Start, Chug, And Stall. Any Ideas?

[GuyTelefunken]

The patient: 1998 GS300, 105K miles

The symptoms: The cold crank time is noticeably longer than usual. To avoid undue wear on the recently replaced starter, I will crank the engine for approx. 5 seconds. On the second attempt, the engine will start, idle poorly and at reduced rpm, chug and die. On the third attempt the engine will start and idle normally.

The engine can now be restarted an infinate number of times without any problem. After a cool down period, say >2 hours, the problems will occur again.

The car is regularly serviced so items like filters, fluids, battery state of charge, etc. are in good shape.

Since I running short on ideas, I thought I'd give a Lexus dealer a shot at this. They started with the computer, of course no codes or misfires recorded. They cleaned the throttle body (even though I had it !Removed!'n'span). They removed and inspected the spark plugs. They were original and in top shape but nevertheless replaced as it was time per normal maintenence. Spark plug well #6 was found to be partially filled with engine oil which ran down from the oil fill area. The thought was that maybe spark plug #6 was being affected by the pooled oil. At this point, they thought the problem was licked.

The car started fine 3 times over a 6 hour period. The next morning, crank, chug, die.

Any suggestions or similar experiences would be appreciated. Currently, the car is still at the stealer....er...you know. Therefore, I need to make a decision to let them continue guessing or to bring it back and study it myself.

I'm thinking that residual fuel pressure is dropping but they claim they checked it and it's A.O.K. I am somewhat familiar with the returnless fuel system having only one fuel line from the pump to the rail but I do not know where the check valve is to prevent drainback to the tank. Is the check valve integral to the replaceable pump or is it in the pump housing?

[SRK]

The symptoms you describe are that of a failing fuel pump. Certainly in domestic cars that is exactly what happens. Overnight or long term cold, fuel pressure builds more slowly than the engine demands it. Multiple cranks engages the pump by bypassing the safety relay, and pressure builds before cranking, which produces the improvement you have noted.

I am surprised that the dealer would think a single spark plug could cause this. A blown head gasket could short out one cylinder, and it would clear up in a few seconds, but the oil would give that one away, as would the unexplained loss of coolant.

Fuel pressure gauges should be installed, engine run briefly, and then allowed to sit overnight. The pressure rise can be observed on the first cold start and during cranking.

I suppose the rule of thumb is that a good fuel pump can meet both pressure and volume demands, whereas a failing pump can do one, but not the other, at the same time.

I don't know where the residual pressure valve is - on Bosch systems it is on the pump itself, and we know how the Japanese copied Bosch.

[SKperformance]

I would have said a dirty throttle body or an improperly functioning IACV.

If they cleaned it then there should be no problem with them. I would see if turning the key to run 3-5 times before trying to start it helps. If it does help prime the pump then you would know you have a fuel delivery problem.

[GuyTelefunken]

Replaced the fuel pump. Problem remains. I'm back to square one.

[SKperformance]

When it runs it is fine right?

MAF?

damaged wire from a bad timing belt change on a cam or crank sensor.

Still a dirty IACV which would need to be removed to validate it is working properly. As the symptoms point at the iacv clearly to me. It may look clean but if it is sticking in one spot it takes a bit to finally open and then close after shut down.

[GuyTelefunken]

My understanding of the IACV is that is has little do due with idle or starting (see attached service data.) I can command the VSV on or off with my scantool. I can clearly hear it open and close.

As I understand it, idle speed is controlled by the throttle plate angle, not the IACV - Intake Air Control Valve.

Furthur: The timing belt has not been changed so I can rule out accidental damage in that area.

A major obsticle in diagnosing this problem is th inability of my scan tool (Snap-On Modis) to record live data while the engine is starting. It seems that most/all? scan tools cannot communicate with the ECM during starting; only before of after the engine is running. I talked at length to the Lexus technician and he concurs that the Lexus scan tool suffers the same issue.

I have confirmed that when the no start condition is present, the engine will crank for an infinate time on the first try. That is to say that holding the key to start for 30 seconds or more will not produce even a hint of combustion. If there was a fuel pressure problem I would think that the engine would attempt to start and stumble. If there was a problem with the MAF, I would expect the same.

The data I get from the MAF both before and after starting is consistantly stable.

IACV.pdf

[SRK]

That's definitely crappy that the new pump didn't fix it. Damn.

