be possible for it to control any one much less all 3 of these outputs to make the vehicle accelerate. It seems that a hybrid needs much more active control from the system to accelerate then a simple gas engine does.
In order for a TCH to accelerate, the engine throttle must be increased at the same time voltage to MG1 and MG2 is increased. Since MG1 and MG2 are 3 phase, this involves switching power in a timed sequence. If the ECU were "locked up", I wonder if it would even be possible for it to control any one much less all 3 of these outputs to make the vehicle accelerate. It seems that a hybrid needs much more active control from the system to accelerate then a simple gas engine does.
#43
02-27-2010, 09:06 AM
Active Enthusiast
Join Date: Nov 2008
Posts: 67
Re: Brake Throttle Override
True, but it is possible (not sure) that high speed, hard real time functions such as injector firing and motor phase sequencing could be handled by auxiliary processors, who get their data from memory locations shared with the main processor. This way those functions could still be happening even if the main cpu were frozen. An analogy in the PC world is if the x86 cpu has crashed, the monitor screen still continues to display the (frozen) text, because the video processor continues to run, getting the screen data from the shared video memory.
EDIT: Duhhh... I see what you're saying now. In order for the vehicle to accelerate, these data values would have to increase in synchronization with each other, which means the main cpu would have to be running properly. However, it still may be possible that: A) the external sensor inputs are controlled by an auxiliary processor which is crashing, or B) the main cpu is getting locked out of the routines which read these sensors by a bug in the code. I don't know, just guessing.
EDIT: Duhhh... I see what you're saying now. In order for the vehicle to accelerate, these data values would have to increase in synchronization with each other, which means the main cpu would have to be running properly. However, it still may be possible that: A) the external sensor inputs are controlled by an auxiliary processor which is crashing, or B) the main cpu is getting locked out of the routines which read these sensors by a bug in the code. I don't know, just guessing.
Last edited by GreenRedTCH; 02-27-2010 at 09:14 AM.
#44
02-27-2010, 10:13 AM
Enthusiast
Join Date: Feb 2010
Posts: 41
Re: Brake Throttle Override
The wide open throttle problem described on ABC news by Dr Gilbert of Southern
Illinois University is created by shorting the two outputs of the accelerator pedal.
http://abcnews.go.com/Blotter/toyota...ory?id=9909319
The story however may have misinterpreted/misrepresented 2 other points.
1- Did the brakes actually fail, or was their application reduced by the speed/power of the engine - the driver appears to pump the brake pedal, not press it all the way down and keep it there
2- even more important - they never said anything about shifting to NEUTRAL, and if they did, what happened ??
Toyota accelerator pedals create 2 different voltages on 2 separate wires from the pedal to the ECU, by shorting those two wires with a resistance that mimics the resistance/voltage at full throttle, the ECU thinks the pedal is all the way down. It is possible, while extremely remote, that this could happen in the real world. The problem is with the wiring, not the ECU itself which is why no error code is generated, there is no error, the ECU is responding to the input from the pedal, the ECU has not failed in any way.
So, assuming the ECU is still functioning, but the input is bad, the other functions of the ECU -- MG1/MG2 control should continue to respond as usual to other inputs such as the shift selector being moved to Neutral. Once the HSD is in neutral no power is going to the wheels, so braking should function as it normally would from whatever speed you got up to before putting the car in neutral.
toyota camry hybrids also do have brake override, which has been reported to function correctly in other threads.
There seems to be several possible situations of UA -
1- Floor mat entrapment, the mats are under the brake pedal and over lower edge of the accelerator pedal - preventing the brake pedal from being fully depressed, and pressing down on the accelerator pedal when the brake pedal is pressed. On a Camry hybrid this could cause the brake override to not function if the brake pedal can not be depressed more than halfway, or on other models not allow enough braking force to stop the car from full throttle.
2- wiring short as described above -
3- Human error - driver intended to press brake pedal but pressed accelerator. there is no way the car can know what you "wanted " to do. Unless auto manufactures program the ECU so when it sees a full throttle condition, allow the vehicle to accelerate for a short period of time and then stop responding until the accelerator is fully released and repressed. This would be very annoying in normal use, as you would no longer be able to accelerate for more then a few seconds, at least at full throttle.
3a- Cruise control misapplication - did the driver possibly have the cruise control on, but not activated and bumped resume button returning the car to a previously set high speed. Again this is not a failure, but the drive not understanding what is happening and taking proper action to correct it, ie hit the brake, or turn off cruise control
The question remaining is if it would ALWAYS be possible to shift to neutral, and then either coast to a stop if the brakes can not be applied fully (floor mats stuck), or using the brakes normally to come to a stop. Some of the reports claim that shifting to neutral did nothing, is this another electronic problem or driver error ?
