2013 dies repeatedly while driving down the road

[CharlieS]

FWIW - I have a 95 Porsche and my DME relay went out on me the other day with very similar symptoms. I was on the highway and the car lost power, shut off and the lights on the dash went on. I would pull over and the car would restart and then die after a few seconds.... Apparently, once the contacts (solder joints) in the relay cracked internally, the heat (and expansion) would cause the contacts to stop...um....contacting. When the car shut down, it would cool just enough to contract and reestablish contact until it heated up again and lost it.

My point is that the relay's you checked (like my DME relay) would checkout as operational when cool, but not when up to temperature..... Could you subject the relay(s) to heat while testing them?

Yes, as you say, that could definitely happen and I suppose might be my issue.

My dad knows a lot about electronics and he said it could even be the tracing on a circuit board that heats up just enough to break contact when energized and then cools just enough to make contact again. I sure hope that's not what I'm dealing with... Although it is a possibility.

My logic (although very flawed) is that this seems to have a time component. Along this path are several plausible scenarios:

• Actual time, like clock time on an ECM/ECU
• Some sort of mode switching that happens at a certain pivotal moment (say OL to OL drive) - when the ECM tries to read some sensor/sensors and the input results in a decision to shut down
• A byproduct of a mechanical issue (like for example, the fuel pump runs, but can't spin fast enough and that is how long it takes to consume the properly pressurized fuel in the fuel lines).
• A boundary condition in a sensor - if for example the voltage drops below a threshold value and the processor drops into a failure mode

If I get stuck much beyond the weekend, it is going to the dealer. I will have exhausted my ability and willingness, and will pass it on to the experts. I'm afraid of what that might cost if it comes to that, but at some point, I'll need to accept it.

What kind of Porsche? I'm a huge fan. I have two 911 (996) coupes right now (one that I'm trying to sell). I'm chasing CELs on one of those too (probably a vacuum leak).

As an aside, we had three automobile related failures in our household in the last week. This one, my wife's car had a cell in the battery die, and the 911 I mentioned above. I'm not feeling especially lucky. At least my wife's car was easily fixed with a new battery...

 

[NLScooby]

I admire the effort you are putting into discovering/solving the issue you are having with the 200. And I feel for you!
My Porsche is a '95 993 C2. It was a dedicated track car (and RS tribute of sorts) and is now again "street-able"......meaning it now has a license plate and two seats instead of a solo configuration! I've actually put a decent looking interior back in and spent some time on the details. She's a beautiful car and allows me to get some track time in while also still being able to cruise around on the weekend (when there is no chance of rain, road construction, or of it being too hot, or too bumpy, or too many birds, or idiot drivers...….)

 

[amagnono]

Yes, as you say, that could definitely happen and I suppose might be my issue.

My dad knows a lot about electronics and he said it could even be the tracing on a circuit board that heats up just enough to break contact when energized and then cools just enough to make contact again. I sure hope that's not what I'm dealing with... Although it is a possibility.

My logic (although very flawed) is that this seems to have a time component. Along this path are several plausible scenarios:

• Actual time, like clock time on an ECM/ECU
• Some sort of mode switching that happens at a certain pivotal moment (say OL to OL drive) - when the ECM tries to read some sensor/sensors and the input results in a decision to shut down
• A byproduct of a mechanical issue (like for example, the fuel pump runs, but can't spin fast enough and that is how long it takes to consume the properly pressurized fuel in the fuel lines).
• A boundary condition in a sensor - if for example the voltage drops below a threshold value and the processor drops into a failure mode

If I get stuck much beyond the weekend, it is going to the dealer. I will have exhausted my ability and willingness, and will pass it on to the experts. I'm aftaid of what that might cost if it comes to that, but at some point, I'll need to accept it.

What kind of Porsche? I'm a huge fan. I have two 911 (996) coupes right now (one that I'm trying to sell). I'm chasing CELs on one of those too (probably a vacuum leak).

