AFR lean on hot idle

[Davezj]

After a drive home from work, not booting it, switch engine off and leave it for 10 mins let's say, start it back up and it will run on petrol with about AFR 18.
If I rev it quite a few times it will drop back to 14.7 and idle. If I drive off it will drop back to normal 14.7 AFR when it idles again.

I am not worried about this just interested on what it could be. I have not yet tried putting my stock map back on to ensure it is nothing I have modified in the rom.

But it is only hot idle, it is fine from cold and it is fine when hot after a drive, but switch the engine off and then back on 10 mins late and it is lean.
It is the se on LPG as you would expect.
So let me eliminate the LPG for starters.

[Confused]

Mine idles like crap when it's been restarted a few minutes after it's hot.

I've always attributed it to heat soak affecting the inlet air temperature sensor - revving it a little or driving it a short distance sorts it out - the air flow over the temperature sensor brings it down to the correct temperature again. (But, when leaving it idling and not switching off after a drive, this heat soak doesn't occur because of the continual air flow)


Well - that's my theory anyway. Maybe logging the temperature sensors via EvoScan might help prove or disprove this theory?

[Davezj]

Good idea. That definitely sound plausible. I will try that just to see if it is heat soak.

[ersanalamin]

hI Dave i have same problem, when cold start, afr seems normal, but tend to lean when engine idle and hot. any idea how to fix this?

[fassi1]

https://www.clubvr4.com/forum/showth...elerator-pedal

That may help you.

[ersanalamin]

https://www.clubvr4.com/forum/showth...elerator-pedal

That may help you.

Yep, I ve read it, but i dont think faulty ICV lead to / has direct relationship with lean AFR on hot idle.

[kc427]

I have the same problem for many years and still cannot fix that. I checked no vacuum leak(there was leak on throttle body shaft, but changed a 2nd hand one, and no more leak), changed new fuel pressure regulator, new PCV value, new BOV, new 255 Walbro fuel pump, new fuel filter, new air filter, tried 2 different used air flow meters, but until now, still cannot find what problem it is. The AFR is about 16-18. But if I turn on A/C the AFR is back to 14.7, but once I turn off A/C, AFR back to 16-18.

[Davezj]

I believe this to be active and short term Fuel trim issue. It active fuel trim gets to it maximum value and after a certain time period about 15 seconds it just switched off. It is only when you blip the throttle again that the active fuel trim springs back into life and starts to richen up the idle mix. If you blip it a few times it will bring it back down but as soon as you stop bliping it will go lean again.
I did a test on my car and plotted active fuel trim, short term fuel trim, O2 sensor and it all makes sense when you look at the together on a graph.

Sent from my MI 5s using Tapatalk

[kc427]

yes, and I step on the throttle smoothly to around 2000rpm, ARF is still lean, seems that fuel trim not triggered, and keeps lean

[Davezj]

Best thing to do is monitor them with evoscan to get a good understanding of how they effect each other.

Oh and watch the video I linked to that explains fuel trim in my other thread.
I can't remember which thread it was though.

Here is a link to the video.
Very comprehensive
https://youtu.be/oRX2V6_a3dc
Sent from my MI 5s using Tapatalk

[kc427]

I am thinking to lower the injector size in the rom in another thread sometimes ago, so that the AFR can be richer a little bit, otherwise, I can't pass the emission test for the high rpm lambda and CO

[Nick Mann]

You can lower the CO by getting the catalyst VERY hot. Drive it like you stole it! Then test the CO.

[Confused]

Yup, the cat is a LONG way back in the exhaust, and the turbos take a lot of energy and heat out of the exhaust. You need to drive it like you stole it, then have the emissions tested as soon as possible.

Or, you need to rev it at around 3-4k for about 5 minutes to get it warm enough - then the readings will come right down, and it'll pass.

But, many MOT testers don't know these cars, so try to test them like a modern car, which has the cat right up hanging off the heads, where they get nice and warm very quickly.

