2013 dies repeatedly while driving down the road

[CharlieS]

Yeah, what you're seeing is definitely one of the most obscure issues I've heard of with these trucks. I hear you with the "while I'm in there" argument for replacing the fuel pump but from a debugging perspective you're introducing new variables into the equation. Not saying I wouldn't consider the same thing, just that my personal preference while debugging is to eliminate variables, not increase them. Since you tested the pump and verified it's working I'd leave it alone from there. Instead, before you close up shop, go back and replace the fuel pump once you sort out and address the root cause. That takes advantage of the knowledge and experience while minimizing rework. Probably not a huge deal, though, either way. I guess I've just been burnt by buying new parts that turn out to be defective too many times.

It sounds like at this point you're down to the ECM, the fuel pump ECU, and the harnesses related to those?

Yes exactly. I would not touch the pump until I have the root cause addressed. Just something I'd do before buttoning up the 2nd row seat area, after this is fixed.

I think so: the fuel pump ecu, the ECM, or some input to the ECM that is causing it to cut fuel. Without more facts, that third one will be hard to figure out.

I'm not leaning toward the wiring harness unless I get new facts that push me in that direction.

 

[CharlieS]

Do you think the modulation of the fuel pump speed is governed by the pump ecu? Based possibly on an air/fuel mix value from the engine ecu. That was a question, I am just looking for a culprit that could do what you witness.

Yes, based on the fsm wiring diagrams, and other Toyota documentation, I do believe that ECM signal to to the fuel Pump ECU results in the modulation of the voltage sent to the fuel pump.

The information I've seen doesn't go into great detail about which inputs are used by the ECM to determine the fuel pump needs (volume?) to keep the injectors well supplied. Most of what I've seen before has more to do with ignition and fueling maps, which impact timing and fuel injector opening dwell time.

Power enters that Fuel Pump ECU at battery voltage and comes out stepped down. I'm not a electronics guy, but I imagine a variable voltage signal from the ECM that drives a variable voltage resistor which steps down the input 12V+ (on the black wire) to the voltage going out (on the black wire with white tracer) to the fuel pump (10v, 8v, whatever).

Side note for people finding this thread in the future: I'm not sure how much of this is model year specific. The change from 3 speed to modulated pump happened for the 2013 model year. It looks like maybe '13-14 specific. Prior years likely have a different setup.

 

[speyrod]

I haven't read this entire thread and thought I posted this earlier. I recall reading in the Owner manual that there is an emergency cut off for the fuel pump. I don't know what triggers it to shut off and this may have been address, but add incase it hasn't been. Good luck!

 

[CharlieS]

Thanks. I'll see if I can find out more. Leave no stone unturned!

 

[bloc]

FYI if you get into the tank for the pump toyota says the module gasket is one-time use.

2013 fuel pump module gasket: 77169-02050

 

[CharlieS]

FYI if you get into the tank for the pump toyota says the module gasket is one-time use.

2013 fuel pump module gasket: 77169-02050

Thanks!

 

[biphnest]

A long shot... try driving with the gas cap removed.

 

[chilibit]

I suspect that the fuel pump shutdown function is to prevent consequences after an accident.

 

[CharlieS]

Just to be sure I am thorough: Some of the wiring diagrams and troubleshooting docs call for testing relays that aren't in the engine bay relay/fuse box (or perhaps are labeled differently?). For example, there is a EFI relay and an A/F relay in the EWD fuel pump circuit.

I can see an EFI NO 2 relay, but think that may be different (it is two different relays in the EWD).

The SFI System > ECM Power Source Circuit tech doc shows a box for the "integration relay", inside which it shows IG2 and EFI relays; and an IG2 fuse and a EFI fuse. The first inspection procedure is to inspect the integration no. 1 relay (step a. Remove integration relay from engine room relay block. Where is this - I don't see it labeled as such?

Are there relay boxes located elsewhere in the system, like for example adjacent to one of the other fuse boxes inside the vehicle?

I want to test and rule these out too.

This is important, since according to the 2013 EWD, the power to Fuel Pump Control ECU comes through the EFI NO 2 relay; and the power to energize the EFI NO 2 relay comes through the EFI relay which is activated by a signal from the ECM (coming from pin 34 labeled "MREL").

I suppose I could remove this EFI relay circuit from the equation for troubleshooting purposes (replace EFI NO 2 with a jumper wire), to see if it keeps running. That would help steer me in one direction or the other.