As an aside, we had three automobile related failures in our household in the last week. This one, my wife's car had a cell in the battery die, and the 911 I mentioned above. I'm not feeling especially lucky. At least my wife's car was easily fixed with a new battery...

Just a totally random observation from someone with absolutely zero electrical or motor experience - but, as you've alluded to, the precise timing of the shutdown seems to me to indicate something other than a component heating up and causing problems - especially since the truck starts back up immediately when - you would have to assume - the component in question would still be relatively hot. The fact that it is ~1:15 every time smells like some sort of software logic kicking in. The truck senses something it doesn't like after a per-determined amount of time and the wires cross or something intentionally shuts down. I think you're on the right path and agree that you shouldn't get sidetracked absent finding a lead from someone else who has experienced this issue.

Watching and wishing you the best!

 

[Phalen302]

Wow. Would recommend take it to Lexus dealer for the diagnosis - they can figure easily.

 

[indycole]

Just curious... where did you land with that sensor you replaced recently and how far in advance of the problem did that happen?

 

[linuxgod]

Sorry I’m not near you or id offer my truck up to swap a few components like the ECU temporarily to see if we could rule out the more obtuse bits (or luck into the solution).

I know it’s just an example but I doubt it’s the fuel pumps ability to maintain pressure as I’d expect revving the engine to consume the pressurized fuel faster and thus your 1:15 run time to decrease

Is the 1:15 runtime before shutdown true both cold start and warmed up? If so then I doubt it’s an issue with the emissions going into a closed loop.

As an IT engineer turned tech manager with 20 yeas debugging system issues (hardware, networks, operating system, etc) I definitely agree you need to be very systematic in how you approach this. In particular the more you can rule in or out with each test the better.

 

[CharlieS]

Just curious... where did you land with that sensor you replaced recently and how far in advance of the problem did that happen?

I replaced the coolant temperature sensor the day after the issue started, and then yesterday, I put the original coolant temperature back in to eliminate it as a variable. There was no difference in behavior with either the new or old sensor.

Just to add to this: the current setup of the car has not changed - all sensors, relays, fuses, Ecus, pumps, etc are the same as they were when this all started.

The only difference is that I have a full tank of fuel now, and it has a can of dry gas in it.

 

[CharlieS]

Sorry I’m not near you or id offer my truck up to swap a few components like the ECU temporarily to see if we could rule out the more obtuse bits (or luck into the solution).

I know it’s just an example but I doubt it’s the fuel pumps ability to maintain pressure as I’d expect revving the engine to consume the pressurized fuel faster and thus your 1:15 run time to decrease

Is the 1:15 runtime before shutdown true both cold start and warmed up? If so then I doubt it’s an issue with the emissions going into a closed loop.

As an IT engineer turned tech manager with 20 yeas debugging system issues (hardware, networks, operating system, etc) I definitely agree you need to be very systematic in how you approach this. In particular the more you can rule in or out with each test the better.

Thank you. You've been a great help already. I appreciate the kind thoughts.

I'll do some fresh tests to see if changing the rpms alters the runtime before the stall. I don't have any good data to offer based on prior tests, so let me gather some fresh data points. And document them!

I would say the 1:15 is a rough number. I didn't time (and definitely document) the runs. It seems to me that I had several cold starts that went over the 1:30 mark.

The runtimes are not "like clockwork", there is some variability.

Sometimes, the vehicle nearly stalls and then catches itself and revs up to idle again, it might do this a couple of times in a row before eventually stalling out. While driving, a couple of times I kept my foot on the accelerator pedal, and it restarted, revved and then drove for a bit longer.

I also have a background as a software developer for enterprise scale systems, so I suspect we approach problem solving from a similar perspective. I didn't approach the initial problem solving logically and systematically, so I need to reset, and give another go. Basically debugging! Wish I could set some breakpoints and interrogate some variables.

I'll report back as I learn more.

