[FIXED] 2007 Lexus ES350 won't stay running

I'll try to describe the problem from the beginning.
2007 Lexus ES350. 240,000 km (148,000 mi)
With no warning, the engine started to run rough, giving my wife just enough time to pull over to the curb before dying. It would not crank over - low battery. I boosted it to get going, and about 10 minutes later, it died again. Would not start, even with boosting. I disconnected the battery posts, cleaned them, and boosted it to get going. I drove it home & did a thorough job of cleaning the posts, checked the charging voltage (14.5 volts) and left it over night.
The next day, after starting 4 times during errands, it died for good. I had to get it towed home, where it sits today, 2 weeks later. I keep a charger on it to ensure battery voltage for troubleshooting purposes.
It starts quickly, runs smoothly for 1 second, then shuts down as though someone turned off the ignition. Each time I clear the codes and try it again, it does exactly the same thing and throws the same codes - P0351 to P0356. If I try to start it immediately a second time, it misfires as though it has bad coils. If I wait a few minutes, it starts perfectly, runs perfectly for 1 second, and then dies.
I put a Tee in the fuel line and measured the fuel pump pressure. It's right on spec - 49 psi.
I have an Innova 3150d scanner. Each time it captures the codes, it does a freeze frame on a different coil, but the memory dump is essentially the same:
Fuel Sys 1 OL
Fuel Sys 2 OL
Calc Load (%) 40.4
ECT(degF) 51.8
STFT B1(%) 0.0
LTFT B1(%) 0.0
STFT B2(%) 0.0
LTFT B2(%) 0.0
Eng RPM 1448
Veh Speed(mph) 0
Spark Adv(deg) 21.5
IAT(degF) 57.2
MAF(lb/min) 1.3
TPS(%) 19.6
O2S B1 S2(V) 0.015
STFT B1 S2(%) 99.2
O2S B2 S2(V) 0.015
STFT B2 S2(%) 99.2
Run Time(sec) 0
LAMBDA B1 S1 0.998
O2S B1 S1(V) 3.297
LAMBDA B2 S1 0.998
O2S B2 S1(V) 3.3
Command EVAP(%) 0.0000
Warm-up DTC Clr 0
Clr Dist(miles) 0
BARO(InHg) 28.3
LAMBDA B1 S1 0.998
O2S B1 S1(mA) 0.00
LAMBDA B2 S1 0.998
O2S B2 S1(mA) 0.00
CAT Te 11(degF) 55.6
CAT Te 21(degF) 55.6
CAT Te 12(degF) 55.6
CAT Te 22(degF) 55.6
ECU Volts(V) 13.964
Load Value(%) 19.2
EQ Ratio 0.682
Rel TPS(%) 0.4
TPS B(%) 51.8
ACC Pedal D(%) 15.7
ACC Pedal E(%) 31.8
Command TAC(%) 19.6
Clr Time 0h0m
I have a basic repair manual with all the schematics for this car, along with information showing how to remove and replace parts.
Not knowing the starting process, I would speculate that the engine starts with memory data instead of sensor data, and when it's running, it tries to get the sensor data to make adjustments, and encounters some bad data which causes the engine to shut down. I just don't know which sensors would cause the engine to shut down.
Can someone point me in the right direction?

Starting, running for one second, and dieing is is a sign of security issues. Although I am not real familiar with Lexus' immobilization system, I would bet that the computer is killing injector pulse. After it starts, runs for one second and dies, WITHOUT turning the key off, do you have injector pulse if you try cranking, again?

Any sign of a security light?

Thanks Pole71. You may be on to something.
I haven't been able to check the injectors for pulses, but I have observed the following:
1. The security light does not come on when the engine stops.
2. The security light starts to flash, and continues to flash when I turn the ignition off. 10 minutes later, with the key fob NOT in the vicinity, it was still flashing. The doors were NOT locked.
3. Engine starts, runs for 1 second, then stops.
a. If I turn the ignition off, then on immediately, it repeats the same performance.
b. If I attempt to start it again without turning the ignition off, it sputters, but will not start.
Does anyone know how to reset the engine immobilizer?

Again, I am not very familiar with Lexus, but on other makes, if you turn they key on, the security light will come on. Leave the key in the "ON" position for 10 minutes. The light will go out. Turn the key off. Turn the key back to "ON" posistion. Wait 10 minutes, the light will turn off. Turn the key off. Turn key On. Wait 10 minutes, the light goes out. Turn the key off. Now, start the car.

Repeat the "Key on, light goes out, key off" cycle a total of 3 times (30 min). On the 4th time, start the car.

Be sure to have a battery charger connected.

Thanks pole71. I'll try that & let you know how it works.

I do not think you're dealing with an immo problem because :

1 - on this system immo problem will result in crank no start
2 - a related code would be set
3 - the codes you get point into another direction

You might want to start with an inspection of the ignition system, the coils for example share a common power feed.

Tried the 10 minute thing 3 times & it didn't work. The security light goes out as soon as I turn the ignition on. Regardless, I went through with the entire 30 minute process.
My Lexus Service Manager returned my call & told me that if it was a security issue, I would get no starting at all, just cranking.

Thanks Juergen. My Lexus Service Manager confirmed that I would just get cranking, and no starting if it was security related. But he's puzzled at the starting and shutting down after 1 second.
I'll start chasing the coils & injectors to see just what is shutting it down and let you know.

I attached a scope image of a coil IGT signal along with the ECM signal to an injector. You can see that the ECM stops sending signals to the injectors long before the coil IGT signal stops. From this I think I can conclude that the ECM has detected something that is causing it to shut the engine down by cutting off the fuel supply.
I apologize for the poor image, as I don't have an automotive scope, but used a lab scope and took a picture of the display using my iPhone.
Any ideas as to where to look next?

In the first place have a close look at the code setting criteria and the freeze frame data just to get a general idea of what's going on. ..

Then have a look into the IGF signal and if/when it cuts out. The coils` IGF signals feed into a common wire that runs to the pcm. Without a valid IGF signal the ecm may shut down the injectors to prevent damage to the catalytic converter.

I've got a better scope trace attached. Trace 1 (yellow) is the IGT signal for coil 6. Trace 2 is the Injector signal for what I believe is cylinder 6. Trace 3 is the IGF signal common to all coils. It looks non-existent! What should it look like? Could it have high resistance short to ground?

HA! Problem solved!
When I couldn't get a signal on the IGF line, I unplugged the ignition coils one at a time and measured the resistance from the IGF pin to ground. The first 2 were in the megohm range, but the 3rd was 16.5 kohms. I left it unplugged and the car started & ran on 5 cylinders. Obviously that one was dragging the IGF line down causing the ECU to shut down the engine.
Thanks to all for your help and suggestions.

"Don't be a parts changer" Analyse!!!

The IGF signal is a 0 to 5 volt square wave similar to the IGT signal. It should start almost immeaditly after the IGT signal starts to rise, maybe 2 ms later or so.

As each coil produces it's IGF signal on its own it is very improbable that all coils started to fail at once- check the IGF wire for a short to ground and also the continuity of this wire to the ecm (although an open in this wire would not take out the signal...) Also have a close look at the respective grounds for the coils.

There is a chance that one coil is shorted internally and hence the IGF circuit is dead. You can disconnect the coils one by one and check wether the IGF signal comes back .

Another option is to cut an IGF signal wire at one coil and then check with the scope on the coil side of the cut wire for a IGF signal. That way u can test coil by coil just connecting one by one to the connector with the cut wire.

Hey, just read your fix, glad you made it .

Thanks again. You pointed me in the right direction!