Was residual fuel pressure checked overnight?

You mention that the first cranking attempt will NOT start the engine, and the key MUST be released and another attempt made before it will? Perhaps the fuel pump relay isn't closing -which a pressure check during cranking would reveal. Maybe there is another relay that powers the ECM or the ignition, that wants to see another cranking attempt to close. I think that engine is coil on plug? Then a coil could be removed, and another loose plug inserted and carefully grounded to see if spark is available during first crank.

As to the IAC valve confusion, Lexus is using a term now reserved in ODBII for "Idle Air Control" valve, for "Intake Air Control Valve" which changes the inlet tract efficiency. The regular IAC valve is a bypass around the throttle plate. Base idle is set by opening the throttle valve a set amount. Any further air needed to maintain idle during warm-up, or AC on or off, or auto trans in drive, is set by that valve. Also, the IAC valve causes a "flare" during starting to ensure that all cylinders light quickly and keep emissions low, and keeps a slightly higher RPM during warmup to heat the cats and O2 sensors. All manufacturers have slightly different strategies, so I am not saying this is exactly what Lexus does, but I think its close.

What ever this is, the fact is isn't setting a code is still a clue to it.

Good luck.

SRK

[mehullica]

Are you using an OEM master key when this problem occurs? I have seen when 2 or more key are on the same keyring the engine won't always start, or dies before the key is relaesed from the run position.

[SKperformance]

Odd it has no mention of the throttle plate being kept open or a bypass chamber being used.

I was thinking not of multiple keys but just an issue with the transponder ring around the ignition.

[GuyTelefunken]

One OEM master key on my keyring, as usual.

I'm going by way of CKP or MAF sensor. Of course, the engine will not start or run with the CKP disconnected. Now the MAF sensor is another story.

According to spec., the car should start and run with substituted values when the MAF sensor is not connected. Here's what I have found:

Starting attempts: 50

Start and stall: 9

Start and continue: 31

No start: 10

Interesting. I would have assumed that it would start each and every time. I just can't figure that one out.

---------------------------

The explaination of idle control, from service data:

Idle_Speed.pdf

[98GS400Pearl]

Forgive me for opining among the gurus...If you are open to blue sky ideas from a novice...

Any chance it is a failed cold start valve or pressure regulator that is leaking and causing the engine to flood when it sits for long spells?

I am assuming the scan tool would catch a failed thermostatic switch/thermo time switch that feeds the cold start logic.

Could you have an oil pressure problem that is causing slow oil pressure build up, such that an oil pressure switch is not giving the OK signal to the ECU until a few seconds after initiating the start sequence?

[SKperformance]

Good thoughts , i would assume the ECU would store a code for any oil issue.

Any thoughts to see if the fuel pump relay is in good condition?

[GuyTelefunken]

I have the lab scope hooked up to the camshaft position sensor, crankshaft position sensor, ignition coil (cylinder 1) and fuel injector (cylinder 1).

Finally, the car decides to not start while connected to the scope. All signals are good except the injection. The injectors are not firing. Ahh, the question is why?

[SRK]

I am out of my depth on the Lexus injection control strategies, but on the Volvo/Bosch LH there is a "radio suppression relay" that helps to somehow control or form the current pulses that trigger the injectors. When it fails, the injectors receive no signal at all, while everything else is fine. Perhaps Lexus uses something similar.

[GuyTelefunken]

The injection system is rediculously simple. When the problem manifests itself, I have 12V at EA2 (see .pdf). Therefore, I must assume that the ECM is not commanding the injectors on because some condition of starting is not being met.

What I need now is a definitive list of things that must be present for the ECM to command the injectors on. I have confirmed the following:

Crankshaft Position Sensor - OK

Camshaft position Sensor - OK

Fuel Pump Control Module - OK

Fuel Pressure - OK

DTC's - None

What else is there?

Injection.pdf

[mehullica]

No fuel sounds like the immobilizer is not recoginzing the key. The ECU will shut down the fuel system if the immobilzer system sees a problem. I'd check out the immobilzer ring aroung the ignition.

[Limabean]

I'm having the exact same problem. However, I thought it might have been a clogged fuel filter - which is an integral part of the fuel pump housing. But I replaced it and that wasn't the problem. This thread seems to have gone dead without providing a solution. Does anyone know how this problem was or could be solved?