With any luck, this process will help some future 'mudder down the line.

 

[Hothonda]

FWIW

1st Gen NSX's.....many worn out ECU's caused unexplainable problems.

New ECU - problems gone.

 

[bloc]

I think I remember @cruiseroutfit having a throttle pedal failure on his race truck caused by dust contamination at some point.. maybe he could have some insight on how it acted.

 

[linuxgod]

I think I remember @cruiseroutfit having a throttle pedal failure on his race truck caused by dust contamination at some point.. maybe he could have some insight on how it acted.

IIRC his was sticking wide open

 

[TeCKis300]

Applaud you stepping back and taking a fresh look at this. I have moonlighted tuning engines for many a turbo and engine swaps including using OE engine management, piggybacks, and standalones. Long time software and systems developer. Sometimes what makes the difference to solving week long riddles is to...sleep on it. A path forward magically congeals in the morning after having beat heads on things for extended periods.

I should probably read your factual symptom again, but as this thread is now pretty long...

1) Fueling
A fueling issue would reflect with symptoms in closed loop, showing ST/LT fuel trims, including fuel CELS. As it doesn't manifest with those signs in closed loop, you've also explored open loop fueling by pulling the MAF and an O2 sensor. No difference in symptoms. My judgement based on your description is that this is not a fueling issue. Though couple more questions:
- When it stalls, does it stall cleanly or does it run rough before stall?
- If it runs rough before stall, watch the ST trims as it stalls
- I believe you said the engine also stall at speed (not just idle)?

2) Ignition
From your symptoms, it sounds more like missed ignition. Problem is, you are not getting misfire codes and/or misfires, which would suggest something in the ignition subsystem. Nor cyclical misfires. It takes more than a single misfire on a V8 to stall the motor. I say missed ignition more broadly as it doesn't seem to be isolated to a single plug on coil/single cylinder. A miss there would definitely set a CEL. It could be the ignition module/driver, yet that should set a CEL too. It seems to me the ECM is never firing that event for whatever reason, as if it did request an ignition event and didn't see the it happen, it would set a CEL.

3) Major engine sensors including crank angle sensor and cam angle sensors
An ECM is reliant on very few key sensors to drive a motor. Not coolant temp. Not MAF as it can fall back on speed density base maps. It fundamentally only needs the cam and crank angle sensors. These generally would be key culprits for a situation like yours. Yet they tend to cross check each other, and if there is a fault in one, it again will trigger a CEL.

I believe there is a fault in the ECM/ECU of the vehicle or somewhere in the power input of that component. Which is why it's not capturing a CEL for a fault in some other sensor or sub-system. It only knows it's seen stalls without an engine off request, and missed-starts.

These cars are pretty complex too. So it's wholly possible it's anywhere. With OBD-II, it's generally able to give more clues via CELs which is what is frustrating here. Which leads me back to suspecting ECM/ECU.

 

[CJ3Flyer]

The plot thickens... For CharlieS’s sake I hope this ends soon but I feel like I’m reading a murder mystery.

 

[indycole]

I believe there is a fault in the ECM/ECU of the vehicle or somewhere in the power input of that component. Which is why it's not capturing a CEL for a fault in some other sensor or sub-system. It only knows it's seen stalls without an engine off request, and missed-starts.

This is what I've kept coming back to as while glancing through the diagnosis procedures. Why no CEL? It'd make me want to cause a CEL prior to the stall. It'd be hard not to focus on the ECM/ECU at that point.

 

[CharlieS]

Applaud you stepping back and taking a fresh look at this. I have moonlighted tuning engines for many a turbo and engine swaps including using OE engine management, piggybacks, and standalones. Long time software and systems developer. Sometimes what makes the difference to solving week long riddles is to...sleep on it. A path forward magically congeals in the morning after having beat heads on things for extended periods.

I should probably read your factual symptom again, but as this thread is now pretty long...