The pedal shorted issue is a real, and serious problem that needs to be corrected. My point here is that the driver should always be able to safely stop the vehicle assuming they have more then a few seconds of open road ahead of the vehicle when the short does take over the throttle.
Illinois University is created by shorting the two outputs of the accelerator pedal.
http://abcnews.go.com/Blotter/toyota...ory?id=9909319
The story however may have misinterpreted/misrepresented 2 other points.
1- Did the brakes actually fail, or was their application reduced by the speed/power of the engine - the driver appears to pump the brake pedal, not press it all the way down and keep it there
2- even more important - they never said anything about shifting to NEUTRAL, and if they did, what happened ??
Toyota accelerator pedals create 2 different voltages on 2 separate wires from the pedal to the ECU, by shorting those two wires with a resistance that mimics the resistance/voltage at full throttle, the ECU thinks the pedal is all the way down. It is possible, while extremely remote, that this could happen in the real world. The problem is with the wiring, not the ECU itself which is why no error code is generated, there is no error, the ECU is responding to the input from the pedal, the ECU has not failed in any way.
So, assuming the ECU is still functioning, but the input is bad, the other functions of the ECU -- MG1/MG2 control should continue to respond as usual to other inputs such as the shift selector being moved to Neutral. Once the HSD is in neutral no power is going to the wheels, so braking should function as it normally would from whatever speed you got up to before putting the car in neutral.
toyota camry hybrids also do have brake override, which has been reported to function correctly in other threads.
There seems to be several possible situations of UA -
1- Floor mat entrapment, the mats are under the brake pedal and over lower edge of the accelerator pedal - preventing the brake pedal from being fully depressed, and pressing down on the accelerator pedal when the brake pedal is pressed. On a Camry hybrid this could cause the brake override to not function if the brake pedal can not be depressed more than halfway, or on other models not allow enough braking force to stop the car from full throttle.
2- wiring short as described above -
3- Human error - driver intended to press brake pedal but pressed accelerator. there is no way the car can know what you "wanted " to do. Unless auto manufactures program the ECU so when it sees a full throttle condition, allow the vehicle to accelerate for a short period of time and then stop responding until the accelerator is fully released and repressed. This would be very annoying in normal use, as you would no longer be able to accelerate for more then a few seconds, at least at full throttle.
3a- Cruise control misapplication - did the driver possibly have the cruise control on, but not activated and bumped resume button returning the car to a previously set high speed. Again this is not a failure, but the drive not understanding what is happening and taking proper action to correct it, ie hit the brake, or turn off cruise control
The question remaining is if it would ALWAYS be possible to shift to neutral, and then either coast to a stop if the brakes can not be applied fully (floor mats stuck), or using the brakes normally to come to a stop. Some of the reports claim that shifting to neutral did nothing, is this another electronic problem or driver error ?
The pedal shorted issue is a real, and serious problem that needs to be corrected. My point here is that the driver should always be able to safely stop the vehicle assuming they have more then a few seconds of open road ahead of the vehicle when the short does take over the throttle.
Last edited by BiggieJohn; 02-27-2010 at 10:39 AM.
#45
02-27-2010, 11:33 AM
Enthusiast
Join Date: Feb 2010
Posts: 41
Re: Brake Throttle Override
I checked the BTO out a little more thoroughly, and it appears that the engine does continue to rev under light braking with throttle applied. However, when the brake peddle was pushed further, the engine did return to idle. Felt like it took about 1/2 of the brake peddle travel to engage BTO.
Accelerated past 45mph so ICE was running, pressed the brakes about halfway, car slowed as usual, continue to press the accelerator and brake together, engine clearly cut out and brake assist kicked in and worked very well.
From my tests, the brakes clearly override the accelerator. If this ever happened to me without pressing the accelerator, i would probably also shift to neutral just to be sure nothing was powering the wheels.
#46
02-27-2010, 02:07 PM
Ridiculously Active Enthusiast
Join Date: Sep 2006
Location: Beautiful Pacific NW
Posts: 1,678
Re: Brake Throttle Override
The wide open throttle problem described on ABC news by Dr Gilbert of Southern
Illinois University is created by shorting the two outputs of the accelerator pedal.
http://abcnews.go.com/Blotter/toyota...ory?id=9909319
The story however may have misinterpreted/misrepresented 2 other points.
1- Did the brakes actually fail, or was their application reduced by the speed/power of the engine - the driver appears to pump the brake pedal, not press it all the way down and keep it there
2- even more important - they never said anything about shifting to NEUTRAL, and if they did, what happened ??
Toyota accelerator pedals create 2 different voltages on 2 separate wires from the pedal to the ECU, by shorting those two wires with a resistance that mimics the resistance/voltage at full throttle, the ECU thinks the pedal is all the way down. It is possible, while extremely remote, that this could happen in the real world. The problem is with the wiring, not the ECU itself which is why no error code is generated, there is no error, the ECU is responding to the input from the pedal, the ECU has not failed in any way.
