Disable EGO correction during ASE?

[elaw]

In general my car runs very well, but unless it's run rich for a few moments after a cold start it stumbles a little. I think that could be corrected by adding a bit of ASE and delaying EGO correction... but as I'm running a cat I like to have EGO correction come in as early as possible, and the optimal delay would depend on temperature.

So... how about an option to disable EGO correction when ASE is active? That way it would be disabled for longer when the engine is colder. of course the "EGO delay after start" setting would also have to be honored, to allow time for the sensor to warm up when restarting a hot engine.

Edit: thinking about this a little more, the above might not delay EGO correction for long enough. Maybe allow making the "delay after start" timer start after ASE ends, rather than upon engine start?

[piledriver]

How long does it take before your sensor/controller give sane readings?

I can only see waiting for that, and for most setups that delay is FAR longer than ASE on an already warm engine.
(Think about it--- what would having EGO (potentially) fight the trailing end of ASE for a few seconds really hurt?)

[elaw]

The sensor warms up pretty fast... about 10 seconds.

The problem is that using the sensor in closed-loop will only get me the target AFR from the table which at low load is 14.7. When I've just started the engine (and this pertains more to cold than hot), I'd like it to run a little richer than that for just a few moments.

Actually, maybe the ideal way to achieve that would be for the MS to take ASE into account when computing the AFR target. So supposing at some given moment, ASE is calling for 10% extra fuel and the entry in your AFR table calls for 14.7, make the actual AFR target 13.4 (if I did the math right).

[piledriver]

IIRC TS accounts for WUE and ASE, perhaps the FW could do it as well?

I haven't run into an issue, but I have EGO on a 30 second delay.

[elaw]

Bringing this back from the dead...

During the summer the above has not been much of an issue for me. But now that it's getting colder the problem is back.

This morning when I started the car, ASE was in effect for about 30 seconds. But the sensor came online after 20, so between the 20 and 30-second marks, EGO correction was fighting ASE and the mixture ended up leaner than I want it to be.

Part of the problem here is quirk of how my engine runs. After a cold start, the first time you try to accelerate it goes very lean and bogs. That does not happen on subsequent accelerations. My theory is it's because the port walls are dry, and putting in a lot of ASE seems to help the problem, probably by depositing extra fuel on the walls. But of course that won't work if EGO correction is fighting ASE.

And yeah, I know using EAE is probably the "right" solution, but I've tried and have never been able to get it tuned well.

[elaw]

Hey I was thinking about this a little more this morning and came up with a potentially better idea: how about giving the option of applying ASE to AFR target instead of directly into the fueling equation?

As in, if ASE is 20% at any given moment, make AFR target equal to the target from the table divided by 1.2 (1 plus 20%). So a 14.7 AFR target would become 12.25 with 20% ASE. Obviously this would only work if "incorporate AFR target" is enabled.

The cool thing is this would completely fix any conflict between ASE and EGO correction, as EGO correction would automatically target the ASE-adjusted AFR value.

[krisr]

This is how OEM's do it for ls1/e38 pcms. Problem with us as you said is the option to include afr into the fuelling equation but I guess that would mean that all corrections like this could be optioned too. I for one would prefer CLT warmup be a multiplication of afr but maybe an 8x8 table so there's more granularity for guys with big cams that need more idle fuel when cold but not as much when moving as the cam is working resulting in a really fat cold mixture.

[jsmcortina]

I like that idea for ASE and WUE. If "include AFR" is on then the adjustment is applied to AFRtarget instead of "Totalcor". The pulsewidth result is the same, but EGO correction has a better chance of working.

James

[krisr]

Yeah, for guys with full time widebands we could just run closed loop from the get go once the sensor is warm. Might be able to avoid that situation I mentioned above then..

[elaw]

I was thinking about this a little more now that cold weather is again approaching...

Maybe the most flexible way to do this would be to offer the user 3 choices of "ASE affecting EGO correction":
1) Off
2) ASE factored into AFR target
3) EGO correction disabled until ASE=0

All the conditions for enabling EGO correction would still apply. If option 3 was chosen, EGO correction would be inhibited until a) the "EGO delay" timer runs out or b) ASE=0, whichever is longer.

[subwoofer]

I may be thick, but my idea of the purpose of ASE and WUE was to add the fuel required to reach the programmed AFR target, not to shift it?

[elaw]

I think it can be either one. At least in theory if you've got EGO correction set up right and all you wanted was to reach target, EGO correction would do the job and there would be no need for WUE and possibly ASE.

But in my opinion there are two other aspects to the problem. First, whatever O2 sensor you're using will take a finite time to become active after starting so another mechanism is needed to add fuel before correction is possible based on sensor output.

The other aspect is that (and this is just a theory of mine) when the engine is first started, the intake port walls have no fuel on them. At least on my engine, the result is the first time you put a load on the engine after a non-hot start, it stumbles a bit because the mixture goes lean. In theory EGO correction might fix that but in my experience it can't react fast enough. So at least on my car, I want it to run rich for a few seconds after starting to deposit additional fuel on the walls and combat that problem.

And yeah I know the *right* answer to issue #2 is EAE but I've never been able to get that even close to tuned properly. And IMHO the effort required to do that is not sensible just to alleviate a problem that exists for about 2 seconds each time the car is started.

[krisr]

Option 2 would get my vote but wouldn't just stop at ASE, I'd have all forms of correction manipulate the target AFR/lambda/eqratio with the exception of MAT correction. There is some validity to keeping that one as is.