1) Fueling
A fueling issue would reflect with symptoms in closed loop, showing ST/LT fuel trims, including fuel CELS. As it doesn't manifest with those signs in closed loop, you've also explored open loop fueling by pulling the MAF and an O2 sensor. No difference in symptoms. My judgement based on your description is that this is not a fueling issue. Though couple more questions:
- When it stalls, does it stall cleanly or does it run rough before stall?
- If it runs rough before stall, watch the ST trims as it stalls
- I believe you said the engine also stall at speed (not just idle)?

2) Ignition
From your symptoms, it sounds more like missed ignition. Problem is, you are not getting misfire codes and/or misfires, which would suggest something in the ignition subsystem. Nor cyclical misfires. It takes more than a single misfire on a V8 to stall the motor. I say missed ignition more broadly as it doesn't seem to be isolated to a single plug on coil/single cylinder. A miss there would definitely set a CEL. It could be the ignition module/driver, yet that should set a CEL too. It seems to me the ECM is never firing that event for whatever reason, as if it did request an ignition event and didn't see the it happen, it would set a CEL.

3) Major engine sensors including crank angle sensor and cam angle sensors
An ECM is reliant on very few key sensors to drive a motor. Not coolant temp. Not MAF as it can fall back on speed density base maps. It fundamentally only needs the cam and crank angle sensors. These generally would be key culprits for a situation like yours. Yet they tend to cross check each other, and if there is a fault in one, it again will trigger a CEL.

I believe there is a fault in the ECM/ECU of the vehicle or somewhere in the power input of that component. Which is why it's not capturing a CEL for a fault in some other sensor or sub-system. It only knows it's seen stalls without an engine off request, and missed-starts.

These cars are pretty complex too. So it's wholly possible it's anywhere. With OBD-II, it's generally able to give more clues via CELs which is what is frustrating here. Which leads me back to suspecting ECM/ECU.

Replies inline:
- When it stalls, does it stall cleanly or does it run rough before stall? ->It stalls cleanly, like it slowly idles down and stops. No stumble like I would associate when you run out of gas in the tank (haven't seen that in a car in decades, maybe they stall more nicely now?). It just peacefully drifts off over a couple of seconds and stops.
- If it runs rough before stall, watch the ST trims as it stalls ->Will do, but it doesn't seem to run rough.
- I believe you said the engine also stall at speed (not just idle)? -> Yes! This was the original symptom - I was driving down the road and it stalled. It happened again a couple of times, so we turned around and went back home. I haven't driven it on the road since last Saturday though (about 5 days).

Definitely no CELs, on either the OBD reader or the Techstream. There are some DTCs (P1603, P1604, P1605), but no CELs.

You wrote " I believe there is a fault in the ECM/ECU of the vehicle or somewhere in the power input of that component. Which is why it's not capturing a CEL for a fault in some other sensor or sub-system. It only knows it's seen stalls without an engine off request, and missed-starts" .

This seems like a really good theory. Any thoughts about how I can diagnose it (other than swap in a known good ECM)?

 

[CharlieS]

No troubleshooting, just more data before heading to work:

I spent a bit of time this morning doing test runs (parked in the driveway for safety sake). It was cool this morning (59 degrees Fahrenheit).

All sensors are reconnected. The vehicle is in the condition it was in when this problem emerged, with the exception of more fuel in the tank (was at 1/4 when this started, full tank now added Saturday to confirm that it wasn't out of fuel), and a can of Heet dry gas in the tank (added Saturday as a troubleshooting step in case of water in the tank). I erased the OBDII codes at the beginning of the session.

I had a watch with a second hand, the Techstream connected, and a pad to take notes.

Unfortunately, I couldn't figure out where to see freeze frame data in Techstream, so I couldn't see some of what I wanted to check. Any pointers to find the freeze frame data would be super helpful.

I did 30 runs and the results were all over the place, and different than what I've seen previously (time wise). Of those runs I did 22 with my foot off the gas, and 8 with varying RPMs.