Not quite the way I heard Dr Gilbert describe what he was doing. He first shorted the two gas pedal sensor output signals together on the expectation that would (and SHOULD) result in a system fault and a resulting "limp home" mode.
According to the factory shop/repair manuals the two signals , while both track the gas pedal position in a fairly linear fashion, should always be about 0.80 volts different. The factory specification most definitely indicates that these two voltages are continuously monitored to be sure the two signals are not too close together in voltage, indicating a short circuit between the two signals...P2138 code.
The problem is that the factory specification for a short is that the voltage difference must be less that 0.02 volts in order to detect a short circuit. That detection voltage, clearly, should be more in the range of 0.20 volts, if not even as high as 0.40 volts.
Typo...??
Perused multiple Toyota/etc factory manuals, different models and model years, and even different dual sensors, and all indicate the use of 0.02 volts...only 20 millivolts.
Unless Dr. Gilbert shorted the two signals together quite SOLIDLY, which it appears he did not, the two signal differences might have very well remained above 0.02 volts.
This, in reality, is the point that Dr. Gilbert needed to make and IMMHO his mistake was going forward and then shorting both sensor output to a ~5.00 volt source. That, of course, would result in a WOT condition.
So, assuming the ECU is still functioning, but the input is bad, the other functions of the ECU -- MG1/MG2 control should continue to respond as usual to other inputs such as the shift selector being moved to Neutral. Once the HSD is in neutral no power is going to the wheels, so braking should function as it normally would from whatever speed you got up to before putting the car in neutral.
toyota camry hybrids also do have brake override, which has been reported to function correctly in other threads.
There seems to be several possible situations of UA -
1- Floor mat entrapment, the mats are under the brake pedal and over lower edge of the accelerator pedal - preventing the brake pedal from being fully depressed, and pressing down on the accelerator pedal when the brake pedal is pressed. On a Camry hybrid this could cause the brake override to not function if the brake pedal can not be depressed more than halfway, or on other models not allow enough braking force to stop the car from full throttle.
2- wiring short as described above -
3- Human error - driver intended to press brake pedal but pressed accelerator. there is no way the car can know what you "wanted " to do. Unless auto manufactures program the ECU so when it sees a full throttle condition, allow the vehicle to accelerate for a short period of time and then stop responding until the accelerator is fully released and repressed. This would be very annoying in normal use, as you would no longer be able to accelerate for more then a few seconds, at least at full throttle.
3a- Cruise control misapplication - did the driver possibly have the cruise control on, but not activated and bumped resume button returning the car to a previously set high speed. Again this is not a failure, but the drive not understanding what is happening and taking proper action to correct it, ie hit the brake, or turn off cruise control
The question remaining is if it would ALWAYS be possible to shift to neutral, and then either coast to a stop if the brakes can not be applied fully (floor mats stuck), or using the brakes normally to come to a stop. Some of the reports claim that shifting to neutral did nothing, is this another electronic problem or driver error ?
The pedal shorted issue is a real, and serious problem that needs to be corrected. My point here is that the driver should always be able to safely stop the vehicle assuming they have more then a few seconds of open road ahead of the vehicle when the short does take over the throttle.
Illinois University is created by shorting the two outputs of the accelerator pedal.
http://abcnews.go.com/Blotter/toyota...ory?id=9909319
The story however may have misinterpreted/misrepresented 2 other points.
1- Did the brakes actually fail, or was their application reduced by the speed/power of the engine - the driver appears to pump the brake pedal, not press it all the way down and keep it there
2- even more important - they never said anything about shifting to NEUTRAL, and if they did, what happened ??
Toyota accelerator pedals create 2 different voltages on 2 separate wires from the pedal to the ECU, by shorting those two wires with a resistance that mimics the resistance/voltage at full throttle, the ECU thinks the pedal is all the way down. It is possible, while extremely remote, that this could happen in the real world. The problem is with the wiring, not the ECU itself which is why no error code is generated, there is no error, the ECU is responding to the input from the pedal, the ECU has not failed in any way.
Not quite the way I heard Dr Gilbert describe what he was doing. He first shorted the two gas pedal sensor output signals together on the expectation that would (and SHOULD) result in a system fault and a resulting "limp home" mode.
According to the factory shop/repair manuals the two signals , while both track the gas pedal position in a fairly linear fashion, should always be about 0.80 volts different. The factory specification most definitely indicates that these two voltages are continuously monitored to be sure the two signals are not too close together in voltage, indicating a short circuit between the two signals...P2138 code.
