Stop responding to gas pedal when cold

[Arkis]

Hello, as the temperatures dropped, the car developed an interesting characteristic. Normally, it starts with the first key turn. It pulls out of the driveway, and then stops responding to the throttle after about 200 metres. Usually it just gets to idle, but today it stalled the engine while it was still in D (drive). I can step on the accelerator for a second or two as much as I want and it doesn't do anything. And then suddenly it catches up like nothing happened and goes completely smoothly. It only does it on completely cold starts when the car has been parked for days. For example, if we go somewhere, the engine cools down and we start it up the same day, it doesn't do it. It really has to stand for a day. What could it be doing?

I was worried I screwed up head gasket job because the car was smoking white-grey on first start after the service, but the cylinder pressure test came up good (13, 11.5, 13 and 14 bar). All cylinders hold pressure. The coolant is not going down so I guess it was just cold start combined with rich fuel mixture.

I measured the fuel pressure on the rail and it's 3,5 bar so it should be good.

The throttle position sensor (TPS) sensor has readings too - 0,68V - 3,88V depending on which position the throttle is in.

I cleaned the MAP sensor, which was pretty oiled.

Lambda probes:
pre-cat 0.1-0.9V (quite a jumping reading, but that may be because it was cold at 10°C)
post-cat 0.4V while idling. Jumping 0.1-0.9V while driving depending on the engine load.
Engine pressure 17.8-18.1 in/Hg (60.3-61.3 kPa) at idle when cold. That should be fine too. I guess it's not pulling air anywhere.
Intake manifold pressure 5.1 psi (about 0.35 bar) at idle (1100 rpm). That should be ok too.
Intake air temperatures shows good data.

At idle like this, everything seems fine.

What else could be affecting the first few hundred metres of cold run? The water temp sensor is new and diagnostics says it works, the data seems believable. I replaced the first lambda sensor before, because the original one wasn't warming up. I didn't replace the second one, and I'm pretty worried that it was a mistake and should have been replaced as well to make it a pair from the same manufacturer.

[Vladimir1977]

Have you plugged in lexia? Any faults related to injection?
You could try driving with cruise control. If any difference, than the suspected area is throttle pedal.
Also the problem could be caused from bad crankshaft (engine speed) sensor or camshaft sensor.

[Arkis]

Unfortunatelly no codes. I tried with Lexia and generic OBD ELM tool. No codes on both of them.
The car does not have a cruise control. It has cable driven throttle body.

The crankshaft/camshaft sensors are one of the things that did not go through my mind. It will be quite tricky to test them out so I am considering just swapping them. I am only sceptical about the timing of the problem. If the sensor was failing, wouldn't it be at random times? My problems seems to be related to cold start and happens only after about few hundred meters. Then the car runs without problems.

[Vladimir1977]

You coud also try live data reading, focusing on air and fuel parameters and cranc-cam comunication.
Prepare everything, and your wife or friend drives, and you following lexia with phone on camera.

[Arkis]

I would like to renew this post. The problem still occurs from time to time. I have never managed to capture in on live data. Even when the problem occurs, the sensors seems to be working normally.

Via live data I only saw O2 sensors (lambdas) throw bad data so I changed them both.

Next weekend I will try to measure the injectors if the resistance is ok and if o-rings on them are good. Today I didn't have much time so I wanted to check at least if some of the injectors were leaking or not. So I pulled out plugs to see if they are wet and 3 out of 4 plugs were oiled. But no sign of fuel. The oil level is constant so it's not leaking into the cylinder from the head. I think the oil could have poured to the sparkplug wells when I was setting the valve clearence. At least I hope that is the case because I really don't want to know how else it got there. The cover gasket was changed last year when I was doing headgasket job. So it's not from valve cover.

But seeing this I started to think if this isn't my problem origin. The oil might have been slowly seeping onto the plugs and saturating them. The problem only occured when the car was not driven for a while. So the oil had time to seep into the plug and then it was burned off during first few minutes of the ride. The problem was only happening during first few hundred meters and never after that. Could oil seeping into the plugs make the engine "cought" so it does not respond to the throttle?

On the pictures are the spark plugs. They are NGK 4619 around 5k kms old. I think the color of the plugs is ok?

[Arkis]

Looks like new idle control valve, cleaning whole throttle body and resseting learned parameter solved the problem. Two weeks after changing it the problem did not come back.

Oil problem was irrelevant to this. Oil was from when I was doing headgasket. After cleaning it, it never came back.