I'll touch upon my final runs first, since I found this interesting. I did 8 runs, two at each RPM: 2000, 3000, 4000, 5000. All of the 2000, 3000, 4000 RPM runs died quickly at 27 seconds. The 5000 RPM runs both died cleanly at 29 seconds.

Contrast this with my prior runs (lets say they were frequently, although not always) about 1:15 or so (not all were timed, some were not timed well, none were documented well).

The runs at idle were a surprise too.

The first run, cold start, it started right up, no hesitation, idled just north of 1000 rpm. Idle dropped to an even 1000 at around 32 seconds, at 45 seconds the idle smoothed out (it seemed a little quieter) but the RPMs were still at 1000, some valve audibly closed at about 1 minute (sounded like a servo operated valve or door) and at 1:15 it started to stall, slowly died over 2 seconds and stopped.

The next several runs were lost - my timing technique wasn't perfected/I got distracted with Techstream. They weemed to be around the 1 minute mark, but I need to discard them due to operator error.

Run 4 was a hard cranking start for about 3 seconds. The rest of the starts were all immediate with no perceptible long cranking.

Runs 7-12 and 14-19 and 21-22 all were in the 45-48 second range when the finally died. Of that time, the last 3 seconds or so were the slowing of idle prior to stall.

Run 13 was an outlier. The stall was at 1:07 and it took about 5 seconds to go from reduced idle to slowed to stall.

The one thing I saw in Techstream that I don't exactly know how to interpret is some misfire data. There are misfires detected by the ECM - there are 4 values with data: misfire rate shows a value of 3, and then misfires are detected on various cylinders (4, 5, 7 and 8) . Cylinder 4 and 8 show a value of 2 while 5 and 7 show a value of 1. These values neither increased nor decreased during the session, at least to the best of my knowledge.

Also, I'll reiterate that the truck seems to run perfectly prior to the stalls. It seems like it is in tip top running condition at that time. No rough running, no hard starting, no stumbling or hesitation, no lack of power.

The countdown clock just started. The local Toyota dealer's next open slot is Sept 10th. I'll have it towed to them on the 9th. I have a week to figure it out, before it goes to them. They have two very senior techs that have worked on my other land cruisers who are amazing, so I have no doubt that they'll figure it out if I can't.

 

[CharlieS]

I admire the effort you are putting into discovering/solving the issue you are having with the 200. And I feel for you!
My Porsche is a '95 993 C2. It was a dedicated track car (and RS tribute of sorts) and is now again "street-able"......meaning it now has a license plate and two seats instead of a solo configuration! I've actually put a decent looking interior back in and spent some time on the details. She's a beautiful car and allows me to get some track time in while also still being able to cruise around on the weekend (when there is no chance of rain, road construction, or of it being too hot, or too bumpy, or too many birds, or idiot drivers...….)

Nice! Sounds like a fun toy.

 

[TeCKis300]

I don't believe it's a fuel pump issue based on you inputs. Yet it still could be upstream such as the fuel pump driver that for whatever reason is completely stopping the party whenever whatever is tripping. I would expect it to not shut down cleanly every time if it were the fueling system however. Or run completely cleanly when running. A diagnostic fuel pressure gauge on the rail would conclusively say whether it is or is not fueling related.

The misfires are curious. So you're seeing live misfires in Techstream yet no CELs latched? I haven't looked at detail at the 200-series electronics architecture. The various misfires across several cylinder shows it's upstream. I don't know if this vehicle has an ignition control module or the coils are driven directly from the ECM/ECU.

So everytime the engine stalls, you're able to maintain comms with techstream? Rules out power issues to the ECM at least.

Too bad there's no logs or freeze frame data. That would help diagnose.

From your data, it's not a timing issue. Nor is it related to open/closed loop operation. There's a fault in the primary components that run the motor and not an ancillary subsystem.

