heat soak / injector dead time / system non-linearities

[flaco]

After the engine has had a good highway run, if I park and go back out in say 30-45 mins, the AFR's go crazy high (~18-19). So high that it is pretty much un-driveable in stop-go traffic (stick shift). I think heat soak is the culprit, and have tried to decrease the IAT correction factor, increase after-start enrichment values, etc.

However, if I give it a WOT run, the AFR's are normal. So the problem only happens in the low MAP bins. This has led me to believe the problem lies elsewhere. Particularly in the injector open/close time. Correct me if Im wrong, but everything else in the injector pulse time calculation is linear. It makes no sense that AFR's go high ONLY at low MAP values (~ <50 kpa). Again, they remain lean for a while until I drive the car on the highway for like 3-4 mins. It will NOT correct itself during stop/go traffic. In fact, the issue also shows up if driven in stop/go only.

To test my idea, I've changed from 1 simultaneous injection pulse to 5 simultaneous (only 'other' option available: engine is a turbo 5 cyl, VR on crank, hall on cam). When the setting is changed to 5 simultaneous, I try to adjust the injector dead time to get the same AFR's while idling. AFR's go lean when changed to 5 simultaneous. However I cannot get the AFR's back to where they were. I must change the VE map for it to just idle!

Haven't tried driving with 5 sim, since Im pretty sure it will result in not enough injection time / reaching 100% duty cycle.

So Im stuck with lean AFR's when hot. Injector dead time is set at 0.8ms right now. Injectors are bosch: http://www.witchhunter.com/flowdatapix/b0280155766.jpg

Any ideas how to fix this problem?

Am I on the right track here? Any other suggestions what to try?

[pigga]

Hi Flaco.
I have exact the same issue and could not figure out the root cause for this.
This mainly happens on hot days (so what's "hot" in Germany, let's say 30°C). If the car parks for a while (~30 Minutes) and I start after this, the engine runs really really lean @lambda 1.1x. I tried to correct this as far as I could by the non linear MAT correction table, but this didn't cure the problem 100%. So the Engine always runs crappy for 30 seconds until EGO correction kicks in.
I am quite sure that this has to do with heat soaking of the manifold...at least I thought.
The funny thing is that at WOT it acts similar to yours: The AFRs are matching. Or other way round: If I screw up non linear MAT correction, I will get rich AFRs at WOT for a while until it's back to normal.
Considering the fact that my (OEM) MAT sensor is placed quite exposed in the front far apart form the engine and if I think on it...I suppose it should be the other way round: If the MAT sensor measures lower °C than there are inside the manifold, the engine should run richer?
I know that the discussion was here some time ago.
The only Idea I had was to do some "non linear" MAT correction.
If several user had this problem it might be a solution to scale the Engine load depending on MAT like the baro correction does?
Thomas
P.S.: With injector opening time I had similair problems as well. It was possible to my app to get 80% matching opening times but they were never 100% dead on. I supposed that the OEM pressure regulator is the cause? It varies the pressure between 2.2 and 3 bar.

[svocapri]

I have seen similar issues with my 85 Capri TBI. At first I had the MAT sensor in the air cleaner, then I tried moving it into one of the runners on the manifold, finally I put it in the right front fender well inside the air intake pipe that leads to the air cleaner. This has provided the best results for me.

As for the difference in AFR between one and five squirts, I think you may be seeing and incorrectly set injector open time. I played with my injector open times on my car until I had it very close between two squirts simultaneous and four squirts simultaneous.

From what I recall, if it went rich going from 2 squirts to 4 squirts, then the injector opening time was too long. If it went lean going from 2 squirts to 4 squirts then the injector opening time was too short.

By the way, the injector opening time is actually a combination of the injector opening and closing time as there is a period of time from when the computer sends the signal to the injector during which the injector is not flowing fuel; but then after the computer drops the signal, the injector is still flowing fuel for some period of time before it closes.

Julian

[gurov]

i've been dealing with this for a while.

one thing to do is put the "gair" or "air correction" percentage out on the dashboard and see how that reacts. with high IAT's.

i've gone as far as to modify the INI file to let me set the mat correction value to 1, which effeectively gives absolutely no fuel corrections based on air temp by default, then i can dial in my own corrections using the non-linear MAT correction.

also, do not forget the voltage fuel correction, if the alternator isn't putting out enough juice one it's really hot (been sitting out for a while) that will affect your fueling (although, in this case it would be the other way)

[hakeem]

Blast from the past I know, but did anyone resolve these issues? I am having the exact same problem on my RX7 racecar.