Rough start when warm

[ThomHov]

Hi everyone,

I have a Clio MK2 1.4 16v that I'm really happy with. There are however a few issues with the car that I want to solve.

One of them is the rough start when the engine is warm. It usually takes 2-3 key turns and 10 seconds of cranking to get the engine running. When i mean warm, i mean at least 30+ minutes of engine running, so it's almost operating temperature. When is does run, the first minute or so it runs rough. Almost on 3 cilinders... When this stops, the car runs great.

I already when to the local mechanic to assess the issue. He did not experience the fault like I did, so that's weird...

Things I already checked:
- Fuel pump electrical feed
- Fuel pump flow (only starts on the second turn of the key?!?)
- State of injectors, like resistance and obvious defects.
- Replaced crank sensor
- Live data shows good engine temperature, even when not running.
- Spark plugs replaced
- Coils replaced with new Bosch (valeo was sold out)

What I am guessing is that the car runs on the cold map before the lambda has reached a good temperature, sprays too much fuel in the port and makes it hard to start.

Does anyone have a suggestion for solving this issue? I will soon make a video for better explanation.

Thanks in advance,
Thom Hovestad.

 

[botfch]

Need to plug it in and have a look at the live data when the faults happening ideally.

 

[ThomHov]

So, after work I've had time to do it.

I did a live read of the data stream. My crappy obd reader doesn't have a plot function, so I can't determine good polling speed of the sensors. did a test at idle and at 2000 rpm. These are the thing I found:

- Lambda pre-cat between 0.01 and 0.9, jumping each second or two up and down.
- Lambda post-cat steady at +0.45. at idle around 0.57, and at 2000rpm around 0.75.
- Short fuel trim bouncing all over the place
- Long fuel trim at around 17%
- Map 45kpa at idle. Did not look at 2000rpm.
- Crank sensor reading seems correct. the parameter goes along with the rpm.

The starting issue: of course it did not do it when I plugged the obd in. Tonight I will try again when the motor has cooled down a bit.

My evaluation:
The post cat lambda (+0.45v) says its a rich mixture, while the fuel trim is positive. So the ecm adds fuel, while the lambda already says it's too rich?!?

 

[ThomHov]

Motor has cooled down to 65 degrees water temp, after an hour. This time the Clio did start very rough. Seems like there needs to be some time between shutting off when warm, and starting it to get the effect.

Lambda goes to 0.99 v. This is the pre-cat one. Rpm during cranking reads 300 rpms, which seems good. After 20 seconds of running, there is a really pungent smell from the exhaust.

Greetings!

 

[ThomHov]

Still not solved.

I did do some more research. The lambda and fuel trim values do not make sense, so I suspect the pre-cat lambda to be faulty. The post-cat lambda confirms a rich mixture, even while the trim is already evened out. I will replace this soon. Don't know if it has already have been done yet in the lifetime of the car.

Also, I have ridden for a while and let the car sit for an hour like last time. Then unscrewed the spark plugs for ruling out a flooded engine (leaking injectors or somesort). they look like this:

Seems to be fine. No wetness on the electrodes and ok color. Any suggestions?

 

[ThomHov]

Here a example of the problem