The problem is that the factory specification for a short is that the voltage difference must be less that 0.02 volts in order to detect a short circuit. That detection voltage, clearly, should be more in the range of 0.20 volts, if not even as high as 0.40 volts.
Typo...??
Perused multiple Toyota/etc factory manuals, different models and model years, and even different dual sensors, and all indicate the use of 0.02 volts...only 20 millivolts.
Unless Dr. Gilbert shorted the two signals together quite SOLIDLY, which it appears he did not, the two signal differences might have very well remained above 0.02 volts.
This, in reality, is the point that Dr. Gilbert needed to make and IMMHO his mistake was going forward and then shorting both sensor output to a ~5.00 volt source. That, of course, would result in a WOT condition.
So, assuming the ECU is still functioning, but the input is bad, the other functions of the ECU -- MG1/MG2 control should continue to respond as usual to other inputs such as the shift selector being moved to Neutral. Once the HSD is in neutral no power is going to the wheels, so braking should function as it normally would from whatever speed you got up to before putting the car in neutral.
toyota camry hybrids also do have brake override, which has been reported to function correctly in other threads.
There seems to be several possible situations of UA -
1- Floor mat entrapment, the mats are under the brake pedal and over lower edge of the accelerator pedal - preventing the brake pedal from being fully depressed, and pressing down on the accelerator pedal when the brake pedal is pressed. On a Camry hybrid this could cause the brake override to not function if the brake pedal can not be depressed more than halfway, or on other models not allow enough braking force to stop the car from full throttle.
2- wiring short as described above -
3- Human error - driver intended to press brake pedal but pressed accelerator. there is no way the car can know what you "wanted " to do. Unless auto manufactures program the ECU so when it sees a full throttle condition, allow the vehicle to accelerate for a short period of time and then stop responding until the accelerator is fully released and repressed. This would be very annoying in normal use, as you would no longer be able to accelerate for more then a few seconds, at least at full throttle.
3a- Cruise control misapplication - did the driver possibly have the cruise control on, but not activated and bumped resume button returning the car to a previously set high speed. Again this is not a failure, but the drive not understanding what is happening and taking proper action to correct it, ie hit the brake, or turn off cruise control
The question remaining is if it would ALWAYS be possible to shift to neutral, and then either coast to a stop if the brakes can not be applied fully (floor mats stuck), or using the brakes normally to come to a stop. Some of the reports claim that shifting to neutral did nothing, is this another electronic problem or driver error ?
The pedal shorted issue is a real, and serious problem that needs to be corrected. My point here is that the driver should always be able to safely stop the vehicle assuming they have more then a few seconds of open road ahead of the vehicle when the short does take over the throttle.
#47
02-27-2010, 02:10 PM
Ridiculously Active Enthusiast
Join Date: Sep 2006
Location: Beautiful Pacific NW
Posts: 1,678
Re: Brake Throttle Override
I just tested this on my 09 TCH, Japanese built (JT vin), purchased feb 09.
Accelerated past 45mph so ICE was running, pressed the brakes about halfway, car slowed as usual, continue to press the accelerator and brake together, engine clearly cut out and brake assist kicked in and worked very well.
From my tests, the brakes clearly override the accelerator. If this ever happened to me without pressing the accelerator, i would probably also shift to neutral just to be sure nothing was powering the wheels.
Accelerated past 45mph so ICE was running, pressed the brakes about halfway, car slowed as usual, continue to press the accelerator and brake together, engine clearly cut out and brake assist kicked in and worked very well.
From my tests, the brakes clearly override the accelerator. If this ever happened to me without pressing the accelerator, i would probably also shift to neutral just to be sure nothing was powering the wheels.
And if the engine/transaxle control firmware happened to be not watching, polling, the shifter position...??
#48
02-27-2010, 02:30 PM
Enthusiast
Join Date: Feb 2010
Posts: 41
Re: Brake Throttle Override
Then we have yet another major issue that I hope would result in a global recall. So far no one has discussed any failure of the ECU in any vehicle, just the brake pedal/wiring.
#49
02-27-2010, 03:30 PM
Active Enthusiast
Join Date: Nov 2008
Posts: 67
Re: Brake Throttle Override
Interesting post, WWest. Yeah, 0.02, or even 0.2 volts is way too close. I know of one manufacturer (can't say their name) that uses 0 - 5 volts for one line and 0 - 2.5 volts for the other. Thus, at max throttle, there is 2.5V difference. Another company uses a PWM signal, so if a short or open occurs, it is easily detected and the engine can shut down. Sounds like Toyota went the cheap route (as Mr. Toyoda alluded to) in order to save the cost of running two V+ lines to the throttle assembly.
Thread
Topic Starter
Forum
Replies
Last Post
haroldo
Toyota Camry Hybrid
1
02-19-2010 08:35 AM