Update, I just jumped the EFI NO 2 relay and the car still stalled and died. This points away from the EFI relay and power to the Fuel Pump Control ECU as an issue.

Time for some back probing of Fuel Pump Control ECU connector wires...

PS. I'd still like to know where those other relays are.

 

[CharlieS]

A long shot... try driving with the gas cap removed.

For the sake of being thorough, I just tried running the car with the gas cap off, and nothing changed: it still died after a bit over a minute.

 

[CharlieS]

So, I back probed the signal wire from the ECM to the Fuel Pump Control ECU.

I'm not getting conclusive evidence, but I'll share what I think I learned. I need to redo the fuel pump voltage tests to see if it correlates, or if it is a red herring.

Note: The voltage signal from the ECM to the Fuel Pump Controller ECU appears to be continuously variable.

I made a video so I can reference back to actual values and the timestamp.

What I got from back probing the signal wire:

• At start the ECM signal to the Fuel Pump Controller ECU starts at 8.12 V then briefly spikes to ~10.63 V for a fraction of a second, and then drops down to ~7.83 V.
• At almost exactly a minute in, the signal reduces to around 6.33 V.
• At 1:15 in, it bumps up to 6.68 V, and quickly tails off: 6.66, 6.69, 6.55, 6.54, 6.12, 6.13, 6.09, then goes offline, over the course of 5 seconds.
• The stall sounds like it happens at 6.66 V (which is 1:18 since starting).
• The signal wire stays reading voltage for two seconds past when the engine dies out.

I'm left with more questions than answers:

• Is 6.6 volts a threshold value where either the pump has too much internal resistance and quits?
• Is there an internal failure in the FPC ECU: a threshold value under which it fails?
• Is there some other as yet undiagnosed issue on the ECM side where it purposefully cuts fuel?

The expensive way forward would be a new Fuel Pump Controller ECU and a new fuel pump, but I HATE just throwing parts at issues.

Any recommendations or suggestions for next steps?

I'll repeat these tests a few more times to be certain. Maybe I'll borrow another multimeter so I can measure both the pump and the ECU at the same time. Ugh.

 

[speyrod]

Where are you located? If not resolved over the weekend maybe someone local can add hands/eyes and perhaps a beer for assistance.

 

[CharlieS]

Where are you located? If not resolved over the weekend maybe someone local can add hands/eyes and perhaps a beer for assistance.

I'm in Charlotte, Vermont, USA. I'd gladly provide beer, food and bad jokes to anyone that wants to swing by. PM me if you're in the area.

 

[indycole]

Seems like you already sorted out the relay piece but here you go for location reference. I didn't see any relevant relays in other locations.

 

[CharlieS]

So, I back probed the signal wire from the ECM to the Fuel Pump Control ECU.

When the car starts, it sends

Seems like you already sorted out the relay piece but here you go for location reference. I didn't see any relevant relays in other locations.

Thank you again.

Any idea where that "Unit C" is located in the vehicle (it has the A/F relay, EFI Main relay, Horn relay, Ignition Main relay) - page 2 of the ERJB-Inn,pdf file.

 

[drftsub]

Just to throw this out there for a thought. I have had 2 race vehicles have similar issues, (one Porsche, one Subaru rally car) .

One was caused by an O2 signal that was erroneously telling the ECU that the mixture was excessively rich, and as part of the reduction of injector duty cycle the ECU reduced the PWM pump output and thus cut pressure.

The second was caused by a bad relay in the primary fueling control circuit. There was a relay that controlled all power to the pump etc at initial start and after 1 minute (ish) it was supposed to cut over to ECU control so that it could run the vehicle with feedback. The primary relay was bad, so as soon as the cold start cycle was done, vehicle would die.

Now I doubt either of these are your actual issue, but more a representative anecdote of it being things that could be outside the direct control circuit you are looking at.

 

[indycole]

The SFI System > ECM Power Source Circuit tech doc shows a box for the "integration relay", inside which it shows IG2 and EFI relays; and an IG2 fuse and a EFI fuse. The first inspection procedure is to inspect the integration no. 1 relay (step a. Remove integration relay from engine room relay block. Where is this - I don't see it labeled as such?

Are there relay boxes located elsewhere in the system, like for example adjacent to one of the other fuse boxes inside the vehicle?

I think unit C is technically part of the main fuse box passenger side fuse box near the air filter box. Check out page F of ERJB. You'll note connector 1B on there. The FSM directions you're working through mentions that 1B comes out of the integration relay. That should get you in the ballpark.