[kc427]

yes, the CO is another issue the I knew I need to make the CAT hot enough to pass. But I have another problem is, once in hot idle, it is as lean as 17, and then even I rev to about 2000-3000rpm, the AFR is still around 17, and hardly drop back to 14.7, and that's the weird to everyone. That's why I am thinking to lower the injector size so the it won't as lean as 17. Also, I am thinking to remove the injectors and clean them, you know, almost 20 years old car and 200K on the clock, the injectors have never been removed and cleaned, so I am not surprised that have a little bit clogging.

[Nick Mann]

Are you running the original lambda sensor? If so I would thoroughly recommend changing to a wideband setup with narrowband emulation.

[kc427]

Are you running the original lambda sensor? If so I would thoroughly recommend changing to a wideband setup with narrowband emulation.

I had that problem when using original lambda sensor. After that, I re-wired it to use the simulated narrow band from the PLX gauge, and the result is exactly the same. So narrow band from the PLX and original lambda sensor have the same result. Now, I keep using the narrow band from the PLX gauge, and also tested with the emission test machine from garage workshop, the lambda readings are accurate. So, it shows that the exhaust gas is lean.

[crazydriver81]

But I have another problem is, once in hot idle, it is as lean as 17, and then even I rev to about 2000-3000rpm, the AFR is still around 17, and hardly drop back to 14.7, and that's the weird to everyone. .

Now that the temperatures re going up I start to enter similar Symptoms. I have a PLX wideband with Narrowband Emulation to ECU, cleaned and flow matched injectors installed 10k kms ago

When hot, car idles at 16.0-16.5 ish. When really hot, it eventually gets 17ish but usually gets back to 16.5ish then. Reving the car does not Change much, even driving in first gear (if you don't Floor it) does not get it back to 14.7ish.Only if I continue to drive like 2nd gear, then it turns back to normal. I am not an ECU expert so I don't know, which values the ECU Needs to trigger the different fuel trims.

Given, that even when cold idling, the car leans out to 15ish AFRs pretty quickly (ambient temperature around 20°C) I assume that this is Kind of normal behavior.

Did anybody tried yet, to drive with AFM disconnected, so that ECU has no values for airflow and intake temp? Does it Change anything in AFR's, when engine hot?

[kc427]

Now that the temperatures re going up I start to enter similar Symptoms. I have a PLX wideband with Narrowband Emulation to ECU, cleaned and flow matched injectors installed 10k kms ago

When hot, car idles at 16.0-16.5 ish. When really hot, it eventually gets 17ish but usually gets back to 16.5ish then. Reving the car does not Change much, even driving in first gear (if you don't Floor it) does not get it back to 14.7ish.Only if I continue to drive like 2nd gear, then it turns back to normal. I am not an ECU expert so I don't know, which values the ECU Needs to trigger the different fuel trims.

Given, that even when cold idling, the car leans out to 15ish AFRs pretty quickly (ambient temperature around 20°C) I assume that this is Kind of normal behavior.

Did anybody tried yet, to drive with AFM disconnected, so that ECU has no values for airflow and intake temp? Does it Change anything in AFR's, when engine hot?

Yes, almost same as you, revving the car does not change much, however, if I put in 1st gear and start moving the car, everything back to normal, ie, 14.7 again.

[Nick Mann]

I am also running PLX with narrowband emulation and I don't have this issue. My car idles at 14.7 reliably. I am not sure why?! I have a Walbro 255, a Sard FPR (I understand it was set up to use slightly higher than stock pressure, but I'm not sure what) and a 7202 re-mapped by Jason.

[Anderz]

This seems to happen when the signal from the lambda is static (zero) for too long, ECU seems to interpret it as lambda failure and go to open loop.
You can check with evoscan that "fueltrim in use" and "O2 correction" goes to zero when this happens, as Dave said
The problem is fuel scaling is too far off at idle, maxing out short term fuel trim. You should change scaling so that it will idle near 14,7 in open loop

Changing values in fuel map won't totally fix it, as they are not read in closed loop. But of course if it idles near 14,7 in open loop, the lambda signal wont freeze.
Closed loop AFR is changeable as well, but don't remember address now, it is 14,7 from factory