 

[CharlieS]

I have a variable voltage driver (or whatever it is called) coming today, and I ordered the parts to tap into the fuel system so I can do a fuel pressure test (in their infinite wisdom, Toyota doesn't provide an easy way to tap into the fuel rail).

I plan to monitor fuel pressures while running once those parts arrive (Monday/Tuesday).

I want to run the fuel pump directly bypassing the whole fuel pump control ECU and ECM circuit for another test. If the car continues to run while the fuel pump is powered directly, it helps me reduce likelihood that the issue is the fuel pump. I can also simulate the reduced voltage coming from the FPC ECM to see if the lower voltage (< 6.5V) causes the fuel pump to stall/stop. It doesn't give me an answer, but might help me feel confident that the fuel pump isn't the issue (I don't think it is, but i want to feel sure of that). I would feel better if I ruled out fuel pump as a root cause with diagnostics (in addition to logically).

If I combine those tests for a third test, I can see whether the fuel rail remains pressurized with the pump being driven directly. That might help eliminate in-tank blockage as a culprit.

II just verified that the ignition is driven directly by the Engine Control Module (ECM). The EWD show ECM pins 33-40 as going direct to the ignition coils. (and incidentally the injectors are also driven by the ECM, pins 83-86 and pins 106-109)

 

[CharlieS]

there are other possibilities based on if it decelerates slowly or rapidly before stalling

Romer (and others): I'm reading through the FSM diagnostic procedures and as you pointed out, they use the concept of "slowly" and "rapidly". Is there any place in the FSM that quantifies these terms?

Is the 2-3 seconds from when the engine starts to die to when it completely stalls slowly or rapidly? While observing, it seems slowly, but when I type it out, 2 or 3 seconds sounds rapid.

There are also branches where it says "normal" and "abnormal". As in the question "Was the air-fuel ratio abnormal or normal?". Does anyone know how Toyota means this? Is that short term fuel trims (SHRTFT1/SHRFT2)\ in a particular range? Or is there some other observable value that I can watch (and what is the "normal" value)?

It matters a bit because the troubleshooting steps branch based on this determination.

The first question is "While idling or while driving" so maybe it doesn't matter. It happens both while idling and driving. Driving only has 3 options: "sensor malfunction", "injection and ignition stoppage" and "fuel supply problem". So let's say it needs to be one of those three.

Now looking at the "while idling" branch, I need to decide if it decreased slowly or decreased rapidly when stalling. Lets say I say it decreased rapidly (2-3 seconds is pretty rapid, I guess), it narrows it to two options: "injection and ignition stoppage" and "load from external parts". The commonality in these two branches of the tree is "injection and ignition stoppage".

If both of my answers were correct, than the only commonality between the idling and driving branches is "injection and ignition stoppage".

This leads me to the next table, which says the suspected area is "power temporarily cut". It says the primary parts to inspect are the Power supply circuit (fuel injector, ignition coil assembly), and points me to procedure 62 and 63.

If, on the other hand, I say the decreased slowly during the stall, I need to decide whether the air-fuel ratio was normal or abnormal. if I knew what to interrogate to get this answer, I could look at the appropriate branch.

Let's say it is "normal" - then there are only two options: "Intake air volume insufficient" and "ignition timing incorrect". No commonality/overlap with the "running" brach possibilities. Let's rule that out.

If it is "abnormal" - then there are three options - "intake air leak", "sensor malfunction" and "fuel supply problem". The two commonalities with the driving branch are "sensor malfunction" and "fuel supply problem". If we have a sensor malfunction is suggest that the primary parts to inspect are: mass airflow meter, engine coolant temperature sensor, air fuel ratio sensor, and water inlet sub assembly with thermostat (procedures 8-20 and 36-48). If it is a fuel supply problem then the suspect parts are the fuel pump control system, purge VSV system, fuel line or ECM (procedures 21-34).

So logically this narrows down what I need to look at, in addition to the p1603 diagnostics. I guess I know what I'm doing all weekend.