 

[TeCKis300]

Just to throw this out there for a thought. I have had 2 race vehicles have similar issues, (one Porsche, one Subaru rally car) .

One was caused by an O2 signal that was erroneously telling the ECU that the mixture was excessively rich, and as part of the reduction of injector duty cycle the ECU reduced the PWM pump output and thus cut pressure.

The second was caused by a bad relay in the primary fueling control circuit. There was a relay that controlled all power to the pump etc at initial start and after 1 minute (ish) it was supposed to cut over to ECU control so that it could run the vehicle with feedback. The primary relay was bad, so as soon as the cold start cycle was done, vehicle would die.

Now I doubt either of these are your actual issue, but more a representative anecdote of it being things that could be outside the direct control circuit you are looking at.

@CharlieS , from the timing of things with the engine cutting off roughly a minute after start, it does seem related to some switchover. Perhaps switching over to closed loop as @drftsub points out.

You've already tried starting the car without MAF and that doesn't seem to have effected the behavior.

It may be worthwhile to disconnect the primary O2 sensors in each bank to see if that changes any behavior. Disconnecting the O2s will remove the feedback loops such that it utilizes stable base maps.

Gotta give you credit for methodically working through this. It's a particularly difficult one.

 

[CharlieS]

Thanks.

New data point, maybe red herring: I did a test starting without the MAF plugged in. I knew it would throw two MAF related codes (P1102 and P0113), but when I went to read the codes, there was a third that wasn't there last time I disconnected the MAF sensor: P1118 Engine Coolant Temperature Sensor 1 Circuit High.

In the EWD there is a EFI Engine Coolant Temperature Sensor (C15) that is an input into the ECU (pin 75 labeled "THW"). The symbol (I think) indicates that it is a resistor.

The other day, I had what looked like an out of range temp reading** in the OBD reader, so swapped out the engine coolant temperature sensor on the front of the engine with another off brand from a local parts store. The OBD reader read 133 degrees Fahrenheit (and slowly increased to 136 degrees).

**The out of range temp thing is bugging me. The OBD reader has white text for values, but the ECT readout was blue, which I interpreted as out of range. Is that true, no idea... I observed/videoed it today when it went from 129 deg F to 131 deg F, the temp value text turned from white to blue. No other values are in blue. I don't know WHY this happens. It is a Foxwell NT-510 unit, and I am using it in generic OBDII mode.

So, to minimize variables, I put the original ECT sensor back to see if I did get this code again. I did get the code again (with the original sensor). The OBD reader said the temp was 142 degrees Fahrenheit (it went up to 155 while I was monitoring).

I also tried to start and run the car with this sensor unplugged and it was not very happy. It idled around 500 rpm, smoothly, but it was on the verge of too little RPM to keep it going. It still stalled at 1:15. With the sensor disconnected, the OBD reader said -40 deg Fahrenheit for ECT, so this has to be the right sensor. It also resulted in trac and low range flashing lights on the dash.

The plug to the sensor reads 5v on one terminal and has continuity to ground on the other.

Any chance there is another coolant temperature sensor that I missed? Or perhaps I need to grab a genuine part?

I've chatted with my father and think I'm going to pick up a variable voltage power supply to see if I get the same stalling behavior if I direct drive the fuel pump. I'm nervous what might happen if I put 12 volts to it for an extended period (recalling the warning from the fsm), so with this power supply, I can reduce the input voltage. That would bypass the entire ECM/PFC ECU circuit...

 

[CharlieS]

@CharlieS , from the timing of things with the engine cutting off roughly a minute after start, it does seem related to some switchover. Perhaps switching over to closed loop as @drftsub points out.

You've already tried starting the car without MAF and that doesn't seem to have effected the behavior.

It may be worthwhile to disconnect the primary O2 sensors in each bank to see if that changes any behavior. Disconnecting the O2s will remove the feedback loops such that it utilizes stable base maps.

Gotta give you credit for methodically working through this. It's a particularly difficult one.

Thank you.

I like the open loop/closed loop theory, but if I am to trust the code reader, that changeover happens quickly. It starts out on open loop (OL) and goes to closed loop (CL) around 10 seconds, and dies before I see another change in this status. Maybe that is just a problem with the reader I use (it only works when the car is running...).

I went to try the idea of disconnecting the O2 sensors and didn't see the connectors handy. Any pointers to where they are? I didn't get underneath yet.